Content uploaded by Marc Steinberg
Author content
All content in this area was uploaded by Marc Steinberg on Nov 29, 2024
Content may be subject to copyright.
54
A SERIES OF READERS
PUBLISHED BY THE
INSTITUTE OF NETWORK CULTURES
ISSUE NO.:
IN/
CONVENIENCE:
INHABITING THE
LOGISTICAL
SURROUND
EDITED BY
JOSHUA NEVES &
MARC STEINBERG
IN/
CONVENIENCE
INHABITING THE
LOGISTICAL SURROUND
Theory on Demand #54
In/Convenience: Inhabiting the Logistical Surround
Edited by: Joshua Neves and Marc Steinberg
Contributors: Neta Alexander, Armin Beverungen, Darren Byler, Liza Rose Cirolia, Maya Indira
Ganesh, Tomasz Hollanek, Orit Halpern, Mél Hogan, Tung-Hui Hu, Steven Gonzalez Monser-
rate, Rahul Mukherjee, Joshua Neves, Susanna Paasonen, Andrea Pollio, Marc Steinberg
Cover design: Katja van Stiphout
Design and EPUB development: Salome Berdzenishvili
Published by the Institute of Network Cultures, Amsterdam 2024.
ISBN: 9789083412559
Contact:
Institute of Network Cultures
Amsterdam University of Applied Sciences (HvA)
Email: info@networkcultures.org
Web: www.networkcultures.org
Order a copy or download this publication at: www.networkcultures.org/publications
This publication is licensed under the Creative Commons Attribution NonCommercial Sha-
reAlike 4.0 Unported (CC BY-NC-SA 4.0). To view a copy of this license, visit www.creative-
commons.org/licences/by-nc-sa/4.0./
CONTENTS
Introduction 5
Joshua Neves and Marc Steinberg
In Convenience 11
Joshua Neves and Marc Steinberg
Convenient-for-the-People Policing, Protected
Consumer-Citizens and Infrastructures of
Disposability in Northwest China 34
Darren Byler
The Resilient Situation: Adaptive Management,
Finance, and Environment 48
Orit Halpern
Collectivizing Convenience? From Delivery to Logisticality 66
Armin Beverungen
Clouded Conveniences 82
Mél Hogan and Steven Gonzalez Monserrate
Waiting for ‘Day Zero’ 88
Tung-Hui Hu
Platform Economies, Reputational Stains, and
the In/conveniences of Porn 100
Susanna Paasonen
Theorizing ‘Anti-Content’: On Sleep Apps and
Horizontal Media 115
Neta Alexander
In/Conveniences of Mobile Payments: “Alternative Data”
and the Distribution Infrastructures of Loan Apps 132
Rahul Mukherjee
Beyond Inclusion: Glitchy Economies and the Promise of
Platformization in African Cities 149
Andrea Pollio & Liza Rose Cirolia
Easy Wins and Low Hanging Fruit. Blueprints, Toolkits,
and Playbooks to Advance Diversity and Inclusion in AI 163
Tomasz Hollanek and Maya Indira Ganesh
Biographies 177
5
IN/CONVENIENCE
INTRODUCTION
JOSHUA NEVES AND MARC STEINBERG
In a Netflix comedy special Asian Comedian Destroys America!, Ronny Chieng asks a ques-
tion that goes to the heart of this volume: ‘How much more convenience can we get?’ It’s also
a slightly disingenuous question, asked of the many services offered by Amazon in the US
(with Chieng nodding to Netflix as well but leaving his sponsor off the hook). Disingenuous
because while America and the overdeveloped ‘West’ is certainly the imagined homeland
of a certain era of conveniences—washing machines, microwaves, frozen pizzas (known,
significantly, as ‘convenience foods’)—the place of convenience today is decidedly more
global, and less domestic (at least in the narrow imagination of home appliances and pro-
cessed foods).
More global in the sense that the frontlines of what one business text calls the ‘convenience
revolution’ are arguably now the fintech experiments in South Asia, Southeast Asia, and
Africa; the e-commerce innovations of East Asia and Latin America; global and regional
streaming platforms; and, of course, sprawling cable and shipping networks, data cen-
ters, orbiting satellites, rare earth mines, express delivery, and so much more. With all due
respect to Chieng (and other America-focused accounts of convenience), we need a more
dynamic frame to understand the contemporary geographies and experiences of conve-
nience—including how convenience consolidates the platformization of culture, surveillance
capitalism, data regimes, conscription into debt, among other infrastructural expediencies.
Less domestic in the sense that if the locus of convenience was once imagined to be the
home, and gendered as the domain of women’s work, now convenience describes a choice
made about work: ‘ Drivers Choose Uber for its Flexibility and Convenience’ reads a 2015
Uber press release.1 Convenience blurs the earlier, artificially neat separation of production,
consumption, and distribution. It signals changing ideas about work, where labor often takes
place outside the home, office, or factory. Indeed, the hailing of convenient work accompa-
nies an understanding of the feminization and racialization of labor more generally. Gig work
and most notably food and grocery delivery are described as a new form of outsourcing the
work of social reproduction. What matters here is that the boundaries between prior sites
of labor and gig work blur, leading to what scholars point to as the gendered economies
of digital labor or a digital politics ‘where platform capitalism and racial capitalism meet’.2
The spaces and itineraries of convenience are themselves in flux. Platforms and the digital
shift have played major roles in this transformation. Put simply, we inhabit an increasingly
logistical surround.
1 Jessica, ‘Drivers Choose Uber for its Flexibility and Convenience’, Uber Newsroom, 7 December 2015,
https://www.uber.com/newsroom/driver-partner-survey/.
2 Melissa Gregg and Rutvica Andrijasevic, ‘Virtually Absent: The Gendered Histories and Economies
of Digital Labour’, Feminist Review 123.1 (2019): 1–7; Tressie McMillan Cottom, ‘Where Platform
Capitalism and Racial Capitalism Meet: The Sociology of Race and Racism in the Digital Society’,
Sociology of Race and Ethnicity 6.4 (2020): 441-49.
6THEORY ON DEMAND
But what does convenience mean? How should we understand it, in its proliferation of uses,
sites, objects, and experiences? Convenience is perhaps one of the most overused terms to
describe where we are going, why we are using a given digital tool, and what the world of
smart devices offers in exchange for our data, our money, our lives. Overused, and under-the-
orized.
In April 2022, amid a late pandemic thaw, we held a hybrid workshop on ‘In/Convenience’ at
Concordia University in Montréal to begin to explore such questions. The event, which was
informed by digital life during ‘lockdown’, was the culmination of a yearlong working group
exploring the cultural politics of convenience. It included eight presentations on themes
ranging from logistics, waste, water, disability, democracy, and policing. One presentation
began by sharing an n-gram of ‘convenience’ (see figure 1). The English-language graph
showed that the term had peaked in 1939 and, though still part of everyday language, has
steadily declined for decades. This downward slope was both expected—given convenience’s
association with postwar consumer culture—and a bit of surprise. Surely, we are living in a
moment of unprecedented access to and demand for conveniences of all kinds. Convenience,
it seems, has become so infrastructural to contemporary life that we barely need to call atten-
tion to it. But unlike the common refrain that infrastructure is only noticeable when it fails,
convenience seems to come sharply into view when it becomes unavoidable or compulsory,
as with gig labor, data plans, police checkpoints, water shortages, transportation apps, and
even access to basic services like health care.
Figure 1. Visualization of key aspects constituting convenience in the age of platforms, from Google n-gram
This book is a rejoinder to such uncritical ubiquity, offering a multi-situated view of convenient
practices and theories that are rooted in the present. In fact, rather than declining in usage—
as a quick glance at an n-gram might suggest—convenience has proliferated a spectrum of
meanings, practices, and technologies.3 Consider, for example, current synonyms, including
3 The rise or fall of its usage also changes according to language and geography. If the English
“convenience” seems on the decline, the Japanese Diet Library n-gram shows the Japanese term
riben (利便) to be sharply on the rise before the n-gram’s cut-off date of 1999: https://lab.ndl.go.jp/
ngramviewer/?keyword=利便&size=100&from=0&materialtype=full. This reminds us of the need for
site-specific analyses of convenience and its comparable terms.
7
IN/CONVENIENCE
terms like: surveillance, platforms, logistics, automation, on demand, fintech, or gig economy.
These entangled yet distinct vectors both constitute some of convenience’s crucial dynamics
and indicate the concept’s current unruliness or fecundity. If technologies change—the prem-
ise of convenience’s attachment to super apps, ride-hailing, digital money lending services,
and so on—so too does convenience itself. We glean insights from earlier theorizations of
convenience including Elizabeth Shove’s prescient observation that ‘the tools of convenience
have escalatory consequences’; as well as an earlier genealogy of convenience offered by
Thomas Tierney, including an emphasis on how capitalist modernity transforms luxury into
necessity.4 We also engage recent turns in anthropology and media studies, as well as pod-
casts, streaming, novels, and other popular practices.5 Yet the authors in this volume locate
convenience in other contexts, geographic and medial and classificatory, interrogating how it
can become a justification for policing; a soporific; a relation to infrastructure; an enticement
to debt; or a promise of infrastructural reliability.
In doing so we seek both to understand the political, even moralizing, undertones of critiques
of convenience—as a banal evil or ecological disaster that is ‘destroying us’—and also the
ways convenience may function as an imperative rather than a choice. It may begin as a minor
time-saving ritual or an aspiration but quickly becomes common sense, ‘just the way things
are’. A luxury for some that is a requirement for others. It is a feeling or atmosphere that we
need to interrogate in order to understand the appeal of platforms, the uneven demands of
‘on demand’, and the many other habituations structuring of digital life.6 Exploring the com-
plexity of living in convenience—a relational dynamic that is crucial to contemporary life—is
the jumping off point for this collaborative volume. That we inhabit the slash or solidus (“/”)
marking the relation of in/convenience as a structure of experience; and that this habitus is
logistical in nature are two insights that the chapters in this book probe.
There is an escalating need to theorize convenience to meet the escalating demands for
convenience. So too the ways the convenient intersects with deeply felt obstacles or untime-
liness. The need to scrutinize these relations is the starting point for this book, and hopefully
other work that will follow from it, exploring convenience in the age of algorithms, platforms,
and other logistical operations. One provocation taken up by several contributions to this
volume is Lauren Berlant’s posthumous monograph, On the Inconvenience of Other People
(2022). For Berlant, inconvenience describes ‘the affective sense of the familiar friction of
being in relation’.7 While their provocation—other people—is decidedly (anti)social in framing,
4 Elizabeth Shove, Comfort, Cleanliness and Convenience: The Social Organization of Normality, Oxford:
Berg, 2003, p. 182; Thomas F. Tierney, The Value of Convenience: A Genealogy of Technical Culture,
SUNY Press, 1993, p. 29.
5 Emily West, Buy Now: How Amazon Branded Convenience and Normalized Monopoly, Cambridge:
The MIT Press, 2022; Jenny Huberman, ‘Amazon Go, Surveillance Capitalism, and the Ideology of
Convenience’, Economic Anthropology (2021): 337-349; Ahmed Ali Akbar, ‘The Cost of Convenience’,
Land of the Giants podcast, 22 June 2021, https://podcasts.apple.com/us/podcast/the-cost-of-
convenience/id1465767420?i=1000526373677; Sayaka Murata, Convenience Store Woman, trans.
Ginny Tapley Takemori, New York: Grove Press, 2018.
6 Here we find inspiration from the turn to consider the structure of feeling of platforms in Geert Lovink,
Sad by Design: On Platform Nihilism, London: Pluto Press, 2019.
7 Lauren Berlant, On the Inconvenience of Other People, Durham: Duke University Press, 2022, p. 2.
8THEORY ON DEMAND
the focus on inconvenience’s profound attachments can also be extended to the kinds of
technologized life worlds taken up in this book: policing platforms, smart systems, logistics
and logisticality, cloud infrastructures, water shortages, sex media, sleep apps, micro-credit,
urban optimization, and AI ethics, among other issues. What unites these diverse essays is
a focus on the ways that proliferating digital conveniences also convene us. As Tung-Hui Hu
reminds us in his chapter, Foucault invoked convenience ‘to describe how similar things
were brought together in the 16th century to form the great chain of being: “Those things are
‘convenient’ which come sufficiently close to one another to be in juxtaposition”’.
While such juxtapositions vary in intensity and are unevenly recognized, they suggest a desire
or demand for social, political, and technical harmonization. As Berlant puts it:
The minima of inconvenience can go under the radar, or not, but it does not register
at first as a traumatic or transformative event. At maximum intensity, though, the
aective sense of inconvenience is harder, less easy to shake o or step around...
[T]he strong version of inconvenience points to forced adaptation to something
socially privileged or structurally pervasive.8
Berlant’s point about (in)convenience’s ordinary and organizational effects/affects is well
taken, especially its emphasis on how such relations are not only compositional but may be
compulsory. But as Liza Rose Cirolia and Andrea Pollio suggest in their contribution to this
collection, there is also something unsatisfying about critical framings based on inclusion
or exclusion—especially when it comes to questioning techno-politics in the Global South.
Instead, ‘moving away from a binary of in/exclusion’, they suggest, convenience may help
us ‘to see the multiple displacements of effort and work now held by new bodies, systems,
and processes’.
Such tensions both bring into relief a set of key themes animating this book and help us to
understand digital convenience as more than a question of consumer choice, a problem of
inside/outside, or something that can simply be resisted by turning off our devices or not
clicking ‘buy now’. We offer an extended theoretical grounding for this book in chapter one,
including a focus on convenience’s shifting temporal, spatial, and affective registers. There
we observe that more than the demand people place on platforms—a common critique of
lazy or selfish consumers—convenience is increasingly a demand placed upon people by
platformization; a condition in which we live. Responding to this dynamic is one of the shared
projects of this volume. From Darren Byler’s analysis of surveillant policing in occupied Xin-
jiang, Armin Beverungen’s call to ‘collectivize convenience’, or Orit Halpern’s suggestion
that we have moved away from convenience to an era of resilience, to Susanna Paasonen’s
exploration of the de-platformization of sex workers, Neta Alexander’s cripping of convenience,
and Rahul Mukherjee’s engagement with digital lending and fintech infrastructure, our col-
lective effort is to question how in/conveniences enable or inhibit certain forms of relation,
juxtaposition, or assembly. How, that is, that a certain type of world emerges or is foreclosed.
8 Berlant, On the Inconvenience of Other People, p. 4.
9
IN/CONVENIENCE
As suggested at the outset, we also aim to situate and follow these entanglements beyond
the familiar focus on a few national contexts, corporations, or habits of living. This book thus
offers a multi-situated view onto convenience as it has been transformed by digital media,
smartphones, and platform capitalism. Like all global views it is partial of necessity, and
penumbral, to return to a term used by one of us to describe something just coming into
view.9 Orit Halpern, in this collection, argues that we have moved from an era of stability and
equilibrium to one of complexity and instability; an era of convenience to one of resilience.
We might read this epistemological shift another way still: as a new description of how conve-
nience is transformed by logistics, supply chains, on-demand services, and data economies.
Such uncertainties resonate across the chapters, from the security systems managing data
centers explored by Gonzalez and Hogan, to the novel ‘convenience delivered’ model Bever-
ungen outlines. Emergent epistemologies are as crucial to map as the local sites where con-
venience is produced. We need to shuttle, too, between macro-level epistemologies and
the day-to-day objects and feelings with which we live. Here ‘convenient noodles’ (方便面),
as instant noodles are known in China, are as ripe an object of analysis as the convenience
stores where they are sold. Convenience today requires an understanding of attention, gov-
ernance, pleasure, logistics, optimization, and the platform-mediated movement of people
and goods that was incidental to earlier theorizations. This includes, of course, the need to
respond to convenience’s demands. As Tomasz Hollanek and Maya Indira Ganesh put it in
their analysis of AI ethics: ‘If convenience is the “condition we inhabit within contemporary
capitalism”, the key question is how? How can inconvenient questions about the trade-offs
and conflicts of interest be posed in ways that are both legible and bearable to those in the
position to transform the development pipeline?’ And by the rest of us as well.
Acknowledgments
We wish to thank GEM Lab and the Platform Lab for their support of this project; the graduate
student participants in the “In Convenience” reading group of 2021-2022; and the contrib-
utors for generously taking the time to write for this book. We also want to thank the Institute
for Network Cultures and Geert Lovink in particular for their warm support of this project
from the outset. Sudipto Basu was a great interlocutor throughout and we appreciate his
incredibly careful editing in the final preparation of the collection. Marc Steinberg would like
to acknowledge the support of the Volkswagen Foundation’s ‘Smartness as Wealth’ grant
and the support of the Social Sciences and Humanities Research Council of Canada. Joshua
Neves would like to acknowledge the Social Sciences and Humanities Research Council of
Canada and the Canada Research Chairs program.
References
Akbar, Ahmed Ali. ‘The Cost of Convenience’, Land of the Giants podcast, 22 June 2021, https://pod-
casts.apple.com/us/podcast/the-cost-of-convenience/id1465767420?i=1000526373677.
Berlant, Lauren. On the Inconvenience of Other People, Durham: Duke University Press, 2022.
9 Joshua Neves and Bhaskar Sarkar (eds) Asian Video Cultures: In the Penumbra of the Global, Durham:
Duke University Press, 2017.
10 THEORY ON DEMAND
Gregg, Melissa and Andrijasevic, Rutvica. ‘Virtually Absent: The Gendered Histories and Economies of
Digital Labour’, Feminist Review 123.1 (2019): 1-7.
Huberman, Jenny. ‘Amazon Go, Surveillance Capitalism, and the Ideology of Convenience’, Economic
Anthropology (2021): 337-349.
Jessica, ‘Drivers Choose Uber for its Flexibility and Convenience’, Uber Newsroom, 7 December 2015,
https://www.uber.com/newsroom/driver-partner-survey/.
Lovink, Geert. Sad by Design: On Platform Nihilism, London: Pluto Press, 2019.
McMillan Cottom, Tressie. ‘Where Platform Capitalism and Racial Capitalism Meet: The Sociology of
Race and Racism in the Digital Society’, Sociology of Race and Ethnicity 6.4 (2020): 441-49.
Murata, Sayaka. Convenience Store Woman, trans. Ginny Tapley Takemori, New York: Grove Press,
2018.
Neves, Joshua and Sarkar, Bhaskar (eds). Asian Video Cultures: In the Penumbra of the Global,
Durham: Duke University Press, 2017.
Shove, Elizabeth. Comfort, Cleanliness and Convenience: The Social Organization of Normality, Oxford:
Berg, 2003.
Tierney, Thomas F. The Value of Convenience: A Genealogy of Technical Culture, SUNY Press, 1993.
West, Emily. Buy Now: How Amazon Branded Convenience and Normalized Monopoly, Cambridge: The
MIT Press, 2022.
11
IN/CONVENIENCE
IN CONVENIENCE
JOSHUA NEVES AND MARC STEINBERG
The felt sense that we inhabit a convenience economy and culture is by now widespread.
Nested in this understanding are ideas about ease and comfort, perpetually new technologies,
and empowered consumers, on the one hand, and growing inequalities and frictions between
the speed and exhaustion that convenience engenders, on the other. Popular critics of Big
Tech such as Tim Wu name this the ‘tyranny of convenience’, where the adoption of modern
conveniences like the washing machine or the smartphone has the ‘ability to make other
options unthinkable’.1 Academic and journalistic assessments of the rise of platforms like
Amazon, Netflix, and Uber, but also Meituan, Grab, Jio, LINE, WeChat, Gozem, and Flipkart,
among many others, paint a similar portrait. Amazon’s conflation of speed with convenience,
Sara Jones argues, ‘is destroying us’. She adds, ‘Someone has to pay for speed, and it will
either be the customer or the worker. Amazon, like most companies, decided to shift the cost
to workers’.2 Another study finds that convenience outstrips commodities themselves, noting
that viewers subscribe to platforms like Netflix for the ‘convenience of on-demand streaming
programming’3 and not because of the rather narrow content offerings. Arjun Appadurai and
Neta Alexander similarly note the ‘prominent promise of convenience, with its emphasis on
immediacy and instant gratification’ at the heart of the appeal of both Silicon Valley and
Wall Street.4 This provision of total convenience, comedian Ronny Chieng jokes in his Netflix
special,5 is key to the lure and excess of the American dream and its global cognates: ‘How
much more convenience can we get?’ Convenience is so pervasive that it has become the
object of parody.
A striking aspect of such assessments is their focus on speed, the reduction of trouble or
work, and ease of access or personal comfort. But they also suggest a surfeit of convenience.
A willingness and meritocratic pretense to encourage or require some among us to do the
heavy lifting in order to create time for privileged others. This includes gig economy services
like: on-demand delivery, shopping, laundry, driving, and much else. This familiar division
of labor and social relations is exacerbated by networked devices and organization, which
are understood to disrupt prior inconveniences by making them smarter. Yet, while we agree
that conveniences involve the social production of inequality, in what follows we argue that
ease, time, and technologized efficiency are not sufficient to grasp and critique this shared
sense of a divided world. Convenience instead resonates with Frederic Jameson’s account
of postmodernism as the cultural logic of an epoch—a constellation of ‘aesthetics, knowl-
1 Tim Wu, ‘The Tyranny of Convenience’, The New York Times, 16 February 2018, https://www.nytimes.
com/2018/02/16/opinion/sunday/tyranny-convenience.html.
2 Sarah Jones, ‘Convenience Is Destroying Us’, Intelligencer, 2 April 2021, https://nymag.com/
intelligencer/2021/04/amazons-convenience-is-destroying-us.html. Emily West, Buy Now: How Amazon
Branded Convenience and Normalized Monopoly, Cambridge: The MIT Press, 2022.
3 Amanda D. Lotz, Portals: A Treatise on Internet-Distributed Television, Ann Arbor: Michigan Publishing,
2017, p. 30.
4 Arjun Appadurai and Neta Alexander, Failure, Medford: Polity Press, 2020, p. 21.
5 Ronny Chieng, Asian Comedian Destroys America!, Netflix Comedy Special, 2019.
12 THEORY ON DEMAND
edge, and political economy’.6 Paraphrasing Jameson, we might say: ‘if [convenience] is a
historical phenomenon, then the attempt to conceptualize it in terms of moral or moralizing
judgements must finally be identified as a category mistake’.7 Convenience is a condition we
inhabit within contemporary capitalism, and must be submitted to rigorous analysis, historical
and conceptual. That even proponents of radical politics assume that convenience will be part
of a post-capitalist society, as exemplified by Aaron Bastani’s promethean treatise on ‘fully
automated luxury communism’, suggests the relational nature of what we term in convenience
in this chapter and book.8 In-convenience bears something of privilege and even boredom,
something of the compulsory, and something of the ‘predatory inclusion’ Tressie McMillan
Cottom finds at work in Internet-accelerated racial capitalism. Responding to this condition
requires us to think beyond simply not clicking ‘buy now’.9
To say that convenience is a condition is also to underscore its affective dimensions. Like
Jameson’s account of postmodern ‘euphoria’, Sianne Ngai’s post-Fordist ‘zany’, or Anna
Kornbluh’s ‘immediacy’, convenience is a privileged form of experience under data capital-
ism, including its platformed iterations. At the economic level, writers such as Nick Srnicek
describe platform capitalism as a moment when ‘capitalism has turned to data as one way
to maintain economic growth and vitality’,10 or where, as The Economist puts it, data is the
new oil and platforms name ‘a new business model, capable of extracting and controlling
immense amounts’ of it.11 Convenience is an implied and under-examined user-side driver
of this shift, even if it’s ultimately folded back into production, creating new demands on
workers. Such conveniences come in app-mediated services from food delivery and taxis to
therapy and autopay. As commonly noted, these perks come with tradeoffs, such as one’s
data being tracked for a faster search result. Draper and Turow call our acquiescence to
networked surveillance ‘digital resignation’;12 we name this relationship to platforms in con-
venience. The state of living in convenience shapes the protocols that make everyday life
‘smart’, wherein ‘each small moment of convenience—be it answering a question, turning
on a light, or playing a song—requires a vast planetary network, fueled by the extraction of
6 Nico Baumbach, Damon R. Young, and Genevieve Yue, ‘Introduction: For a Political Critique of Culture’,
Social Text 34.2 / 127 (2016): 2.
7 Fredric Jameson, Postmodernism, or, The Cultural Logic of Late Capitalism, Durham: Duke University
Press, 1991, p. 46.
8 We render the relation ‘in convenience’ by the hyphenated ‘in-convenience’ when grammar requires it.
9 West, Buy Now, pp. 110-11. West offers an important consideration of this compulsory aspect of
convenience, especially in relation to the model of subjectivity she calls the ‘served self’. Yet, in our
view, her return to consumer activism by way of conclusion, as a presumed counter to the passivity
of the served self, assumes a model of the autonomous, consuming, liberal subject that we argue the
compulsory nature of ‘in convenience’ makes untenable.
10 Nick Srnicek, Platform Capitalism, Malden: Polity Press, 2017, p. 6.
11 ‘The World’s Most Valuable Resource is No Longer Oil, but Data’, The Economist, 6 May 2017, https://
www.economist.com/leaders/2017/05/06/the-worlds-most-valuable-resource-is-no-longer-oil-but-data.
Given this close relation between data and platforms we use ‘platform capitalism’ interchangeably with
‘data capitalism’ in what follows, with the caveat that data capitalism is more capacious in describing a
longer set of transformations.
12 Nora A. Draper and Joseph Turow, ‘The Corporate Cultivation of Digital Resignation’, New Media and
Society 21.8 (2019): 1824-1839.
13
IN/CONVENIENCE
non-renewable materials, labor, and data’.13 In today’s platform capitalism, convenience is
the often unstated explanation and material organization for why things are as they are; why
user-citizens understand data tracking, express delivery, global supply chains, climate-warm-
ing energy consumption, rare mineral mining, waste, toxic working conditions, and much else,
as a necessary evil, the infrastructure supporting everyday work, leisure, and self-fulfillment.
Our basic argument in this essay is that convenience’s consequence stems from its perceived
inconsequence—which is significant precisely because of the relationships to inconvenience
that it consolidates. Convenience is boring (habitual, just the way things are), imperceptible
(like infrastructure, and often as infrastructure, it is most noticeable when it fails), or downright
embarrassing (we lie about subscriptions to Amazon Prime, for instance). It also exacerbates
existing inequalities by further partitioning society. This includes the ways that our relationship
to convenience shifts over the course of a day or week inasmuch as we are workers, consum-
ers, or (non)citizens. Drawing out this tension, we approach convenience as a peculiar con-
stellation of service, logistics, and affect that exceeds narrow approaches centered on either
political economy or cultural practices and artifacts. Indeed, our aim is to bring discussions of
data capitalism more squarely into conversation with everyday calculations and experience.
We locate these shifts not in the glamorous industries of high-tech and finance, but instead
in examples like home appliances, the convenience store, and the endless Netflix or Tik Tok
scroll—examples that ground each of the following sections. With this focus on retail and the
ordinary objects and spaces of convenience, we aim to add nuance to recent interventions
that emphasize only the most conspicuous forces of networked life and industry.
Convenience has appeared as a chief value under many guises, in many eras; below we
account for one such lineage. Contemporary platformed convenience, we suggest, partici-
pates in a broader shift to the service economy and its retrofitting by just-in-time manufac-
turing and distribution, such that all work is reconfigured as part of a service-logistics-af-
fect logic.14 In what follows, we trace a particular genealogy of the emergence and place of
convenience in data capitalism today, focusing on three entangled axes and transitions: (1)
timing and individuation via the home appliance during industrial modernity and especially
the postwar period; (2) spacing and optimization via retail convenience and the service sector
from the 1970s onwards; (3) feeling and logistical form via the platform as enclosure, since
the 2000s. Finally, we conclude by observing how the relation of in-convenience is not simply
a luxury we can choose to indulge or not; it is the normative condition of life and politics today.
13 Kate Crawford and Vladan Joler, ‘Anatomy of an AI System: The Amazon Echo as an Anatomical Map
of Human Labor, Data and Planetary Resources’, AI Now Institute and Share Lab (2018): 3, https://
anatomyof.ai.
14 For our approach in what follows, it matters that logistics is taken up by a wide range of fields, from
business studies and operations research to geography and to Black studies. For some key texts
informing our approach here, see Deborah Cowen, The Deadly Life of Logistics: Mapping Violence in
Global Trade, Minneapolis: University of Minnesota Press, 2014; Ned Rossiter, Software, Infrastructure,
Labor: A Media Theory of Logistical Nightmares, New York: Routledge, 2016; Stefano Harney and Fred
Moten, The Undercommons: Fugitive Planning & Black Study, Wivenhoe: Minor Compositions, 2013;
Kee-hung Lai and T.C. Edwin Cheng, Just-in-Time Logistics, London: Routledge, 2009.
14 THEORY ON DEMAND
Timing: Home and the Rise of ‘Personal Logistics’
Why, we might ask, is convenience so often recognized as a basic value of modernity,
associated at once with the fruits of hard work and self-actualization and, at the same
time, tied to banality, guilt, cynicism or even desperation: ‘the future we all chose, but that
nobody seems to want’.15 Before turning to more recent scholarship about convenience,
it’s worth remembering etymologies that precede its twentieth century associations with
appliances and effortlessness. As Thomas Tierney describes in The Value of Convenience:
A Genealogy of Technical Culture, prior to the 17th century, understandings of conve-
nience in English remained linked to their Latin roots, indicating a sense of agreement,
conformity or harmony; a coming together (as in: to convene). In modernity, this semantic
link shifted, such that ‘the value of technology in modernity is centered on technology’s
ability to provide convenience’.16 What matters here is both the persistence and rupture
between convenience as structures of agreement and proximity, on the one hand, and
the contemporary provisioning of the individual self and population, on the other. Rather
than social harmony, modern convenience is about ‘personal comfort or ease’.17 It is, for
Tierney, a process that is always about making life easier; a configuration that spreads
from the West to the Rest and assumes a neo-Heideggerian view of technology as inher-
ently dominating nature.
Tellingly, Tierney begins his examination with the modern household: the sphere where
‘convenience reigns’.18 This insight, echoed by later scholarship, also relies on a con-
ceptual shift from production to consumption as what matters for understanding the
work of convenience in technical culture. More than the expansion of this realm, Tierney
emphasizes how it ‘becomes narrow and pointed’, penetrating modern individuals and
their values.19 This narrowing, however, is tied to an entangled expansion of necessity that
parallels concerns about modern trespasses into the private domain. Such problems are
central to Hannah Arendt’s critique of private and public realms in The Human Condition,
especially her concern that their boundaries are blurred by capitalism and the rise of the
social.20 Two points about convenience and the household are worth emphasizing here.
First, as many commentators have noted, that Arendt’s distinction necessarily brackets
how the overcoming of necessity relies on slavery and rigid gendered divisions, among
other violences, is hardly inconsequential. Indeed, this elision remains central to how
convenience organizes society. Further, the tension between classical and contemporary
formulations suggests a related shift in the imagination of the human body. This distinction
moves from understandings of ‘ancient necessity [as] primarily concerned with satisfying
15 Colin Horgan, ‘The Tyranny of Convenience’, OneZero, 29 April 2019, https://onezero.medium.com/
the-tyranny-of-convenience-2e7fa145ab4.
16 Thomas F. Tierney, The Value of Convenience: A Genealogy of Technical Culture, Albany: State University
of New York Press, 1993, p. 6.
17 Tierney, The Value of Convenience, p. 39.
18 Tierney, The Value of Convenience, p. 11.
19 Tierney, The Value of Convenience, p. 4.
20 Hannah Arendt, The Human Condition, trans. Margaret Canovan, Chicago: University of Chicago Press,
1958.
15
IN/CONVENIENCE
the demands of the body’, to a modern preoccupation with ‘overcoming [its] limits’. If the
former stresses basic needs like food, clothing and shelter, the latter seeks to command
and reorder the body in order to minimize or eradicate ‘inconveniences, obstacles or
annoyances’.21 In other words, convenience, by now, convenes us (and the more-than-
human world) differently.
In the so-called developed world in the 20th century, the mitigation of inconveniences
became the task of labor-saving domestic appliances such as washing machines and
electric ranges, televisions and toasters, refrigerators and freezers, that promised con-
venience by reducing work, stretching time, and providing for new desires. The restruc-
turing of daily rhythms and the reduction of time spent on ordinary tasks is essential to
the promise of convenience.22 In Comfort, Cleanliness, and Convenience, Elizabeth Shove
shows how prior associations of convenience with ease and the saving of time are, in the
latter half of the twentieth century, supplanted by the new ‘capacity to shift, juggle and
reorder episodes and events’.23 Here, ‘things that are “convenient” are those that enhance
peoples’ control over the scheduling of activity’.24 In light of growing temporal pressures
and the historical rise of the consumer society, Shove continues:
contemporary usage relates convenience to the scheduling and co-ordination
of people and objects in time and space. Understood in this way convenience is
about timing, that is, the ability to shift and juggle obligations and to construct and
determine personal schedules. From this point of view, the really important benefit
of convenience food is not that it saves time but that it makes it possible to prepare
and eat a meal at very short notice.25
Convenient devices allow for better management of time. They are not about saving time
as such but rather about the ways modern conveniences allow for a control over the timing
of domestic activities. Timing over time.
The rise of convenience as a value is also synchronous with shifts in labor and productiv-
ity over the course of the 20th and 21st centuries. Control over one’s time or ‘temporal
sovereignty’ is, as Melissa Gregg writes, ‘a historically specific form of freedom’.26 This
freedom is deeply enmeshed with shifts in work time and place, aligning with the rise
of the neoliberal subject who must manage themselves and take a new responsibility
21 Tierney, The Value of Convenience, p. 30.
22 This promise was often structured around women’s work in the home. And yet these new conveniences,
as Ruth Cowan points out, often resulted in ‘more work for mother’. See Ruth Schwartz Cowan, More
Work for Mother: The Ironies of Household Technology from the Open Hearth to the Microwave, New
York: Basic Books, 1983.
23 Elizabeth Shove, Comfort, Cleanliness and Convenience: The Social Organization of Normality, Oxford:
Berg, 2003, p. 170.
24 Shove, Comfort, Cleanliness and Convenience, p. 170.
25 Shove, Comfort, Cleanliness and Convenience, p. 171.
26 Melissa Gregg, Counterproductive: A Brief History of Time Management, Durham: Duke University
Press, 2018, p. 7.
16 THEORY ON DEMAND
for their temporal resources. Gregg calls attention to two aspects of this shift that help
draw out the implications of timing. First, she tracks how practices of time-management
emerge not only from the default sites of workplace efficiency—the office and factory—
but also, crucially, from the management of the home. This is suggestive of a longer history
of ‘women’s role as managers and efficiency engineers’ that remains vital to the ways
that home economics feed into models of efficiency.27 Convenience’s facilitation of time
management parallels, we note, the flexibilization of work that begins with the postwar
restructuring of women’s employment on a temporary or contingent basis (e.g. in the
US and Japan), suggesting a prehistory of the gig economy’s destructive flexibilization
of labor, which is so often framed as convenient for workers. Second, these shifts drive
an emergent sense of timing that Gregg calls a ‘new kind of personal logistics’.28 Amidst
the rise of new forms of precarious work, the temporalities of convenience also shift
from the household to become services and service jobs, including those now provided
by companies like Amazon and UberEats. This is to link the gigification of labor and the
flexibilization of work schedules to contemporary habits, aspirations, and bodily norms.
Convenience, in other words, is not equally available to all. The rise of timing and personal
logistics relies on, indeed generates, peculiar forms of social untimeliness. Put simply,
convenience is produced by inconvenience—a term that needs to be re-operationalized
to capture its present purchase as much more than minor discomfort or frustrating ineffi-
ciency; signifying social inequity, exploitation, and oppression. Sarah Sharma makes this
point sharply in In the Meantime. There she interrogates the myth of the culture of speed,
pointing to how speed is unevenly distributed and inhabited by those who labor to create
it, including taxi drivers who must wait in their cars at the airport for hours to produce the
sense of timeliness and convenience for those who can afford it.29
To put it in Shove’s terms, ‘scheduling and co-ordination of people and objects in time and
space’ relies on complex inconveniences that extend far beyond appliances or current
fears of automation. It also anticipates the increasing emphasis on scheduling associated
with the shift to ‘on-demand’ services, hybrid work, and so on, wherein the on-demand
is itself an iteration of the paradigm of convenience as a matter of scheduling and timing
of activities.
Spacing: The Convenience Store
Beyond timing, platformed convenience relies on a logistical mastery of space and spac-
ing. The convenience store, especially in its development in Japan from the 1970s to the
present and subsequent re-exportation around the world from the 1980s onwards, is, in
our view, an emblematic example of this spatialization of convenience.
27 Gregg, Counterproductive, p. 34.
28 Gregg, Counterproductive, p. 129 (emphasis ours). An important work that engages with QR codes
and their facilitation of ‘convenient efficiency’, also engaging with the history of productivity, is Dang
Nguyen, ‘Convenient efficiency: A media genealogy of QR codes’, New Media & Society (2022): 1-21.
29 Sarah Sharma, In the Meantime: Temporality and Cultural Politics, Durham: Duke University Press,
2014, p. 56.
17
IN/CONVENIENCE
While first established in the US South, in Texas, in the 1930s, what became 7-Eleven
established the convenience store format that would have its heyday in the US in the
1960s. This attracted the interest of both large Japanese retailers and the Japanese
government, both captivated by the idea of rationalizing retail and reorganizing under
a single franchise banner the many small stores across Japan. Over the 1980s and into
the 1990s, the Japanese convenience store chains overtook their US models to become
the cutting edge of spatialized convenience. Hence, global chains like 7-Eleven offer a
pre-history of logistically enabled forms of convenience that transform and intensify the
time-centered approaches noted above. They continue to influence the present in signif-
icant ways, focusing our attention on ordinary sites and experiences of convenience that
both condition and are often overlooked by digital and platform studies. 7-Eleven, among
similar examples, also reminds us of the longer histories of both retail and automobile
manufacture (a precursor to platform capitalism),30 as well as the continued economic
and social impact of service- and logistics-driven ‘lean’ management techniques that
integrate retail ‘store operations, product development, distribution and information
systems’.31 As such, the global-Japanese convenience store is our point of departure.
To focus on the Japanese convenience store means emphasizing franchises as a crucial
origin in the consolidation of convenience culture.32
The Japanese convenience store impacts retail convenience worldwide, from East and
Southeast Asia to the North American 7-Eleven stores remodeled to resemble their Jap-
anese counterparts (Figure 1). Its mundane amplification of convenience as a structure
of experience serves as a model for understanding the emergence of platforms like Alib-
aba and Amazon, iQiyi and Netflix. A focus on retail allows us to emphasize the regular,
repeat visits that have as much of a role in the transformation of habits and expectations
of convenience as the more rarefied platforms and e-commerce sites that are touted as
their replacements. The convenience store makes visible a shift from home appliances to
neighborhood services as a locus of convenience, which is in turn part of a larger transfor-
mation of manufacture in the image of the service sector.33 If approaches to timing in the
previous section suggest a shift in attention from production to consumption, our interest
here is the way that service industries fuel an emergent organizational sphere—and a
production of distribution—that brings into view a new set of concerns. Notably, the place
of convenience shifts from the domestic sphere to public and semi-private spaces, and
with a corresponding shift of emphasis from timing to spacing as convenience’s crucial
offering.
30 Marc Steinberg, ‘From Automobile Capitalism to Platform Capitalism: Toyotism as a Prehistory of Digital
Platforms’, Organization Studies 43.7 (2022): 1069-1090.
31 David Marutschke, Continuous Improvement Strategies: Japanese Convenience Store Systems,
Basingstoke: Palgrave Macmillan, 2011, p. 23.
32 This, of course, is not to forget about the many informal corner stores (from the bodega to the
dépanneur) or automobile-centric gas stations that are also part of this story but beyond the scope of
this chapter.
33 Rutvica Andrijasevic, et al., Media and Management, Lüneburg: Meson Press, 2021.
18 THEORY ON DEMAND
Figure 1: Map of 7-Eleven stores worldwide as of January 2020. 7-Eleven Inc., ‘7-Eleven is 70,000
Stores Strong’, PR Newswire, 23 January 2020, https://www.prnewswire.com/news-releases/7-eleven-
is-70-000-stores-strong-300992154.html.
Consider the global career of 7-Eleven, which now operates 71,000 stores around the world,
only 9300 of which are in the United States.34 Founded in the United States in 1927, it was
imported to Japan in 1973, with the first store launched in 1974. 7-Eleven played a key role in
reinventing the convenience store in Japan alongside chains like Lawson and Family Mart. By
the 1980s, the renovated Japanese convenience store model was re-exported around the world,
particularly within East Asia, but also to the US and European markets. The chain’s success
led 7-Eleven Japan’s parent company to purchase most of the Southland Corporation in 1991,
making it a fully owned subsidiary of the Japanese firm in 2005. With this takeover of the original
US company, the ascendancy of the Japanese convenience model was complete. At the heart
of its success was a new mode of spatial organization that reaches far beyond the store itself.
This spatial organization starts with the ubiquity of convenience stores in Japan, with over 50,000
outlets nationwide that receive an estimated 16.7 billion visits per year.35 Wherever they are
located, they are hubs of daily life, providing fresh prepared foods and a variety of basic ameni-
ties, as well as array of services including, as Marutschke enumerates, ‘bank, postal and delivery
services, acting as ticket agents, accepting utility payments and even handling laundry, home
cleaning services, printing services, garbage pick-up tickets and online shopping’.36 Fresh foods
account for both a large amount of sales and a large proportion of the chains’ product devel-
34 Ron Chang, ‘7-Eleven Opens 71,100th Store in S. Korea’, TBS, 9 July 2020, http://tbs.seoul.kr/
eFm/newsView.do?typ_800=J&idx_800=3395420&seq_800=20387997. 7-Eleven is the largest
convenience store operator in the world by far. Japanese FamilyMart, with a large footprint in Japan and
elsewhere in Asia, comes in second place with around 24,000 stores, followed by Mexican OXXO. Of the
top 10 chains worldwide, it is significant that Japanese-owned chains account for 3 of the top 5 and 4 of
the top 10.
35 Mieko Shirai, Takeshi Kojima, and Masashi Oguri, ‘Konbini Wo Kagaku Suru’ (Doing the Science of
Convenience Stores), Shūkan Diamond (Diamond Weekly), 29 October 2016, 28.
36 Marutschke, Continuous Improvement Strategies, p. 5.
19
IN/CONVENIENCE
opment strategies.37 Because of their small footprints and prepared fresh food sales, stores
receive at least 7-10 deliveries per day,38 and their model and margins require an immense
circulation of goods and people into and out of the store. This distribution is known as logis-
tics. Most discussions of logistics focus on the transformations in production they enable, with
‘transportation conceptualized as a vital element of production systems rather than a separate
domain or the residual act of distributing commodities after production’.39 7-Eleven stores,
among others, give further texture to this perspective by focusing on how distribution not only
transforms production but the very idea and experience of convenience.
A signal, if under-theorized, aspect of convenience today is hence the reorganization of space or
spacing. Spacing calls attention to the positioning of a store in a neighborhood or city, its prox-
imity to distribution hubs, the management of logistical networks supplying just-in-time delivery
of fresh foods and inventory, as well as the layout and design of stores for ease of customer use.
The aspects of convenience crucial to 7-Eleven all speak to a conjugation of timing with spac-
ing: long opening hours; proximity of stores to consumers; the ability to ‘buy all essential goods
in just one place’; and ‘quick shopping’ wherein the ‘layout of the store is ideal for customers
to locate their required products easily’.40 Convenience is treated as a logistical problem—a
problem of optimizing distribution in real time.41 Spacing is, then, always also about timing.
Inspired by the Toyota Production System and its emphasis on just-in-time delivery (which has
always been about the optimizing of space via as-needed delivery of auto parts),42 7-Eleven
and other convenience stores in Japan elaborated the principles of auto production into retail
empires. Dependent not upon a network model of infinite connectability, the convenience store’s
convenience is a kind of proto-platform premised upon proprietary logistical and information
systems, point-to-point transportation services, total coordination of circulation of people and
things, and the enclosure of the store space nested within the larger enclosure of the logistical
system itself: the 7-Eleven franchise.
Like Amazon, Alibaba, or Walmart, 7-Eleven is a logistics company as much or more than a
retail enterprise.43 It focuses on the planning of stores, the development and procuring of mer-
37 Tai Negō and Kyōichi Hiraki, Konbini Gyōkai No Dōkō to Karakuri Ga Yoku Wakaru Hon (A Book for
Really Understanding the Trends and Mechanisms of the Convenience Store Industry), Tokyo: Shuwa
Shisutemu, 2015, p. 122.
38 Akira Ishikawa and Tai Nejo, The Success of 7-Eleven Japan: Discovering the Secrets of the World’s Best-
Run Convenience Chain Stores, Singapore: World Scientific, 2002, p. 55.
39 Cowen, The Deadly Life of Logistics, p. 40.
40 Ishikawa and Nejo, The Success of 7-Eleven Japan, p. 14.
41 As Halpern and Mitchell’s account of ‘smartness’ teaches us, optimization is a future-oriented,
open-ended, never-ending process; one that, like convenience, relies on a particular epistemology
of smartness. Orit Halpern and Robert Mitchell, The Smartness Mandate, Cambridge: The MIT Press,
2023.
42 Shinji Naruo and Sorin George Toma, ‘From Toyota Production System to Lean Retailing. Lessons from
Seven-Eleven Japan’, in Jan Olhager and Fredrik Persson (eds) Advances in Production Management
Systems, New York: Springer, 2007, pp. 387-395.
43 Jesse LeCavalier quotes a Walmart manager making a similar point: ‘The misconception is that we’re
in the retail business, [but really] we’re in the distribution business’. See Jesse LeCavalier, The Rule of
Logistics: Walmart and the Architecture of Fulfillment, Minneapolis: University of Minnesota Press, 2016,
p. 11. On Amazon and logistics, see Armin Beverungen, ‘Remote Control. Algorithmic Management of
20 THEORY ON DEMAND
chandise, and most especially the delivery of this merchandise to its owner-operated stores.
The emphasis on distribution affects the very organization of stores within city space, with store
locations planned according to delivery routes to ensure savings on delivery costs and times.
The centrality of logistics and optimization also determines the layout of stores. Like Walmart,
convenience stores are designed to ensure that consumers circulate as easily as the daily deliv-
eries that restock them. Store layout is itself optimized for legibility. All stores, even across chains,
adopt a familiar inverted ‘C’ layout, with magazines at the entrance, drinks at the back, and fresh
foods across from the entrance near the cash register. While store products change every year,
the layout is constant over decades.44 As a result, the average consumer spends less than three
minutes in a convenience store. Convenience fades into the background as a vague feeling, with
muscle memory and habit guiding users as they shop.
Convenience stores are themselves a response to both the expansion of work hours and the
destructuring of work routines in recent decades. Consider that the name 7-Eleven, which clearly
states the original store hours, is already anachronistic. Its hours have long since extended
to meet the demands of a 24/7 lifestyle. The stores both rely on part time and temp workers
for staffing and also cater to the irregularities of the increasingly precarious workforce that
makes up its consumer base.45 ‘Starting in the latter half of the 1990s’, Gavin Whitelaw notes,
convenience stores ‘have been referred to as “life infrastructure” (seikatsu infura), akin to crit-
ical public services such as water, gas, and electricity’.46 7-Eleven even brands itself as an
infrastructure: ‘Electricity, gas, water, and 7-Eleven’, reads the company’s landing page. This
recognition of convenience as infrastructural to daily life is crucial to our understanding of the
timing and spacing of the present, but also asks new questions about convenience as a cultural
and logistical form.
Figure 2: Convenience stores are widely associated with immigrant labor in North America and Europe, as
captured in the Toronto-based sitcom, Kim’s Convenience. The production itself was marred by ‘diversity
issues, unfair pay and racist storylines’.47 Promotional material from the television show.
Circulation at Amazon’, inMarcus Burkhardt, Mary Shnayien, and Katja Grashöfer (eds) Explorations in
Digital Cultures, Lüneburg: Meson Press, 2021, pp. 5-18.
44 Shirai, Kojima, and Oguri, ‘Konbini Wo Kagaku Suru’, 32.
45 Ishikawa and Nejo, The Success of 7-Eleven Japan, p. 5.
46 Gavin H. Whitelaw, ‘Konbini-Nation’, in Katarzyna J. Cwiertka and Ewa Machotka (eds) Consuming Life
in Post-Bubble Japan, p. 79.
47 Chris Gardner, ‘‘Kim’s Convenience’ Stars Simu Liu and Jean Yoon Open Up on “Painful” Lack
21
IN/CONVENIENCE
Feeling: Convenience as Logistical Form
If the prior section understands convenience as a fusing of timing and spacing—which togeth-
er define logistics—our interest here is to examine convenience as a cultural and aesthetic
form. That is, a mode of address, an affect, a feeling, and a judgment about our encounters
with an accelerating techno-economic world. The convenience store offers one suggestive
genealogy of this contemporary sensorium. A bright white cube with unchanging hours, recog-
nizable design and reliable offerings, it is peculiarly antiseptic and homey. In Japan, it is the
rare place where one can go in pajamas. And yet despite this, store space is also distinctly
depersonalized. The habitual irasshaimase greeting called out to customers as they enter the
store—a ‘Welcome’ that is decidedly not a ‘Hello’—is distance-producing and unidirectional.
Here the banality of convenience as a structure of feeling or experience begins to come into
view. It is at once a social infrastructure and, at the same time, a kind of calculated un-care
(like ‘contactless’ delivery) that drives immaterial and affective labor, or what we call logistical
form. This phrase points to the ways cultural forms are themselves determined by logistical
systems under platform capitalism, shifting the very place of analysis from the form of a
content to the form of distributed experience.
Our discussion of logistical form draws inspiration from Sianne Ngai’s crucial expansion
and reformulation of aesthetic categories with particular attention to minor or compromised
aesthetics. No longer limited to purified or sublime encounters, Ngai suggests that in late
capitalism aesthetics become ‘part of the texture of everyday social life’.48 In particular, she
focuses on a set of aesthetic categories or feelings like the zany, the cute, the interesting,
and more recently, the gimmick. These categories, she argues, are two-sided and mediated,
including ‘the judgment we utter, a way of speaking; the form we perceive, a way of seeing’
and are ‘sutured by affect into a spontaneous experience’.49 Put simply, an aesthetic category
is composed of a felt sensation and a verbal response that elicits some agreement from one’s
interlocutor. At the same time, these minor aesthetics tend not to move observers much, and
are instead characterized by a ‘deficit of power’.50 This last point is key to Ngai’s intervention:
aesthetic categories are interested and are deeply informed by economic processes; classical
disinterest is displaced by ordinary attention and proximity.51 In many respects, convenience
foregrounds the weakness and banality described in Ngai’s aesthetic categories, its very
ordinariness shaped by the dulling effects of social standardization and expanded calculation.
This includes conflicting senses of ease and exhaustion, comfort and cynicism, mundane
habitation and righteous condemnation.
of Diversity, “Overtly Racist” Storylines’, The Hollywood Reporter, 6 June 2021, https://www.
hollywoodreporter.com/tv/tv-news/kims-convenience-netflix-cancelation-1234963806/.
48 Sianne Ngai, Our Aesthetic Categories: Zany, Cute, Interesting, Cambridge: Harvard University Press,
2012, p. 29.
49 Sianne Ngai, Theory of the Gimmick: Aesthetic Judgment and Capitalist Form, Cambridge: Harvard
University Press, 2020, p. 1.
50 Ngai, Our Aesthetic Categories, p. 18.
51 Ngai, Our Aesthetic Categories, p. 27.
22 THEORY ON DEMAND
Yet as we suggest in what follows, if convenience starts as an aesthetic category—a punctual,
momentary experience that may be followed by the declaration: ‘this is so convenient’—it
also very often recedes into the calculative background. It becomes an ambient feeling. In this
way convenience offers something like Brian Massumi’s early account of fear as an affect: not
an emotion but rather ‘the objectivity of the subjective under late capitalism’.52 Convenience
undergoes a phase shift from being a momentary experience (‘this is convenient’) to being
an underlying, ongoing structure of feeling or a cultural logic. Here we are interested in both
the punctuality of convenience as an aesthetic category and the way in which it becomes a
cultural and economic background—feeling and form. Here we might recall Shove’s point,
above: that the production of conveniences is never finished. Once habituated to a conve-
nience, lives become dependent on it, expectations are heighted, and new conveniences must
be invented to allow us to cope with the increased temporal pressures put on all of us. In this
sense, the convenient is also tied to Massumi’s account of fear—the fear of falling behind, or
of falling out of time in late capitalism.
Murata Sayaka’s award-winning novel Convenience Store Woman (Konbini ningen) offers one
point of departure for understanding convenience as an aesthetic or vernacular style in the
sense Ngai describes. The novel opens:
A convenience store is a world of sound. From the tinkle of the door chime to the voices
of TV celebrities advertising new products over the in-store cable network, to the calls
of the store workers, the beeps of the bar code scanner, the rustle of customers pick-
ing up items and placing them in baskets, and the clacking of heels walking around
the store. It all blends into the convenience store sound that ceaselessly caresses my
eardrums.53
Told from the first-person perspective of a shop employee, the framing description both
illuminates the convenience store’s sensorium as well as its dependable conventions and
address. Working there is a bodily experience, a set of routines embedded and embodied as
habits: ‘Speed is of the essence, and I barely use my head as the rules ingrained in me issue
instructions directly to my body’.54 Tired of struggling with social cues and implied norms
in the outside world, for example, the narrator is relieved to find a place where all behav-
iors and social interactions are prescribed by management—allowing her to ‘transform into
the homogeneous being known as the convenience store worker’.55 She finds comfort in
its peculiar repetitions: ‘we had greeted the same morning 6,607 times’.56 In this way, the
novel is a strange ode to the standardization and social legibility that the convenience store
as a life infrastructure provides. But it also suggests that convenience is produced through
routine, even cold intimacy. This allows customers to experience each store as the same,
52 Brian Massumi, ‘Everywhere You Want to Be: Introduction to Fear’, in Brian Massumi (ed) The Politics of
Everyday Fear, Minneapolis: University of Minnesota Press, 1993, p. 12.
53 Sayaka Murata, Convenience Store Woman, trans. Ginny Tapley Takemori, New York: Grove Atlantic,
2018, p. 1.
54 Murata, Convenience Store Woman, p. 2.
55 Murata, Convenience Store Woman, p. 16.
56 Murata, Convenience Store Woman, p. 73.
23
IN/CONVENIENCE
and to navigate them as efficiently as the store workers trained to recognize their most min-
ute gestures, if not to learn their names. Like the soundscape, bright fluorescent lights, and
familiar layout, the feeling of convenience suggests the inverse of anxiety or uncertainty: it is
habitual, reliable, efficient.
While Murata’s novel offers an important aperture, the focus on the convenience store as
distribution network above and the platform in what follows necessitates a methodological
expansion. Understanding logistical form requires that we look beyond interfaces of consump-
tion and discrete modernist texts or works of art, like the novel, film, or video game.57 Here
we build on Patrick Jagoda’s engagement with network aesthetics, which both provocatively
theorizes the emergent ‘sensibilities of distribution’ tied to the rise of the internet, among other
network imaginaries, and yet takes as its evidence ‘artworks that experiment with network
aesthetics’ rather than everyday logistical space or affective relations.58 At stake here are the
very parameters for what constitutes the ordinary, including its spatial, temporal and sensory
form or arrangement, and their relations to critique. The point is not that everyday sites or
texts are somehow unmediated, or that we ignore novels or artworks, but rather, by focusing
only on particular kinds of mediation—and the familiar or privileged objects of film and media
studies—we fail to appreciate what is distinct about the feeling of convenience, and platform
aesthetics more generally. More to the point: logistical form suggests that existing understand-
ings of late capitalism and network cultures are out of sync with our present challenges and
current configurations of logistically-informed platform capitalism in consequential ways that
a grappling with convenience brings into view.
This is to underscore an alternate history of the platform—traced through service, logistics
and retail, rather than histories of cybernetics, TV networks, net art and social networks—and
calls into question media studies’ continued fascination with particular network diagrams and
aesthetics. What matters here is that while actual and imaginary networks may take many
forms, received understandings have sedimented into inert images and interventions. This
includes the persistent fascination with distributed networks, web 2.0, and the residual claims
of cyberspace, as well as the influence of certain critical responses like Deleuze’s ‘control
society’ or Jameson’s ‘cognitive mapping’, including the latter’s claim that the complexity of
the world system overwhelms our sensorium, making it difficult if not impossible to grasp its
totality and prepare as political actors. Our aim is to hold onto these problems while shifting
attention away from the infinite network, the rhizome, and the fetish of (dis)connectivity, and
toward the platform’s dominant logic of standardization, habituation, and enclosure. That
these platform logics are deeply bound up with their modes of value creation and political
economy, should also remain top of mind.
The idea that platforms standardize and enclose is hardly novel. Many scholars have noted
this tendency and discussed its impact on existing understandings of the internet, its material
57 A break with this focus on discrete texts separates our consideration here from Ngai’s otherwise
generative approach to everyday aesthetics.
58 Patrick Jagoda, Network Aesthetics, Chicago: University of Chicago Press, 2016, pp. 18-19.
24 THEORY ON DEMAND
form and everyday practices.59 The present emphasis includes Joss Hands’ claim that ‘[t]he
Internet is vanishing’. He continues, ‘as its ubiquity increases, it has also become less and less
visible in the production and experiences of network culture. Indeed, many of the operations
that used to typify the Internet are now funneled through so-called “platforms”’.60 Drawing on
such typical accounts, our argument here is that, contrary to familiar assumptions, platform
enclosures operate by standardizing the experience or feeling of convenience. They also do
so to explicitly economic ends; enclosure is a means of value creation. Following on the 19th
century standardization of time, and the 20th century standardization of space,61 we argue
that 21st century standardization takes feeling, experience, and affect as its object.62 The
implications of this for our aesthetic sensibilities and political orientations are manifold.
If the rise of the convenience store offers one example of how distribution, countability, and
calculation transforms everyday experience—constituting a life infrastructure—the rise of
Netflix, among other video platforms (from Showmax to Douying/Tik Tok), offers an instructive
example of how they change culture. Growing out of a landscape of video shaped by brick-
and-mortar VHS and DVD rental stores, Netflix, founded in 1997, initially distinguished itself
with its online catalogue, flat rate subscriptions, and DVD-by-mail service, before launching
its streaming services in 2007. The company can both be understood to emerge from the
logistical space of the convenience store (being a distribution firm much like 7-Eleven), and
to amplify its material and affective transformations. Despite its shift into streaming and its
investment in producing or licensing so-called Netflix Originals, what remains constant from its
days as a DVD distributor is both its interface and its concern with distribution. This very shift
is itself narrated in terms of the augmented provision of convenience, wherein the ‘collapse
of Netflix’s [DVD] browsing interface into a viewing interface removed even the trip from the
computer to the mailbox’.63 While too easily framed in terms of the laziness critics of conve-
nience decry, this should be viewed as a persistent focus on distribution over content, and
the prioritizing of convenience as a cultural form or structure of experience.
Consider the suggestion that people subscribe to Netflix for the convenience it offers rather
than its small, and by some accounts, shrinking content catalog. This simple observation,
familiar to users through the experience of endlessly scrolling or ‘watching’ the site’s interface,
only to find the same titles repeated in different categories, brings into focus a key change
associated with logistical form. Logistics is the primary emphasis of Netflix from its mailed-
DVD days to its ‘Netflix Originals’. And while its programming remains important, we take Ed
Finn’s position that ‘[r]eading Netflix as a series of algorithms, interfaces, and discourses is
59 The enclosure of the internet is actually ongoing, growing out of 1990s ‘walled gardens’ like AOL and
i-mode. Marc Steinberg, The Platform Economy: How Japan Transformed the Commercial Internet,
Minneapolis: University of Minnesota Press, 2019.
60 Joss Hands, ‘Introduction: Politics, Power and ‘Platformativity’’, Culture Machine 14 (2013): 1.
61 Nigel Thrift, Knowing Capitalism, London: Sage, 2005.
62 Joshua Neves, ‘Social Media and the Social Question: Speculations on Risk Media Society’, in Bhaskar
Sarkar and Bishnupriya Ghosh (eds) The Routledge Companion to Media and Risk, New York: Routledge,
2020, pp. 347-361.
63 Colin Jon Mark Crawford, Netflix’s Speculative Fictions: Financializing Platform Television, Lanham:
Lexington Books, 2021, p. 48.
25
IN/CONVENIENCE
far more instructive for understanding its role as a culture machine than reading the cultural
products produced by the system’.64 To this we might add distribution systems. Netflix videos
load instantaneously because of its proprietary content delivery network, ‘which stores vid-
eo and audio content in servers located close to end users’.65 This network also transforms
the very nature of content by gathering viewing data at a scale hitherto unimaginable in the
network TV era, and customizing content based on viewing habits. Patterns of distribution
hence work back into content. Standardization and calculability are here crystalized as a set
of generic codes, including some ‘76,897 genres’ identified by Netflix, many ‘still waiting
for content’.66 This generic quality across categorical differences is captured in a statement
by a former Vice President of product engineering: ‘Netflix seeks the most efficient content.
Efficient here meaning content that will achieve the maximum happiness per dollar spent’.67
Here we come close to a definition of logistical form: calculation, standardization, and the
primacy of distribution animate Netflix, among many other platforms, as a particular kind of
culture machine. Companies like Netflix hence both expand the discourse and offerings of
personal logistics, noted above, and transform this individuating tendency into a widely shared
form of logistical experience.
Streaming platforms like Netflix now account for a large percentage of global internet band-
width,68 and signal an intensification of what Raymond Williams’ called ‘mobile privatization’
to describe a new way of living in post-war industrial nations that was ‘at-once mobile and
home-centered’.69 In subsequent years, Williams’ observation was intensified by expanding
capacities for flexible distribution, from the Walkman to the smartphone, now the center of
on-demand cultures especially in the Global South. Platformed convenience now both per-
meates domestic life in new and old ways, and signals a diffusion of mediation. As one recent
article puts it, ‘Home is where your Netflix is’.70 In this sense, global streaming services are
neither domestic nor public technologies but rather organize and permeate sociality on a
planetary scale. This is also to build on Thomas Lamarre’s claim, itself drawing on Williams,
that distribution, far from being neutral or simply secondary to production, is itself productive
and creates new formal and sensorial relations. It ‘entails a sense of affective possession’ that
emerges ‘in conjunction with the mapping of the transmedial onto a geopolitical domain’.71
Beyond distribution capacities or infrastructures, in other words, the production of distribution
demonstrates novel aspects of the platform economy, which, like Jameson’s cultural logic,
may help us to apprehend the work of form and feeling under data capitalism. Put simply,
64 Ed Finn, What Algorithms Want: Imagination in the Age of Computing, Cambridge: The MIT Press, 2017,
p. 103.
65 Ramon Lobato, Netflix Nations: The Geography of Digital Distribution, New York: NYU Press, 2019, p.
94.
66 Finn, What Algorithms Want, p. 94.
67 Finn, What Algorithms Want, p. 108.
68 Lobato, Netflix Nations, p. 88.
69 Raymond Williams, Television: Technology and Cultural Form, London: Routledge, 2003, p. 18.
70 Barbara Maly-Bowie, ‘‘Home is where your Netflix is’ – From Mobile Privatization to Private
Mobilization’, Literary Geographies 5.2 (2019): 216-233.
71 Thomas Lamarre, ‘Regional TV: Affective Media Geographies’, Asiascape: Digital Asia 2.1-2 (2015): 94.
26 THEORY ON DEMAND
ordinary experience is reorganized and becomes newly logistical. Lamarre’s emphasis is on
the creation of affective media geographies—a pre-personal feeling of being-in-common
that anchors experience into zones of affiliation that precede and exceed physical geography
(like the fandoms of Hana yori dango or BTS). By contrast, the affective sense produced by
the distributive systems of Netflix—not to mention WeChat, Grab, Doordash, etc.—is one of
convenience. Like many platforms, this feeling starts as a conscious experience, or aesthetic
judgment: the novelty of viewing at whim from the content library available is convenient.
But this immediate aesthetic experience quickly shifts to the background or default. Further,
as video stores have long since disappeared in most cities, and pirate networks become
increasingly specialized and difficult to access, the default of media experience becomes
compulsorily convenient. Convenience moves from being a punctual, conscious feeling—that
which is felt when one first clicks on a Netflix title and a video begins to load—to being infra-
structural to experience in toto. As Anna Kornbluh summarizes the style of too late capital-
ism: ‘immediacy swallows everything’.72 For many there is no alternative to home streaming,
just as one cannot request 10-day Amazon delivery rather than 2-day, 1-day, or 1-hour, or
whatever the standard of convenience may now be. Convenience stops being a demand con-
sumers place on platforms, a content of experience as many frame it,73 and becomes instead
a demand on consumers placed by platforms. The experience of this logistical sensorium is
one of in-convenience. In-convenience names the very form or atmosphere of platformed
experience, at least in an era when convenience is the default timing, spacing, and feeling
of the world. If platforms enclose the web, the subjective feeling of inhabiting this enclosure
is one of in-convenience. To paraphrase Massumi, it is the objectivity of the subjective in
platform capitalism. This may lead us to ask not only what is convenient, but when is conve-
nience perceived as such and when is it merely the infrastructure or the ‘affective surround’
of life today?74
Like the convenience store, whose experience of convenience is predicated on its becoming a
social infrastructure or habitual encounter, platforms produce convenience as a technologized
background. What we have called logistical form is generated by computational distribution
and an emergent logic of standardization, ease, and reliability that constitutes the vague
feeling of platform capitalism itself. It is a structuring of experience based on the potential
for an endless accounting and reorganization of time, space, and sensation. It signals the
porousness of what is home, when is public, and where the feeling of convenience mediates
and shifts these categories. Acknowledging that distribution produces so much more than
the movement of people and things—and instead moves the very capacities to sense or
feel—means understanding that what logistical systems like convenience stores, Netflix, and
Amazon create is a particular distribution of convenience. Indeed, if platform capitalism were
to have an aesthetic category, it would be: the convenient. Convenience is part of the total
72 Anna Kornbluh, Immediacy: or, the Style of Too Late Capitalism, London: Verso, 2024, p. 9.
73 Robert M. Pallitto, Bargaining with the Machine: Technology, Surveillance, and the Social Contract,
Lawrence: University Press of Kansas, 2020.
74 Brian Massumi, ‘Fear (The Spectrum Said)’, Positions: East Asia Cultures Critique 13.1 (Spring 2005):
41.
27
IN/CONVENIENCE
aestheticization of life in late capitalism, at once immediate and infrastructural. While such
experiences no doubt differ across platforms—including home delivery (Amazon, Alibaba,
Flipkart, Rakuten), personal mobility (Uber, Didi), video delivery (Netflix, Hotstar, Tudou),
super apps (WeChat, LINE, Grab), and social media platforms (Facebook, Instagram, Tik
Tok)—they produce and are produced by the experience of convenience. To take this argu-
ment one step further: platforms sell convenience, not products.
In Convenience: The Cultural Logic of Platforms
Convenience names the normative timing, spacing, and feeling of data capitalism. It is also a
relational concept: what we term in convenience in our title. Our interest here is less to reiter-
ate the claim that contemporary conveniences rely on and generate deep inconveniences,
though that is certainly the case. Instead, in convenience describes a sensorium shored up
by smartphones, logistics, and a swelling service sector that shapes the charm and demands
of the present. It is how the world shows up, forms the boundaries of what is possible, and
establishes new thresholds for living and working (recall that gig work was initially presented
as convenient for the worker). In this context, we both want to take seriously the popular
recognition of convenience’s explanatory power—that we inhabit a convenience culture and
economy; that convenience is killing us; etc.—and challenge the assumption that it can be
explained away as mere laziness, or a problem of desire or ideology.75 Convenience, we have
begun to sketch in this essay, is not simply a consumer choice or an indulgence that can be
shirked by putting down one’s phone or by purchasing a new green product. It is experience,
datafied. Building on this and by way of conclusion, we suggest three ways the above account
of convenience requires us to reshape our approach to digital convenience and contemporary
critique.
First, convenience is the ground of politics today. It is a lure to a kind of living, a call to life,
and beyond this the background environment of what Peter Sloterdijk has called pampering,
a form of living in a ‘gigantic hothouse of relaxation’ that he associates with the welfare state,
a ‘relieving process’ that only becomes visible in ‘the age of the radical de-scarcification
of goods’.76 Sloterdijk’s reactionary politics and framing of the Global North as the norm of
experience aside, his treatment of capitalism from the angle of consumption is helpful as it
acknowledges the crucial place of convenience (as a kind of pampering; and a living inside).
Further, it relies on the promise that ‘comfort and convenience will never stop flowing and
growing’.77 Inhabiting convenience has also become part of the promise of platformization; in
Kenya or India as much as the United States or Japan—perhaps more so as these are often
the test beds for new kinds of digital convenience; what Orit Halpern and Robert Mitchell call
the ‘smartness mandate’.78 Such presumptions have become the background to radical poli-
tics today, which are sometimes themselves informed by fantasies of anti-capitalism without
75 For an account of Amazon Go stores via ideology critique, see Jenny Huberman, ‘Amazon Go, Sur veillance
Capitalism, and the Ideology of Convenience’, Economic Anthropology 8 (2021): 337-349.
76 Peter Sloterdijk, In the World Interior of Capital: For a Philosophical Theory of Globalization, trans.
Wieland Hoban, Cambridge: Polity Press, 2013, p. 171, p. 212.
77 Sloterdijk, In the World Interior of Capital, p. 171.
78 Orit Halpern and Robert Mitchell, The Smartness Mandate, Cambridge: The MIT Press, 2023.
28 THEORY ON DEMAND
inconveniences. Even: convenience communism. Such is the undertone of works like Fully
Automated Luxury Communism among other visions of automated post-scarcity or post-work
politics.79 That critiques of these works underscore the simple fact that automation is powered
by millions of laborers working as Mechanical Turks, ‘human-as-a-service’,80 or a ‘surrogate
humanity’,81 among other forms of alienation, often in the Global South, speaks to the relation
‘in convenience’ we highlight here. What’s required is a rethinking of how convenience trans-
forms work, alongside leisure and consumption, and shapes emergent aspirational horizons.
While we have focused on consumption and distribution here—and a genealogy linking the
convenience store and streaming platforms—a more triangular approach to convenience
requires bringing these observations to bear on the productivity softwares and work cultures
that also presume convenient lifeworlds.
Second, this also means recognizing how convenience as we unfold it here pushes us to
rethink the descriptions and models of politics we inherit from critical theory and media stud-
ies, among other fields. The network, the unrepresentability of transnational capital, cognitive
mapping, surveillance capitalism, and the like, may reach certain limits as they jostle against
supply chains, logistical form, and platformed affect. Enclosure as border displaces the net-
work as utopia guiding our methods and politics. This is implicitly recognized in platform stud-
ies’s tendency to undertake close analyses of a single platform, whether Instagram, Twitch, or
Twitter. These studies recognize that there is no longer a network, there are only segregated
platforms or homophilic worlds, each of which operate with their own sets of policies, politics,
resistances, and technocultures.82 If networks signaled open borders, global interconnection,
and presumptions about the movement of goods, information, and people, platforms signal
closure, national boundaries, geoblocking, redlining, the return of the locked-in model of the
internet, and global delinking. To these, we hope to offer a productive counterpoint, suggesting
that convenience allows us to think across platforms, across sectors (like retail and streaming),
and, most crucially, across geographies.
Third, convenience operates through inclusion and exclusion; animating a particular distri-
bution of in-convenience. Convenience is not only produced by exploited, abandoned and
inconvenienced workers and groups, but constitutes a kind of threshold for legitimately inhab-
iting society and its benefits. This was devastatingly captured during the COVID-19 pandemic,
when employees at firms like Amazon and Meituan were classified in many jurisdictions as
essential workers. Ordering via online services was revalued from convenience to neces-
sity, and indeed civic duty.83 The racialization and economic marginalization of necessary
workers—workers in inconvenience—reminds us of Tressie McMillan Cottom’s call for a
79 Aaron Bastani, Fully Automated Luxury Communism, London: Verso, 2019.
80 Phil Jones, Work without the Worker: Labour in the Age of Platform Capitalism, London: Verso, 2021.
81 Neda Atanasoski and Kalindi Vora, Surrogate Humanity: Race, Robotics, and the Politics of Technological
Futures, Durham: Duke University Press, 2019.
82 Wendy Hui Kyong Chun, Discriminating Data: Correlation, Neighborhoods, and the New Politics of
Recognition, Cambridge: The MIT Press, 2021.
83 Joshua Neves and Marc Steinberg, ‘Pandemic Platforms: How Convenience Shapes the Inequality of
Crisis’, in Pandemic Media: Preliminary Notes Towards an Inventory, Lüneburg: Meson Press, 2020, pp.
105-112; Andrijasevic et al., Media and Management.
29
IN/CONVENIENCE
bringing together of platform studies with racial capitalism as an analytic, this time applied to
platform services globally.84 Unequal distribution of convenience applies to racialized workers
in the U.S. context and to the migrant workers in the Chinese context—both the backbone of
the essential workforce. This brings to light the deeply unequal distribution of inconvenience,
including how in convenience consolidates modes of predatory inclusion.85 Inconvenience at
once describes, and fails to register, the condition of those who labor to produce the conve-
niences of others. From fulfillment center workers who rush to complete orders to delivery
personnel dragging large handtrucks through city streets, human workers are a crucial part
of the infrastructure of convenience.
Turning our attention from appliances to services underlines the unequal distribution of con-
venience, including the complex ways it is generated and consumed. These positions are
not mutually exclusive; a convenience store employee or Meituan food delivery worker is
still a consumer at the end of their shift. The reformulation of labor and exploitation in terms
of inconvenience also draws on a lineage of ideas that frame, if somewhat ironically, global
climate change (The Inconvenient Truth) and settler colonialism (The Inconvenient Indian)
as inconvenient.86 What it means to adopt the language of inconvenience in doing so requires
further exploration, but signals the political hold that convenience has on the imagination of
counter-politics. We cannot think of convenience without its counterpart. Living today means
inhabiting the hyphenated relation that we call in-convenience. To be outside of convenience
is either an ephemeral privilege—the yuppie who abandons their smartphone—or an extreme
form of precarity or abandonment. At an everyday level, convenience can no longer be limited
to coziness, ease or comfort; it is quite simply the price of admission. Like the protagonist of
Convenience Store Woman: embracing convenience is by now a survival strategy.
Acknowledgments
We wish to thank all the contributors in this volume for their valuable feedback on this chapter.
We thank David Humphrey, Diane Wei Lewis, and Sarah Sharma for their suggestions. Finally
we thank Jacqueline Ristola for her early research assistance on this project.
References
Andrijasevic, Rutvica; Chen, Julie Yujie; Gregg, Melissa; and Steinberg, Marc. Media and Management,
Lüneburg: Meson Press, 2021.
Appadurai, Arjun and Alexander, Neta. Failure, Medford: Polity Press, 2020.
Arendt, Hannah. The Human Condition, trans. Margaret Canovan, Chicago: University of Chicago
Press, 1958.
84 Tressie McMillan Cottom, ‘Where Platform Capitalism and Racial Capitalism Meet: The Sociology of
Race and Racism in the Digital Society’, Sociology of Race and Ethnicity 6.4 (2020): 441-449.
85 McMillan Cottom, ‘Where Platform Capitalism and Racial Capitalism Meet’.
86 Al Gore, An Inconvenient Truth: The Planetary Emergency of Global Warming and What We Can Do About
It, New York: Rodale Press, 2006; Thomas King, The Inconvenient Indian: A Curious Account of Native
People in North America, Toronto: Doubleday Canada, 2012.
30 THEORY ON DEMAND
Atanasoski, Neda and Vora, Kalindi. Surrogate Humanity: Race, Robotics, and the Politics of Technolog-
ical Futures, Durham: Duke University Press, 2019.
Barbara Maly-Bowie, ‘‘Home is where your Netflix is’ – From Mobile Privatization to Private Mobiliza-
tion’, Literary Geographies 5.2 (2019): 216-233.
Bastani, Aaron. Fully Automated Luxury Communism, London: Verso, 2019.
Baumbach, Nico; Young, Damon R.; and Yue, Genevieve. ‘Introduction: For a Political Critique of
Culture’, Social Text 34.2 / 127 (2016): 1-20.
Beverungen, Armin. ‘Remote Control. Algorithmic Management of Circulation at Amazon’, inMar-
cus Burkhardt, Mary Shnayien, and Katja Grashöfer (eds) Explorations in Digital Cultures, Lüneburg:
Meson Press, 2021, pp. 5-18.
Chang, Ron. ‘7-Eleven Opens 71,100th Store in S. Korea’, TBS, 9 July 2020, http://tbs.seoul.kr/eFm/
newsView.do?typ_800=J&idx_800=3395420&seq_800=20387997.
Chieng, Ronny. Asian Comedian Destroys America!, Netflix Comedy Special, 2019.
Chun, Wendy Hui Kyong. Discriminating Data: Correlation, Neighborhoods, and the New Politics of
Recognition, Cambridge: The MIT Press, 2021.
Cowan, Ruth Schwartz. More Work for Mother: The Ironies of Household Technology from the Open
Hearth to the Microwave, New York: Basic Books, 1983.
Cowen, Deborah. The Deadly Life of Logistics: Mapping Violence in Global Trade, Minneapolis: Universi-
ty of Minnesota Press, 2014.
Crawford, Colin Jon Mark. Netflix’s Speculative Fictions: Financializing Platform Television, Lanham:
Lexington Books, 2021.
Crawford, Kate and Joler, Vladan. ‘Anatomy of an AI System: The Amazon Echo as an Anatomical Map
of Human Labor, Data and Planetary Resources’, AI Now Institute and Share Lab (2018): 3, https://
anatomyof.ai.
Draper, Nora A. and Turow, Joseph. ‘The Corporate Cultivation of Digital Resignation’, New Media and
Society 21.8 (2019): 1824-1839.
Finn, Ed. What Algorithms Want: Imagination in the Age of Computing, Cambridge: The MIT Press,
2017.
Gardner, Chris. ‘‘Kim’s Convenience’ Stars Simu Liu and Jean Yoon Open Up on “Painful” Lack of
Diversity, “Overtly Racist” Storylines’, The Hollywood Reporter, 6 June 2021, https://www.hollywoodre-
porter.com/tv/tv-news/kims-convenience-netflix-cancelation-1234963806/.
Gore, Al. An Inconvenient Truth: The Planetary Emergency of Global Warming and What We Can Do
About It, New York: Rodale Press, 2006.
Gregg, Melissa. Counterproductive: A Brief History of Time Management, Durham: Duke University
Press, 2018.
Halpern, Orit and Mitchell, Robert. The Smartness Mandate, Cambridge: The MIT Press, 2023.
Hands, Joss. ‘Introduction: Politics, Power and ‘Platformativity’’, Culture Machine 14 (2013): 1-9.
Harney, Stefano and Moten, Fred. The Undercommons: Fugitive Planning and Black Study, Wivenhoe:
Minor Compositions, 2013.
Horgan, Colin. ‘The Tyranny of Convenience’, OneZero, 29 April 2019, https://onezero.medium.com/
the-tyranny-of-convenience-2e7fa145ab4.
Huberman, Jenny. ‘Amazon Go, Surveillance Capitalism, and the Ideology of Convenience’, Economic
Anthropology 8 (2021): 337-349.
31
IN/CONVENIENCE
Ishikawa, Akira and Nejo, Tai. The Success of 7-Eleven Japan: Discovering the Secrets of the World’s
Best-Run Convenience Chain Stores, Singapore: World Scientific, 2002.
Jagoda, Patrick. Network Aesthetics, Chicago: University of Chicago Press, 2016.
Jameson, Fredric. Postmodernism, or, The Cultural Logic of Late Capitalism, Durham: Duke University
Press, 1991.
Jones, Phil. Work without the Worker: Labour in the Age of Platform Capitalism, London: Verso, 2021.
Jones, Sarah. ‘Convenience Is Destroying Us’, Intelligencer, 2 April 2021, https://nymag.com/intelli-
gencer/2021/04/amazons-convenience-is-destroying-us.html.
King, Thomas. The Inconvenient Indian: A Curious Account of Native People in North America, Toronto:
Doubleday Canada, 2012.
Kornbluh, Anna. Immediacy: or, the Style of Too Late Capitalism, London: Verso, 2024.
Lai, Kee-hung and Cheng, T.C. Edwin. Just-in-Time Logistics, London: Routledge, 2009.
Lamarre, Thomas. ‘Regional TV: Affective Media Geographies’, Asiascape: Digital Asia 2.1-2 (2015):
93-126.
LeCavalier, Jesse. The Rule of Logistics: Walmart and the Architecture of Fulfillment, Minneapolis:
University of Minnesota Press, 2016.
Lobato, Ramon. Netflix Nations: The Geography of Digital Distribution, New York: NYU Press, 2019.
Lotz, Amanda D. Portals: A Treatise on Internet-Distributed Television, Ann Arbor: Michigan Publishing,
2017.
Marutschke, David. Continuous Improvement Strategies: Japanese Convenience Store Systems, Bas-
ingstoke: Palgrave Macmillan, 2011.
Massumi, Brian. ‘Everywhere You Want to Be: Introduction to Fear’, in Brian Massumi (ed) The Politics
of Everyday Fear, Minneapolis: University of Minnesota Press, 1993, pp. 3-37.
_____. ‘Fear (The Spectrum Said)’, Positions: East Asia Cultures Critique 13.1 (Spring 2005): 31-48.
McMillan Cottom, Tressie. ‘Where Platform Capitalism and Racial Capitalism Meet: The Sociology of
Race and Racism in the Digital Society’, Sociology of Race and Ethnicity 6.4 (2020): 441-449.
Murata, Sayaka. Convenience Store Woman, trans. Ginny Tapley Takemori, New York: Grove Atlantic,
2018.
Naruo, Shinji and Toma, Sorin George. ‘From Toyota Production System to Lean Retailing. Lessons
from Seven-Eleven Japan’, in Jan Olhager and Fredrik Persson (eds) Advances in Production Manage-
ment Systems, New York: Springer, 2007, pp. 387-395.
Negō, Tai and Hiraki, Kyōichi. Konbini Gyōkai No Dōkō to Karakuri Ga Yoku Wakaru Hon (A Book for
Really Understanding the Trends and Mechanisms of the Convenience Store Industry), Tokyo: Shuwa
Shisutemu, 2015.
Neves, Joshua and Steinberg, Marc. ‘Pandemic Platforms: How Convenience Shapes the Inequality of
Crisis’, in Pandemic Media: Preliminary Notes Towards an Inventory, Lüneburg: Meson Press, 2020,
pp. 105-112.
Neves, Joshua. ‘Social Media and the Social Question: Speculations on Risk Media Society’, in
Bhaskar Sarkar and Bishnupriya Ghosh (eds) The Routledge Companion to Media and Risk, New York:
Routledge, 2020, pp. 347-361.
Ngai, Sianne. Our Aesthetic Categories: Zany, Cute, Interesting, Cambridge: Harvard University Press,
2012.
32 THEORY ON DEMAND
_____. Theory of the Gimmick: Aesthetic Judgment and Capitalist Form, Cambridge: Harvard Universi-
ty Press, 2020.
Nguyen, Dang. ‘Convenient efficiency: A media genealogy of QR codes’, New Media & Society (2022):
1-21.
Pallitto, Robert M. Bargaining with the Machine: Technology, Surveillance, and the Social Contract,
Lawrence: University Press of Kansas, 2020.
Rossiter, Ned. Software, Infrastructure, Labor: A Media Theory of Logistical Nightmares, New York:
Routledge, 2016.
Sharma, Sarah. In the Meantime: Temporality and Cultural Politics, Durham: Duke University Press,
2014.
Shirai, Mieko; Kojima, Takeshi; and Oguri, Masashi. ‘Konbini Wo Kagaku Suru’ (Doing the Science of
Convenience Stores), Shūkan Diamond (Diamond Weekly), 29 October 2016.
Shove, Elizabeth. Comfort, Cleanliness and Convenience: The Social Organization of Normality, Oxford:
Berg, 2003.
Sloterdijk, Peter. In the World Interior of Capital: For a Philosophical Theory of Globalization, trans.
Wieland Hoban, Cambridge: Polity Press, 2013.
Srnicek, Nick. Platform Capitalism, Malden: Polity Press, 2017.
Steinberg, Marc. The Platform Economy: How Japan Transformed the Commercial Internet, Minneapo-
lis: University of Minnesota Press, 2019.\
_____. ‘From Automobile Capitalism to Platform Capitalism: Toyotism as a Prehistory of Digital Plat-
forms’, Organization Studies 43.7 (2022): 1069-1090.
‘The World’s Most Valuable Resource is No Longer Oil, but Data’, The Economist, 6 May 2017, https://
www.economist.com/leaders/2017/05/06/the-worlds-most-valuable-resource-is-no-longer-oil-but-
data.
Thrift, Nigel. Knowing Capitalism, London: Sage, 2005.
Tierney, Thomas F. The Value of Convenience: A Genealogy of Technical Culture, Albany: State Univer-
sity of New York Press, 1993.
West, Emily. Buy Now: How Amazon Branded Convenience and Normalized Monopoly, Cambridge: The
MIT Press, 2022.
Whitelaw, Gavin H. ‘Konbini-Nation’, in Katarzyna J. Cwiertka and Ewa Machotka (eds) Consuming Life
in Post-Bubble Japan, pp. 69-88.
Williams, Raymond. Television: Technology and Cultural Form, London: Routledge, 2003.
Wu, Tim. ‘The Tyranny of Convenience’, The New York Times, 16 February 2018, https://www.nytimes.
com/2018/02/16/opinion/sunday/tyranny-convenience.html.longer-oil-but-data.
Thrift, Nigel. Knowing Capitalism, London: Sage, 2005.
Tierney, Thomas F. The Value of Convenience: A Genealogy of Technical Culture, Albany: State Univer-
sity of New York Press, 1993.
West, Emily. Buy Now: How Amazon Branded Convenience and Normalized Monopoly, Cambridge: The
MIT Press, 2022.
Whitelaw, Gavin H. ‘Konbini-Nation’, in Katarzyna J. Cwiertka and Ewa Machotka (eds) Consuming Life
in Post-Bubble Japan, pp. 69-88.
33
IN/CONVENIENCE
Williams, Raymond. Television: Technology and Cultural Form, London: Routledge, 2003.
Wu, Tim. ‘The Tyranny of Convenience’, The New York Times, 16 February 2018, https://www.nytimes.
com/2018/02/16/opinion/sunday/tyranny-convenience.html.
34 THEORY ON DEMAND
CONVENIENT-FOR-THE-PEOPLE POLICING,
PROTECTED CONSUMER-CITIZENS AND
INFRASTRUCTURES OF DISPOSABILITY IN
NORTHWEST CHINA
DARREN BYLER
Since 2014 authorities in the Xinjiang Uyghur Autonomous Region in Northwest China have
introduced a ‘seamless’ ‘people’s convenience’ (bianmin) checkpoint, passcard and face
surveillance system that tracks and inhibits the movement of targeted minority citizens while
providing ‘green lanes’ and greater feelings of security to system-approved Han settler citizens.
The system draws in part on business management logistics in data capitalism, a Japanese
and Singaporean model of policing infrastructure, and new developments in surveillance
technology to produce a limit case in contemporary population management.
Since China’s turn to a market-oriented economy in the 1990s, the Uyghur Region has
become a domestic center of an extractivist oil, gas, coal and cotton economy driven by
the settlement of millions of non-native people—figured as ‘The People’ in the convenience
surveillance system—from other parts of the country. Over the 2010s, Uyghur resistance
to land dispossession and the settler take-over of Uyghur institutions, reached a threshold
moment resulting in the buildout and implementation of the ‘people’s convenience’ system.
While aspects of this resistance featured Uyghur violence toward Han civilians in the form
of suicide attacks, the system was used as part of the People’s War on Terror to assess the
past ‘terrorist and extremist’ digital activities of not just the hundreds of Muslims connected
to violent crime, but rather the entire population of 15 million Uyghurs, Kazakhs and other
Muslims who were native to the region. This resulted in the removal of more than one million
Muslims from urban and rural contexts into a system of high-tech internment camps.1 Millions
more were placed on finely graded watchlists, while their movements and digital behavior
were monitored by automation-assisted surveillance and state workers. A significant portion
of both former detainees and those on watchlists were assigned to years of labor in highly
surveilled industrial parks in the region as underpaid workers in factories, plantations, and
undesirable service sector jobs such as street sweeping.2
This chapter explores the valences of the people’s convenience system to understand the
way policing and policy discourses and technologies conflate citizenship with particular types
of disciplined, racialized, convenient consumption. I argue that this mode of sociality, which
1 Darren Byler,In the Camps: China’s High Tech Penal Colony, Columbia Global Reports, New York:
Columbia University Press, 2021.
2 Darren Byler,Terror Capitalism: Uyghur Dispossession and Masculinity in a Chinese City, Durham:
Duke University Press, 2022; ‘关于印发新疆维吾尔自治区’十四五’就业促进规划的通知 [Notice on Printing and
Distributing the ‘14th Five-Year’ Employment Promotion Plan of Xinjiang Uyghur Autonomous Region]’,
Xinjiang Government, 14 December 2021, https://web.archive.org/web/20230324214327/https://
archive.ph/PFzsb.
35
IN/CONVENIENCE
Neves and Steinberg refer to as being ‘in-convenience’, is premised in part on spending the
lifetimes of disposable others.3 By tracing the origins of ‘Convenient-for-the-People’ thinking
through the customer-to-customer business practices of information-centric capitalism, Mao-
ist political practice, and contemporary global city models, the chapter shows how designs
for efficiency in economic consumption coexist, or are ‘stacked’ on top of, digitized urban
governance platforms and networked checkpoints. This application creates a form of state
power that begins to normalize and automate differential forms of governance, producing a
form of sociality that can be characterized as convenient disposability.
Discourses of Convenience
Variations on the term ‘convenience’ appears thousands of times in internal police reports
from the capital of Xinjiang, Ürümchi—a city of more than 2 million mostly non-Muslim Han
settlers in China’s vast Muslim majority frontier. These reports from 2017 to 2019 were drawn
from a mobile policing system that was built by a state-contractor Landasoft, which brands
itself as ‘China’s Palantir’.4 The files were part of the base data set that forms the People’s
Convenience system. This aspect of the system relied on a network of tens of thousands of
formal and informal informants to file intelligence reports regarding the effects of policy and
infrastructure implementation of the people’s convenience system.5
The discussion of ‘convenience’ in the files centered around consumption and mobility. Initial-
ly, in early 2017 as the mass internment campaign was just beginning, the reports catalogued
Uyghur complaints regarding how inconvenient it was for most Muslims in the city to find
jobs, rent apartments, or stay in hotels anymore. The police reported that Han settlers were
likewise complaining that nearly all the Uyghur bakers and butchers had been detained or
expelled from the city, so Han urbanites could no longer buy bread fresh out of a clay oven.
Over and over, the reports noted Han citizens’ complaints about how inconvenient the mass
internment of the Uyghurs was for the day-to-day business of their lives. But as the months
documented in the reports wore on, discussion of inconvenience began to fade into a state
manufactured ‘digital resignation’.6 By the end of 2017, both Muslim and Han citizens no
longer complained. It appeared as though Uyghur citizens realized that the intensification of
3 See the introduction to this book, Joshua Neves and Marc Steinberg, ‘In Convenience’.
4 The internal police files examined in this chapter were obtained by the news journal The Intercept
through the investigative journalist Yael Grauer. See Yael Grauer, ‘Millions of Leaked Police Files
Detail Suffocating Surveillance of China’s Uyghur Minority’, The Intercept, 29 January 2021, https://
theintercept.com/2021/01/29/china-uyghur-muslim-surveillance-police/. I assisted Grauer in assessing
the millions of files included in the policing system and am now working with a team of researchers
to make a subset of these documents publicly available through the Xinjiang Documentation Project,
housed at Simon Fraser University and the University of British Columbia.
5 For a discussion of the history on internal information systems in China, see Martin K.Dimitrov,
Dictatorship and Information: Authoritarian Regime Resilience in Communist Europe and China, London:
Oxford University Press, 2023.
6 For a parallel discussion of the way corporations in Europe and North America cultivate a similar
resignation among consumers, see Nora A. Draper and Joseph Turow, ‘The Corporate Cultivation of
Digital Resignation’, New Media and Society 21.8 (2019): 1824-1839; and Neves and Steinberg, ‘In
Convenience’, p. 3.
36 THEORY ON DEMAND
the People’s Convenience policing system was intended to restrict or ban Uyghur movement
and consumption. Han citizens likewise realized that bread should now be bought through
app-based delivery services or in barcoded packages at convenience stores. The reports also
note that Muslim run convenience stores that had catered before to the halal standards of
Muslim customers now sold beer and cigarettes—making them spaces more convenient for
settler consumption. Daily face scans were routinized and over time Han people learned that
they could just walk or drive through ‘green lanes’ based on the way the phenotypes of their
faces were recognized by the system and security workers. The region after all was understood
by Han settlers to be a new frontier of the global economy. Many of them had been drawn
there because of the lucrative resource sector and the promise of a convenient middle-class
lifestyle. A new consistency of being ‘in convenience’ reemerged as they realized anew that
they were ‘the People’ the governance system was designed to serve.
The police were observing these shifting dynamics as well. In a report from mid-2017 an
officer surnamed Li notes:
Since they are facing such high-pressure from the system, we should really screen
Uyghur-speaking beggars in the city. Many Uyghur-speaking people from Southern
Xinjiang dare not enter and exit places that have ID check points, so they are likely to
be in public spaces that oer shelter from the wind and rain such as ATM booths.7
The People’s Convenience surveillance system, it appeared, was pushing Uyghurs from their
homes and jobs in the service and construction sectors of the city to repurpose an older plat-
form on which the People’s Convenience system was literally stacked (or positioned directly
beside)—the distributed network of on-demand cash withdrawal banking booths built to
shelter computerized banking equipment. The layers of the city—financial infrastructure
meeting security infrastructure—formed a new segment in the enmeshed convenience city;
a material and digital instantiation of what the theorist Benjamin Bratton might refer to as a
megastructure of stacked digital and material platforms.8 Uyghurs were pushed to reclaim
spaces in the platform stack where they waited to be stumbled over by system-approved
citizens and swept up by the police who followed.
However, as I observed when I visited the region in 2018, these spaces of temporary respite
from life on the run would soon be fitted with surveillance equipment. Similarly, parks, under-
passes, and construction sites were either boarded up, or outfitted with ID checkpoints or
cameras. There were no spaces for Uyghurs to hide in this system. A temporary inconvenience
for Han citizens—a logistical transformation to the process of buying bread or mutton—was
met by a permanent position of disposability for Uyghurs. This framing, which I will return to
in the final section of the chapter, is a way of thinking about a particular type of a racialized
inconvenient other—a living form of human waste—that Neferti Tadiar and Melissa Wright
identify as a feature of the contemporary global economy.9
7 Xinjiang Documentation Project, “Internal Police Files,” Unpublished Manuscript, 2024.
8 Benjamin H. Bratton,The Stack: On Software and Sovereignty, Cambridge: The MIT Press, 2016.
9 Neferti X.M. Tadiar,Remaindered Life, Durham: Duke University Press, 2022; Melissa
Wright,Disposable Women and Other Myths of Global Capitalism, New York: Routledge, 2013.
37
IN/CONVENIENCE
Scholarship on the affective politics of counter-terrorism and governance has shown that fear
of the other as a structure or atmosphere that pervades popular discourse is a prime motiva-
tor of technological interventions in contemporary public life.10 This chapter takes a similar
approach to show that within this broader structure an infrastructure of feeling—to borrow a
framing from Lauren Berlant—appears under the sign of being ‘in convenience’.11 That is to
say, as fear is operationalized, it does more than produce a structure of feeling. It is more than
an ‘atmospherically felt but unexpressed class-based affect’, rather it ‘confirms and solidifies
the sediment of many proximate kinds of sociality, including pasts and futures as they express
themselves in the present’.12 Berlant is thinking about the way the global pandemic and
protests against police brutality toward Black Americans sanctioned corresponding types of
social and material infrastructure—defined as technological and organizational systems of
management and communication—intervening into American life in the late 2010s. These
moments of heighted state intervention and intersubjective experience—mass protests in
the streets, mass quarantines, corresponding police brutality, and systematic public health
interventions—produced a kind of ‘glitch’ in life lived as normal, that allowed patterns, habits,
and norms to be shown in sharper relief.13 By focusing attention on infrastructural interven-
tions and their norm-shaping effects in these moments, worldbuilding is shown in motion.
Inconvenience, the friction that emerges from being in relation with others, challenges the
sanctioned temporal and spatial sovereignty of privileged citizens. The disappearance of
inconvenience in the discourse of the privileged—as shown in the police reports I described
above—is thus an indicator of political and economic power being routinized. It marks the
consistency-making that makes the disposability of the ethno-religious other common sense.
Spending the lifetimes of Muslims, by using convenience stations to trap them, and in turn,
locking them up and putting them to work, became normal.
This discourse and infrastructure of feeling was not entirely new. Over the course of 24 months
of ethnographic fieldwork between 2011 and 2020 in the Xinjiang Uyghur Autonomous Region
and in nearby Kazakhstan, and in reviewing thousands of police files, I came to understand
the particular infrastructural valences of ‘convenience’ (fangbian) in relation to politically and
economically sensitive topics in Northwest China. As in Euro-American etymologies of the
term, in the Chinese context, perhaps influenced by Confucian conservatism, convenience
often referred to speech and behaviors that would not destabilize the social order.14 In my
conversations with Han interlocutors and government workers, I often found conversations
redirected. For instance, I was told that it was ‘not convenient’ (bu fangbian) to talk about
the history of the independent East Turkistan Republic or the way Turkic Muslims such as
Uyghurs and Kazakhs were prevented from practicing aspects of their faith. For my Muslim
friends this political use of the term extended far beyond vague conversation, they were often
told that their halal lifestyle itself was ‘inconvenient’ (bu fangbian).
10 Joseph Masco,The Theater of Operations: National Security Affect from the Cold War to the War on
Terror, Durham: Duke University Press, 2014; Brian Massumi,Ontopower: War, Powers, and the State of
Perception, Durham: Duke University Press, 2015.
11 Lauren Berlant,On the Inconvenience of Other People, Durham: Duke University Press, 2022.
12 Berlant,On the Inconvenience of Other People, p. 20.
13 Berlant, p. 95.
14 For an etymology of the term, see Neves and Steinberg, ‘In Convenience’, p. 14.
38 THEORY ON DEMAND
For instance, a young Kazakh woman told me how the term was used when she applied for a
job at a newspaper soon after she graduated from college in the late 2000s. She said:
At first they were welcoming and interested, but then they called and rejected me.
When I asked why (the Han manager) provided a bunch of excuses, but the main
reason was that ‘it will be “inconvenient” (bu fangbian) for you due to your diet restric-
tions’. I said that shouldn't be a problem and I was still interested in the job. I pressed
and pressed him, but he kept talking about ‘inconvenience’ over and over. Even when I
said I was not that religious, he still said no.
The young woman said that it ‘felt terrible’ to be rejected like that simply because of her eth-
nicity and presumed religious values. It would have been easier to accept if the manager would
have admitted that it just was ‘not “convenient” for them’. It was the fact that ‘they made it
about me, “it's not convenient for you”’ that made her most angry. She said that the paternal-
ism in this language game of ‘fake politeness’, where Han authority figures ‘pretend to care’
about minorities, but actually just want to put inconvenient Muslims out of sight and mind, is
a common feature of daily life for Uyghurs and Kazakhs in China’s internal colonial project. As
a result, she saw her lifepath rerouted into work environments where the inconvenience of
Muslimness did not interfere as deeply with Han convenience. The theater of inconvenience
that confronted the Kazakh woman demanded a price from her—a life lived as an inconvenient
object, but a life nonetheless. However, when the social friction of inconvenience intensified
as a result of the People’s War on Terror in the mid-2010s, producing a ‘glitch’ in the normal
Muslim inconvenience, the inconvenient Muslim demanded a more durable infrastructural fix.
Convenience and State and Economic Power
When it comes to state power, as in data capitalism itself, the concept of convenience in service
provisioning aligns closely with efficiency in production. Building on Neves and Steinberg’s
framing of being ‘in convenience’ as a form of contemporary life,15 it can be said that when it
comes to governance and development of states, convenience and efficiency are focused on
controlling the timing and spacing of power reproduction. As the anthropologist Akhil Gupta
has argued in his study of bureaucratic power and social welfare in contemporary India, power
cannot be understood as merely ‘distributed’ from a centralized reserve.16 Rather it requires
a disaggregated theory of the state which focuses on the mundane everyday practices and
infrastructures of governance and market development. Power, characterized by the ability
to impact others and be impacted, is enacted not by a unitary state or corporate class but
through discrete acts within the infrastructure of a state and economy. Identification of groups
of individuals and their property with different degrees of protection, and the convenience of
mobility and power, fundamentally constitutes the modern state form; this isthecentral daily
business of policing, its infrastructure and being ‘in-convenience’. As the intensification of
such a system in contemporary Northwest China demonstrates, the material experience of
15 Neves and Steinberg, ‘In Convenience’.
16 Akhil Gupta,Red Tape: Bureaucracy, Structural Violence, and Poverty in India, Durham: Duke University
Press, 2012.
39
IN/CONVENIENCE
the state form and its economy gains structure and significance at locations such as ‘Con-
venient-for-the-People’ banking access points, call centers, and surveillance hubs. These
locales regulate intra-urban movement and control the social environment. The structures and
infrastructures of convenience both facilitate and restrict certain forms of movement, acting
as both visible and invisible walls that categorize populations and permitted actions.
Convenience architectures, symbolic of and supporting policing and economic systems, draw
citizens into ‘scripts of action’ based on gendered and ethno-racial guidelines.17 They observe
and constrain their targets while extending the power of those who design and control these
technologies. Collectively, infrastructures of convenience shape dispositions and patterns
of propensity that influence life itself. Despite attempts to provide technologists and urban
managers with an omniscient view of the city, the complexity of these systems results in power
being enacted in small doses at discrete locations—checkpoints, surveillance cameras, traffic
meters—often overseen by city workers or police with limited perspectives on the entire system.
This complexity mirrors the distribution of information in a platform economy, dominating global
production since the late twentieth century.18
As Marc Steinberg has shown, platform economies emerge out of Taylorist modes of produc-
tion management in the form of Toyotism: a mode of production that focuses on just-in-time
manufacturing utilizing data analytics, supply chain and delivery logistics, standardization,
and temporary workers. China’s adoption of platform economy principles and infrastructure
borrowed largely from Japan, and spaces with significant Japanese influence, like Taiwan and
Hong Kong, in the 1990s, began to organize social life. Housing, infrastructure, and logistics
came to be centered around standardization, just-in-time efficiency, and hundreds of millions
of temporary migrant workers.19 By the 2000s, much of China’s social life revolved around
platform-like systems, evolving further with the integration of digital platforms, automated
inventory systems, and logistics interfaces in the 2010s.
In a platform economy, workers engage in simplified, highly specialized tasks, promoting effi-
ciency but reducing individual knowledge of overall productive activity. Similarly, a ‘Conve-
nient-for-the-People’ city in a platform economy isolates state workers, preventing them from
comprehending the full effects of their actions. From the perspective of low-level workers
and the governed, these systems appear as black boxes generating statistics, taxation, and
regulation. For city workers, it translates into banal, bureaucratic work, maintaining system
equilibrium by ticking boxes. Over the 2000s, as China moved rapidly into a market-driven data
capitalist system, the policing of political speech and action was given a different gloss through
the term bianmin or ‘Convenient-for-the-People’. That is to say, the discourse of convenience
17 Bryan Pfaffenberger, ‘Social anthropology of technology’,Annual review of Anthropology21.1 (1992):
491-516.
18 Marc Steinberg, ‘From automobile capitalism to platform capitalism: Toyotism as a prehistory of digital
platforms’,Organization Studies43, no. 7 (2022): 1069-1090; Marc Steinberg,The Platform Economy:
How Japan Transformed the Consumer Internet, Minneapolis: University of Minnesota Press, 2019.
19 Jacky FL Hong, Robin Stanley Snell, and Mark Easterby-Smith, ‘Cross-cultural influences on
organizational learning in MNCS: The case of Japanese companies in China’,Journal of International
Management12.4 (2006): 408-429.
40 THEORY ON DEMAND
was not limited to discussions of ‘sensitive’ historical and life practice concerns, it was also a
major part of the development of contemporary, logistics-driven capitalism.
In its most general sense, the term bianmin is linked to on-demand 24-hour service. Drawing
on Singaporean models of urban design and governance, which were in turn modeled on
Japanese infrastructural systems, the term began to appear in China first in the 2000s in the
banking system.20 This was one of the first spaces to be digitized with ‘Convenient-for-the-Peo-
ple’ ATM services that could be accessed any time of day at convenient locations aggregated
throughout urban space.21 This convenience-as-a-public-utility then moved from the financial
sector and spread throughout the service economy with on-demand call centers and on-line
message boards providing instant service to their customer base. In the 2010s it also began
to cross over into the public sector in the naming of hospital payment processing centers and
postal service centers. At the same time, it became synonymous with the national level 120,
119 and 110 emergency phone line system—which linked citizens with emergency health
care, fire fighters and the police.22
At its core, the logic of the ‘Convenient-for-the-People’ infrastructure systems appeared to
provide ordinary people with a full range of autonomous action, to provide self-service to cit-
izens framed as consumers of goods and services. In the policing literature, this is framed as
good for the consumer and the producer since direct access helps both to meet each other’s
needs.23 And while ‘Convenient-for-the-People’ services do indeed make banking and health
emergency services easier to access, what is missing here, in the context of data-driven cap-
italism and governance, is the way information flows are typically asymmetrical. The service
provider actually gains a great deal more knowledge about the consumer’s behavior than the
reverse, particularly when the relationship is digitized, lurking in the background in apps and
surveillance tools. While being ‘in convenience’ is often framed around economic production
or consumption, it is also, in step with and independent of these, a form of governance, stan-
dardization, or world configuration. The producer-consumer relation is not only about market
optimization, it is also about the optimization of power.
The Power of Convenience Stations in a Colonial Context
In 2014 the People's War on Terror positioned the Uyghur region at the forefront of both ‘smart
cities’ and ‘preventative policing’ that were articulated to nationwide ‘people’s convenience’
20 Lu Peng and Bi Wenni, ‘Comparative study of smart city security governance: Taking Singapore and
Shanghai as examples’,Journal of Sinology17.1 (2023): 128-144. For a discussion of Singapore as a
model for urban planning in China, see also Ananya Roy and Aihwa Ong (eds) Worlding Cities:Asian
Experiments and the Art of Being Global. Studies in Urban and Social Change. Malden: Wiley-Blackwell,
2011.
21 Li Yuhua, ‘Atm防护便民亭(Automatic teller machine (ATM) protective convenience-for-people booth)’,
PRC Patent Announcement Number CN 201891278 U, filed 13 December 2010, and issued 6 July
2011, https://patentimages.storage.googleapis.com/17/66/43/5a1adfff9b2071/CN201891278U.pdf.
22 ‘Convenient Police Service in the Internet Age’, Modern World Police 3 (2016): 12-14.
[现代世界警察 2016, (03),12-14互联网时代的便民警务 吴尚 初福善 王晓燕 顾庆忠 谢红 郭广生 陈川]
23 ‘Convenient Police Service in the Internet Age’.
41
IN/CONVENIENCE
services. This frontier intensification was shaped by the construction of more than 9,000
surveillance hubs known as People's Convenience Police Stations and the recruitment of
10,000s of assistant data police (xiejing) as policing ‘grid workers’ (wangge renyuan) that
implement the hubs and associated checkpoints.24 These stations, like the design and logis-
tics of Chinese cities themselves, were modeled on police booths used in Singapore and
Japan. Both emerge out of imperial legacies, from the expansionist Meiji era in Japan and
the British police booth model in Singapore. Similarly, in those national contexts, both were
folded into platform economies that transformed their respective economies in the 1980s. In
both places, police booths are frequently positioned next to convenience stores as a parallel
form of provisioning. But while the material infrastructure of Chinese police booths may be
borrowed from the platformization of other Asian cities, the function of the police booths in
Northwest China—the site of their most intensive use in China, focused on a racialized colonial
project of policing inconvenient Muslim bodies. The remainder of this chapter will focus on
the functions and shape-shifting effects of these convenience stations.
The stations, which were built every 200-300 meters, function as surveillance hubs in the
segmented ‘grid’ of the policing logistics system that strove to achieve knowledge of any
abnormalities in the lives of residents and non-residents of the grid as they moved between
checkpoints. By breaking down the population of the city into quantifiable numbers they made
the task of tracking the behavior of individuals feasible. For many in the protected population
of residents, the convenience stations appear to be a silent presence, perhaps even a wel-
come presence, since their ubiquity means that state workers are always nearby. According
to an interview with a Ürümchi police chief named Lu Wenlong, the People’s Convenience
Police Stations provide ‘The People’ with 24-7 access to public toilets, wireless Wi-Fi, first aid
kits, mobile phone charging stations, escort services for the elderly and children, and legal
publicity work. That is, they facilitate a relationship of being ‘in convenience’ with on-demand
data and care services. The convenience stations attempt to increase the efficient circulation
of the things and people that are wanted by local authorities, while decreasing the circulation
of things and people that are unwanted. Local authorities refer to this as an important aspect
of ‘stability work’ in fighting the ‘three evil forces’ of religious extremism, ethnic separatism,
and violent terrorism. Put simply, the convenience stations are there to make settlers feel
safe from violence in a space where they are in fact the violent presence.
As part of an infrastructure of feeling, the convenience stations interpellate both the ‘consum-
ers’ of policing services, and the Muslims that they target, pulling them into an ongoing rela-
tionship with the aesthetics and gaze of the state. The blocky concrete stations often occupy
sections of previously open sidewalk or parking lots near the checkpoints at the entrances of
institutions. They are often two stories high and feature a tower or pole topped with red and
blue flashing lights. No matter where a pedestrian is in the city, they are often able to spot
the lights of a convenience station down the street. If there is none within sight, one can rest
assured that there is one right around the corner. The stations often have a horizontal red
24 Vicky Xiuzhong Xu, James Leibold, and Daria Impiombato, Architecture of Repression: Unpacking
Xinjiang's Governance, Australian Strategic Policy Institute, 2021, https://www.aspi.org.au/report/
architecture-repression.
42 THEORY ON DEMAND
LED sign that scroll slogans from the most recent national Party congress, or, failing that, a
red banner with similar phrases about social stability and fighting terrorism. The stations are
built with pedestrians in mind. The convenience stations are meant to be seen as a positive
sign that the state, like a benevolent bank distributing ATMs throughout the topography of
its customer base, was there to provide services to the protected majority. They remind the
protected population of the care of the state, and to stay vigilant, while at the same time they
send a message to Muslims that they are always being watched. It was clear to everyone I
spoke with, when I conducted field research in the region in 2018, who exactly the ‘Conve-
nient-for-the-People’ stations were convenient for. Convenience is far from a neutral descriptor.
It naturalizes the consumer-citizen as carriers of value and simultaneously produces Muslims
as separatists, extremists, terrorists as carriers of threat.
To understand the arrival of convenience stations, checkpoints and other tools of sorting and
atomization in the ‘doing’ or ‘producing’ of infrastructural power in the city, I want to return
here to the plan and function of the platform economy and its intersection with the smart
city.25 The notion of the ‘smart city’ can be theorized as an analogous construct to a stack of
platforms whose primary objective is the enclosure of docile yet productive citizens. In this
context, convenience stations are analogous to workstations or platform distribution hubs
wherein state workers actively engage in the sorting and shaping of citizen behaviors in the
most efficient way possible. The five zeroes of the Toyotist model can be thought of in this
context as: zero delay in detection, zero unaccounted-for bodies, zero unneeded information,
zero crime, and zero breakdowns in the system.26 The stations strategically positioned within
the city and at its perimeters function as pivotal nodes within a vast surveillance apparatus.
These checkpoints sort the population: matching biometrics to official identification docu-
ments at turnstiles. Additionally, the multiple checkpoints associated with the stations are
equipped with datadoors and metal detectors, developed by the China Electronics Technol-
ogy Corporation: a military contractor and parent company to HikVision, the world’s largest
camera manufacturer. These devices were specifically designed to detect and register the
MAC addresses of smartphones, further contributing to the surveillance regime.27
This process, particularly in its initial implementation phase, was time-intensive and indis-
criminate, affecting all citizens. However, as narrated by a Han interviewee, a noticeable
shift occurred over time, matching the changes in discourses of convenience discussed at
the beginning of this chapter. Initially, all passengers, regardless of ethnicity, were subjected
to these checks, resulting in significant delays. However, a subsequent modification in the
operational protocol led to the exclusion of Uyghurs from public transportation services, as
25 For a further exploration of this analogy see Darren Byler, “Producing ‘Enemy Intelligence’: Information
Infrastructure and the Smart City in Northwest China,” Information & Culture 57.2 (2022): 197-216.
26 The five zeros of Toyotism are famously: Zero delay: products are manufactured in just-in-time; Zero
stock: no overproduction is tolerated; Zero paper: paper consumption is reduced to a minimum; Zero
defects : no product must be defective; and Zero breakdowns: no machine defects are tolerated.
See Ricardo Silva et al., ‘Active Learning ‘Factory of Boxes’ in the Teaching-Learning Processes in
Engineering and Entrepreneurship’,Journal of Technical Education and Training13.3 (2021): 1-14.
27 ‘China: Big Data Fuels Crackdown in Minority Region’, Human Rights Watch, 26 February 2018, https://
www.hrw.org/news/2018/02/27/china-big-data-fuels-crackdown-minority-region.
43
IN/CONVENIENCE
buses began to depart checkpoints, leaving Uyghur individuals behind while transporting Han
passengers directly to urban centers. This change was perceived by the Han interviewee’s
relatives as an improvement, indicating a societal adaptation to, and acceptance of, the eth-
noracially biased operational mechanisms of the convenience stations.
The stations and checkpoints formed what the Ürümchi police chief purported to be a ‘seam-
less’ system28—a response to Xi Jinping’s 2014 call to build ‘walls of steel’ and a ‘net over the
sky’ to defend against Muslim terrorism.29 The tasks of the data police on the production line
consisted of ‘fixed duty, video patrol, car patrol, foot patrol, and plainclothes patrol’.30 Based on
prior research, it is clear that much of the work of police assistants focused on the first two tasks,
sorting populations at fixed checkpoints and watching banks of video monitors.31 In some areas
such as mosques and train stations, face recognition-enabled cameras would issue alarms if
someone identified by a watchlist walked in front of them or scanned their ID at a turnstile.32
A Kazakh woman I interviewed in Kazakhstan in January 2020, soon after she fled across
the border, said that it ‘became normal’ to have her face and ID scanned 10 times during
an average work day.33 Her phone was also scanned with a plug-in device on a regular basis.
The city itself effectively became what might be thought of as platform stack for producing
protected citizens and enemy ‘terrorists’—with automated dataveillance scans and biometric
surveillance grids fitted on top of material checkpoints run by neighborhood level informants,
all of which were constantly feeding data into an integrated region wide system. The form and
function of the convenience stations and checkpoints separated movement and assessment
of individuals into discrete procedures which were replicated across the entire population of
city blocks in sequence. That is, the arrangement of the stations and checkpoints—the aggre-
gation of blocky gray concrete buildings at traffic intersections and the chrome-barred, cam-
era-studded turnstiles in entrances to public buildings, housing complexes and jurisdictional
boundaries—amplified the inconvenient presence of some people and behaviors, justifying
their removal, and thus smoothing out the space for the convenience of the protected majority.
The police records I have reviewed show that the default, non-Muslim body who moves through
the checkpoints and automated surveillance systems without targeted assessment is unnoted
even as their movements are also recorded.
Much of this work of producing disposable others and protected citizens was done using the
parameters of smartphone apps and digital forensics tools. For instance, on a weekly basis
28 Zhang Xinde, ‘Wulumuqi shi dajie xiao xiang jiang jian 949 ge bianmin jingwuzhan [949 People’s
Convenience Police Stations have been built in the streets of Ürümchi]’, Yaxin Net, 2016, https://
kknews.cc/society/2a4mrng.html.
29 ‘Xi urges anti-terrorism ‘nets’ for Xinjiang’, Xinhua News, May 29 2014, http://www.chinadaily.com.cn/
china/2014-05/29/content_17552457.htm
30 Zhang Xinde, ‘Wulumuqi shi dajie xiao xiang jiang jian 949 ge bianmin jingwuzhan’.
31 Byler, In the Camps.
32 Chinese Government Procurement Network, ‘Xinjiang Shawan County Smart (Safe) Project Feasibility
Study’, ChinaFile, 2017, https://www.chinafile.com/library/reports/xinjiang-shawan-county-smart-safe-
project-feasibility-study.
33 Byler, In the Camps.
44 THEORY ON DEMAND
at half a dozen convenience stations and assorted checkpoints, around 40 officers scanned
the phones of more than 2000 people using a digital forensics tool called an ‘Anti-Terrorism
Sword’. These devices made by a range of companies used software from the company Mei-
ya Pico and the Urumchi Public Security Bureau. The tools were themselves adapted from
devices developed by the Israeli company Cellebrite and were used to search for more than
53,000 unique identifiers of religious and violent activity. In addition to scanning phones, the
convenience station workers also manually scanned the faces of more than 900 people using
face recognition software installed on their smartphones. Throughout 2018 the weekly reports
present slight fluctuations in these numbers, some weeks the station workers scanned slightly
more, some weeks slightly less. In an average week it is likely that close to half of the adult
Muslim population in the jurisdiction was subjected to phone scans. My interviewees told me
that they began to anticipate that their phone could be scanned at any time.34 The consistent
aggregation and repetition of these digital and the biometric scans train the Muslim population
in how to use their phones and stay within their allotted place. Information that should have
been disregarded as noise was instead detected via the parameters of assessment as signals
of being too Muslim. Ethnic and religious difference was detached from the lifeworld of the
surveilled, becoming a technology—a multi-step algorithmic assessment scored on a smart
phone interface—in the hands of a convenience station employee.35
Disposability
‘Convenient-for-the-People’ policing stations begs the question: convenient for who? By exam-
ining the material and political effects of convenience stations themselves, this chapter has
explored the intimate relationship between market-driven logics and governance infrastruc-
tures. But what ultimately is the goal of this marriage? This system is of course not only a prod-
uct of lessons learned from Japanese and Singaporean models of global capitalism, since the
political will to hire 10,000s of grid workers to work in the surveillance stations comes from a
socialist legacy of mass mobilization and Maoist neighborhood committees that policed coun-
terrevolutionary thought.36 But the goal of the current system is not to reform the thoughts of
Muslims by teaching them socialist critiques of class privilege. The goal now is to smooth out
space, making it consistent, predictable. Seamless. It is attempting to manage and, ultimately,
eliminate the inconvenience of Muslim bodies out of place. Muslim communities that remain
rooted in their native land and traditions need to be standardized, enclosed, reformed and
spent in service to the convenience of the settler population. To put this simply, Muslim lives
are made disposable.
In her framing of disposability, Neferti Tadiar highlights the way Latina, Chinese and Filipina
migrant women, and the settler colonized, such as Palestinians, are pushed into a spectrum
of ‘used up’ life forms as a feature of the contemporary global economy. The lives of these
inconvenient others are spent serving or waiting according to the convenience of protected
34 Byler, In the Camps.
35 For a parallel discussion of ‘race-as-algorithm’, see Sareeta Amrute, ‘Bored techies being casually
racist: race as algorithm’,Science, Technology, & Human Values45.5 (2020): 903-933.
36 Byler, Terror Capitalism.
45
IN/CONVENIENCE
citizens in such a way that the intrinsic value of their lives decline irreversibly over time.37
The Uyghurs and other Muslims who are made the object of the ‘Convenient-for-the-People’
system I have described in this chapter are also forced into such conditions of disposability.
Like Palestinians they are forced by a carceral settler state to wait for a life worth living that
will never come. Like guest worker populations they are forced to spend their socially repro-
ductive lifetimes to reproduce the convenience of lives that are not their own. That is to say,
the spent lives of colonized Others, like guest workers and other precarious or temporary
workers, is crucial to the global economy of convenience. In the Uyghur case, this spending
of life appears in fortified factories and plantations—where their time and labor are stolen
from them in service to the production of cheap fast fashion products for the domestic and
global market. In still other instances their unfree labor is spent in the reproduction of people’s
convenience as an urban infrastructure of feeling itself.
In a video obtained by the New York Times in 2019, Uyghur workers are shown wearing orange
sanitation uniforms in a locked dormitory.38 By day they are forced to sweep the streets of a
Chinese majority city in Northern Xinjiang, at night they are forced to study Chinese language.
One of the workers furtively interviewed in the video says that he is being paid a third of what
he made before he was swept up by the system. The compound where they have been for
over a year is guarded by a People’s Convenience Police Station. It is what stands between
them and their families. The system it represents is what monitors them during the day as they
sweep the streets, creating the convenient consistency that settlers around them desire. On
the wall of the dormitory a sign instructs the workers on how to maintain the convenient neu-
trality that the system is designed to produce when they are outside of the compound. It says:
Treat people with politeness, be neither servile nor overbearing.
When receiving help, say ‘Thank You’.
When being misunderstood, say ‘Sorry’.
As human infrastructure in the affective system of being ‘in-convenience’, Muslim workers it
appeared were to strive to be unnoticed, ‘neither servile nor overbearing’, when interacting
with ‘the People’. That is, they were to pretend that they had not been placed in servile posi-
tions. Instead, they were to take on an affect of on-demand customer service in deference to
the People’s Convenience brand. Their job was to reproduce the enjoyment of non-Muslim
convenient life.
When asked if he can ever return home to see his family, the Uyghur man interviewed in the
video responds, ‘return home? No, no, no’. In the refrain of the ‘no, no, no’ of Uyghur-accented
Mandarin there is a feeling that even thinking about this out loud might be read as a violation
37 Tadiar, Remaindered Life, p. 91.
38 Chris Buckley and Austin Ramzy, ‘Inside China’s Push to Turn Muslim Minorities Into an Army of
Workers’, New York Times, 30 December 2019, https://www.nytimes.com/2019/12/30/world/asia/
china-xinjiang-muslims-labor.html.
46 THEORY ON DEMAND
of the role he has been forced to play. The infrastructure of feeling produced by ‘convenience
for the people’ is everywhere. It appears that playing his role, cleaning the streets of a con-
venience-oriented platform city, working everyday to make up for his inconvenience, is the
only way the Uyghur man can spend his lifetime.
Over the 2010s the people’s convenience system transformed life in Northwest China. Ordi-
nary experience was reorganized. For Han people, buying bread became newly logistical. For
Uyghurs, this transformation placed them in service to the promise that Han convenience
would continue to grow. The streets would always be clean. The products would always be
cheap. The temporal and spatial sovereignty of Han citizens would always be enforced by the
People’s Convenience police stations. The convenience infrastructure of feeling routinized
the disposal of Uyghurs and legitimated Han citizenship. Even in China’s internal settler col-
ony—the site of a People’s War on Terror, life was once again reassuringly convenient. Being
‘in-convenience’ was guaranteed.
References
‘China: Big Data Fuels Crackdown in Minority Region’, Human Rights Watch, 26 February 2018,
https://www.hrw.org/news/2018/02/27/china-big-data-fuels-crackdown-minority-region.
‘Convenient Police Service in the Internet Age’, Modern World Police 3 (2016): 12-14. [现代世界警察
2016, (03),12-14互联网时代的便民警务 吴尚 初福善 王晓燕 顾庆忠 谢红 郭广生 陈川]
‘Xi urges anti-terrorism ‘nets’ for Xinjiang’, Xinhua News, May 29 2014, http://www.chinadaily.com.cn/
china/2014-05/29/content_17552457.htm
‘关于印发新疆维吾尔自治区’十四五’就业促进规划的通知 [Notice on Printing and Distributing the ‘14th Five-
Year’ Employment Promotion Plan of Xinjiang Uyghur Autonomous Region]’, Xinjiang Government, 14
December 2021, https://web.archive.org/web/20230324214327/https://archive.ph/PFzsb.
Amrute, Sareeta. ‘Bored techies being casually racist: race as algorithm’,Science, Technology, &
Human Values45.5 (2020): 903-933.
Berlant, Lauren.On the Inconvenience of Other People, Durham: Duke University Press, 2022.
Bratton,Benjamin H. The Stack: On Software and Sovereignty, Cambridge: The MIT Press, 2016.
Buckley, Chris and Ramzy, Austin. ‘Inside China’s Push to Turn Muslim Minorities Into an Army of
Workers’, New York Times, 30 December 2019, https://www.nytimes.com/2019/12/30/world/asia/
china-xinjiang-muslims-labor.html.
Byler,Darren. In the Camps: China’s High Tech Penal Colony, Columbia Global Reports, New York:
Columbia University Press, 2021.
Darren Byler, “Producing ‘Enemy Intelligence’: Information Infrastructure and the Smart City in North-
west China,” Information & Culture 57.2 (2022): 197-216.
Byler,Darren. Terror Capitalism: Uyghur Dispossession and Masculinity in a Chinese City, Durham:
Duke University Press, 2022.
Chinese Government Procurement Network, ‘Xinjiang Shawan County Smart (Safe) Project Fea-
sibility Study’, ChinaFile, 2017, https://www.chinafile.com/library/reports/xinjiang-shawan-coun-
ty-smart-safe-project-feasibility-study.
Dimitrov, Martin K. Dictatorship and Information: Authoritarian Regime Resilience in Communist
Europe and China, London: Oxford University Press, 2023.
47
IN/CONVENIENCE
Draper, Nora A. and Turow, Joseph. ‘The Corporate Cultivation of Digital Resignation’, New Media and
Society 21.8 (2019): 1824-1839.
Grauer, Yael. ‘Millions of Leaked Police Files Detail Suffocating Surveillance of China’s Uyghur Minori-
ty’, The Intercept, 29 January 2021, https://theintercept.com/2021/01/29/china-uyghur-muslim-sur-
veillance-police/.
Gupta,Akhil. Red Tape: Bureaucracy, Structural Violence, and Poverty in India, Durham: Duke Univer-
sity Press, 2012.
Hong, Jacky FL; Snell, Robin Stanley; and Easterby-Smith, Mark. ‘Cross-cultural influences on
organizational learning in MNCS: The case of Japanese companies in China’,Journal of International
Management12.4 (2006): 408-429.
Li, Yuhua. ‘Atm防护便民亭(Automatic teller machine (ATM) protective convenience-for-people booth)’,
PRC Patent Announcement Number CN 201891278 U, filed 13 December 2010, and issued 6 July
2011, https://patentimages.storage.googleapis.com/17/66/43/5a1adfff9b2071/CN201891278U.pdf.
Lu, Peng and Bi, Wenni. ‘Comparative study of smart city security governance: Taking Singapore and
Shanghai as examples’,Journal of Sinology17.1 (2023): 128-144.
Masco,Joseph. The Theater of Operations: National Security Affect from the Cold War to the War on
Terror, Durham: Duke University Press, 2014.
Massumi,Brian. Ontopower: War, Powers, and the State of Perception, Durham: Duke University Press,
2015.
Neves, Joshua and Steinberg, Marc. ‘In Convenience’, introduction to this book.
Pfaffenberger, Bryan. ‘Social anthropology of technology’,Annual review of Anthropology21.1 (1992):
491-516.
Roy, Ananya and Ong, Aihwa (eds). Worlding Cities:Asian Experiments and the Art of Being Global.
Studies in Urban and Social Change. Malden: Wiley-Blackwell, 2011.
Silva, Ricardo; Garcez, Ana; Gomes, Maria; Morais, Antònio; and Charrua, Fernando. ‘Active Learning
‘Factory of Boxes’ in the Teaching-Learning Processes in Engineering and Entrepreneurship’,Journal of
Technical Education and Training13.3 (2021): 1-14.
Steinberg, Marc. ‘From automobile capitalism to platform capitalism: Toyotism as a prehistory of
digital platforms’,Organization Studies43, no. 7 (2022): 1069-1090.
Steinberg,Marc. The Platform Economy: How Japan Transformed the Consumer Internet, Minneapolis:
University of Minnesota Press, 2019.
Tadiar,Neferti X.M. Remaindered Life, Durham: Duke University Press, 2022.
Wright,Melissa. Disposable Women and Other Myths of Global Capitalism, New York: Routledge, 2013.
Xinjiang Documentation Project, “Internal Police Files,” Unpublished Manuscript, 2024.
Xu, Vicky Xiuzhong; Leibold, James; and Impiombato, Daria. Architecture of Repression: Unpacking
Xinjiang's Governance, Australian Strategic Policy Institute, 2021, https://www.aspi.org.au/report/
architecture-repression.
Zhang, Xinde. ‘Wulumuqi shi dajie xiao xiang jiang jian 949 ge bianmin jingwuzhan [949 People’s Con-
venience Police Stations have been built in the streets of Ürümchi]’, Yaxin Net, 2016, https://kknews.
cc/society/2a4mrng.html.
48 THEORY ON DEMAND
THE RESILIENT SITUATION: ADAPTIVE
MANAGEMENT, FINANCE, AND ENVIRONMENT
ORIT HALPERN
Convenience appears to be a memory of the past. Climate crisis, geo-political instability, eco-
nomic volatility, and technical change all make the ideals of labor, lifestyle, and comfort that
originally defined the term ‘convenience’ seemingly no longer achievable. According to the
admittedly limited google n-gram, convenience has fluctuated, but in the United States the
1950s appeared to herald the most recent height of its use. Convenience store interestingly
peaks in 2002. The relationship thus with convenience as an aspiration, and convenience
as a logistical reality are apparently separated; but both appear closely linked to economy
and technology. The decline of convenience stores in n-gram results might be understood
as related to the rise of the internet and other forms of consumption.
The original rise of convenience in the post-war period occurred within a context of rapid
suburbanization, the rise of a popular consumer society. By deduction, one might assume that
convenience would be tied to American imaginaries of nuclear families, unlimited economic
growth, prosperity, and technological advancement. The convenience of consumption and the
ideals of comfort in lifestyle reflected and refracted social values concerning accessible con-
sumption, the demand for new infrastructures of mobility, and the clear spatial organization of
gendered labor, and racial apartheid, that is often related to the ideals of Cold War America.1
Today, however, we appear to be in a new situation. One that is replacing the ideals and logics
of Keynesian Cold War America and even the globalized consumer with a new logic. Terms
such as ‘new normal’ and even ‘next normal’ are appearing in the discourses of major poli-
cy making, corporate, and governmental institutions.2 In turn, policy makers, psychologists,
ecologists, and business strategists all urge making systems, institutions, and even human
subjects, resilient; capable of enduring and even profiting from change and crisis. A new
expectation has emerged that disruption and volatility are to be expected. Moreover, this new
situation appears to have an immanent relationship to smartness, AI, and digital technologies.
Ever smarter and data-driven systems are imagined to be the technical managers of these
new levels and scales (temporal, geological, spatial) of volatility.3
Whereas convenience implies comfort, stability, even stasis, in its most popular uses, resil-
ience thus defines radically different ideas of life, economy, technology, and futurity. Emerging
from both recognition of anthropogenic changes to the environment and the concept of the
information economy, resilience, I argue, may be the dominant logic of contemporary digital
1 Melinda Cooper, Family Values: Between Neoliberalism and the New Social Conservatism, New York:
Zone Books, 2017.
2 Aongus Hegarty, ‘Digital Resilience: Building the Economies of Tomorrow on a Foundation of
Cybersecurity’, World Economic Forum, 20 May 2022, https://www.weforum.org /agenda/2022/05/
digital-resilience-building-the-economies-of-tomorrow-on-a-foundation-of-cybersecurity/.
3 Orit Halpern and Robert Mitchell, The Smartness Mandate, Cambridge: The MIT Press, 2023.
49
IN/CONVENIENCE
media systems. This article posits that many of us (of course not all and not everywhere) live
in a resilient condition.
Resilience and convenience have, however, critical genealogical interactions. Resilience is in
many ways a logic arriving from the infrastructures and attention economies first produced
for the fantasy of globalization and convenience. Marc Steinberg and Joshua Neves argue
that convenience is about the management of time and space. They map convenience to
certain infrastructures of both logistics and attention (immediate gratification, on-demand
fulfillment, and immediacy) seemingly necessary for economic survival and impossible to
contest. Perhaps, one might deem this convenient condition ‘natural’.4
This naturalization involves also the obfuscation of the extreme violence of the logistical labor
arrangements necessary for delivering so many products and streaming so much content ‘just
in time’.5 The strained and overworked warehouse employees, the toxic labor conditions of
mineral and energy extraction, the securitization of borders and migration, and the precarity
of gig-laborers all exemplify the costs of ‘convenience’.6 Understanding what might be at
stake in this emerging reformulation of convenience into resilience is important for thinking
how we might intervene in, but also re-imagine, our relations through and to contemporary
digital media infrastructures.
The Resilient Situation
For example, at the last Davos Conference held in May 2022, the World Economic Forum
(WEF) announced that resilience would become one of the central concepts organizing global
response to the war in the Ukraine, climate change, and the aftereffects of the COVID pan-
demic. The Forum identified a series of central challenges or ‘frontier risks’ that accompanied
these events. Rapid technological change, climate disturbances, economic volatility and
disparity, and balkanization rose to the top. These risks, if not managed correctly or even
capitalized upon, would pose dangers to globalization and plural democracy. Institutions
therefore need to become ‘resilient’. As the President of International Markets at Dell Tech-
nologies, Aongus Hegarty put it at the Davos conference in 2022, we need ‘digital resilience’.
This digital resilience emerges from the unprecedented acceleration of ‘innovation’ in the
sector from the COVID-19 pandemic, and ‘positions an enterprise to pivot fast, adapt to
fluid conditions, maintain seamless business continuity, and capitalize on opportunities’.7
Such agility and fluidity will be accomplished, Hegarty argued, by centering cybersecurity
accompanied with ubiquitous computing (we may presume that this is a central concern for
the Dell corporation).
4 Joshua Neves and Marc Steinberg, ‘In Convenience’, in this volume.
5 Emily West, Buy Now: How Amazon Branded Convenience and Normalized Monopoly, Cambridge: The
MIT Press, 2022.
6 Deborah Cowen, The Deadly Life of Logistics: Mapping Violence in Global Trade, Minneapolis: University
of Minnesota Press, 2014; Ned Rossiter, Software, Infrastructure, Labor: A Media Theory of Logistical
Nightmares, New York: Routledge, 2016.
7 Hegarty, ‘Digital Resilience’.
50 THEORY ON DEMAND
While resilience might commonly be thought of as a psychological or environmental attri-
bute, the WEF highlights the new centrality of this term to organizing political economy. As
discourses of resilience imply, the idea that we now live in a world of both constant technical
innovation and regular trauma appears natural. What is at stake in this discussion is funda-
mentally how we understand social, natural, and technical change and how corporations,
governments, and other institutions plan and respond to change. While the management
of supply chains, finance, and natural sciences are not always thought together, this essay
intends to preliminarily map out notions of this ‘new normal’ and the changing models of both
nature and economy that underpin it.
Replacing notions of comfort, convenience, and control, ‘resilience’ has become a valued
quality in both supply chains and human psychologies. It might then appear un-intuitive
that this article is in an issue dedicated to ‘in/convenience’. Hardly. This new condition is
not autonomous from the history of convenience and logistics; in fact it is grounded in the
infrastructures and even cannibalization of earlier ideals of immediate access to consumer
items, and speed of global delivery of goods. In fact, as I will show, most of the ideas and
management tactics that are now labelled ‘adaptive’ or ‘resilient’ initially emerged under
the conditions that Neves and Steinberg articulate as critical to convenience. However, it
has critical divergences as well that I would like to put forward for political reasons. I insist on
the separation from the language of convenience and even in/convenience because those
conditions still maintain fantasies of comfort as an aspiration and ideal, and still focus on
individual consumption as the key site of governmentality.
Resilience, on the other hand, is a language resolutely entrenched in biological concepts
of species survival, not in individual and liberal ideals of choice and consumerism. This is
a central transformation that is embedded in new geo-political realities where the smooth
imagined space of globalization and its supply chains has been transformed into new ideas
of planetarity, fragmentation, borders, and constant precarity for many people and more-
than-human forms of life.
Resilience, I will argue, consumes the infrastructures of convenience, but now reformulates
the systems dedicated to ease, leisure and comfort in terms of survival. There are both pos-
itive and negative features to this move. Fundamentally, discourses of resilience recognize
the imbrication of human systems with earth systems (therefore the switch from globaliza-
tion to planetarity). Resilience comes attached to new forms of ‘adaptive management’ that
articulate and co-produce geopolitical and environmental bordering and disruptions. Finally,
resilience has different ideals of future making, and new reactionary political-economic imag-
inaries of a lost ‘stable’ past. Yet resilience is not a discourse only adopted by corporations
and management discourses, it also appears in other places. For example, in the concept of
‘after comfort’ in architecture, articulating a demand to construct housing beyond mechanized
forms of climate control; or in re-imagining spatial relations by engaging the impossibility of
eliminating toxicity in our environment and assuming that we need new relations to atmo-
sphere and environment that accept the impossibility of “purity” and demand learning to live,
51
IN/CONVENIENCE
or perhaps one progressive form of resilience for existence.8 The Black Lives Matter organiza-
tion too adopts resilience in its mission statements, implying the ongoing violence of racism,
as well as the ability to withstand and overcome such structural and infrastructural politics.9
We witnessed the supersession of resilience discourses throughout the COVID pandemic,
when convenience was recast as necessity. Disease, but also increased stresses and tempo-
ralities of labor, were re-articulated as demanding at-home delivery and streaming services.
It was no longer convenience, it was survival. People could not go to stores, theaters, or
restaurants; yet this was not framed as a question of merely choice, but an environmental and
economic necessity. Health risks, time constraints, excesses of workload, and transformations
in the work environment all demanded home delivery and online services. Such survivalist and
biological, even racial arguments can of course also be negative, re-affirming and solidifying
essentialist and bio-deterministic conceptions of identity, politics, and society. 10 This is a
survivalism affirmed by the fact that resilience, as I will show, has a history related to ecology,
modelling extinction and population change over time, and concepts of how to model nature.
Becoming Resilient
Concepts of resilience emerged precisely at the moment when Fordist production and Keynes-
ian economics were being supplanted by new attitudes to both nature and economy. Resil-
ience emerged first as an idea in ecology mirroring the very rise of the convenience economy
and platforms of just-in-time production and globalization that has been documented by
many authors.11 In 1973, the Canadian ecologist C.S. Holling introduced this new concept of
resilience to the discourse on nature, evolution, and extinction:
Individuals die, populations disappear, and species become extinct. That is one view
of the world. But another view of the world concentrates not so much on presence or
absence as upon the numbers of organisms and the degree of constancy of their num-
bers. These are two very dierent ways of viewing the behavior of systems and the use-
fulness of the view depends very much on the properties of the system concerned.12
Holling posits a world where change, even catastrophic change, is the norm and heralds
not the end of systems but evolution. Extinctions happen but systems, ‘degrees’ and evolu-
8 Daniel A. Barber, Modern Architecture and Climate: Design before Air Conditioning (Princeton:
Princeton University Press, 2020); Daniel A. Barber et al., ‘Editorial’, E-flux Architecture: ‘After Comfort:
A User’s Guide’ (October 2023), https://www.e-flux.com/architecture/after-comfort/568230/editorial/;
Nerea Calvillo, Aeropolis: Queering Air in Toxicpolluted Worlds, New York: Columbia University Press,
2023.
9 Black Lives Matter, ‘About’, https://blacklivesmatter.com/about/.
10 Lydia Polgreen, ‘Restoring the Past Won’t Liberate Palestine’, The New York Times, 18 February 2024,
https://www.nytimes.com/2024/02/18/opinion/israel-gaza-palestine-decolonization.html.
11 Marc Steinberg, The Platform Economy: How Japan Transformed the Consumer Internet, Minneapolis:
University of Minnesota Press, 2019; Anna Tsing, Friction: An Ethnography of Global Connection,
Princeton: Princeton University Press, 2004.
12 C.S. Holling, ‘Resilience and Stability of Ecological Systems’, Annual Review of Ecological Systems 4
(1973): 1.
52 THEORY ON DEMAND
tion continue. Rather than focusing on the event of extinction, or the numbers of animals or
humans in an ecosystem, Holling argued that ecologists should instead think about the rela-
tionships in a system. The environment itself had to become a system which had ‘properties’
that, like programs, could be maintained irrespective of the life or death of individuals. This
new concept of resilience thus posited a new idea of both change and event.
Ecology and economy have of course long been linked both in etiology and in ideology. State-
ments emerging from ecology were anticipated in economics. In 1971 Chicago School neo-
liberal economist Milton Friedman made a seemingly similar pronouncement, this time in
relationship to currency markets. He announced a ‘major need for a broad, widely based,
active, and resilient futures market’.13 Counter to standard understandings of the economy
at the time he projected a positive valence for active, volatile markets. For Friedman, the col-
lapse of the Bretton Woods system was not a calamity, but an opportunity for creating a new,
what he labelled ‘resilient’ system of international currency exchange. In an article titled ‘The
Need for Futures Markets in Currencies’, Friedman acknowledged that, in the absence of an
international system of currency controls, exchange rates would shift constantly in relation-
ship to one another. The architects of Bretton Woods had seen such volatility as a problem,
since it meant that those engaged in foreign trade would have to take significant risks that the
currency in which a trade was negotiated would depreciate by the time payments were to be
made. Bretton Woods thus sought to institute a ‘system of rigidly fixed [exchange] rates that
do not change’. However, as Friedman noted, they ended up with a ‘system of rigidly fixed
rates subject to large jumps from time to time’, and these large jumps eventually broke what
was designed to be a rigid system of control.14 Friedman argued that the solution could not be
another rigid centrally-controlled system, but instead a resilient futures market for currencies:
that is a system that might allow those engaged in foreign trade to hedge the risks associated
with currency exchange changes.
For Friedman, ‘resilience’ was to be understood as the opposite of ‘rigidity’, and would mean,
in practice, something like the oxymoronic notion of ‘stable change’. More specifically, curren-
13 Milton Friedman, ‘The Need for Futures Markets in Currencies’, Cato Journal 31.3 (2011): 637.
Friedman stressed that this market ‘cannot depend solely on hedging transactions by persons involved
in foreign trade and investment’; in addition, the ‘market needs speculators who are willing to take
open positions as well as hedges. The larger the volume of speculative activity, the better the market
and the easier it will be for persons involved in foreign trade and investment to hedge at low costs and
at market prices that move only gradually and are not significantly affected by even large commercial
transactions’ (638). The terminology of ‘resilience’ seems not to have been Friedman’s innovation, as
other economists had also used this term in the late 1960s when discussing the need for Bretton Woods
reform.
14 Friedman, ‘The Need for Futures Markets in Currencies’, 636. This article is a reprint of a December 20,
1971 report to the Chicago Mercantile Exchange, and as Donald MacKenzie notes, this was an ‘article
for hire’, as Friedman was paid by the head of the Chicago Mercantile Exchange to write the article,
which was intended to (and did) pave the way for federal approval of precisely such a market. Hence,
the article advocates not only for a futures currency market, but for its location in the United States.
Friedman contended, ‘[a]s Britain has demonstrated in the nineteenth century, financial services of
all kinds can be a highly profitable export commodity’, and proposed that a U.S.-based futures market
would strengthen the American position while also maintaining the stability and expansion of global
trade.
53
IN/CONVENIENCE
cy markets would change in response to global events, but nevertheless continue to protect
international trade, the international global political order of the West, and the primacy of the
United States within that order.
These new ideas of nature came then within a context where older models of political economy
were also in flux. The end of Bretton Woods, decolonization, post-Fordism, and the OPEC oil
crisis, to name a few of the transformations at the time, induced extreme volatility in politics,
currency, and commodity markets. New financial technologies and institutions, such as deriv-
ative pricing equations and hedge funds, emerged in order to hedge bets. These technologies
literally produced ways to insure that risks were reallocated, decentralized and networked.
Dangerous bets would be combined with safer ones and dispersed across multiple territories
and temporalities (consider short bets, credit swaps, and futures markets). Corporations,
governments, and financiers flocked to these techniques of uncertainty management in the
face of unnamable, and unquantifiable, risks.15 Epistemologically ecology and finance would
then come to share a model of a world of ceaseless volatility and uncertainty.
Cybernetic Systems
Volatility and uncertainty were not always considered the norms of nature. Since the Second
World War, cybernetically informed ecologists had built models that understood the world in
terms of homeostatically organized networked systems.
Fig. 1: Schema of biogeochemical processes from G. Evelyn Hutchinson, ‘Circular Causal Systems in Ecology’,
Annals of the New York Academy of Science 50 (1948): 223.
15 It is worth noting that the Black Scholes Derivative pricing equation inaugurating the financialization
of the global economy was introduced in 1973. For an excellent summary of these links and of the
insurance and urban planning fields, please see Kevin Grove, Resilience, New York: Routledge, 2018.
54 THEORY ON DEMAND
Initial models grounded in communication sciences and tested on the landscape of nuclear
blast sites, valorized stability. Ecosystems were supposed to be made of feedback loops that
aspired to balance, much like the early models of a homeostat coming from the sciences of
communication and control.
Fig.2: The World Model, from Donella H. Meadows et al., The Limits to Growth, New York: Universe
Books, 1972, p. 102.
Imbalance was to be avoided, and systems would have to be managed for stability. The most
extensive efforts at computing the future of the planet and its populations, The Limits to Growth
report of 1972, modelled, to cite Paul Edwards, such a ‘closed’ world with limited resources
that had to be kept in balance.16 The clarion call to an emergent environmental movement,
this computerized report saw a world in need of balance, one where change was an anomaly
not a norm. Computer scientists modelled human behavior and populations as aberrations
producing terminal traumas on the environment that would lead to catastrophe. The answer
was to restore the balance of the planet through the careful management of feedback loops
and return it to a sustainable state.
Such notions of cybernetically informed balanced systems were also readily found in business
manuals and management approaches. While this article is too short to outline the centrality
of ideals of operations research, cybernetics, and game theory on economics and logistics, a
wealth of research has done so.17
16 Paul N. Edwards, The Closed World: Computers and the Politics of Discourse in Cold War America,
Cambridge: The MIT Press, 1997.
17 Cowen, The Deadly Life of Logistics; Jesse LeCavalier, The Rule of Logistics: Walmart and the
55
IN/CONVENIENCE
Volatility and Adaptation
But many ecologists, environmentalists, and economists did not agree with the report. Eco-
systems, they argued, did not appear to stabilize after suffering disruption. There could be
no going back historically to a less “damaged” planet. DDT had demonstrated destructive
results impacting systems far outside the immediate locus of intended insect elimination in
agriculture and for purposes of public health. Agent Orange, heavily used in the Vietnam War
as a defoliant, and related dioxins were demonstrated to produce long ranging impacts on
humans and ecosystems. And the list goes on. Just ceasing the use of a toxin or attempting
to reseed an environment did not return systems to their pasts. Even seemingly environmen-
tally friendly actions, such as lowering fishing quotas or replanting trees would be found to
return little result once certain levels of disruption to the ecosystem were surpassed.18 Nature
appeared to constantly be evolving.
Fig 3: Future Population Projections from C.S. Holling, ‘Resilience and Stability in Ecological Systems’,
Annual Review of Ecological Systems 4 (1973): 10.
Architecture of Fulfillment, Minneapolis: University of Minnesota Press, 2016; Steinberg, The Platform
Economy.
18 ‘Agent Orange’, History.Com, 2 August 2011, updated 16 May 2019, https://www.history.com/topics/
vietnam-war/agent-orange-1; Jacob Darwin Hamblin, Arming Mother Nature: The Birth of Catastrophic
Environmentalism, Oxford: Oxford University Press, 2013.
56 THEORY ON DEMAND
Resilience
In response, a new discourse began to emerge in ecology—resilience. Countering the dis-
course of Limits to Growth, C.S. Holling developed the concept of resilience to contest the
premise that ecosystems were most healthy when they returned quickly to an equilibrium
state after being disturbed. His argument, first cited at the beginning of this article, was that
over-emphasis on predator-prey relationships often ignored more complex interactions, and
over-valued equilibrium. Nitrogen, carbon, and other cycles, interactions of mutual aid, col-
laboration, or competition between many species not structured as predator-prey relations,
and myriad other such factors might permit ecosystems to persevere in their functions even
if in mutated or varied forms. Extinction might not be the limit to the growth or change of a
system, unless it fundamentally transforms a complex web of interactions that sustains life.
The seeming absolute limit to life—extinction—could therefore be extended by factoring in
complexity and a new value for biodiversity.
Fig 4: Topological models generated from historical data since 1951 of budworm population densities in
space. It is also worth noting that these new forms of dynamic maps and capacities to compare data sets
came with the introduction of digital computation and new platforms such as the Canadian Geographic
System (CGIS) considered the root of contemporary GIS systems in the early 1970s.19
19 C.S. Holling (ed) Adaptive Environmental Assessment and Management, New York: John Wiley and Sons,
1978, pp. 164-165.
57
IN/CONVENIENCE
If sustainability was the language of stable systems in a cyclical economy, resilience is the
language of volatility. In an early critique of industrial fishery and forestry management, Holling
argued that the focus on using insecticides, re-seeding lakes with fish, or attempting to simply
replant one type of tree would not work over extended periods of time. Managing ecosystems
with a focus on stability was an error. Managers, he suggested, had to cease counting and
taxonomically placing populations in boxes and flow charts, and needed to realize that positive
feedback is dynamic and produces change. Populations are not static numbers but ongoing
processes. The important thing is to maintain the process, not the steady state of the system.
For example, in the case of the boreal forest the absolute number of spruces is not important,
what is important is the ability of the forest to rejuvenate and continue growing trees, which
depends on fluctuating numbers of populations and constant variations between spruce,
fir, birch and budworms. The system regularly changes. In general, this allows the forest as
a forest to continue existing. Better ecological management might also apprehend the fact
that systems ultimately change. For example, since forests in Ontario are increasingly used
for leisure and vacationing, then forestry management must change accordingly. For other
systems, one might identify different processes defining them. Today we deploy the term
‘ecosystem services’ to describe this process of identifying and managing processes rather
than discrete numbers.
Resilience, by contrast with sustainability, denoted for Holling the capacity of a system itself
to change in periods of intense external perturbation, as a mode of persistence. The concept
of resilience enabled a management approach to ecosystems that ‘would emphasize the
need to keep options open, the need to view events in a regional rather than a local context,
and the need to emphasize heterogeneity’.20 Managers had to create multiple strategies for
future actions, think ‘regionally’ which is to say in terms of networks and connections across
different territories and times, and emphasize heterogeneity or biodiversity in order to secure
more possible routes for adaptation in case of unanticipated shocks. Holling would later label
this form of management ‘adaptive management’, arguing that it necessitated the constant
feedback of data to respond to constant changes.21
Holling also underscored that the movement from valuing stability to valuing resilience
depended upon an epistemological shift: ‘Flowing from this would be not the presumption
of sufficient knowledge, but the recognition of our ignorance: not the assumption that future
events are expected, but that they will be unexpected’.22 In short, expect the unexpected. Plan
for extreme events without any conception of absolute prediction.
There are three summary points I want to underscore. The first is that resilience within this
genealogy assumed uncertainty and volatility as common, perhaps even ‘normal’, conditions.
Stability and resilience are not correlated. As a corollary, the life and death of individuals or
20 Holling, ‘Resilience and Stability of Ecological Systems’.
21 For a summary of strategies in adaptive management see: C.S. Holling (ed), Adaptive Environmental
Assessment and Management.
22 Holling, ‘Resilience and Stability of Ecological Systems’.
58 THEORY ON DEMAND
even populations became secondary to the ongoing evolution of systems. Second, resilience
was a new way to model systems and therefore measure them. Instead of taxonomizing and
organizing populations into stable categories, one must define systems in terms of processes,
and measure the relationships between populations and potentially other factors (nitrates,
carbon, energy, etc.). A corollary of this new approach is that past data can be used to build
concepts but can never actually predict the future. Probabilities have to intervene. Finally,
ecologists emphasized ‘heterogeneity’ and diversity as important to facilitating resilience.
Systems without a surplus of functions and populations could not adapt. Perfectly optimized
systems would collapse when change happened.
Resilience thus possesses some curious features. On one hand, the focus on processes and
what are today labelled ‘ecosystem services’ means that some lives and populations are
acceptably sacrificed as long as the system continues to operate; and trauma is a regularized
and normalized event. On the other hand, environmental managers recognize that only sys-
tems with robust diversity, redundancy, and supplemental capacities might survive abrupt
and catastrophic events. Resilience fluctuates between the two poles of Darwinian evolution-
ary theory—survival of the fittest, and the need for variety and diversity within and between
populations to allow for adaptability. Perfect optimization might come at the cost of adaptation.
Resilient Speculation
Markets have also long been modelled on ideas of nature, adaptation, fitness, and evolution.
In his 1974 Nobel Memorial Prize in Economic Science speech, the economist Friedrich
Hayek disparaged The Limits of Growth report as part of a more general plea, addressed to
both mainstream economists and their leftist critics, for a more modest epistemology that
would give up on the dream of complete control over the future. Hayek noted drily that the
recent creation of the Nobel Memorial Prize in Economic Science was itself testimony to the
‘propensity [of economists] to imitate as closely as possible the procedures of the brilliantly
successful physical sciences’, but stressed that, in economics, this often ‘led to outright error’.
Hayek stressed that economies were not equivalent to the isolated systems of physics. This
was in part because a social science such as economics focused on the behavior of large
populations of different agents, with the result that:
like much of biology but unlike most fields of the physical sciences, [economics has]
to deal with structures of essential complexity, i.e. with structures whose character-
istic properties can be exhibited only by models made up of relatively large numbers
of variables. Competition, for instance, is a process which will produce certain results
only if it proceeds among a fairly large number of acting persons.23
Rather than pretending to be able to replicate the kinds of discoveries about the natural world
available to physicists, economists should instead accept a biology-like world of uncertainty,
23 Friedrich August von Hayek, ‘The Pretence of Knowledge’, The Nobel Prize, Lecture delivered for The
Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, 11 December 1974, www.
nobelprize.org/prizes/economic-sciences/1974/hayek/lecture/.
59
IN/CONVENIENCE
chance, and large populations of different individuals. This would in turn mean relinquish-
ing the goal of planning and turning instead to the more modest goal of managing. For
Hayek, societies emerge from decentralized networks of information coordinated through
markets, which meant that seeking to plan or regulate the economy—by, for example,
limiting or eliminating growth—could only end in disaster.
Hayek suggested that mainstream economists, by seeking to emulate the physical sci-
ences, had in fact given encouragement to precisely that fantasy of control that he saw
as central to The Limits of Growth. He suggested that:
It is often dicult enough for the expert, and certainly in many instances impossible
for the layman, to distinguish between legitimate and illegitimate claims advanced
in the name of science. The enormous publicity recently given by the media to a
report pronouncing in the name of science on The Limits to Growth, and the silence
of the same media about the devastating criticism this report has received from the
competent experts, must make one feel somewhat apprehensive about the use to
which the prestige of science can be put. But it is by no means only in the field of
economics that far-reaching claims are made on behalf of a more scientific direction
of all human activities and the desirability of replacing spontaneous processes by
“conscious human control.”24
For Hayek, systems self-organize from the ‘free efforts of millions of individuals’, and not
the conscious decision-making power of the few. As a consequence, control—understood
as the prediction of future events, whether by mainstream economists or the Club of
Rome—was impossible. For Hayek, though, this was not cause for despair. Rather, it was
grounds for hope, provided that those populations of millions were allowed to engage new
and unanticipated problems flexibly by means of unrestricted market activity. Hayek’s
speech reflects the epistemic ‘modesty’ or ignorance that resilience managers also
espoused. Uncertainty was the only certainty in this worldview.
This reminds us of Friedman’s observations at the start of this essay. Though Friedman
was one of those economists chastised by Hayek in his lecture as overly committed to
‘scientific’ models of economics, Friedman’s proposal for a resilient futures markets nev-
ertheless exemplified Hayek’s image of markets that flexibly managed, rather than rigidly
controlled or planned, an always uncertain future. These conceptions of not only managing,
but actually arbitraging uncertainty would find actualization in technology. While risk has
always perhaps been necessary for profit, never before had it been so clearly demarcated
as a site of technological innovation and intervention.
Hayek’s lecture focused primarily on the rather abstract realm of epistemology, and pro-
vided relatively little guidance as to what this approach might look like in practice. How-
ever, in the 1970s, several economists and ecologists turned to concepts of flexibility and
‘resilience’ to explain how the epistemological modesty valorized by Hayek could generate
24 Hayek, ‘The Pretence of Knowledge’.
60 THEORY ON DEMAND
solutions to specific new and unanticipated problems while at the same time avoiding
system collapse.
At the center of these approaches was the introduction of algorithmic trading and deriva-
tive instruments to the market. Such technologies might permit the ‘hedging’ of risks that
were increasingly difficult to calculate with certainty, while also permitting management of
volatile currency changes resulting from the end of Bretton Woods and geo-political con-
flicts impacting energy and commodity markets. The computer scientist turned financial
guru, Fischer Black, one of the creators of the automated derivatives market, wrote an
important essay on noise that summarizes this new resilient view of market technologies.
At the center of his new vision of options markets and futures was the idea of entropy and
noise borrowed from cybernetics:25
The eects of noise on the world, and on our views of the world, are profound. Noise
in the sense of a large number of small events is often a causal factor much more
powerful than a small number of large events can be. Noise makes trading in financial
markets possible, and thus allows us to observe prices for financial assets.26
Fischer Black’s famous article ‘Noise Trading’ formalized a new discourse in finance and
posited that we trade and profit from misinformation and information overload. In this new
embrace of automated financial trading, what no longer existed was the problem of equilib-
rium or a concern for entropic disorganization. If 19th and earlier 20th century economists,
even Hayek, worried about the maintenance of the market itself, and of the stability of value,
i.e. about entropy and the tendency of systems (whether political or economic) to degrade,
now that concern had been deferred, and even capitalized upon. Noise in communication
theory is directly correlated with increases in entropy. Options trading makes volatility and
speculation—an excess of information in the market—a site of extracting value. Arbitrage.27
The significance of this turn towards embracing entropy and noise as a site of value cannot
be overstated. Black embraced the concept, central to resilience, that markets are always
unstable and volatile. Moreover, he recognized that, as a result of this seemingly natural
condition, full prediction is impossible, and therefore new technologies of preemption are
necessary. Hedge funds, and their central technology of derivative pricing, thereby became
key vehicles to monetize this uncertainty and manage the operations of the market while
enduring constant evolutionary stresses.
The question ecologists and economists turned to asking, then, was: if prediction of the future
was impossible, how were the models of ecology failing? More importantly, how can these
seemingly un-anticipatable events be dealt with? How does one manage for radical uncer-
25 Fischer Black studied at MIT and finished a degree at Harvard, initially under the guidance of Marvin
Minsky, and read cybernetic texts by Norbert Wiener throughout his high school and early college
education. See Perry Mehrling, Fischer Black and the Revolutionary Idea of Finance, New York: John
Wiley and Sons, 2005, p. 30.
26 Fischer Black, ‘Noise’, The Journal of Finance 41.3 (1986): 529.
27 Mehrling, Fischer Black and the Revolutionary Idea of Finance.
61
IN/CONVENIENCE
tainty and change? The response was to create a new class of technologies including adaptive
management techniques, increasingly data driven and automated management of supply
chains, and computerized systems for logistics, to regularly command and manage this future
without ever having to predict it. These instruments were the automated analogues to resilient
management strategies that reflected similar attitudes to uncertainty and the management
of systems in both economy and ecology.
It is very important to note that financial instruments like derivatives are the computational
mirrors to the general transformation at the time in logistics. As scholars have noted, in the
1960s falling corporate profit rates and the rise of total cost analysis increased focus on dis-
tribution decisions and supply chain management as sites of value production to offset falling
returns. Business management became increasingly professionalized. The MBA degree and
the computer simultaneously arose, and managers could now handle processes previously
viewed as separate—purchasing, manufacturing, transportation, warehousing, returns, and
so forth—as part of one system. These management techniques borrowed from the same
systems and cybernetic sciences underpinning ecology and finance.28
These models offered a concept of systems as capable of purposeful evolution without direc-
tion, and created a language by which to imagine systems whose capacity for change and
adaptation would come through internal mechanisms of feedback and reflexivity rather than
political oversight and the state.29 It is no accident, of course, that the concept of eternal-
ly evolving and unpredictable, and therefore unplannable, systems emerged directly as a
response to both demands for civil rights by disenfranchised and racialized groups and global
decolonization. Derivative pricing markets naturalized and automated crisis while assuming
that planning was counter-evolutionary and forestalled adaptation, and by extension, survival.
In fact, both the transformation in logistics and management and financialization were co-pro-
duced to manage geo-political crisis. The Black-Scholes and other financial instruments
emerged from the end of Bretton Woods and monetization, and became central techniques
for managing events such as the OPEC oil crisis, along with other decolonial events that
resulted in commodity price fluctuation that demanded new ways of distributing risk glob-
ally. While the space to elaborate here is not possible, the key take away, is that resilience
was co-produced with financial and logistic risk management technologies in the 1970’s
and 1980’s, within a particular context of global transformations in political-economy, race
relations, and governmental orders.30
28 Charmaine Chua et al., ‘Introduction: Turbulent Circulation: Building a Critical Engagement with
Logistics’, Environment and Planning D: Society and Space 36.4 (2018): 617-629.
29 Paul Lewis, ‘The Emergence of “Emergence” in the Work of F.A. Hayek: A Historical Analysis’, History of
Political Economy 48.1 (2016): 111-150.
30 Ryan C. Smith, The Real Oil Shock: How Oil Transformed Money, Debt, and Finance, Cham: Palgrave
Macmillan, 2022, pp. 73-76.
62 THEORY ON DEMAND
Conclusion
By the early 2000s following 9/11, the 2008 financial crisis, and climate change, resilience
has taken a central discursive place in fields ranging from business management and logistics
to psychology. ‘Adaptive management’, ‘business continuity management’, and ‘climate
resiliency planning’ and many related terms are all the direct outgrowths from ecological
resilience and largely shape our understanding of how changing climactic and security con-
ditions are to be dealt with.
A search online for resilience in the aftermath of COVID-19, and in the wake of the war in the
Ukraine, reveals a massive number of articles, websites, and consulting services dedicated
to logistics, psychology, and community activism. For managers of supply chains and cor-
porations such as SAP and IBM, resilience is what corporations must do to ensure business
continuity. ‘Just in case’ has become the new mantra, and corporations are urged to increase
their options, to diversify supply chains geographically, and begin thinking about plasticity in
manufacturing infrastructure (being able to make alternative products), and to identify vital
services and processes ahead of time. For many neo-liberal and right-leaning politicians,
resilience is a call to expend populations they do not value—the elderly, people with under-
lying health conditions, people of color—through the ongoing annihilation of environmental
protections, civil rights, and social benefits in the name of saving the economy. Resilience
thus becomes a mode of naturalizing violence for the Right.
Here we must contend with how we understand evolution and genealogy. Financial and
logistical comprehensions of resilience largely assume a world of scenario plans and un-an-
ticipatable futures divorced from historical legacy or context. Resilience suggests that the
contemporary derivatives economy is one of infinite calculation without termination. There is
no final prediction, just constant adjustments, grounded in the calculation of differences. The
racism and injustice of such an economy, we might extrapolate, emerges from the fact that,
as technologies, derivatives but also data driven supply chains consume differences, whether
via correlations between the value of homes in different places, the comparative poverty or
wealth of different populations, the differing cost of labor in different locales, the differentials
in resource landscapes, or differences in the speed by which between investors can buy and
sell options. These differentials become the site of a new form of automated and algorithmic
speculation. Betting on differences in this way has the effect of making the future homogenous
with the present, as it perpetuates contemporary class, racial, and sexual inequities.
More critically, resilience re-invents or returns biological and survivalist discourses. The focus
increasingly is about management of populations. Populations as a resource for machine
learning, AI, and smartification, but also sometimes, as a return to older ideals of nation and
race, as embodied in the alt-right, and in contemporary re-bordering, immigration policy, and
re-nationalization of resources and production.
But there are other options. In turn, some activist movements are also turning to finance as
the centerpiece of environmental activism. Greenpeace has recently refocused its efforts on
financial entities, targeting groups such as BlackRock (the largest investment manager in the
63
IN/CONVENIENCE
world) and the financial systems behind oil pipelines, by shifting protests to the headquarters
of the financial firm, rather than focusing efforts solely at the pipeline.31 While in retrospect
this may seem like common sense, in fact, until recently, the financialization of the energy
sector was not a focus of activism in the way that the actual corporate perpetrators of envi-
ronmental damage were.
Similar evidence can be found in bond markets, where financial instruments can serve as
sources of political action. For example, developing climate preparedness mechanisms for
large American cities is very expensive. While finance is usually depicted in urban history as
an attempt by neoliberalism to bankrupt the public, numerous recent examples suggest other
possibilities. In Houston’s reconstruction after Hurricane Harvey, for example, a unique bill
structuring the bond instruments was passed in 2018. This plan used the money from bonds
to assist poor and minority neighborhoods that were at highest risk for flooding. The bond
was structured so that different forms of risk were assessed differently, rather than merely
evaluating risk through property value and property loss, which would have benefited rich
neighborhoods. While there are many ongoing battles, its success is not yet clear, and Federal
funding was slowed under President Donald Trump, the bond structure and financing became
a site of political action.32 Such efforts demonstrate that new models of what is urban and what
is natural extend the terrain of action. No longer only focused on hard infrastructure such as
roads or sea walls, increasingly urban activists understand many ‘soft infrastructures’ such
as wetlands as part of climate defense. Environmental justice (as in the case of Houston)
becomes integral to understanding what makes a city resilient.
Resilience, therefore, despite all odds and the intellectual genealogy outlined above, might
have positive connotations, or faint messianic capacities to invoke Walter Benjamin. Resil-
ience can be a discourse recognizing the historical situatedness of our ecological relations to
others, the necessity for diversity, and the possibility that the future of a system will never be
its past. Resilience, we might recall from ecology, demands change and diversity.
We must begin to understand resilience for its historically situated reworking of the relation-
ship between nature, ecology, and economy. Such historical consciousness might facilitate
different forms of imagining future institutions, economies, and environments. The future
does not need to replicate the past. Resilience can be a call for multiplicity, and for futures not
yet known, it could yet offer a model of ecological thinking that might defeat the optimizing
demands of capital or conservatism. It might offer the possibility of not a new normal but a
new nature. If resilience has replaced convenience as an imaginary, our project is to imagine
the lives, social structures, and infrastructures that accompany or contest this new logic.
31 Greenpeace, It's the Finance Sector, Stupid, 21 December 2020, https://media.greenpeace.org/
Detail/27MZIFJ8ZCWOC.
32 Catrin Einhorn and Christopher Flavelle, ‘A Race against Time to Rescue a Reef from Climate Change’,
The New York Times, 10 December 2020, https://www.nytimes.com/2020/12/05/climate/Mexico-
reef-climate-change.html; Stephen J. Collier, James Christopher Mizes, and Antina von Scnitzler,
‘Preface: Public Infrastructures / Infrastructural Publics’, Limn 7 (2016), https://limn.it/issues/public-
infrastructuresinfrastructural-publics/.
64 THEORY ON DEMAND
References
‘Agent Orange’, History.Com, 2 August 2011, updated 16 May 2019, https://www.history.com/topics/
vietnam-war/agent-orange-1.
Barber, Daniel A. Modern Architecture and Climate: Design before Air Conditioning (Princeton: Prince-
ton University Press, 2020).
Barber, Daniel A.; Kuo, Jeannette; Uduku, Ola; Auer, Thomas; and e-flux Architecture. ‘Editorial’,
E-flux Architecture: ‘After Comfort: A User’s Guide’ (October 2023), https://www.e-flux.com/architec-
ture/after-comfort/568230/editorial/.
Black Lives Matter, ‘About’, https://blacklivesmatter.com/about/.
Black, Fischer. ‘Noise’, The Journal of Finance 41.3 (1986): 528-543.
Calvillo, Nerea. Aeropolis: Queering Air in Toxicpolluted Worlds, New York: Columbia University Press,
2023.
Chua, Charmaine; Danyluk, Martin; Cowen, Deborah; and Khalili, Laleh. ‘Introduction: Turbulent
Circulation: Building a Critical Engagement with Logistics’, Environment and Planning D: Society and
Space 36.4 (2018): 617-629.
Collier, Stephen J.; Mizes, James Christopher; and von Scnitzler, Antina. ‘Preface: Public Infrastruc-
tures / Infrastructural Publics’, Limn 7 (2016), https://limn.it/issues/public-infrastructuresinfrastructur-
al-publics/.
Cooper, Melinda. Family Values: Between Neoliberalism and the New Social Conservatism, New York:
Zone Books, 2017.
Cowen, Deborah. The Deadly Life of Logistics: Mapping Violence in Global Trade, Minneapolis: Univer-
sity of Minnesota Press, 2014.
Edwards, Paul N. The Closed World: Computers and the Politics of Discourse in Cold War America,
Cambridge: The MIT Press, 1997.
Einhorn, Catrin and Flavelle, Christopher. ‘A Race against Time to Rescue a Reef from Climate
Change’, The New York Times, 10 December 2020, https://www.nytimes.com/2020/12/05/climate/
Mexico-reef-climate-change.html.
Friedman, Milton. ‘The Need for Futures Markets in Currencies’, Cato Journal 31.3 (2011): 635-641.
Greenpeace, It’s the Finance Sector, Stupid, 21 December 2020, https://media.greenpeace.org/
Detail/27MZIFJ8ZCWOC.
Grove, Kevin. Resilience, New York: Routledge, 2018.
Halpern, Orit and Mitchell, Robert. The Smartness Mandate, Cambridge: The MIT Press, 2023.
Hamblin, Jacob Darwin. Arming Mother Nature: The Birth of Catastrophic Environmentalism, Oxford:
Oxford University Press, 2013.
Hayek, Friedrich August. ‘The Pretence of Knowledge’, The Nobel Prize, Lecture delivered for The
Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, 11 December 1974, www.
nobelprize.org/prizes/economic-sciences/1974/hayek/lecture/.
Hegarty, Aongus. ‘Digital Resilience: Building the Economies of Tomorrow on a Foundation of Cyberse-
curity’, World Economic Forum, 20 May 2022, https://www.weforum.org/agenda/2022/05/digital-resil-
ience-building-the-economies-of-tomorrow-on-a-foundation-of-cybersecurity/.
Holling, C.S. (ed). Adaptive Environmental Assessment and Management, New York: John Wiley and
Sons, 1978.
65
IN/CONVENIENCE
Holling, C.S. ‘Resilience and Stability of Ecological Systems’, Annual Review of Ecological Systems 4
(1973): 1-23.
Hutchinson, G. Evelyn. ‘Circular Causal Systems in Ecology’, Annals of the New York Academy of
Science 50 (1948): 221-246.
LeCavalier, Jesse. The Rule of Logistics: Walmart and the Architecture of Fulfillment, Minneapolis:
University of Minnesota Press, 2016.
Lewis, Paul. ‘The Emergence of “Emergence” in the Work of F.A. Hayek: A Historical Analysis’, History
of Political Economy 48.1 (2016): 111-150.
Meadows, Donella H.; Meadows, Dennis L.; Randers, Jorgen; and Behrens III, William W. The Limits to
Growth, New York: Universe Books, 1972.
Mehrling, Perry. Fischer Black and the Revolutionary Idea of Finance, New York: John Wiley and Sons,
2005.
Neves, Joshua and Steinberg, Marc. ‘In Convenience’, in this volume.
Polgreen, Lydia. ‘Restoring the Past Won’t Liberate Palestine’, The New York Times, 18 February 2024,
https://www.nytimes.com/2024/02/18/opinion/israel-gaza-palestine-decolonization.html.
Rossiter, Ned. Software, Infrastructure, Labor: A Media Theory of Logistical Nightmares, New York:
Routledge, 2016.
Smith, Ryan C. The Real Oil Shock: How Oil Transformed Money, Debt, and Finance, Cham: Palgrave
Macmillan, 2022.
Steinberg, Marc. The Platform Economy: How Japan Transformed the Consumer Internet, Minneapolis:
University of Minnesota Press, 2019.
Tsing, Anna. Friction: An Ethnography of Global Connection, Princeton: Princeton University Press,
2004.
West, Emily. Buy Now: How Amazon Branded Convenience and Normalized Monopoly, Cambridge: The
MIT Press, 2022.
66 THEORY ON DEMAND
COLLECTIVIZING CONVENIENCE? FROM DELIVERY
TO LOGISTICALITY
ARMIN BEVERUNGEN
Introduction
Amazon’s convenience enchants. In 2019, 68% of US Americans were already Amazon Prime
members. For Germany, Amazon’s second largest market, that figure was 63%.1 In these
countries and many others, as Amazon expands globally and intensifies its grip on delivery
and inroads into streaming, figures have risen and are expected to rise further. Amazon Prime
provides the convenience of streaming media content and of home delivery of items from
its web shop. Amazon has branded convenience, as Emily West suggests, with customers
affectively and intimately enchanted by its brown boxes that seemingly magically appear at
our doorsteps.2 Convenience is thus key to understanding Amazon, and an account of con-
venience today requires making sense of how Amazon has shaped it.
At the same time, a significant part of scholars concerned with socialist or democratic plan-
ning are in thrall of Amazon and its promise of luxury and plenty. Framing Amazon alongside
other companies such as Walmart as a ‘master planner’, which through its ‘logistical and
algorithmic innovations’ provides the kinds of convenience desired by its customers, Leigh
Phillips and Michal Rozworski suggest that ‘Amazon offers techniques of production and
distribution that are just waiting to be seized and repurposed’.3 In putting this concern for
the appropriation of Amazon’s logistical prowess in the context of a debate on convenience,
this contribution asks a simple question: can Amazon’s convenience be collectivized? By
characterizing Amazon’s convenience as logistical, as convenience delivered, the contribution
points to the entanglement between logistics, planning and convenience at Amazon. Where
critical commentary has established the costs of convenience in terms of labor exploitation
and consumer surveillance, the contribution contends that Amazon’s convenience further-
more implies a logistification of life, which largely evacuates collectivity.
1 L. Lohmeier, ‘Umfrage zur Amazon Prime - Mitgliedschaft in den USA 2019’, Statista, 2 January 2024,
https://de.statista.com/statistik/daten/studie/1029566/umfrage/amazon-prime-mitgliedschaft-in-den-
usa/; L. Lohmeier, ‘Amazon Prime - Mitgliedschaft in Deutschland 2019’, Statista, 2 January 2024,
https://de.statista.com/statistik/daten/studie/1029563/umfrage/amazon-prime-mitgliedschaft-in-
deutschland/.
2 Emily West, Buy Now: How Amazon Branded Convenience and Normalized Monopoly. Distribution
Matters. Cambridge, MA: The MIT Press, 2022.
3 Leigh Phillips and Michal Rozworski, The People’s Republic of Walmart: How the World’s Biggest
Corporations Are Laying the Foundation for Socialism, London: Verso, 2019, p. 77. Amazon is, besides
Walmart and Project Cybersyn, perhaps the most important point of reference for recent debates
around democratic planning and socialist calculation, also e.g. in Evgeny Morozov, ‘Digital Socialism?’,
New Left Review 116/117 (2019): 33-67.
67
IN/CONVENIENCE
The contribution subsequently challenges celebrations of Amazon’s logistical convenience,
and suggests that a potential collectivization of convenience demands a more specific reck-
oning with convenience delivered. If Amazon’s convenience is logistics in disguise, and if
the techniques and operations of Amazon’s logistics are fundamentally counter-collective,
then Amazon’s convenience cannot simply be collectivized. Instead, it must be confronted
with logisticality, that is, the collective capacity to organize life without logistical planning.
Logisticality defies logistical convenience, and may bring forth a different kind of convenience.
Logistical Convenience, Convenience Delivered
How Amazon redefines convenience can be situated in a long history of technologies promis-
ing convenience. Thomas F. Tierney traces the emergence of a modern notion of convenience
to the 17th century, where convenience is ‘is no longer a matter of the suitability of something
to the facts, nature, or a moral code’ but instead necessarily refers to a person’s body, so that
something is considered convenient ‘in the modern sense of these words if it is suitable to per-
sonal comfort or ease’.4 This coincides, according to Tierney, with a changed understanding of
the body as imposing limits, and modern technology offering to overcome these: ‘something is
a convenience if it is suitable to the modern task of overcoming the limits which are imposed
by the body’.5 This value of convenience—‘the value of the masses […] who consume the
products of technical culture’—comes to dominate technological development, according
to Tierney.6 Tierney’s subsequently rather static notion of convenience lends itself to a quite
determinist history of technology. In contrast, recognizing the plasticity and historicity of the
notion of convenience puts into focus how convenience develops alongside technologies and
their associated socialites and cultures.
How convenience changes in the 20th century has been shown by Elizabeth Shove in her
account of consumption cultures, highlighting in particular ‘illuminating developments in the
sociotemporal order’.7 Where previously ‘conveniences’ were situated somewhere between
necessity and luxury, at the end of the 20th century ‘hypermodern’ conveniences such as
‘microwave cookers, freezers, answerphones and text messaging facilities’ promise the ability
to affect timing: ‘that is, the ability to shift and juggle obligations and to construct and deter-
mine personal schedules’.8 A broad understanding of convenience as overcoming bodily limits
here gives way to a socially and culturally coded capacity to order life temporally and spatially.
The provision of this capacity is unevenly distributed and highly gendered and racialized, since
many of the conveniences in question center around the household and therefore feminized
and racialized labor, with convenience also always involving a reorganization of such labor.9
4 Thomas F. Tierney, The Value of Convenience: A Genealogy of Technical Culture, Albany: State
University of New York Press, 1993, pp. 39, 91-93. See also: Rahul Oka, ‘Introducing an Anthropology of
Convenience’, Economic Anthropology 8.2 (2021): 188-207.
5 Tierney, The Value of Convenience, p. 40.
6 Tierney, The Value of Convenience, p. 8.
7 Elizabeth Shove, Comfort, Cleanliness and Convenience: The Social Organization of Normality, Oxford:
Berg, 2003, p. 185.
8 Shove, Comfort, Cleanliness and Convenience, p. 186.
9 See, for example, Neda Atanasoski and Kalindi Vora, Surrogate Humanity: Race, Robots, and the Politics
68 THEORY ON DEMAND
At least since the middle of the 20th century, convenience has also become an explicit subject
of marketing, and therefore shaped by the ways in which organizations seek to construct,
promote and sell it, particular in relation to an emerging service sector.10 That is not to sug-
gest that convenience can be reduced to an attribute of a product or service, but to trace
how its social and cultural dominance has been formed also by practices of marketing.11
Amazon perhaps stands at the pinnacle of this development in marketing convenience: as
West convincingly portrays, Amazon has branded convenience. Amazon offers a wide set of
convenient services, such as media streaming and Alexa as personal assistant, which overall
revolve around Amazon’s image as a distribution brand which delivers convenience. As West
puts it: ‘The box encapsulates Amazon’s brand promise to deliver smiles to our doorsteps
– something the company rarely says with words, but communicates on every branded box
and envelope’.12 The ‘everything store’ literally promises—with its swoosh from A to Z—to
make anything available for fast delivery.
Amazon is inscribed in broader shifts in convenience, which impact the spatio-temporal
orderings explored by Shove. The juxtaposition with debates about convenient devices for
the home, the convenience store or convenience food of the late 20th century makes clear
what kinds of shifts have taken place in the meaning of convenience. Convenience stores
respond to time-sensitivities of customer by providing easy ways to shop while on the road,13
and convenience food both offers a reduction in the labour involved in its preparation and
what Alan Warde calls ‘time-shifting’, in this case the ability to quickly and spontaneously
prepare a meal.14 Amazon is also in the business of convenience stores, providing a suppos-
edly new level of convenience in enabling customers to skip queues at checkouts through its
just-walk-out technology deployed in its Amazon Go stores, in North America and the UK.15
And through Amazon Fresh and its takeover of Whole Foods in the USA, Amazon is also in
the business of convenience food. However, Amazon’s convenience pivots around delivery,
and epitomizes the shift from retail to delivery in recent decades.
Focusing on convenience as convenience delivered, as well as on its concomitant spatio-tem-
poral orderings, manifests its logistical character. It also indexes Amazon as a key player in
logistics: according to Clare Lyster, Amazon represents ‘the epitome of contemporary logisti-
cal intelligence’.16 Last mile delivery is key to Amazon’s promise and branding of convenience,
since it is the brown box arriving on our doorsteps which fulfils this promise. In this, Amazon
partakes in broader shifts towards ‘logistical urbanism’, wherein developments particularly
of Technological Futures, Durham: Duke University Press, 2019; Sarah Sharma, In the Meantime:
Temporality and Cultural Politics, Durham: Duke University Press, 2014.
10 Jillian Dawes Farquhar and Jennifer Rowley, ‘Convenience: A Services Perspective’, Marketing Theory
9.4 (2009): 425-438.
11 Oka, ‘Introducing an Anthropology of Convenience’, 204.
12 West, Buy Now, p. 66.
13 Steven M. Graves, ‘Convenience Stores: A Landscape Perspective’, Yearbook of the Association of
Pacific Coast Geographers 79.1 (2017): 134-152.
14 Alan Warde, ‘Convenience Food: Space and Timing’, British Food Journal 101.7 (1999): 518.
15 Jenny Huberman, ‘Amazon Go, Surveillance Capitalism, and the Ideology of Convenience’, Economic
Anthropology 8.2 (2021): 337-49.
16 Clare Lyster, Learning from Logistics: How Networks Change Our Cities, Basel: Birkhäuser, 2016, p. 119.
69
IN/CONVENIENCE
in last-mile logistics have shifted the terrain of how the temporalities and spatialities of cit-
ies are reproduced and repurposed around delivery.17 It is not that convenience stores or
convenience foods weren’t logistical achievements—they merely followed a different logic
oriented around the visit to the store rather than towards delivery.18 Where a general focus
on logistics makes cities appear in many ways always already ordered through the flows of
communication and commerce, bringing forth their own topologies,19 a specific focus on
Amazon’s logistics highlights how it produces particular notions of convenience and coinci-
dent spatio-temporal orderings.
Of note in particular is how Amazon has over the last years extensively developed not only its
network of fulfillment centers, but also its capacities for last mile delivery, principally in North
America and Europe, but also elsewhere such as the UAE.20 This has allowed it to continuously
improve on the speed and flexibility of last-mile delivery, moving from two-day to next-day to
same-day delivery for popular items and customers in select urban areas mostly in the Global
North, and even two-hour delivery for food in particular vicinities of Amazon Fresh stores
in the USA. The urban landscape of fast delivery is crowded with other providers, such as
DoorDash or UberEats for delivery of fresh meals, or Getyr, Zepto or Ola for 10-minute deliv-
ery of a limited basket of everyday goods. However, what qualifies Amazon’s convenience is
that Amazon’s everything store offers a much wider range of goods than 10-minute-delivery
companies, and its development of last-mile delivery infrastructure is matched only by postal
services in its depth within individual countries. Its delivery is also thoroughly integrated with
a broader technological stack, such as its ‘1-click-technology’, easy payment facilities, or the
voice assistant Alexa, framed as the easiest gateway to ordering.21
In sum, Amazon’s convenience combines the breadth of products on offer in the everything store,
ease of ordering and payment through specific technologies provided by or allowing access to
Amazon’s store, and speed of delivery to one’s home. Variations and extensions of these elements
are part of Amazon’s promise of convenience, for example when it expands into retail or allows
other providers to adapt its technologies—such as Amazon Pay or Amazon One for checkout with
one’s palm. Yet the key premise remains that customers are invited to stay at home, and to have
goods delivered to their doorstep. The ‘Amazonification’ of logistics, building on earlier logistical
imaginations like those associated with the Sears mail order catalogue,22 can be understood as
concerned with the consumer home as the end-point of logistics and the effort to dominate last
touch logistics.23 Amazon’s convenience is logistical convenience, convenience delivered.
17 Moritz Altenried, ‘On the Last Mile: Logistical Urbanism and the Transformation of Labour’, Work
Organisation, Labour & Globalisation 13.1 (2019): 114-29.
18 See Joshua Neves and Marc Steinberg, ‘In Convenience’, this volume, for a useful discussion of the
relation between the logistics of convenience stores and of platform capitalism today.
19 Lyster, Learning from Logistics.
20 Martin Kenney, Dafna Bearson, and John Zysman, ‘The Platform Economy Matures: Measuring
Pervasiveness and Exploring Power’, Socio-Economic Review 19.4 (October 2021): 1467; Altenried, ‘On
the Last Mile’, 124.
21 West, Buy Now, pp. 45-46.
22 Matthew Hockenberry, Nicole Starosielski, and Susan Zieger (eds) Assembly Codes: The Logistics of
Media, Durham, NC: Duke University Press, 2021, p. 7.
23 Jake Alimahomed-Wilson, ‘The Amazonification of Logistics: E-Commerce, Labor, and Exploitation in
70 THEORY ON DEMAND
Inconveniences of Logistical Life
A concern with the collectivization of convenience builds on the malleability of technol-
ogy, and therefore the possibility of appropriating the technologies at work in Amazon’s
logistical operations. It articulates a critique of how technology operates within Amazon
today, yet presumes that this technology can operate differently in the context of socialist
or democratic economic planning. For example, Srnicek and Williams argue that logis-
tics will be essential for postcapitalism, that despite its association with the exploitation
of labour, logistics is at the forefront of automation and struggles towards postwork.24
These debates much less challenge Amazon’s notion of convenience, which promises
something close to an imaginary of luxury, of a kind of plenty or post-scarcity associated
with the ’everything store’ that makes goods available at home the next day; a resilient
convenience that even promises to deliver when disaster looms, as Amazon did during
the COVID-19 pandemic.25 The ‘actually existing automation’ at Amazon also serves as
forerunner to post-scarcity in labour, which is a key element of visions of ‘fully automated
luxury communism’.26
Celebrations of Amazon’s logistical operations to be repurposed for mostly centralized
planning presume that the convenience of logistics can be disentangled from its incon-
veniences. Phillips and Rozworski, for example, note that, alongside Walmart, Amazon’s
story ‘is another tale of getting the logistics right—in other words, getting things from
point A to point B as cheaply as possible’.27
In simplest terms, Amazon is a giant planned machine for distributing goods. It
is a mechanism for forecasting, managing and meeting demand for an incredibly
wide array of things we need and want. It is a collection of thousands of interlocking
optimization systems that work together to carry out the deceptively simple task of
moving objects from producers to consumers.28
The authors qualify this adoration, noting how its planning technologies ‘are a way of
meeting a skewed set of social needs—one that ends up enriching a few, misusing sub-
stantial free social labor, and degrading workers’.29 They also list some challenges in
appropriating and collectivizing Amazon, in particular with regards to large-scale techni-
the Last Mile’, in Jake Alimahomed-Wilson and Ellen Reese (eds) The Cost of Free Shipping: Amazon in
the Global Economy, London: Pluto Press, 2021, pp. 69-70.
24 Nick Srnicek and Alex Williams, Inventing the Future: Postcapitalism and a World without Work, Revised
and updated edition, London: Verso, 2016, pp. 150-154.
25 Dave Lee and Patricia Nilsson, ‘Amazon Auditions to Be "the New Red Cross" in Covid-19 Crisis’,
Financial Times, 31 March 2020, https://www.ft.com/content/220bf850-726c-11ea-ad98-
044200cb277f. Amazon has also recently opened disaster relief hubs at various fulfilment centers in
North America and Europe. On the relationship between convenience and resilience, see Orit Halpern’s
contribution in this volume.
26 Aaron Bastani, Fully Automated Luxury Communism: A Manifesto, London: Verso, 2019, p. 88.
27 Phillips and Rozworski, People’s Republic of Walmart, pp. 78, 80.
28 Phillips and Rozworski, People’s Republic of Walmart, p. 92.
29 Phillips and Rozworski, People’s Republic of Walmart, p. 93.
71
IN/CONVENIENCE
cal feasibility, its continuing reliance on the price mechanism of markets, and the dangers
of surveillance.30
Yet their basic utopian premise is that some of these negative aspects of Amazon’s logistical
operations can be disentangled from the planning techniques to be appropriated and col-
lectivized for centralized, democratic planning.31 In contrast, focusing on the specificities of
Amazon’s logistical convenience as convenience delivered emphasizes how closely related
it is to the inconveniences of Amazon’s logistical operations, and how impossible it may be to
recode technologies when they are precisely geared towards logistically managing workers
and consumers. It suggests that Amazon’s innovations may be found less in advanced plan-
ning techniques as in particular forms of automation which tie both workers and consumers
to specific spatio-temporal regimes of control and speed, which exhibit what Neves and
Steinberg characterize as the compulsory aspects of convenience.
The inconvenience of logistical labor at Amazon and elsewhere have been widely docu-
mented.32 Fulfilment centers are spaces carefully designed and technologically equipped
to organize logistical labor whose discrete grammars of action are meticulously captured.33
Alessandro Delfanti describes the kinds of technologically enhanced forms of management
in Amazon’s fulfilment centres as ‘machinic dispossession’, wherein techniques of ‘chaotic
storage’ deprive labour of the knowledge of the whereabouts of things in the warehouse, and
‘augmented despotism’, where machinic control is complemented with autocratic cultur-
al-managerial techniques.34 Beyond the warehouse, in the last-mile labor organized through
the Amazon Flex app, a highly flexible and scalable workforce is algorithmically managed and
directed.35 Labor here is thoroughly coded and grammatized in logistical terms. For example,
Matthew Hockenberry explores the role of the cell phone in constructing what he terms
‘cellular labor’: cellularity ‘enables not just a multiplicity of mobility but a multiplication of
management’.36 As logistical media technologies, the cell phone and the Flex app allow cel-
lular labor to be geolocated and directed, and they also allow the scanning of barcodes as the
quintessential operation of tracking both logistical goods and the operations of logistical labor.
30 Phillips and Rozworski, People’s Republic of Walmart, pp. 93-95.
31 Current planning debates extend beyond rejuvenated proposals for centrally planned economies since
William Paul Cockshott and Allin Cottrell, Towards a New Socialism, Nottingham: Spokesman, 1993.
For an overview, see Christoph Sorg and Jan Groos (eds) Competition and Change: special issue on
‘Rethinking Economic Planning’ (2024, forthcoming); and Campbell Jones, ‘Introduction: The Return of
Economic Planning’, South Atlantic Quarterly 119.1 (2020): 1-10.
32 Jake Alimahomed-Wilson and Ellen Reese (eds), The Cost of Free Shipping: Amazon in the Global
Economy, London: Pluto Press, 2021; Alessandro Delfanti, The Warehouse: Workers and Robots at
Amazon, London: Pluto Press, 2021.
33 Armin Beverungen, ‘The Invisualities of Capture in Amazon’s Logistical Operations’, Digital Culture &
Society 7.2 (2022): 185-202.
34 Alessandro Delfanti, ‘Machinic Dispossession and Augmented Despotism: Digital Work in an Amazon
Warehouse’, New Media & Society 23.1 (2021): 39-55.
35 Altenried, ‘On the Last Mile’, 123-126.
36 Matthew Hockenberry, ‘Cellular Capitalism: Life and Labor at the End of the Digital Supply Chain’, in
Mark Graham and Fabian Ferrari (eds) Digital Work in the Planetary Market, Cambridge, MA: The MIT
Press, 2022, p. 265.
72 THEORY ON DEMAND
The inconveniences of consumption at Amazon are usually discussed in terms of sur-
veillance and capture. Jennifer Huberman, considering the case of Amazon Go stores,
argues that convenience functions as an ideology justifying extraction and control in
the register of surveillance capitalism.37 Yet rather than obscure it, Amazon provides
surveillance as a service: its attraction lies in ‘the brand’s knowledge of the consumers’
and therefore ‘the intimacy of the relationship and the quality of its services’.38 This obser-
vation already points to how surveillance is more than an ideology at Amazon: the data
gathered on consumer behavior constitutes an essential input for Amazon’s predictive
algorithms, which tie its technologies for anticipatory shipping to the optimization of its
logistical operations.39 Despite being sold explicitly as a service, Amazon’s technologies
of surveillance seek ‘to capture forms of behavior that are unaffected by self-conscious
awareness of surveillance’,40 with the Echo and Ring devices enticing us to unconscious
consumption, in a process that David Hill calls ‘the disappearing from consciousness of
“habitual media”’.41
The automation of behaviour and the attendant reduction of liberties are recurring
themes in these critiques of Amazon, whether in the register of a critique of alienation,
of ideology, or otherwise. These critiques, essential as they are, do not suffice to direct
a collectivization of convenience which requires the disentangling of logistical planning
from surveillance and control. On the one hand, operating in what Jean Burgess and her
co-authors have called ‘big critique’, they partly overstate the efficacy of Amazon’s tech-
nologies, for example with regards to the automation of behaviour through an address of
the unconscious.42 In doing so, they potentially reproduce a technological sublime which
also misleadingly fuels the infatuation with Amazon’s planning techniques. On the other
hand, the critiques don’t fully articulate the consequences of convenience in terms of
collectivity. West, Huberman, and more famously Shoshana Zuboff, lament the loss of the
sovereign subject of consumption in the surveilled, served self of convenience.43 Their
analysis implies a politics which seeks a return to the sovereign subject of consumption.
Yet as a political horizon for a collectivization of convenience this seems insufficient,
considering the marketing of convenience is certainly not the starting point of an under-
37 Huberman, ‘Amazon Go, Surveillance Capitalism, and the Ideology of Convenience’, 338, 346. The
media scholar Lauren Bridges in a complementary way recounts how Amazon Ring devices, through
what she calls ‘infrastructural obfuscation’, partakes in broader kinds of surveillance, where
Amazon’s infrastructures of surveillance connect to carceral regimes. Lauren Bridges, ‘Infrastructural
Obfuscation: Unpacking the Carceral Logics of the Ring Surveillant Assemblage’, Information,
Communication & Society 24.6 (2021): 830-49.
38 West, Buy Now, p. 118.
39 Eva-Maria Nyckel, ‘Ahead of Time: The Infrastructure of Amazon’s Anticipatory Shipping Method’,
in Axel Volmar and Kyle Stine (eds) Media Infrastructures and the Politics of Digital Time: Essays on
Hardwired Temporalities, Amsterdam: Amsterdam University Press, 2021, pp. 263-78.
40 Mark Andrejevic, Automated Media, New York: Routledge, 2020, p. 40.
41 David W. Hill, ‘The Injuries of Platform Logistics’, Media, Culture & Society 42.4 (2020): 524-525.
42 Jean Burgess, ‘Everyday Data Cultures: Beyond Big Critique and the Technological Sublime’, AI &
Society 38.3 (2023): 1243-1244.
43 Huberman, ‘Amazon Go, Surveillance Capitalism, and the Ideology of Convenience’; West, Buy Now,
p. 133-137; Shoshana Zuboff, The Age of Surveillance Capitalism: The Fight for a Human Future at the
New Frontier of Power, 1st edition, New York: PublicAffairs, 2019.
73
IN/CONVENIENCE
mining of individual sovereignty, not to mention collectivity. The history of marketing, not
only of convenience, is one traversed by attempts to undermine the sovereign subject.44
These inconveniences are not merely a negative flipside of convenience, to be separated from
it or easily critiqued away. More fundamentally, they demonstrate how logistical convenience,
as convenience delivered, imposes what I would call the logistification of life. Convenience as
logistics in disguise requires the logistification of life, producing what Julian Reid has called
‘logistical life’: ‘a life lived under the duress of the command to be efficient, to communicate
one’s purposes transparently in relation to others, to be positioned where one is required, to
use time economically, to be able to move when and where one is told to’.45 Here the incon-
veniences of labour and consumption become visible as related. For example, in having to
make oneself available for the blocks of delivery offered by Amazon Flex or the changes in
shift work in the fulfilment center, the logistification of labor extends to life. And Amazon’s
logistical convenience invites a personal logistics as much as the calculation of desires to be
fulfilled becomes an essential part of planning and prediction.
The compulsory aspects of convenience that Neves and Steinberg describe are a key aspect
of this logistification, here in the form of logistical convenience, convenience delivered. The
compulsory aspects of logistical convenience become apparent, for example, in the ‘personal
logistics’ described by Melissa Gregg as imposing ‘the labor of synchronizing schedules and
commitments’ onto everyone, also with regards to the power differential between those who
schedule and those who are scheduled.46 They can also can be recognized in what Stefano
Harney and Fred Moten call ‘synaptic labor’, which logistics demands and which they char-
acterize as a ‘capacity for composition given in having been entered, as it were, into the flow
of assembly upon command’: ‘Synaptic labor plugs in anywhere, translates anything, and one
must devise one’s own forms of “queue theory” for the flow of lines that run in every direction,
like a sea’.47 These demands to be available for synchronization and for composition, essential
for logistics, extend from the logistical labor of the warehouse and delivery to the consumer
and citizen in their organization of daily, logistified life.
I want to suggest that the debates around economic planning and the collectivization of con-
venience would benefit from understanding Amazon’s logistical convenience in terms of the
44 The economic historian Philip Mirowski has called ‘murketing’ those practices which play on both a
promise of sovereignty while at the same time undermining it, producing a murky space of decision
and choice. Philip Mirowski, Never Let a Serious Crisis Go to Waste: How Neoliberalism Survived the
Financial Meltdown, London: Verso, 2013, pp. 138-148; also see Stephen Dunne, ‘'Murketing' and the
Rhetoric of the New Sincerity’, Journal of Marketing Management 34.15-16 (2018): 1296-1318.
45 Julian Reid, The Biopolitics of the War on Terror: Life Struggles, Liberal Modernity, and the Defence of
Logistical Societies, Manchester: Manchester University Press, 2006, p. 17. The quote continues: ‘and
crucially, to be able to extol these capacities as the values which one would willingly, if called upon,
kill and die for’. To assess Reid’s analysis of logistics in the context of a biopolitics of war would here
sidestep the more immediate task of asking how Amazon contributes to the logistification of life.
46 Melissa Gregg, Counterproductive: Time Management in the Knowledge Economy, Durham, NC: Duke
University Press, 2018, pp. 129-130.
47 Stefano Harney and Fred Moten, All Incomplete. Wivenhoe New York Port Watson: Minor Compositions,
2021, p. 109.
74 THEORY ON DEMAND
logistification of life. The term highlights how closely Amazon’s planning techniques are nec-
essarily tied to automation, surveillance and capture as essential elements of its logistical
operations; how the inconveniences of labor and consumption must not be tackled separately
but rather be understood in the context of how logistical convenience structures life; and how
this logistification is marked by an evacuation of collectivity. It complements interventions such
as those by Brett Neilson who warns against a mere reverse engineering, and instead calls for
a ‘reverse of engineering’, which formulates a critique of predictive techniques as extractivist
and reliant on ‘merely evidential and measurable’ data, and instead wants to articulate plan-
ning with effective modes of political organization.48 It also complements interventions such
as Max Grünberg’s, which challenges Amazon’s characterization of its predictive analytics
and machine learning capabilities as ‘the state of the art in capitalist demand-forecasting’ by
exploring demand-forecasting not as a technique to be appropriated, but instead one which
involves the modulation of behavior and the logistification of life.49
Logistication as Collectivity Evacuated
The task of collectivizing convenience already seems formidable, considering how Amazon’s
convenience relies on the logistification of life. Amazon’s logistical convenience also implies
a fundamental evacuation of collectivity, which would need to be recuperated in democratic
planning, if planning is not to mean the neutralization of the political.50 First and foremost, the
experience of labor at Amazon is highly individualized, as Amazon deploys standard mana-
gerial techniques derived from Taylorism, cybernetics and behavioral economics which are
geared towards the individual worker, and ties these to algorithmic forms of management
where workers mostly interact with algorithms measuring individual performance.51 Amazon
is also notorious for union busting; recent successes in unionization, such as the establish-
ment of the Amazon Labor Union or increasing strike activities in various countries in Europe
such as Italy, the UK and Germany, point to the discrepancy between the requirements of
Amazon’s logistical operations and the political desires of labor.52 While this may be a price
to pay for socialist planning, it certainly doesn’t bolster the political composition of labour.
There are also specific ways in which Amazon seeks to foreclose a sense of collectivity or
solidarity between its consumers and workers. Amazon, West argues, cultivates what she
calls ‘distribution fetishism’, which means to ‘encourage a personalized, affective relation-
ship between consumer and brand, while discouraging attention to the labor and materiali-
48 Brett Neilson, ‘The Reverse of Engineering’, South Atlantic Quarterly 119.1 (2020): 75-93.
49 Max Grünberg, ‘The Planning Daemon: Future Desire and Communal Production’, Historical
Materialism 31.4 (2023): 115.
50 Matteo Mandarini and Alberto Toscano, ‘Planning for Conflict’, South Atlantic Quarterly 119.1 (2020):
11-30.
51 Armin Beverungen, ‘Automatisiertes Verhalten: Regierungskünste Bei Amazon’, in Georg Toepfer and
Sophia Gräfe (eds) Wissensgeschichte Des Verhaltens. Interdisziplinäre Perspektiven, Berlin: DeGruyter,
2025, forthcoming.
52 Jodi Kantor and Karen Weise, ‘How Christian Smalls and Derrick Palmer Beat Amazon’, The New York
Times, 2 April 2022, https://www.nytimes.com/2022/04/02/business/amazon-union-christian-smalls.
html.
75
IN/CONVENIENCE
ties that underlie heretofore unprecedented short delivery times’.53 Hill similarly argues that
‘unthinking’ consumption conceals ‘the labour that brings our purchases to the doorstep’.54
One particular, and once again essential, technique which optimizes delivery is that of leav-
ing parcels on the porch—a practice which was easily justified and became widespread
during the COVID-19 pandemic, and has become standard for Amazon. The point is that the
relationship between customer and delivery driver becomes mediated through operational
images, as drivers take pictures of parcels on the porch as proof of delivery: ‘As the system
does not intend for the consumer to see the worker, the worker need not see the consumer.
It is the camera—the system—that sees’.55 While making labour invisible doesn’t preclude
solidarity, and the design of unthinking consumption doesn’t preclude thinking about social
relations, this is yet another example of how Amazon’s operational techniques discourage
collectivization.
Consumption at Amazon is also fundamentally personalized. The personalization of the con-
sumer experience mentioned above relies on algorithmic and data operations that are framed
as collective, such as the ‘collaborative filtering’ that is pivotal to Amazon’s recommender
system,56 yet whose purpose is precisely to identify patterns in consumer habits which allow
further personalization. While some other platform enterprises such as Alibaba, Shein or
Pinduoduo have experimented with collective shopping, where the sharing and discussion of
consumer choices is central to the shopping experience, Amazon has largely refrained from
doing so. As the architect Jesse LeCavalier notes, the fulfilment industries ‘foreground the
capacity for individual impulsive choice’ and in doing so ‘claim to free us from confronting
either the abstract but shared responsibilities related to, for example, the “slow violence”
of global warming or the collective immediate action required by contemporary crises of
government, economy, or environment’.57 A democratic planning that builds on these per-
sonalized modes of consumption associated with logistical convenience would need to step
back from the admittedly limited politics of consumption widespread today,58 as much as it
would eschew the possibility for political composition in this realm.
The evacuation of collectivity also becomes apparent in what the architectural theorist Mat-
thew Stewart has termed ‘Amazon urbanism’.59 The kinds of spatio-temporal orderings of the
city that the patents Stewart explores, speculating as they do on drone delivery and flying
warehouses, foresee automated logistical cities largely bereft of sociality.60 The actuality of this
53 West, Buy Now, pp. 62-63.
54 Hill, ‘The Injuries of Platform Logistics’, 5.
55 Hockenberry, ‘Cellular Capitalism’, 273.
56 Brent Smith and Greg Linden, ‘Two Decades of Recommender Systems at Amazon.Com’, IEEE Internet
Computing 21.3 (May 2017): 12-18.
57 Jesse LeCavalier, ‘New Interfaces in the Automated Landscapes of Logistics’, FOOTPRINT 23: The
Architecture of Logistics (Autumn / Winter 2018): 108.
58 Alan Bradshaw, Norah Campbell, and Stephen Dunne, ‘The Politics of Consumption’, Ephemera: Theory
& Politics in Organization 13.2 (2013): 203-16.
59 Matthew Stewart, ‘Amazon Urbanism: Patents and The Totalizing World of Big Tech Futures’, Failed
Architecture, 23 May 2018, https://failedarchitecture.com/amazon-urbanism-patents-and-the-
totalizing-world-of-big-tech-futures/.
60 Amazon’s speculative experiments with drone delivery and robots are far from successful. Cf. Jeff
76 THEORY ON DEMAND
Amazon urbanism manifests itself in the parcels left on porches without human interaction,
and the Amazon Go stores which are meant to require no interaction with a cashier. Amazon’s
logistical convenience is also more broadly reflected in the way Amazon is remaking the city,
in the way its automated delivery builds on an existing ‘urban stack’ for last-mile delivery and
introduces new elements such as Amazon lockers, producing new logistical topologies of the
city largely bereft of human encounter and exchange.61 In scenarios of what Lyster calls the
‘post-human city’, visions of automation perpetuate spatio-temporal arrangements in which
a mix of architectures of convenience enable personalized consumption experiences. Lyster
contends that cities are potentially rescripted today in more equitable ways, since automated
landscapes ‘open up the design of the city to a range of creative stakeholders’.62 Again, though,
most visions of automated logistical cities, and certainly those of Amazon urbanism, largely
discourage collective experience.
Focusing on Amazon’s logistical convenience, as convenience delivered, and on how this
convenience requires the logistification of life, therefore highlights how in the realms of labour
and consumption, as much as in the city, collectivity is eschewed. It also emphasises how con-
venience becomes a demand for a logistified life, which is not only captured and surveilled, but
also thoroughly individualized and personalised. Not only do the inconveniences associated
with Amazon’s convenience appear as essential to it, but Amazon’s operational techniques
push against the collective at every juncture. A recuperation of Amazon’s planning techniques
is faced with the formidable challenge of fundamentally reorienting Amazon’s technologies,
given their articulation with these logics of personalization.
After Logistical Convenience: Counter-Logistics and Logisticalty
Debates on economic planning have somewhat moved on since Phillips and Rozworski’s
intervention, distancing themselves from the distribution fetishism that characterized some
of the earlier debates. For example, recent contributions recognize the need to develop ‘alter-
native socio-technical infrastructures’ and take account of aspects such as care work and
the climate crisis,63 and have explored ideas for distributed planned economies that do not
commence with an infatuation with logistical convenience.64 Jasper Bernes has already earlier
proposed a ‘counter-logistics’ as ‘a proletarian art of war to match capital’s own ars belli’.65
Instead of appropriating logistical techniques they would be turned against logistical capital-
Link, ‘Amazon’s Drone Delivery Dream Is Crashing’, Wired, 4 April 2023, https://www.wired.com/story/
crashes-and-layoffs-plague-amazons-drone-delivery-pilot/; James Vincent, ‘Amazon Stops Field Tests of
Its Delivery Robot Scout’, The Verge, 7 October 2022, https://www.theverge.com/2022/10/7/23392360/
amazon-disbands-delivery-robot-scout-development.
61 Armin Beverungen, ‘Automated Deliver y: Amazon’s Urban Stack’, Navigationen: Zeitschrift für Medien-
und Kulturwissenschaft 24.2 (forthcoming).
62 Clare Lyster, ‘Disciplinary Hybrids: Retail Landscapes of the Post-Human City’, Architectural Design
89.1 (January 2019): 105.
63 Christoph Sorg, ‘Failing to Plan Is Planning to Fail: Toward an Expanded Notion of Democratically
Planned Postcapitalism’, Critical Sociology 49.3 (2023): 478.
64 Jan Groos, ‘Distributed Planned Economies in the Age of Their Technical Feasibility’, BEHEMOTH: A
Journal on Social Dis/Order 14.2 (2021): 75-87.
65 Jasper Bernes, ‘Logistics, Counterlogistics and the Communist Prospect’, Endnotes 3 (2013): 187.
77
IN/CONVENIENCE
ism. More recently, Bernes has rephrased his critique of central planning, noting—in a similar
vein to the analysis above—that the efficacy of central planning requires ‘both surveillance
and automatic coercion’ and thereby ‘reproduces much of what we find intolerable about
capitalism’.66 His contention is that a ‘truly emancipatory revolution’ requires ‘the distribution
of power throughout society’;67 implying that Amazon’s logistical convenience cannot be part
of this politics.
Where counter-logistics is largely conceived as a resistive practice which opposes capitalist
logistics, more recently it has been redefined as an affirmative project to recover the collective
capacities that logistical convenience annihilates, particularly in the context of urbanism.
Moving beyond conceiving of counter-logistics as disruption, Leandro Minuchin and Julieta
Main identify a ‘popular logistics’ developed during the COVID-19 pandemic, which differ-
entially assembles the circulation of resources, solidarities and territorial scales.68 These in
their view could provide ‘a different territorial organisation structured around open and dem-
ocratic supply chains that value environmental resources, cooperative economies and the
sustainment of life’.69 In a similar vein, Matthew Thompson and Yousaf Nishat-Botero contend
that postcapitalist planning requires an urban revolution, which will transform the abstract
space of logistical urbanism into a differential space, wherein planning must be ‘grounded
in the actually existing material struggles and experiments of the “urban everyday”’.70 These
interventions open up a terrain of counter-logistics and alternatives for economic planning
no longer derived from Amazon’s logistical convenience and focused instead on producing
different spatio-temporal orderings of the urban. The analysis of logistical convenience as
convenience delivered shares a concern for the urban while contributing an account of how
Amazon’s logistical convenience relies on an urban stack for last-mile delivery and, more
broadly, the automation of logistical cities.
These affirmative projects of counter-logistics, which do not embrace Amazon’s planning
techniques but may merely appropriate particular elements, such as parts of its technological
stack,71 also align with a politics of what Harney and Moten have called ‘logisticality’. They
define logisticality as ‘the resident capacity to live on earth’, opposed to logistics as ‘the reg-
ulation of that capacity in the service of making the work, the zero-one, one-two world that
pursues the general antagonism of life on earth’.72 This notion of logisticality surfaces from a
more radical critique of logistics as a ‘science of whiteness’ emerging from the slave trade,73
66 Jasper Bernes, ‘Planning and Anarchy’, South Atlantic Quarterly 119.1 (2020): 68.
67 Bernes, ‘Planning and Anarchy’, 69.
68 Leandro Minuchin and Julieta Maino, ‘Counter-Logistics and Municipalism: Popular Infrastructures
during the Pandemic in Rosario’, Urban Studies 60.11 (2023): 2073.
69 Minuchin and Maino, ‘Counter-Logistics and Municipalism’, 2092.
70 Matthew Thompson and Yousaf Nishat-Botero, ‘Postcapitalist Planning and Urban Revolution’,
Competition & Change (2023, online first version): 16.
71 The artist collective knowbotiq, for example, has repurposed Amazon’s Dash buttons for an artistic
project in which conversations with bots around laziness and work refusal are meant to produce
solidarities. See knowbotiq, ‘Amazonian Flesh – how to hang in trees during strike?’, knowbotiq + krcf,
27 April 2019, https://archive.knowbotiq.net/amazonian-flesh/.
72 Harney and Moten, All Incomplete, p. 64.
73 See also Susan Zieger, ‘'Shipped': Paper, Print, and the Atlantic Slave Trade’, in Matthew Hockenberry,
78 THEORY ON DEMAND
and one which represents the degradation of means, where ‘the body is to become a means
only for the smooth flow of transactions […] for the interoperability of all things’.74 Logisticality
for Moten and Harney materializes alongside logistics in ‘the hold, the middle passage’, by
the captured and the fugitive, as ‘the ability to find each other, to move together, to break the
rule of Newtonian time and space, disorder it, and legislate new time and space to disorder, to
gather, stranded into refuge together’.75 This logisticality may seem distant today considering
how logistical convenience or convenience delivered so thoroughly conditions life today. And
yet, it may equally be perceptible in the multiple ways in which life is organized collectively
despite or against logistics.
Logisticality, elusive as the term certainly is, here indexes a more radical politics against
logistical convenience, one which refuses both the spatio-temporal orderings of logistics
and the solutionism of convenience delivered offered by Amazon and others. Instead of col-
lectivizing convenience, it suggests a move away from delivery to logisticality, a refusal of
distribution fetishism and a recognition of the compulsory as much as antagonistic character
of logistical convenience. It also indexes, against the evacuation of collectivity which charac-
terizes the logistification of life underwriting Amazon’s logistical convenience, a concern for
collective capacities which are not tied to centralized planning techniques, but rather rely on
an assembly of a different stack of technologies, capacities and socialities situated in urban
space. Consequently, it also demands an analysis more attuned to antagonism, to the ways
in which the logistical capacities developed by Amazon may imply a denigration of collective
capacities for logisticality. And how logisticality may in turn provide a ground for a different
kind of convenience. What convenience could possibly denote in this context, other than
logistical convenience as convenience delivered, and whether logisticality could point away
from a broader condition of convenience that is compulsory and antagonistic, remains to be
enumerated—not in writing but in the speculative practices associated with logisticality in
urban spaces and beyond.
Acknowledgements
This writing emerges from a research project on ‘Automating the Logistical City: Space, Algo-
rithms, Exploration’, generously funded by the Lower Saxony Ministry for Science and Culture
and the Volkswagen Foundation as part of the zukunft.niedersachsen program. Many thanks
to my collaborators on the project – Ilia Antenucci, Maja-Lee Voigt and Klara Friese – for the
collective research and insights on which I draw here, as well as to Marc Steinberg and Maja-
Lee Voigt for very helpful comments on an earlier version of this paper.
Nicole Starosielski, and Susan Zieger (eds) Assembly Codes: The Logistics of Media, Durham, NC: Duke
University Press, 2021, pp. 34-51.
74 Harney and Moten, All Incomplete, pp. 14, 71.
75 Stefano Harney in Niccolò Cuppini and Mattia Frapporti, ‘Logistics Genealogies: A Dialogue with
Stefano Harney’, Social Text 36.3 (2018): 98.
79
IN/CONVENIENCE
References
Alimahomed-Wilson, Jake and Reese, Ellen (eds). The Cost of Free Shipping: Amazon in the Global
Economy, London: Pluto Press, 2021.
Alimahomed-Wilson, Jake. ‘The Amazonification of Logistics: E-Commerce, Labor, and Exploitation in
the Last Mile’, in Jake Alimahomed-Wilson and Ellen Reese (eds) The Cost of Free Shipping: Amazon in
the Global Economy, London: Pluto Press, 2021, pp. 69-84.
Altenried, Moritz. ‘On the Last Mile: Logistical Urbanism and the Transformation of Labour’, Work
Organisation, Labour & Globalisation 13.1 (2019): 114-29.
Andrejevic, Mark. Automated Media, New York: Routledge, 2020.
Atanasoski, Neda and Vora, Kalindi. Surrogate Humanity: Race, Robots, and the Politics of Technologi-
cal Futures, Durham: Duke University Press, 2019.
Bastani, Aaron. Fully Automated Luxury Communism: A Manifesto, London: Verso, 2019.
Bernes, Jasper. ‘Logistics, Counterlogistics and the Communist Prospect’, Endnotes 3 (2013): 172-
201.
_____. ‘Planning and Anarchy’, South Atlantic Quarterly 119.1 (2020): 53-73.
Beverungen, Armin. ‘The Invisualities of Capture in Amazon’s Logistical Operations’, Digital Culture &
Society 7.2 (2022): 185-202.
_____. ‘Automated Delivery: Amazon’s Urban Stack’, Navigationen: Zeitschrift für Medien- und Kultur-
wissenschaft 24.2 (forthcoming).
_____. ‘Automatisiertes Verhalten: Regierungskünste Bei Amazon’, in Georg Toepfer and Sophia Gräfe
(eds) Wissensgeschichte Des Verhaltens. Interdisziplinäre Perspektiven, Berlin: DeGruyter, 2025, forth-
coming.
Bradshaw, Alan; Campbell, Norah; and Dunne, Stephen. ‘The Politics of Consumption’, Ephemera:
Theory & Politics in Organization 13.2 (2013): 203-16.
Bridges, Lauren. ‘Infrastructural Obfuscation: Unpacking the Carceral Logics of the Ring Surveillant
Assemblage’, Information, Communication & Society 24.6 (2021): 830-49.
Burgess, Jean. ‘Everyday Data Cultures: Beyond Big Critique and the Technological Sublime’, AI &
Society 38.3 (2023): 1243-1244.
Cockshott, William Paul and Cottrell, Allin. Towards a New Socialism, Nottingham: Spokesman, 1993.
Cuppini, Niccolò and Frapporti, Mattia. ‘Logistics Genealogies: A Dialogue with Stefano Harney’, Social
Text 36.3 (2018): 95-110.
Delfanti, Alessandro. ‘Machinic Dispossession and Augmented Despotism: Digital Work in an Amazon
Warehouse’, New Media & Society 23.1 (2021): 39-55.
Delfanti, Alessandro. The Warehouse: Workers and Robots at Amazon, London: Pluto Press, 2021.
Dunne, Stephen. ‘'Murketing' and the Rhetoric of the New Sincerity’, Journal of Marketing Manage-
ment 34.15-16 (2018): 1296-1318.
Farquhar, Jillian Dawes and Rowley, Jennifer. ‘Convenience: A Services Perspective’, Marketing Theo-
ry 9.4 (2009): 425-438.
Graves, Steven M. ‘Convenience Stores: A Landscape Perspective’, Yearbook of the Association of
Pacific Coast Geographers 79.1 (2017): 134-152.
Gregg, Melissa. Counterproductive: Time Management in the Knowledge Economy, Durham: Duke
University Press, 2018.
80 THEORY ON DEMAND
Groos, Jan. ‘Distributed Planned Economies in the Age of Their Technical Feasibility’, BEHEMOTH: A
Journal on Social Dis/Order 14.2 (2021): 75-87.
Grünberg, Max. ‘The Planning Daemon: Future Desire and Communal Production’, Historical Material-
ism 31.4 (2023): 115-159.
Harney, Stefano and Moten, Fred. All Incomplete. New York: Minor Compositions, 2021.
Hill, David W. ‘The Injuries of Platform Logistics’, Media, Culture & Society 42.4 (2020): 521-36.
Hockenberry, Matthew. ‘Cellular Capitalism: Life and Labor at the End of the Digital Supply Chain’,
in Mark Graham and Fabian Ferrari (eds) Digital Work in the Planetary Market, Cambridge: The MIT
Press, 2022, pp. 263-80.
Hockenberry, Matthew; Starosielski, Nicole; and Zieger, Susan (eds). Assembly Codes: The Logistics of
Media, Durham: Duke University Press, 2021.
Huberman, Jenny. ‘Amazon Go, Surveillance Capitalism, and the Ideology of Convenience’, Economic
Anthropology 8.2 (2021): 337-49.
Jones, Campbell. ‘Introduction: The Return of Economic Planning’, South Atlantic Quarterly 119.1
(2020): 1-10.
Kantor, Jodi and Weise, Karen. ‘How Christian Smalls and Derrick Palmer Beat Amazon’, The New York
Times, 2 April 2022, https://www.nytimes.com/2022/04/02/business/amazon-union-christian-smalls.
html.
Kenney, Martin; Bearson, Dafna; and Zysman, John. ‘The Platform Economy Matures: Measuring
Pervasiveness and Exploring Power’, Socio-Economic Review 19.4 (October 2021): 1451-1483.
knowbotiq, ‘Amazonian Flesh – how to hang in trees during strike?’, knowbotiq + krcf, 27 April 2019,
https://archive.knowbotiq.net/amazonian-flesh/.
LeCavalier, Jesse. ‘New Interfaces in the Automated Landscapes of Logistics’, FOOTPRINT 23: The
Architecture of Logistics (Autumn / Winter 2018): 105-113.
Lee, Dave and Nilsson, Patricia. ‘Amazon Auditions to Be "the New Red Cross" in Covid-19 Cri-
sis’, Financial Times, 31 March 2020, https://www.ft.com/content/220bf850-726c-11ea-ad98-
044200cb277f.
Link, Jeff. ‘Amazon’s Drone Delivery Dream Is Crashing’, Wired, 4 April 2023, https://www.wired.com/
story/crashes-and-layoffs-plague-amazons-drone-delivery-pilot/.
Lohmeier, L. ‘Amazon Prime - Mitgliedschaft in Deutschland 2019’, Statista, 2 January 2024, https://
de.statista.com/statistik/daten/studie/1029563/umfrage/amazon-prime-mitgliedschaft-in-deutsch-
land/.
____. ‘Umfrage zur Amazon Prime - Mitgliedschaft in den USA 2019’, Statista, 2 January 2024, https://
de.statista.com/statistik/daten/studie/1029566/umfrage/amazon-prime-mitgliedschaft-in-den-usa/.
Lyster, Clare. Learning from Logistics: How Networks Change Our Cities, Basel: Birkhäuser, 2016.
Lyster, Clare. ‘Disciplinary Hybrids: Retail Landscapes of the Post-Human City’, Architectural Design
89.1 (January 2019): 100-105.
Mandarini, Matteo and Toscano, Alberto. ‘Planning for Conflict’, South Atlantic Quarterly 119.1 (2020):
11-30.
Minuchin, Leandro and Maino, Julieta. ‘Counter-Logistics and Municipalism: Popular Infrastructures
during the Pandemic in Rosario’, Urban Studies 60.11 (2023): 2073-2095.
Mirowski, Philip. Never Let a Serious Crisis Go to Waste: How Neoliberalism Survived the Financial
Meltdown, London: Verso, 2013.
81
IN/CONVENIENCE
Morozov, Evgeny. ‘Digital Socialism?’, New Left Review 116/117 (2019): 33-67.
Neilson, Brett. ‘The Reverse of Engineering’, South Atlantic Quarterly 119.1 (2020): 75-93.
Nyckel, Eva-Maria. ‘Ahead of Time: The Infrastructure of Amazon’s Anticipatory Shipping Method’,
in Axel Volmar and Kyle Stine (eds) Media Infrastructures and the Politics of Digital Time: Essays on
Hardwired Temporalities, Amsterdam: Amsterdam University Press, 2021, pp. 263-78.
Oka, Rahul. ‘Introducing an Anthropology of Convenience’, Economic Anthropology 8.2 (2021): 188-
207.
Phillips, Leigh and Rozworski, Michal. The People’s Republic of Walmart: How the World’s Biggest
Corporations Are Laying the Foundation for Socialism, London: Verso, 2019.
Reid, Julian. The Biopolitics of the War on Terror: Life Struggles, Liberal Modernity, and the Defence of
Logistical Societies, Manchester: Manchester University Press, 2006.
Sharma, Sarah. In the Meantime: Temporality and Cultural Politics, Durham: Duke University Press,
2014.
Shove, Elizabeth. Comfort, Cleanliness and Convenience: The Social Organization of Normality, Oxford:
Berg, 2003.
Smith, Brent and Linden, Greg. ‘Two Decades of Recommender Systems at Amazon.Com’, IEEE Inter-
net Computing 21.3 (May 2017): 12-18.
Sorg, Christoph. ‘Failing to Plan Is Planning to Fail: Toward an Expanded Notion of Democratically
Planned Postcapitalism’, Critical Sociology 49.3 (2023): 475-493.
Sorg, Christoph and Groos, Jan (eds). Competition and Change: special issue on ‘Rethinking Economic
Planning’ (2024, forthcoming).
Srnicek, Nick and Williams, Alex. Inventing the Future: Postcapitalism and a World without Work,
Revised and updated edition, London: Verso, 2016.
Stewart, Matthew. ‘Amazon Urbanism: Patents and The Totalizing World of Big Tech Futures’, Failed
Architecture, 23 May 2018, https://failedarchitecture.com/amazon-urbanism-patents-and-the-totaliz-
ing-world-of-big-tech-futures/.
Thompson, Matthew and Nishat-Botero, Yousaf. ‘Postcapitalist Planning and Urban Revolution’, Com-
petition & Change (2023, online first version): 1-20.
Tierney, Thomas F. The Value of Convenience: A Genealogy of Technical Culture, Albany: State Univer-
sity of New York Press, 1993.
Vincent, James. ‘Amazon Stops Field Tests of Its Delivery Robot Scout’, The Verge, 7 October 2022,
https://www.theverge.com/2022/10/7/23392360/amazon-disbands-delivery-robot-scout-develop-
ment.
Warde, Alan. ‘Convenience Food: Space and Timing’, British Food Journal 101.7 (1999): 518-527.
West, Emily. Buy Now: How Amazon Branded Convenience and Normalized Monopoly. Distribution
Matters. Cambridge: The MIT Press, 2022.
Zieger, Susan. ‘'Shipped': Paper, Print, and the Atlantic Slave Trade’, in Matthew Hockenberry, Nicole
Starosielski, and Susan Zieger (eds) Assembly Codes: The Logistics of Media, Durham: Duke University
Press, 2021, pp. 34-51.
Zuboff, Shoshana. The Age of Surveillance Capitalism: The Fight for a Human Future at the New Fron-
tier of Power, 1st edition, New York: PublicAffairs, 2019.
82 THEORY ON DEMAND
CLOUDED CONVENIENCES
MÉL HOGAN AND STEVEN GONZALEZ MONSERRATE
‘The data center pretty much runs itself’, Noah, a fifty-something-year-old facility manag-
er tells us, indicating the three monitors where he spends most of his twelve-hour shifts
troubleshooting servers and resolving issues through a virtual interface. ‘If things are set up
right, there is little you have to do out there on the floor.’
‘And if they aren’t?’, we ask. Noah frowns. By now he is used to our endless questions and
fathomless curiosity about every minute detail of life in the Cloud’s windowless halls. He
still doesn’t get what ethnography is, why anthropologists and social researchers might find
him fascinating, instead of ‘bones and apes’, as he put it, but he plays along now, no longer
threatened by our spy-like presence.
‘Then you have problems’, Noah says. He grabs a bundle of ethernet cables as we set out
into the ‘out there’ of glittering servers, neatly arrayed on alternating metallic racks like
library books on shelves. We stride through a meshwork of locked metal cages, sections
of the data center securely partitioned for clients, their servers and cables off-limits to
ensure that the traders or business owners that rely on them don’t experience costly
interruptions to their connectivity. To us, they look like chicken coops, though strangely
odorless and without any trace of life. In fact, it is difficult to imagine anything thriving in
this frigid, mechanical labyrinth, where fans whir so loudly that we can barely hear Noah
explaining how the cooling system works. Cold and hot are quarantined in opposing glass
gable structures that prevent hot exhaust from mixing with the refrigerated air that is drawn
into the blinking faceplates of servers in the racks. Butcher slats and blanking panels fill
in any gaps where air might leak and impact cooling efficiency. ‘It starts with cooling. This
is foundational. It’s our biggest expense and priority, because if the servers overheat, it’s
game over for us.’
Noah explains that the seamless user experience of internet access and cloud services
relies on technicians like him ensuring constant, meticulous, thermal regulation. Were it
not for the unseen hands of people like Noah, the Cloud would melt away in a cataclysmic
heat death. For us, this sobering detail feels like a total contradiction of his earlier remark
that the Cloud essentially ‘runs itself’.
‘How do you know how much to cool?’, we ask, wondering if the troubling statistics we’ve
encountered about data centers’ carbon footprint had anything to do with this practice,
which sometimes seemed almost like an art as Noah described it.
‘It depends on what resources you have, your experience, and well…math’, Noah answers,
explaining some basic electrical engineering principles to us, something about matching
kilowatt hours (kwH) to British thermal units (BTUs).
‘So there is more to it than the math?’
Noah brings us to a white machine attached to the ceiling’s ventilation ducts and a pres-
surized underfloor plenum beneath our feet. ‘Some techs have a CRAC addiction.’
We stare at each other in puzzled amusement.
‘Don’t get excited, it’s shorthand for Computer Room Air Conditioner. C-R-A-C.’
‘What do you mean by addiction?’
Noah sighs, hands rifling through the greying wisps of his receding hairline.
83
IN/CONVENIENCE
‘They think that adding more cooling solves the problem. But it’s not that simple. Adding
more traps doesn’t catch a wily mouse. Sometimes it’s about using the right kind of trap.’
‘So there’s more to it than that BTU to KwH equation?’
Noah gestures for us to feel the faceplate of the CRAC unit; a subtle film of condensation
has slicked its metallic surface.
‘Cooling is also about managing humidity.’ Noah peers into a handheld environmental sen-
sor. ‘Too little of it, and a flash of static electricity from your ugly Christmas sweater could
wipe out a drive.’
‘And if there’s too much?’
Noah adjusts one of the dials on the CRAC unit. ‘Pop!’ We chuckle awkwardly at his stilted
onomatopoeia.
‘You know what happens when electronics are submerged in water?’, he asks, not expect-
ing us to answer. We’re grateful he doesn’t add any more dramatic popping sounds as he
explains the adversely conductive properties of water. ‘That’s why Hurricane Sandy was so
catastrophic. The flooding destroyed the equipment.’
If flooding is a threat from below, humidity, we learn, is a threat from above. As water mol-
ecules thicken the air and temperatures rise or fall, condensation occurs. Clouds can form
in the Cloud, and if they go unchecked, rain will fall in the server halls. This is the paradox
of the data center as both a fortress and a crossroads. Too much permeability allows the
outside element of chaos to seep in, as it did for a Meta data center in the cold, wet biome
of Oregon.1 Data centers have a tricky relationship with water. In some facilities we toured,
we knew that water, not air, was the primary cooling agent. A hydraulic ‘irony’.2
Our tour of the cooling system closes on the rooftop, where Noah is eager to introduce us
to the ventilation units jutting up from the ceiling to exchange air. To Noah’s dismay, our
attention is drawn instead to the decaying carcasses and wiry skeletons of pigeons littered
across the roof. We peer at the grisly remnants of these avian pests, wondering if Noah had
anything to do with this archaeology of violence.
‘It’s like a giant pigeon barbeque!’, Noah exclaims, lighting a cigarette.
From this vantage point, we can see the rooftops of other buildings nearby. We notice that
only this one has a crown of ductwork poking up toward the sky, a tell that this nondescript
building is a data center, a vital node of cloud infrastructure built on the bones of an old
factory. It has been ‘slotted in’ to existing urban circuits of electricity, coaxial cabling, and
sewers, structurally robust enough to hold hundreds of server racks, each of which weighs
more than two thousand pounds, as Noah revealed on a previous visit.
‘The falcons are eating good these days’, Noah chuckles, peering at the cranium of a
pigeon, picked clean by the offending osprey. ‘Maybe the ventilation units are affecting the
bird ecosystem.’
Or perhaps pigeons enjoy the computational warmth these units vent out. Life always finds
niches. The Cloud is no different. Despite their attempts to keep living things out—the
spikes poking out from the edges of the roof to dissuade these pigeons from roosting, or
1. Everest Pipkin, It was Raining in the Data Center, PhD diss., Carnegie Mellon University, Pittsburgh,
2018.
2. Jeffrey Moro, Atmospheric media: Computation and the environmental imagination, PhD diss., Uni-
versity of Maryland, College Park, 2022.
84 THEORY ON DEMAND
the mouse traps we noticed in the underfloor plenum protecting vulnerable cables from
gnawing vermin, or the hand sanitizer dispensers that target invisible invaders like bacteria
and the COVID-19 virus—the Cloud is not impenetrable.
We are reminded of this permeability on a return visit, led by a younger tech named Tony.
Tony’s second life as a bassist has so calloused his fingers that he has to ‘borrow’ a finger-
print from a colleague to get us through the biometric checkpoint guarding the ‘core cage’
where the servers are. We find this breach of the center’s otherwise spy-film-level security
apparatus amusing, given the extraordinary screening and vetting measures we under-
went to get here: no parking in the primary lot, to mitigate the risk of vehicular explosives;
surrendering personal electronics into locked bins at the reception; emptying pockets
and walking through a metal detector to ensure that we don’t have hidden USB drives to
introduce bugs into their network, or worse, weapons; proof of our credentials and identifi-
cation submitted well in advance of our visit; non-disclosure agreements to prevent us from
leaking sensitive ‘proprietary business advantage’ details after our visit; the prohibition of
photography or recording of any sort while on the premises. All of these features and more,
including the use of fobs and biometric scans for ingress and egress through the series
of double-locking rooms or ‘mantraps’ between the facility entrance and the server halls,
are deterrents for would-be saboteurs. But, as Tony’s flagrant violation of protocol reveals,
they are also performative. A carefully orchestrated security theater designed to attract
prospective clients or assure existing ones that their ‘mission-critical’ data is ‘safe’.
Tony leads us past a series of recently emptied cages, where scraps of cabling still coil
around racks without servers. Maybe those former clients found out about Tony’s habit of
‘borrowing’ fingerprints. Maybe they left because they felt their data was unsafe. We find
this hard to believe, as we notice the spherical cameras installed at every intersection, a
thousand glass eyes capturing and storing every moment at every rack on a hard drive
somewhere. With this level of constant surveillance, it’s hard to believe that security was
their motive for leaving. The Cloud is embedded in capitalism; it is not merely infrastructure,
but a service and a ‘product’ that is marketized like anything else.
‘Noah said you might be interested in checking out the generators.’
We shout our assent over the din of ventilation and computation. Soon we are outdoors,
relieved by the relative quiet of the cityscape and the gentle, warming presence of the
morning sun on a clear spring day. Tony directs us to a fleet of diesel generators, one of
them flashing in a state of hot-standby.
‘This is our security.’
We frown, wondering why this and not the fingerprint scanner and the rest of the gauntlet
of screenings is considered ‘security.’
Tony reads the confusion in our faces.
‘Security is more than just “security”. We have so many threats to deal with. Yeah, the
human threat is real. But there are worse things. Overheating. Flooding. And, the worst of
the worst, a power outage.’
‘So, these generators keep you running if the power goes out.’
‘They’re our lifeline’, Tony says, as the sun creeps towards the horizon behind him. ‘Even if
the city goes dark, we’ll still be in business.’
We imagine the fleet of generators sputtering to life, their choking stench and toxic plumes
rising up to blot the sun from view. The Cloud emits carbon directly as well as indirectly. We
know from Tony that this single, 80,000 square foot facility draws as much electricity as
a big town or small city (the equivalent of 50,000 homes or more). If the grid goes down,
these generators will keep it chugging along, a special ‘uninterruptible power supply’
85
IN/CONVENIENCE
routed throughout the data center to support its most crucial systems. This process occurs
automatically. A pre-programmed fail-safe. A chain of redundancies that prevent down-
time, a hiccup in digital capitalism, those remarkably rare moments when users encounter
a ‘server error’ or ‘server unavailable’ message. Disruptions cost companies thousands of
dollars per minute. If only they could harness the sun.
‘We follow strict EPA guidelines when we use the generators. Of course, we’d rather not
use them, but this is why data centers are so resilient. Come hurricanes or wildfires or grid
disruptions, we can keep this place going with these generators.’
‘And why can’t you use renewables like solar or wind to power the data center?’
Tony scoffs, and his response sounds more like a direct quote from a marketing brochure
than an off-the-cuff answer from the human being we were speaking to just seconds earlier.
‘We invest in green energy. We care about that, for sure. But you can’t run the Cloud on
renewables.’
‘Why not?’
A cloud passes over the sun, momentarily dimming it.
‘It’s not reliable. Wind fluctuates. The sun’s brightness is variable. You can’t control for
those things. Data centers have to be ultra reliable. This is like a space station; a single
crack, a single breach can mean catastrophe. We have to avoid every risk, every opportuni-
ty for failure.’
Later, over coffee, we learn how the employees internalize this risk-averse attitude. Tony
tells us about the ‘reality’ of human error – the fallibility of techs like him or Noah who are,
after all, human beings with limited capacities – and their constant fear of those mistakes.
‘It’s a stressful job’, Tony admits, ‘but the challenge is exhilarating. It keeps you on your
feet. Every day is different. There are days when I want to strangle myself with one of these
cables and other days where I feel like I’m a conductor, orchestrating a beautiful symphony.
On those days, it feels like the Cloud is just running itself, and we’re along for the ride.’
Tony repeats Noah’s claim that the data center is nearly automated, but we are skeptical.
We see the cascading redundancies, control measures always on the brink of failure, nest-
ed complexities that require constant care and finessing. The constancy and convenience
of our online experiences are anything but given, we realize. Contrary to Tony and Noah’s
unshakable optimism, the Cloud’s ‘reliability’ is the outcome of myriad convergences, an
ever-hungering beast devouring precious resources: water, electricity, land.
We thank Tony as we set off into the ‘out there’, the world beyond the data center, the false
fortress built atop the bones of a factory from a long-vanished era of Midwestern industrial
splendor. We look back as we make our way to the exit, at the marble sprawl of the lobby
and its vaulted-stone ceilings. The building is strong – made to survive industry turnover
through decades, possibly centuries, with steel-reinforced columns, thick flat slab con-
struction, and floors built as plates. The engineers ensured that the floors could withstand
250 pounds per square foot, withstanding the weight of heavy industrial equipment. Only
now, this structural robustness supports server racks (2000 lbs each) rather than the print-
ing presses and other equipment that populated old space.
The main entrance’s two-story arched doorway depicts, in relief, an indigenous prairie per-
son in stereotypical ‘traditional’ garb and a white frontiersman. It reminds us that the Cloud,
too, is an accomplice to the theft of indigenous land. Rather than a twenty-first century
novelty, this node of Cloud, wrought from the same red brick of the factories that preceded
it, speaks to continuity rather than representing something radical, or even something
new. We stroll past the terracotta shields of past industrial workers that adorn the lime-
86 THEORY ON DEMAND
stone-trimmed walls. We see, in their severe expressions, echoes of Tony and Noah’s faces
as they rummage through cables or pry open floor tiles to feed more air to hungry servers.
Are they so different, we wonder, from these turn-of-the-century factory workers from
Chicago’s industrial heyday?
Is the Cloud, as Nathan Ensmenger suggests, nothing more than an informatic ‘factory’?3
These thoughts fade behind us as we make our way back to the train, returning to the
universities where we work, borne up by an infrastructure as old as the factories that have
become data centers. Maybe the Cloud isn’t a factory. Maybe it’s more like a train station,
a node in a network of connections that enable the rapid movement of people, things, and
stowaways like the rats and pigeons we found in the data center, the ‘vermin’ that have for
centuries hijacked human infrastructures like ships to propagate themselves. Like a train,
the Cloud breaks down. Though far less visible, weather-related delays or service disrup-
tions, the expansion of connections to new nodes or networks, still impact these centers.
Like train workers, the Cloud’s technicians must overcome a thousand catastrophes, but
unlike displaced travelers who cannot reach their destination in a timely fashion, most
of the Cloud’s users are never alerted to its everyday breaks and ruptures. The Russian
doll of redundancies that Tony and Noah described assure that our everyday experiences
online are insulated from the behind-the-scenes ruptures and catastrophes that constantly
threaten the Cloud.
Susan Leigh Starr famously observed that when working properly, infrastructure is
seldom visible.4 Like the air we breathe, it retreats into the background, bursting into the
foreground of our awareness only when it fails us. Has the Cloud proved to be so elusive
because its visible failures are so infrequent? As we type up field notes on our laptops
hours later in a shared cloud doc, we wonder about the failures that might occur behind
the scenes, and all the nested measures enlisted to ensure that we never experience the
failures that are endemic to data centers. We think of Tony and Noah as we comment
and copy and paste and spell check. Far from automated, far from futuristic and science
fictional, the Cloud is more of the same, albeit obscured from view, by the metaphor of its
name as much as the security apparatus designed to prevent most of us from ever getting
a glimpse of its inner workings.
As ethnographers largely focused on the internet as a site of inquiry about society and
culture, we’re compelled to think about the stories told through promises made manifest
in technological and infrastructural forms. Specifically, on this tour, we wondered—and
uncovered—what the materiality of a promise like ‘convenience’ might look like. We could
be at once impressed with The Cloud as a feat of engineering, and highly skeptical of it as
anything more than a vehicle for scaled-up, ambiguated capitalist exploits. Our takeaway
is that the data center—and all it encompasses—is an illusion that is hard to uphold, an
infrastructure requiring immense human energy and labor to maintain. All of this becomes
exponentially more difficult, and more problematic, on a struggling planet, drained by the
promises of The Cloud.
References
Bridle, James. Ways of Being: Animals, Plants, Machines: The Search for a Planetary Intelligence,New
York: Farrar, Straus and Giroux, 2022.
3. Nathan Ensmenger, ‘The Cloud is a Factory’, in Thomas Mullaney et al. (eds) Your Computer is on Fire,
Boston: MIT Press, 2021, pp. 29-50.
4. Susan Leigh Starr, ‘The ethnography of infrastructure’, American behavioral scientist 43.3 (1999):
377-391.
87
IN/CONVENIENCE
Ensmenger, Nathan. ‘The Cloud is a Factory’, in Thomas S. Mullaney, Benjamin Peters, Mar Hicks,
and Kavita Philip (eds) Your Computer is on Fire, Boston: MIT Press, 2021, pp. 29-50.
Moro, Jeffrey. Atmospheric media: Computation and the environmental imagination, PhD diss., Univer-
sity of Maryland, College Park, 2022.
O'Gieblyn, Meghan. God, Human, Animal, Machine: Technology, Metaphor, and the Search for Mean-
ing, London: Doubleday, 2021.
Pickren, Graham. ‘‘The global assemblage of digital flow’: Critical Data Studies and the Infrastructures
of Computing’, Progress in Human Geography 42.2 (2016): 1-19.
Pipkin, Everest. It was Raining in the Data Center, PhD diss., Carnegie Mellon University, Pittsburgh,
2018.
Starr, Susan Leigh. ‘The ethnography of infrastructure’, American behavioral scientist 43.3 (1999):
377-391.
Velkova, Julia. ‘ Data Centers as Impermanent Infrastructures’, Culture Machine 18: The Nature
of Data Centers (2019), https://culturemachine.net/vol-18-the-nature-of-data-centers/data-cen-
ters-as-impermanent/.
88 THEORY ON DEMAND
WAITING FOR ‘DAY ZERO’
TUNG-HUI HU
In January and February 2018, during a third year of severe drought, I was living in Cape
Town, one of the largest modern cities to come close to running out of water.1 They called
it ‘Day Zero’, the day that taps would stop, although technically it referred to the date when
dams would drop below a crucial threshold of 13.5% full. If Day Zero occurred, the plan was
to ration water to 200 water collection points in the city. Farmers half-volunteered their water
reserves and were half-forced to surrender them. Stores quickly sold out of bottled water. The
rich, as I saw in my walks up Signal Hill, predictably filled their rooftop pools, sometimes by
paying water to be trucked in from elsewhere. Others queued at springs with as many plastic
containers as they could carry, sometimes for hours at a time. We were limited to 50 liters a
person per day from the taps, but that was luxurious compared to the threat of 25 liters if Day
Zero occurred. (By comparison, the average American consumes 372 liters per person per
day.) The TV blared with tips on how to flush as rarely as possible, and how to shower within
90 seconds; I remember a playlist of songs that lasted precisely 90 seconds. Paper plates
and cups replaced ceramic at chic restaurants; sanitizer was provided instead of soap. The
flies settled into our bathroom, and the city would cut water pressure at random times, as if
the tap had a mind of its own.
In the time leading up to Day Zero, Cape Town’s water infrastructure had become hypervisible.
Websites for tracking dam levels and expected Day Zero dates were popular, as were google
searches for water tanks. Hedley Twidle noted that ‘reservoirs… have been photographed
from every available angle, surveyed by drones and helicopters, snapped by passing motor-
ists who pulled onto the verge to watch the uncanny spectacle of dust clouds roiling across
Theewaterskloof, the city’s main water reserve’.2
Day Zero is a textbook case for Susan Leigh Star’s much-cited idea that we know infrastructure
only when it breaks.3 And knowledge of infrastructural failure assuredly changed some of the
population’s relationship with water. Indeed, with a dramatic drop in water consumption, Day
Zero was postponed, at first by a week or two, then a month, then, finally, to the rainy season.
Cape Town had made it through, and this success is often described as a triumph: case stud-
ies laud the high dam levels now, the determination and grit of its citizens, and new sources
of water, suggesting that Day Zero is in the rearview mirror. The word most often applied is
resilience: both climate resilience and the behavioral adaptations that city planners spurred
in the population.
1 When São Paulo was down to 40 days of water in 2015, it still had sufficient water in the dams, just not
the pipes to deliver it.
2 Hedley Twidle, ‘Shadow of a Drought: Notes from Cape Town’s Water Crisis’, Interventions 24.3 (2022):
371.
3 Susan Leigh Star, ‘The Ethnography of Infrastructure’, American Behavioral Scientist 43.3 (November/
December 1999): 377-391.
89
IN/CONVENIENCE
But I am left with questions: what did we really learn from this sudden rupture, from a year
of those people with the privilege of having good plumbing suddenly thinking about their
plumbing? Did the threat of sending middle-class inhabitants to communal access points, like
those access points used by the other half of the population—the two million living in informal
settlements, for whom Day Zero water rationing was simply an everyday reality—bring the
city closer together? Did the sight of queueing at springs like Newlands, which, when I went
by, seemed racially and ethnically mixed, cause a slight shift in the sense of the commons, in
the sense of an affective ground between people? Or did it—given that fights also occurred
in those queues, and the fact that police quickly shut down those informal systems—simply
lay bare the divisions that were already there and were already known? What more can we
say about infrastructure in a city that is a textbook example of the infrastructural splitting and
fracturing of apartheid? Some people I spoke with were often fiercely proud of the infrastruc-
ture in Cape Town—to be sure, it was often better than where I live—echoing the political
line of some Afrikaner-first parties which suggested that good roads were the positive legacy
of Afrikaner rule in the postwar era.
The hyper-attention to tap water and dam levels also pushed many other problems to the
side. The Day Zero crisis and the years following, Suraya Scheba and Nate Millington write,
may have produced new possibilities but ‘these new possibilities serve to entrench existing
unequal socio-natural conditions. While we are witnessing an apparently radical reconfigu-
ration in the deployment of techno-managerial instrumentation in crisis conditions, these in
turn serve an existing path dependency predicated on inequality and uneven access’.4 As
evidence, they point out that South Africa has guaranteed a Basic Water Right in its consti-
tution since 2001. However, under the new changes Cape Town instituted post-crisis, an
indigent household had to now ‘“prove its poverty” through extensive documentation’ to
receive this water. What’s worse, that guaranteed supply initially came through ‘smart’ water
management devices, which cut off water flow after their free Basic Water Right ration of 350
liters/day. These water management devices were often used to tie households to regimes
of debt and failed to account for multiple households in the same plot. As one resident told a
researcher, ‘Daily water supply is re-started at 0500 each morning, and on plots with multiple
households residents queue to fill buckets from the external tap until the water expires (often
before 0530)’.5 The city eventually withdrew the devices after a firestorm of protest.
These poor connections to city sewer and water in the settlements, and the distrust those
residents have with the city government, have led residents to tap sewer lines and vandalize
water connections. Steel pipes and taps are sold to junkyards, and so are the power supply
generators that power the pumps. There is often open conflict between the utility and its
4 Nate Millington and Suraya Scheba, ‘Crisis Temporalities: Intersections Between Infrastructure and
Inequality in the Cape Town Water Crisis’, International Journal of Urban and Regional Research,
Spotlight on: ‘Parched Cities, Parched Citizens’ (December 2018), https://www.ijurr.org /spotlight-on/
parched-cities-parched-citizens/crisis-temporalities-intersections-between-infrastructure-and-
inequality-in-the-cape-town-water-crisis/.
5 Charlotte Lemanski, ‘Infrastructural citizenship: The everyday citizenships of adapting and/or destroying
public infrastructure in Cape Town, South Africa’, Transactions of the Institute of British Geographers 45
(2020): 599.
90 THEORY ON DEMAND
customers: ‘from hijackings, armed robberies, to the stoning of water and sanitation vehi-
cles’.6 Charlotte Lemanski writes that ‘citizens essentially “see” the state in infrastructure, and
therefore interpret their everyday encounters with infrastructure as a representation of their
relationship with the state’.7 Is the state transparent, an embedded feeling of convenience or a
source of pride? Or is it opaque, and a source of hostility? If you’re poor, the state is everywhere.
It is commonly now accepted that, in Twidle’s words, that Day Zero was ‘a social and political
fiction… [but] a necessary fiction’ to force change in water consumption.8 But calling out the
ways that private companies profited from the crisis—from private desalinization companies
bidding for contracts, to water giveaways to the beer conglomerate SA Breweries—the water
activist group Water Crisis Coalition has suggested that Day Zero was an event manufactured
for big business to profit, further suggesting the disconnect between the state and its population.
Making water’s infrastructures and its costs visible has thus contributed to the financializa-
tion of water; as Scheba and Millington write, drought resilience ‘is deeply intertwined with…
financial resilience’.9 For after water consumption was cut in half, this left the water utility in
a precarious financial state, causing water tariffs to double after Day Zero. Because water
infrastructure is paid for by usage, this creates a paradoxical incentive to both conserve but
also to use at the same time. A Facebook user wrote on a local water group about this para-
dox: ‘We need to consume more in times of relative plenty (and stabilize City of Cape Town
water revenues) otherwise we will simply keep using less and less while having further levies
introduced’.10 The poster suggests the continuous cycle of crisis that this bizarre system of
financialization produces: in the name of climate (i.e. fiscal) resilience, it asks consumers to
use more water.
Marc Steinberg and Joshua Neves have written that ‘convenience stops being a demand
consumers place on platforms… and becomes instead a demand on consumers placed by
platforms’.11 While they are talking about digital platforms such as Netflix, their insight is all
too applicable here: the convenience of water infrastructure, the easy flow of water from a
tap, is something pushed on water users. The infrastructure demands to be used—so that
it can be paid for.
6 Nomalanga Tshuma, ‘R5k reward for info on theft, vandalism of CoCT's water and sanitation
infrastructure’, Cape Argus, 10 January 2022, https://www.iol.co.za/capeargus/news/r5k-reward-for-
info-on-theft-vandalism-of-cocts-water-and-sanitation-infrastructure-84f3b73d-3408-4df0-ad87-
af3e81f545f2.
7 Lemanski, ‘Infrastructural citizenship’, 602.
8 Twidle, ‘Shadow of a Drought’, 372.
9 Nate Millington and Suraya Scheba, ‘Day Zero and The Infrastructures of Climate Change: Water
Governance, Inequality, and Infrastructural Politics in Cape Town’s Water Crisis’, International Journal of
Urban and Regional Research 45.1 (2021): 127.
10 User from Facebook Watershedding group, December 2018, as quoted by Millington and Scheba, 127.
11 Marc Steinberg and Joshua Neves, ‘Introduction: In/Convenience’, in this volume.
91
IN/CONVENIENCE
Let me describe my own relationship to water. I live in the US state of Michigan, next to a
fifth of the world’s freshwater in the form of the Great Lakes; one popular motto on Michigan
license plates is ‘winter – water wonderland’. And yet I remember the bitter irony of the water
situation in the majority-Black city of Detroit, which declared bankruptcy in 2013; $5.7 billion,
or one third of the debt, was attributed to its water department. I remember reading about the
sudden visits of demolition company trucks to shut off the plumbing for those most indebted
by their water bills, and about one city council member who suggested that residents drink
from the river if they really needed water. As Peter Hammer, director of the Damon J. Keith
Center for Human Rights at Detroit’s Wayne State University, comments, ‘They are also shut-
ting water off not wishing people will pay necessarily, but implicitly hoping people will move’.12
I remember, too, the lead leaching from pipes in Flint the following year. It was caused, in
large part, by the emergency financial managers that had placed the city in receivership;
as a result, ultimate authority in Flint lay not with the elected mayor but with the Michigan
Department of Treasury. At one point, half of the Black population in the state was governed
by an emergency financial manager,13 suggesting the brutal link that right-wing governance
makes between Blackness and ‘the emergency,’ and the sleight of hand that declaring a crisis
or emergency does to disavow structural racism. Though water in Southeast Michigan and
Cape Town have very different histories and contexts, the memory of residents snapping up
bottled water came rushing back to life as I stood in the supermarkets of Cape Town.
The fantasy of water—as flow, as liquidity, even as territory outside of the rule of law—tracks
closely with the fantasies of global finance capital. Yet it is too reductive to understand
water through finance alone. To be sure, freshwater’s seeming fungibility is a function of the
abstracting qualities of finance, but ask residents at Flint about the peculiar smell and metallic
taste, the brown color, of corrosive Flint River water, and one quickly realizes that water is
intertwined with pipes, and pipes with the state. Chandra Mukerji has written about the logis-
tics of water in the 17th century France, in which the king’s Canal du Midi ‘entered politics in
an impersonal way. It was an agent of a state that could not be killed, but still had enormous
influence over local life’.14 The water in Flint, then, represents a second-order displacement
of power, where the state itself is seen by followers of the Austrian school of economics as
the beast to be starved or ‘killed’, the solid to be made mobile and liquid,15 and where the
impersonality of state governance gives way to the impersonality of financial management.
‘They shut off water… hoping people will move’.
I am wondering if water has its own political sense. As this volume asks us to think about con-
venience, I am reminded that the word itself comes from the sense of convening; convenience
12 Peter Hammer, as quoted by Rose Hackman, ‘What Happens When Detroit Shuts Off the Water of
100,000 People’, The Atlantic, 17 July 2014, https://www.theatlantic.com/business/archive/2014/07/
what-happens-when-detroit-shuts-off-the-water-of-100000-people/374548/.
13 Edward Helderop, Elizabeth Mack and Tony H. Grubesic, ‘Exploring the invisible water insecurity of
water utility shutoffs in Detroit, Michigan’, GeoJournal 88 (2023): 4185.
14 Chandra Mukerji, ‘The Agency of Water and the Canal du Midi’, in Kim De Wolff, Rina C. Faletti, Ignacio
López-Calvo (eds) Hydrohumanities, Oakland, CA: University of California Press, 2022, p. 28.
15 Jason Adams, ‘The Liquid State’, Critical Inquiry blog, 1 February 2016, https://critinq.wordpress.
com/2016/02/.
92 THEORY ON DEMAND
used to mean a meeting or agreement (convenientia), a coming together. Foucault uses it to
describe how similar things were brought together in the 16th century to form the great chain
of being: ‘Those things are “convenient” which come sufficiently close to one another to be in
juxtaposition’.16 Today we see convenience in the near merger of water and metal pipe into the
same entity: tap stands for water and vice versa. What this tells us is that when we say water,
we are really talking about its interface effects. Water is the content of the pipes, but water is
also the medium (the pipes) that transports the fluid and the systems of power that sustain
that medium. It is the ability to bring things so closely together that they seem to merge into
one, liquid, thing: ‘water’. Water convenes not just physical infrastructure but also operates
as the place where finance and the state meet, where humans and nonhumans come to an
agreement, and where the individual and the public meet. The convening does not happen
once, but continuously, over and over. This is a better description of infrastructure: not as
working and then broken, broken and then fixed, but as a continual requirement to assemble
things, neighbor with neighbor, state next to debtholder, one future with another.
Water, then, becomes a question of how to convene. As Andrea Ballestero points out in her
ethnography of water, regulatory authorities at Costa Rica’s ARESEP laboriously adjust the
equations of cost to balance a mandate to deliver water affordably with a mandate to produce
a surplus.17 To calculate the cost of living, they must tie water to the shifting basket of com-
modities for an average, ‘unmarked’ consumer household, indexing water to an assembly of
objects such as screwdrivers, video games, pork ribs, and—recursively—bottled water. But
such technical measures often formalize certain practices or qualities at the expense of others.
Other practices, such as tapping a power line or distribution box in a refugee camp or pulling
a self-made spur from formal infrastructure,18 or paying $30 for a plumber to turn the taps
that the city of Detroit has shut off back on, exceed what the state can see as infrastructure,
even as they are themselves ways of convening.
When I say convene, then, I mean more than a neutral, Latourian assemblage of things. I mean
to invoke the relation of water that binds someone, if not to the state, then to the imaginaries
of infrastructure. The state is busily cutting deals to supply water or power to data centers,
semiconductor manufacturing, and other new economy corporations; it is promoting the right-
wing market capture of the state because, in the words of Thatcher, ‘the object is to change
the soul’. But another kind of public is being made at the same time, built out of living within
dispossession, even out of indifference to a state which has abandoned them. For ‘Southern-
ness’ transcends its hemispheric geography and shows up in even the northernmost states,
such as Michigan.19 The legacy of Southernness is a testbed for new methods of accumulation
16 Michel Foucault, The Order of Things: An Archaeology of the Human Sciences, New York: Vintage, 1994,
p. 18.
17 Andrea Ballestero, A Future History of Water, Durham, NC: Duke University Press, 2019.
18 Angela D. Storey, ‘Implicit or illicit? Self-made infrastructure, household waters, and the materiality of
belonging in Cape Town’, Water Alternatives 14:1 (2021).
19 This thought, and others in this paragraph, come out of the collective thinking produced at the Southern
Urbanities workshop, British Academy, London, 15-16 September 2023; my fellow travelers were
Bilgin Ayata, Laura Guimarães Corrêa, Rodrigo Firmino, Rafael Grohmann, Ali Karimi, Jovan Scott Lewis,
Bingchun Meng, Nancy Odendaal, Bhaskar Sarkar, and AbdouMaliq Simone. My gratitude to have had
93
IN/CONVENIENCE
through dispossession. This is the case in the right-wing Mackinac Center for Public Policy’s
experiments with governance, advocating for the then-novel idea of replacing elected city
governments with emergency financial managers,20 or the always-looming proposals to sell
off the Detroit water system to a private company. But it is also a testbed for new forms of
convening. I think of those engaged in direct action, advocacy, and mutual aid, as with the
work of the Detroit Water Brigade or the Peoples’ Water Board, as well as the work by the
Canadian advocacy group Blue Planet Project to invite UN Special Rapporteurs to visit Detroit,
thereby internationalizing its case. If water insecurity is embedded in translocal structures,
tracing these linkages—convening across continents—might be a necessary first step.
*
And so I am suspicious of the narrative of crisis through which Day Zero is framed. To be sure,
when I made trips outside of Cape Town, I remember how good it felt to experience water
again—luxuriating, for example, in the water pressure in a trip to Durban; moments, one could
say, where I was re-experiencing the joys of convenience. Was that the end goal of this crisis?
In a life that is endlessly elongated by waiting and standstill, in a time of lethargy, we might
begin to consider the racialized temporality of nothing changing, whether because of a ‘his-
torical stillness’ (Hortense Spillers) or simply the period of indefinite deferral by which conve-
nience—in the form of modernization or development—is promised but recedes, continually.21
After all, the idea of Day Zero as a stopping or a disruption or a crisis was as open to interpre-
tation as any other temporal event. As I have said, ‘Day Zero’ was already a living reality for
half of the city. The day of Day Zero itself—calculated simply as current date plus water in
reservoirs over 13.5% divided by daily consumption—was constantly changing. First it would
be a month, then, after another week of no rain, a couple of weeks, then, after conservation,
a month again. We watched the sky and the weather, wondering if it would ever rain, but that
was the point, wasn’t it: we were watching weather and, if we were thinking about it, climate
change, the cycle that has made it 100 times more likely that this supposedly one-in-a-thou-
sand-year event would happen again. The temporality of the water outage was postponed,
then closer, then postponed again, a feeling of crisis endlessly stretched out, turning into
something permanent if seasonal in its characteristics.
There was a crisis, and yet there was not. Western Cape’s premier Helen Zille described the
efforts ‘to prevent anarchy’ should Day Zero happen, and newspapers ran with it, describing
an apocalyptic ‘Mad Max scenario’ of resource wars (there’s that word again: apocalypse,
from the Greek, to uncover, reveal). Congregations led mass prayer services for rain, and
even the water department’s website asked for citizens to pray for rain on the weekend of 9
February 2018; tens of thousands of evangelical Christians—some estimates say 250,000—
converged in Mitchell Plain, in a service memorialized a year later. And yet these mass prayers
such brilliant colleagues to think with.
20 Adams, ‘The Liquid State’.
21 Hortense Spillers, Black, White, and in Color: Essays on American Literature and Culture, Chicago:
University of Chicago Press, 2003.
94 THEORY ON DEMAND
suggest how its participants saw the ‘crisis’ in some sense as outside of the framework of the
human. They wanted something outside of the calendrical and calculated timeframe of the
planners; they wanted something outside of the time of the ordinary.
The gathering at Mitchell Plain thus points us to alternative frameworks for considering water.
The UN estimates that in the Western Cape region, water run-off will decline by 13% by 2050
(even as demand for irrigated water is projected to increase by 6.4%) because of climate
change.22 This decline is neither temporary and nor is it reversible on a human timescale. The
sheer magnitude of this decline is hard to fathom, and it presents a methodological prob-
lem: Does it still make sense to consider water alongside digital networks and other forms of
infrastructural media, when water has traditionally represented the idealized limitlessness
that digital media typically tries to mimic through its flows, streaming, and liquidity? And
though the cost of water is often indexed to the Consumer Price Index basket of consumer
prices—Costa Rica, to take Ballestero’s example, includes many modern-day conveniences,
such as the cost of internet service, cable TV, and a movie rental—can water be said to be a
convenience in the same way?
While common sense might hold that water is unlike digital media in that the former is both
singular and a necessity for human life, media scholars have shown that the two are often
interlinked. Water carriers in Zambia, Lisa Parks has argued, are as much part of the internet
infrastructure as the satellite internet systems that physically carry the data packets, because
the water carriers bring life to, and thereby make possible, the schoolhouse where the inter-
net is accessed.23 In much the same way, the hyperscale data centers of today guzzle water
but pay lower prices than humans. Some scientists estimate AI may remove between 4.2 to
6.6 billion liters of water per year by 2027;24 in this way, artificial life competes with and is
invested in the depletion of human life. Thus, water’s differences from digital media are less
important than the new questions that can arise from thinking the two in conjunction. Rather
than scaling one smoothly into another, however, we should instead cast a critical eye on the
frameworks that are increasingly bringing the two closer together.
Take the question of publicness. Catherine Fennell has shown that our tendency to univer-
salize water as opposed to other things considered ‘private’ is itself worthy of analysis: why
is water uniquely ‘public’, she asks, while other forms of infrastructure, such as housing, are
not?25 Mapping the ‘kinds of risks that a far-flung group of citizens can recognize as shared’,
Fennell continues, ‘and thus worthy of collective concern and action’, we might begin to ask
22 UNU-WIDER, Potential Impacts of Climate Change on National Water Supply in South Africa, WIDER
Research Brief, Vol. 2016 Issue 3 (Helsinki: UNU-WIDER, November 2016), https://www.wider.unu.
edu/publication/potential-impacts-climate-change-national-water-supply-south-africa.
23 Lisa Parks, ‘Water, Energy, Access: Materializing the Internet in Rural Zambia’, in Lisa Parks and Nicole
Starosielski (eds) Signal Traffic: Critical Studies of Media Infrastructures, Champaign, IL: University of
Illinois Press, 2016, pp. 115-136.
24 Pengfei Li et al., ‘Making AI Less “Thirsty”: Uncovering and Addressing the Secret Water Footprint of AI
Models’, preprint v1, arXiv, 6 April 2023, https://arxiv.org/pdf/2304.03271.pdf
25 Catherine Fennell, ‘Are We All Flint?’, Limn, special issue: ‘Public Infrastructures/Infrastructural Publics’,
July 2016, https://limn.it/articles/are-we-all-flint/.
95
IN/CONVENIENCE
what other things are seen differently, as a matter for the private sector. Digital platforms
and other matters of personal convenience are often inevitably seen through the lens of tech
companies and private investment, rather than public goods. But what would it mean if they
were publicly maintained and provided?
Reciprocally, the framing of drought as a matter of inconvenience (however serious) has
gone hand in hand with the privatization of water. The disappointment of Day Zero is how
much work went into quick fixes (smart meters, a new app, drones over the reservoirs) that
would have an immediate payoff rather than the work of repairing past harms and rethinking
growth in the context of a permanent decline of water supplies. If there is a lesson here, it
is surely about how urgency itself, when wedded to neoliberal demands, can exacerbate
or crystallize the slow violence of environmental harm and racial capitalism. One public is
assembled out of crisis. But if we are to find another temporality for thinking infrastructure, it
must start from those missing from that assembly. It must start from those who are already
decoupled from urgency.
Day Zero is often seen as a harbinger of the future for other cities, but as Akhil Gupta argues,
infrastructure is better thought of in terms of a future in ruins, a suspension of the future that
was promised at its beginning. Infrastructure inhabits a process of ruination even before it is
built (contractors might have used shoddy material or bribed a government official), and even
after it is built (it is continually under repair). As Gupta writes, ‘Beginning from movement
requires a completely different optic, making us think of periods when infrastructures are “at
rest” as not the normal condition, but as something to be explained’.26 This is a much better
way of understanding water in Cape Town, or elsewhere: not breakage as a disturbance of an
infrastructure ‘at rest’, or a revelation of what infrastructure has been, but rather breakage and
ruination as the ordinary state of things, with the moments of solidity just a social and political
fiction to stabilize it. When did we ever really have that ‘convenience’ that the municipal water
supply promises? And which ideologies (Afrikaner exceptionalism, post-crisis triumphalism)
are allied to it? That the crisis was photographed and visible so often was simply an attempt
to freeze the idea of water usage into place.
At the same time, suspension is also a feeling, a convening. In the mass prayers, in the protests
against the city, in the informal economy of water-haulers, in the fleeting solidarities between
strangers, in hashtags and buses and in the kitchen sinks and bathrooms, there was some-
thing in the air between winter 2017 and summer 2018: a feeling of time suspended. Everyone
was waiting. Lauren Berlant asks: ‘What is in the air to make new genres of convergence?’27
That the net result of Day Zero, a few years on, was arguably to cement further inequalities
does not take away from the fact that social form was shifting and was itself in-progress. Ber-
lant asks us to attend to the affective infrastructure behind events such as Day Zero, even
26 Akhil Gupta, ‘The Future in Ruins: Thoughts on the Temporality of Infrastructure’, in Nikhil Anand, Akhil
Gupta, and Hannah Appel (eds) The Promise of Infrastructure, Durham, NC: Duke University Press,
2018, pp. 73-74.
27 Lauren Berlant, ‘The commons: Infrastructures for troubling times’, Environment and Planning D:
Society and Space 34:3 (2016): 412.
96 THEORY ON DEMAND
if that feeling of ambivalence and worry and fear and resentment has receded, for most of
those residents looking back on it, into something more akin to everyday life. Inconvenience
again seems normal, ordinary, part of the banal structure of convenience: a pumped-up bill;
a sign asking you to save water.
Suspension is the temporality of a feeling that I termed lethargy elsewhere, for the sense that
one is acted upon, rather than the agent of action.28 Lethargic time is something to be wasted
or passed, rather than to be marshalled or saved; it often wears away at the self, rather than
being directed outward. To invoke Stephen Best, this ‘violence “turned inward”’ can offer
some ‘fleeting relief from the pressure to endorse what Kant calls the world “as is.”’29 This
offers a partial explanation of why some residents attacked their own water and sewer infra-
structure: not just because that infrastructure was the embodiment of state disinvestment
and abandonment—or to put it in plain terms, it’s shitty and doesn’t work very well—but
also because, to cite Nikhil Anand’s work on ‘hydraulic citizenship’, infrastructure is also a
metonym for one’s own sense of belonging to a place.30 The tearing away of these forms of
connection was their own form of registering discontent with the world ‘as is’. Paradoxically,
this tearing away was their own form of belonging in dispossession.
*
I began writing this essay skeptical of a certain scholarly (but also critical and journalistic)
move that I found exhausting in digital media: the contrived drama of revealing the hidden
interior of a digital system, which functions by first casting infrastructure as secret, forbidden,
or otherwise ‘black boxed’. It’s a move that plays on the heroism of the critic or journalist, who
parachutes in (often to countries in the global South) to find the hidden source of a system.
Today this revelation is often ‘hidden’ laborers, such as an AI worker or a content moderator.
However well-meaning, in its worst iterations, these stories become an anodyne plea for
empathy, or for a form of white saviorism that implies that only those from the global North
have agency. The cast of the story, this framework implies, can only be found in the North,
whether tech regulators, academics, critics, or the familiar villains of Silicon Valley. To be blunt,
the problems of digital platforms and outsourced work are well-known to those workers; the
revealing is for the benefit of privileged audiences elsewhere, not for them.
When the idea of revelation in infrastructure studies relies on so-called ‘glitches’ or ‘breaks’
or temporary inconveniences to unmask it (think of Al Gore’s climate film The Inconvenient
Truth, which links truth and unconcealment to a temporary but necessary moment of inconve-
nience), it ironically bolsters the monolithic nature of the system. This sort of theory, as I have
been arguing, both assumes the glitch is a temporary event and assumes that infrastructure
itself is a static thing that works, rather than always in ruins. As a result, we come to believe
28 Tung-Hui Hu. Digital Lethargy: Dispatches from an Age of Disconnection, Cambridge, MA: MIT Press,
2022.
29 Stephen Best, None Like Us: Blackness, Belonging, Aesthetic Life, Durham, NC: Duke University Press,
2018, p. 26.
30 Nikhil Anand, Hydraulic City: Water and the Infrastructures of Citizenship in Mumbai, Durham, NC: Duke
University Press, 2017.
97
IN/CONVENIENCE
that infrastructures are largely fixed in place and unable to be fundamentally changed or
redesigned. In turn, they become inevitable, and thereby invisible. These inconveniences,
disavowed as temporary, make infrastructure convenient: at hand when needed, receding
when not. The search for the glitch is tied to, and complicit with, the need to produce the
‘crisis’. Instead, breakdown and suspension are periodic and even continual processes; we
are, in Berlant’s words, ‘always going through a phase’.31
That this feeling of things falling apart has become general, rather than localized, suggests
that we now may need a media theory that revolves around scarcity and decline, around the
ruins of infrastructure that is already obvious. What used to be invisible is quite visible; the
background has already inverted into foreground. In the West, the news stories today are
over consumer goods delays from container ships stuck, spectacularly, in the Suez Canal
and, more recently, in the Chesapeake Bay (coincidentally from the same company), around
broken supply chains and laggy networks. We need a media theory that is less focused on
revelation, that gazes less at the future or at an imagined Day Zero than at the brokenness
of the present.
At the same time, I do not mean to diminish the fundamental desire to understand infra-
structure. The difference, in my mind, is that such knowledge should never be a quick win,
a sudden reversal from invisible to visibility; it lies in the unspectacular and more mundane
acts—such as recalculating the water utility’s index for cost of living—that constitute how
we live with and adjust to today’s technical systems. In a recent workshop I attended in Ber-
lin, Lachlan Kermode expressed disappointment about how artificial intelligence is seen as
an impenetrable black box; what if, he argues, we understood AI as a system more akin to
government?32 We would not shy away from the complexity of an AI model any more than we
should shy away from tracing the passage of a bill through the legislature; indeed, democracy
counts on it.
Infrastructure is, like government, a plural noun, a convening. What makes infrastructure
possible include explicit mediators, such as the financial equations that work to link end-user
and water supplier. What we should also include in this conversation, however, are the many
other forms of assembly. Angela D. Storey has written of the implicit infrastructure that snake
across the Khayelitsha township in Cape Town, the pulled lines of plastic pipe for water access
and sand sinks for wastewater disposal. This self-made infrastructure confuses our normal
distinction between fixed and temporary, formal and informal: as Storey tells it, at one point the
‘official’ tap ran dry, but city workers, realizing it would take too much work to fix it, treated the
pulled tap as a de facto part of the formal city water system. ‘Thus enrolled,’ Storey comments,
for some residents it ‘had become a city tap, made real through the thinking, explanations, and
everyday experiences of residents’.33 I have seen similar ingenuity on Reddit forums, of a ‘smart’
31 Berlant, ‘The commons’, 414.
32 Lachlan Kermode, conversation at ‘Models – Abstraction – Scale: Understanding Historical and
Societal Impacts of Artificial Intelligence’ workshop, American Academy in Berlin, Berlin, 5 December
2023.
33 Storey, ‘Implicit or illicit?’, 86.
98 THEORY ON DEMAND
water meter in Detroit bypassed through some home-made welding. Rather than producing
truth, the question becomes how these stories, explanations, and experiences ‘make real’ the
assembly between neighbors, in part by making the city (or state) into a fiction. They suggest
the extension of infrastructure not just into physical systems on the edge of the urban core
but into affects and imaginaries. A city is built on the sense of finitude and futurelessness and
hope and fiction that mix together: a contemporaneity. Heraclitus has suggested a city is like
kykeon, an ancient Greek drink of wine, cheese, and barley: an emulsion, it ‘must be stirred
up to avoid division’.34 A city separates into its component parts until placed in movement,
until—we could say—it is once again convened.
References
Adams, Jason. ‘The Liquid State’, Critical Inquiry blog, 1 February 2016, https://critinq.wordpress.
com/2016/02/.
Anand, Nikhil. Hydraulic City: Water and the Infrastructures of Citizenship in Mumbai, Durham, NC:
Duke University Press, 2017.
Ballestero, Andrea. A Future History of Water, Durham, NC: Duke University Press, 2019.
Berlant, Lauren. ‘The commons: Infrastructures for troubling times’, Environment and Planning D:
Society and Space 34:3 (2016): 393–419.
Best, Stephen. None Like Us: Blackness, Belonging, Aesthetic Life, Durham, NC: Duke University Press,
2018.
Fennell, Catherine. ‘Are We All Flint?’, Limn, special issue: ‘Public Infrastructures/Infrastructural Pub-
lics’, July 2016, https://limn.it/articles/are-we-all-flint/.
Foucault, Michel. The Order of Things: An Archaeology of the Human Sciences, New York: Vintage,
1994.
Gupta, Akhil. ‘The Future in Ruins: Thoughts on the Temporality of Infrastructure’, in Nikhil Anand,
Akhil Gupta, and Hannah Appel (eds) The Promise of Infrastructure, Durham, NC: Duke University
Press, 2018, pp. 62-79.
Hackman, Rose. ‘What Happens When Detroit Shuts Off the Water of 100,000 People’, The Atlantic, 17
July 2014, https://www.theatlantic.com/business/archive/2014/07/what-happens-when-detroit-shuts-
off-the-water-of-100000-people/374548/.
Helderop, Edward; Mack, Elizabeth; and Grubesic, Tony H. ‘Exploring the invisible water insecurity of
water utility shutoffs in Detroit, Michigan’, GeoJournal 88 (2023): 4175-4188.
Hu, Tung-Hui. Digital Lethargy: Dispatches from an Age of Disconnection, Cambridge, MA: MIT Press,
2022.
Lemanski, Charlotte. ‘Infrastructural citizenship: The everyday citizenships of adapting and/or destroy-
ing public infrastructure in Cape Town, South Africa’, Transactions of the Institute of British Geogra-
phers 45 (2020): 589-605.
Li, Pengfei; Yang, Jianyi; Islam, Mohammad A.; and Ren, Shaolei. ‘Making AI Less “Thirsty”: Uncover-
ing and Addressing the Secret Water Footprint of AI Models’, preprint v1, arXiv, 6 April 2023, https://
arxiv.org/pdf/2304.03271.pdf
34 Heraclitus, fragment DK B 125, as discussed in Nicole Loraux, The Divided City: On Memory and
Forgetting in Ancient Athens, New York: Zone Books, 2002, pp. 108-111.
99
IN/CONVENIENCE
Loraux, Nicole. The Divided City: On Memory and Forgetting in Ancient Athens, New York: Zone Books,
2002.
Millington, Nate and Scheba, Suraya. ‘Crisis Temporalities: Intersections Between Infrastructure
and Inequality in the Cape Town Water Crisis’, International Journal of Urban and Regional Research,
Spotlight on: ‘Parched Cities, Parched Citizens’ (December 2018), https://www.ijurr.org/spotlight-on/
parched-cities-parched-citizens/crisis-temporalities-intersections-between-infrastructure-and-in-
equality-in-the-cape-town-water-crisis/.
_____. ‘Day Zero and The Infrastructures of Climate Change: Water Governance, Inequality, and Infra-
structural Politics in Cape Town’s Water Crisis’, International Journal of Urban and Regional Research
45.1 (2021): 116-132.
Mukerji, Chandra. ‘The Agency of Water and the Canal du Midi’, in Kim De Wolff, Rina C. Faletti, Ignacio
López-Calvo (eds) Hydrohumanities, Oakland, CA: University of California Press, 2022, pp. 23-41.
Parks, Lisa. ‘Water, Energy, Access: Materializing the Internet in Rural Zambia’, in Lisa Parks and Nicole
Starosielski (eds) Signal Traffic: Critical Studies of Media Infrastructures, Champaign, IL: University of
Illinois Press, 2016, pp. 115-136.
Spillers, Hortense. Black, White, and in Color: Essays on American Literature and Culture, Chicago:
University of Chicago Press, 2003.
Star, Susan Leigh. ‘The Ethnography of Infrastructure’, American Behavioral Scientist 43.3 (November/
December 1999): 377-391.
Steinberg, Marc and Neves, Joshua. ‘Introduction: In/Convenience’, in this volume.
Storey, Angela D. ‘Implicit or illicit? Self-made infrastructure, household waters, and the materiality of
belonging in Cape Town’, Water Alternatives 14:1 (2021): 79-96.
Tshuma, Nomalanga. ‘R5k reward for info on theft, vandalism of CoCT's water and sanitation
infrastructure’, Cape Argus, 10 January 2022, https://www.iol.co.za/capeargus/news/r5k-reward-for-
info-on-theft-vandalism-of-cocts-water-and-sanitation-infrastructure-84f3b73d-3408-4df0-ad87-
af3e81f545f2.
Twidle, Hedley. ‘Shadow of a Drought: Notes from Cape Town’s Water Crisis’, Interventions 24.3 (2022):
369-373.
UNU-WIDER. Potential Impacts of Climate Change on National Water Supply in South Africa, WIDER
Research Brief, Vol. 2016 Issue 3 (Helsinki: UNU-WIDER, November 2016), https://www.wider.unu.
edu/publication/potential-impacts-climate-change-national-water-supply-south-africa.
100 THEORY ON DEMAND
PLATFORM ECONOMIES, REPUTATIONAL STAINS,
AND THE IN/CONVENIENCES OF PORN
SUSANNA PAASONEN
Throughout the history of pornography, intermediaries such as distributors and vendors have
reaped notable benefits from the work of performers and producers through hefty markups
in sales of commodities (prints, booklets, books, magazines, Super 8mm films, VHS tapes,
DVDs, movie tickets). The dynamics changed as performers and producers could directly
market goods on their own web sites. Around the mid-1990s, this merely meant mastering
the very basic HTML, even as it took more effort to monetize the content. Yet traffic there was,
leading to the development of online payment system providers (PSPs) that were to a large
extent—but not exclusively—coined for the needs of pornographers.1 The market dominance
of centralized webcam platforms and video aggregator sites during the past fifteen years or
so has marked another drastic shift in how sexual media is distributed and consumed, and
in how content creators are rewarded for their work.2 Centralized distribution platforms have
become a novel branch of intermediaries, the profit margins of which are often opaque. Mean-
while, their cultural visibility has made them targets of financial deplatforming. This line of
developments meets one where payment systems’ obscure terms of use meet cross-platform
techniques of surveillance ousting sex workers from membership in platform economies.3
The entanglements and co-developments of web hosting services, payment system providers,
and software development since the 1990s point to the factual impossibility of plying the porn
industry apart from the rest of the online economy. This has less to do with the permeability of
the industry’s boundaries than its very embeddedness in the development of the World Wide
Web that came about—not least since pornographers were among the first to recognize the
technology’s financial potential. Yet, within the contemporary, sexual content is energetically
ousted by infrastructural gatekeepers such as app stores, centralized social media platforms,
banks, credit card companies, and PSPs.
This chapter inquires after the position and perceived value of sex and its laborers within
contemporary platform economies through the analytical lens of in/convenience. It asks: for
1 Frederick S. Lane III, Obscene Profits: The Entrepreneurs of Pornography in the Cyber Age, New York:
Routledge, 2001; Lewis Perdue, EroticaBiz: How Sex Shaped the Internet, New York: Writers Club Press,
2002; Wendy Hui Kyong Chun, Control and Freedom: Power and Paranoia in the Age of Fiber Optics,
Cambridge, MA: The MIT Press, 2006.
2 Heather Berg, Porn Work: Sex, Labor, and Late Capitalism, Chapel Hill: The University of North
Carolina Press, 2021; Rachel Stuart, ‘Webcam Performers Resisting Social Harms: “You're on the Web
Masturbating… It's Just about Minimising the Footprint”’, International Journal of Gender, Sexuality and
Law 2.1 (2022): 171-198.
3 Lana Swartz, New Money: How Payment Became Social Media, New Haven: Yale University Press, 2020,
pp. 77-79, 96; Tara Siegel Bernard, ‘Sex Workers Have Been Shunned by Banks, Even When Their Work
Is Legal’, The New York Times, 18 November 2023, https://www.nytimes.com/2023/11/18/business/
sex-workers-bank-accounts.html.
101
IN/CONVENIENCE
whom centralized platforms are convenient and how, what vulnerabilities they yield, and what
it means for content creators to be cast as inconvenient enough to be demonetized. To do this,
I build on Lauren Berlant’s work on inconvenience as broadly descriptive of an unpleasant
sense of overcloseness with the world and the people within it as ‘the affective sense of the
familiar friction of being in relation’.4 Berlant frames inconveniences as both intensely felt
and as grounded in social dynamics preceding and enveloping the subject; as entangled
in relations of privilege wherein the subordinated are the ones to be cast as inconvenient.
At the same time, such relations are not merely ones of hierarchical control, given that the
subordinated experience much inconvenience in attachments and encounters where they
are not fully wanted, which can nonetheless also allow for their thriving. Inconvenient attach-
ments and relations are, for Berlant, both an ontological condition and steeped in ambiguity.
Enlivening as much as eating away at capacities to act, they are comforting and irritating,
beneficial and harmful, and certainly a constant.
Berlant’s questioning of inconveniences continues their long-term interest in how affective
attachments (and distances) give shape to sociality and community beyond that which is
seen to comprise the public sphere—the blurriness of the boundaries of intimate and public
spheres.5 The notion of inconvenience then foregrounds frictions and inequalities involved in
the mundane crafting of connections and disconnections that we cannot do without; exclusion
from which comes at a cost.
Applying the notion of inconvenience to the attachments and detachments between online
payment systems, sex work, and platform politics, I suggest, makes it possible to consider
the affective dynamics underpinning corporate policies—the deplatforming of sexual media
resulting from norms particular to US culture yet yielding global power.6 Following Berlant,
affective publics presume shared worldviews and sentiments that may not in fact exist.7 The
same can be said of the operations within the public sphere and capitalist logics, the decisions,
rulings, and policies of which are not simply led by straightforwardly optimized profits but
which also balance taste cultures, moral imaginaries, and reputation economies operating
in realms of affect.
As a dynamic, inconvenience remains contingent so that the perceived gains and damages
among the actors involved do not remain the same, even as their asymmetrical positions do:
for mundane inconveniences bolster systematic, historically construed practices of oppres-
sion.8 Taking on the question, ‘[w]hat price and what kinds of price are being paid in order to
4 Lauren Berlant, On the Inconvenience of Other People, Durham, NC: Duke University Press, 2022, p. 2.
5 Lauren Berlant, The Queen of America Goes to Washington City: Essays on Sex and Citizenship, Durham,
NC: Duke University, 1997; ‘Intimacy: A Special Issue’, Critical Inquiry 24.2 (1998): 281-288; The
Female Complaint: The Unfinished Business of Sentimentality in American Culture, Durham, NC: Duke
University Press, 2008.
6 Walter Kendrick, The Secret Museum: Pornography in Modern Culture, 2nd edition, Berkeley: University
of California Press, 1997, pp. 241–43.
7 Lauren Berlant and Jay Prosser, ‘Life Writing and Intimate Publics: A Conversation with Lauren Berlant’,
Biography 34.1 (2011).
8 Berlant, On the Inconvenience, p. 151.
102 THEORY ON DEMAND
live a life as other people’s inconvenient object?’,9 this chapter explores inconvenience as an
affective dynamic underpinning the presence of commercial sex within platform economies.
As a persistent dynamic, inconveniences associated with sexual media have notable tenacity
as felt and articulated friction embedded in relations of power wherein sex workers are cast
as less-than, even as their labor has been key to how online economies have developed. What
then, is at stake when such a dynamic, rooted in sexual politics and corporate policies particu-
lar to the US, becomes articulated as ethical norms impacting the infrastructural affordances
of mediated sex on a global scale? Spoiler: much.
Oh, very convenient!
First, the obvious utter convenience of sexual content in and for web economies. Throughout
the 1990s, before online gaming picked up and before Amazon ever broke even, porn—
along with eBay—formed an exception by catering content that users were willing to pay for.
Consequently, safe credit card processing systems, hosting services, and streaming video
technologies were first developed for the needs of porn sites.10 Wendy Chun argues that the
commercial success of pay-porn was instrumental to the rise of mainstream e-commerce in
demonstrating both people’s willingness to use credit cards online and the functionality and
safety of these payment options, enabling to the ‘dotcom’ boom of late 1990s.11 In a concrete
sense, porn sites were incubators for payment systems forming the infrastructural backbone
of online economies today.
The profits of pornography were, from the start, highly lucrative for Web hosting firms, yet
business collaboration with pornographers—at this point largely independent entrepreneurs,
established print and DVD brands being reluctant to move online—involved reputational
risk so that their role within the emergent economy was often downplayed despite it being
elementary to how things kept afloat and grew.12 The same logic of in/convenience extended
to payment infrastructures which, by classifying porn sites as high-risk customers, began
charging them higher processing fees.
Independent Sales Organizations (ISOs), third-party payment processing companies autho-
rized to handle merchant accounts for businesses, emerged as new intermediaries serving
‘as the middlemen for risk just as they serve as the middlemen between merchants and pay-
ment acquiring services’.13 As Lana Swartz explains, since the 1990s, we have witnessed a
shift toward payment intermediaries that are not only economic but equally circulatory as
communication infrastructure. The question, then, is, ‘who gets to control and profit from com-
munication infrastructure, who gets to access it and on what terms, what kind of traffic gets to
travel over it’—and, factually, who can get paid and what fields of trade are excluded.14 This has
9 Berlant, On the Inconvenience, p. 5, emphasis added.
10 E.g., Lane, Obscene Profits, p. 70; Perdue, EroticaBiz; Brian McNair, Porno? Chic! How Pornography
Changed the World and Made It a Better Place, London: Routledge, 2013, pp. 27–29.
11 Chun, Control and Freedom, pp. 78–79.
12 Perdue, EroticaBiz, p. 3.
13 Swartz, New Money, p. 87.
14 Swartz, New Money, p. 6.
103
IN/CONVENIENCE
involved the development of financial tech (as a subset of the Silicon Valley business sector)
partly building on the models of social media in person-to-person payment options. Within
this, charging high-risk customers higher processing fees has led to their flat-out exclusion
of from payment platforms. In order to identify such customers, PSPs use data analytics and
AI-driven prediction to troll through merchants’ social media accounts in order to evaluate
their viability and profitability.15 This extension of surveillance effectively ousts sex workers
with social media presence from payment systems and, in a very concrete sense, from mem-
bership in platformed transactional communities where they can make a living.16 A similar
logic of exclusion remains ubiquitous on social platforms where sex workers dependent on
networked visibility ‘are seen to sell their bodies and to advertise their services merely by
virtue of their […] presence’, rendering their accounts subject to removal.17
In a striking example of governmental intervention, the US Department of Justice’s 2013–17
Operation Choke Point classified porn sites, along with firearm dealers and payday lenders, as
potentially illegal businesses involving a high risk of money laundering and fraud, pressuring
banks to not merely charge higher fees but to discontinue their services without proof of
wrongdoing, thus circumventing law enforcement and courts in determining criminal viola-
tions.18 The operation was discontinued as it became evident that its underlying motivation
concerned the unsavory, rather than criminal, nature of the said businesses, the House Over-
sight and Government Reform Committee ‘releasing evidence that “federal regulators are
pressuring banks to terminate relationships with legal yet disfavored industries, without regard
to the legitimacy or risk profile of individual companies”’.19 The aim was to shut industries
down by cutting them off from payment infrastructures. This points to not just how the US
government allows for banks’ unequal treatment of commercial entities on the basis of taste
and value judgements, but also how they prefer to enforce this: ‘In the metaphor of Operation
Choke Point, money is like “air”: those who are denied it can be “choked off”’ (Swartz 2020,
80). This logic of exclusion and erasure prevails.
There is much opaqueness in how financial infrastructures make their decisions concerning
un/viable customers or calculate their degrees of riskiness. Following Frank Pasquale, this is
part and parcel of ‘blackbox society’ within which the operations of Wall Street and Silicon Val-
ley companies are hidden from public scrutiny even as data culture renders the lives of individ-
uals affected by their operations subject to increased surveillance. Such companies ‘organize
the world for us’ within ‘data-driven convenience’ that quickly flips into inconvenience when
services are withdrawn—as they very abruptly can be.20 Beyond the inconveniences articulat-
15 Swartz, New Money, pp. 78–79, 93.
16 Swartz, New Money, pp. 49, 82, 96.
17 Carolina Are and Susanna Paasonen, ‘Sex in the Shadows of Celebrity’, Porn Studies 8.4 (2021): 417.
18 Frank Keating, ‘Justice Puts Banks in a Choke Hold’, Wall Street Journal, 25 April 2014, https://www.wsj.
com/articles/frank-keating-justice-puts-banks-in-a-choke-hold-1398381603; Swartz, New Money, pp.
79–80.
19 Isaac William, ‘Don't Like an Industry? Send a Message to Its Bankers,’ Wall Street Journal, 22
November 2014, https://www.wsj.com/articles/william-isaac-dont-like-an-industry-send-a-message-to-
its-bankers-1416613023.
20 Frank Pasquale, The Black Box Society: The Secret Algorithms That Control Money and Information,
104 THEORY ON DEMAND
ed by the banking and payment sector (here loosely defined) as the reputation stain involved
in working with pornographers, Operation Choke Point speaks loudly of how moral, rather than
legal, concerns drive decision-making in a domino-like effect where the ones falling under
the metaphorical pile of tiles are the ones trading in sexual content. Such policies build on,
and feed, imaginaries wherein the inconvenience of pornography is located in its presumed
miasmic potential to sully (corporate) public images through association. Reputation is an
intangible corporate asset which, if well managed, ‘represents the organization favourably to
its publics and can be particularly valuable in doing so to its customers’.21 It is easily tarnished
and, even more crucially, its tarnishes are differently perceived among corporate publics.
In a 2015 Wall Street Journal interview, Lori Morettini, vice president of Humboldt Merchant
Services—which at the time processed an estimated half of online adult industry’s credit card
transactions totaling a billion USD annually—identified reputational liability of being outed
as an aide to pornographers as their primary risk involved. The main financial risk, again, had
to do with paybacks, as in customers disputing the charges made: ‘If you snuck off to the
dry cleaner and your wife saw that on your charge she’s going to be happy you did your own
laundry, but if the credit card statement comes with an adult merchant on it, you are going
to say it’s fraud.’22 Here, porn site membership subscriptions, and the masturbatory activities
they imply, are framed as inconvenient in terms of heterosexual matrimonial accord—the
risk not involving the business practices but being identified with the customer. Paybacks
in porn consumption are, indeed, an issue, them being fourfold to other online purchases.23
Swartz associates paybacks with ‘a matryoshka doll of rules’ connected to predicting high-
risk customers: some of these are set by banks, others by credit card companies and PSPs,
and many simply carry down:
Transactions that would be considered “high risk” in the market model are simply
banned. This is because PSPs access acquiring banks as master merchants, and in
order to qualify for the lowest rates, PSPs must guarantee that all the transactions
they conduct will be low risk for chargebacks. There are long-standing lists, provided
by regulatory and industry groups, of high-risk merchant categories: time-shares,
home-based charities, herbal remedies, and so on. Most payment start-ups simply
take these lists and drop them into their terms of service as explicitly prohibited.24
This elevated risk of paybacks gestures toward the in/conveniences associated with pornogra-
phy as highly desired and routinely used, yet also as that which consumers do not wish to be
associated with. Such intermeshing of interest and shame, excitement and potential humilia-
tion arguably speaks of affective dynamics wherein secrecy feeds fascination, and where the
Cambridge, MA: Harvard University Press, 2015, p. 6.
21 Gary Davies and Louella Miles, ‘Reputation Management: Theory Versus Practice’, Corporate
Reputation Review 2 (1998): 16.
22 Gregory J. Millman, ‘Managing Credit Card Risk for Sex Businesses’, The Wall Street Journal,
18 February 2015, https://www.wsj.com/amp/articles/managing-credit-card-risk-for-sex-
businesses-1424281370.
23 Swartz, New Money, p. 94.
24 Swartz, New Money, p. 93.
105
IN/CONVENIENCE
sense of the forbidden—of some norm being breached—invests the objects consumed with
the frisson of a forbidden fruit. Porn then invites in/convenient proximities between the bodies
displayed and those watching: bodies on the screen can press inconveniently close even as
such closeness feeds sexual arousal and pleasure.25 A persistent dynamic of in/convenience
underpins the uses of porn, a genre historically defined through the regulation of its allegedly
miasmic impact.26 But this is, of course, one quickly flipping from convenience to its perceived
opposite in ways indicating that these two may not be opposites.
Applying Berlant’s discussion of inconveniences to platform ecologies helps in considering not
merely the fundamental, friction-laden interdependencies between the (corporate, individual,
algorithmic, and individual) actors involved, but also in accounting for the strained instability
of these interrelations owing to their public nature. In order to operate successfully, porn,
webcam, and content subscription sites, as well as individual entrepreneurs, are dependent
on myriad actors for whom such relations present an inconvenience. This was never a level
playing ground.
Enter the Platforms
Launched in 2007, Pornhub was among the first porn video aggregators sites emulating the
operating principles of YouTube (est. 2005) while trading in content banned in its content pol-
icy. Tube sites are pornographers by proxy in not producing the videos they stream (although
Pornhub’s owner, currently known as Aylo, has bought up several production studios that
financially suffered due to the piracy affected by tube sites). As corporate players, these dis-
tribution and advertising platforms have by and large replaced the more independent-yet-net-
worked model of porn distribution comprised of individual sites and partnerships between
sites hosted on diverse servers throughout the 1990s, and slightly beyond. Whereas twenty
years ago users would click from one site to another (web rings being a thing) or make use
of search engines to find compelling content, they are currently more likely to search within
a chosen tube platform.
Nick Srnicek defines platforms as ‘digital infrastructures that enable two or more groups to
interact’: as intermediaries, they bring together ‘customers, advertisers, service providers,
producers, suppliers, and even physical objects’.27 Such ability to link actors and markets,
as well as to constrain the forms and exchanges occurring between them, has been concep-
tualized as platform power.28 Platforms are further characterized by what Srnicek identifies
as ‘network effects’: centralizations on hubs with high volumes of user traffic. A good social
media site, for example, is one with many users (think Facebook, Instagram, or TikTok) just as
a good search engine is one honed through massive scales of queries (think Google), and a
good porn site is one offering a vast range of content (think Pornhub), so that volume breeds
25 For an extended discussion, see Susanna Paasonen, Carnal Resonance: Affect and Online Pornography,
Cambridge, MA: The MIT Press, 2011.
26 Kendrick, The Secret Museum.
27 Nick Srnicek, Platform Capitalism, Cambridge, UK: Polity, 2016, p. 43.
28 Katrin Tiidenberg, ‘Sex, Power and Platform Governance’, Porn Studies 8.4 (2021): 383.
106 THEORY ON DEMAND
volume and mass feeds mass. And since ‘more users beget more users’, this cycle ‘leads to
platforms having a natural tendency towards monopolisation’.29
As centralized distribution platforms, porn tube sites offer the promise of convenience. For
users, this promise is one of abundant content to choose from; for content producers, vast vol-
umes of potential consumers. That the actual deal is less convenient for creators with limited
means to monetize their videos, and for independent sites to compete with hubs is, of course,
a different matter. And while the hubs of online porn may not be convenient customers for
PSPs, they hold a different kind of sway in having access to these infrastructures from which
creators themselves are routinely excluded.30 This development has been accelerated by the
passing of the US SESTA-FOSTA law packages in 2018 which makes service providers liable
for the sexual content that their users publish. Yet it cannot be reduced to it.
Advocated for by the Christian right in alliance with feminist anti-pornography activists, SES-
TA-FOSTA has hurt sex workers in particular while also having detrimental effects on sexual
and gender minorities and the sexual rights of social media users internationally.31 Ostensibly
set up to curb sex trafficking, these bills have in practice targeted online porn and sex work,
and impacted the possibilities of social media users to share sexual content well beyond the
realm of commercial sex—hence Tumblr’s 2018 decision to ban all nudity and sexual content,
or Meta’s gradually tightened content policies.32
The inconvenience of sex workers, porn, and porn consumers for financial infrastructures is
underpinned by a mode of pre-emption—of premeditation concerning normative imaginaries.
As Katrin Tiidenberg points out, ‘sex has an ambivalent relationship with consumer capital-
ism, and thus also platform capitalism, wherein sanitized versions are used to manipulate
user attention, yet many lived, diverse versions of sex are rejected for their assumed lack of
appeal for advertisers’.33 Just as social media platforms have the tendency to over- rather
than under-moderate sexual content out of concern for offending users and advertisers (this,
post SESTA-FOSTA, entailing potential legal liability), PSPs are much more likely to opt for
conservative rather than liberal policies pertaining to the monetization of said content; this
is also articulated as an ethical choice pertaining to corporate values.
Deplatforming Pornhub
Concerns over corporate reputation are very much at the heart of Pornhub’s financial deplat-
forming in December 2020, following the publication of Nicholas Kristof’s emotional New
York Times opinion piece, ‘The Children of Pornhub’. Setting out to reveal the dark side of
29 Srnicek, Platform Capitalism, p. 45.
30 Swartz, New Money, p. 95.
31 Danielle Blunt and Ariel Wolf, ‘Erased: The Impact of FOSTA-SESTA & The Removal of Backpage’,
Hacking//Hustling, 2020, https://hackinghustling.org/erased-the-impact-of-fosta-sesta-2020/.
32 Melissa Gira Grant, ‘Nick Kristof and the Holy War on Pornhub’, The New Republic, 10 December 2020,
https://newrepublic.com/article/160488/nick-kristof-holy-war-pornhub; Tiidenberg, ‘Sex, Power and
Platform Governance’.
33 Tiidenberg, ‘Sex, Power and Platform Governance’, 389.
107
IN/CONVENIENCE
the platform, Kristof dramatically claimed that Pornhub, the then-globally leading porn video
aggregator site, the traffic of which had continued to increase during COVID-19 lockdowns,
was monetizing ‘child rapes, revenge pornography, spy cam videos of women showering,
racist and misogynist content, and footage of women being asphyxiated in plastic bags’.34
Kristof accused Visa and Mastercard for benefitting from rape and (child) abuse materi-
al, very publicly calling the companies out. The piece highlighted well-known problems in
the platform’s principles of operation; expansive piracy, lax moderation practices, and long
response times pertaining to the removal of content. Much of this critique had been voiced
by sex workers, yet it was Kristof who became heard. As an act of reputation management,
Pornhub suspended access to nine million videos—the majority of its content—uploaded
from unverified users accounts. This had little effect on Visa and Mastercard which, along
with the ISO PayPal, discontinued their service with Aylo (then MindGeek) sites, causing a
dramatic rupture in the income of content creators reliant on the platforms (the ban was later
revoked for platforms catering studio content). As intermediaries, Aylo and PSPs build their
brands through relational and reputational labor involving promises of smooth and glitch-
free operability, professionalism, and efficiency. Before the company’s renaming/rebrand-
ing, MindGeek’s website did not mention its business being connected to porn but rather
promoted itself as a lucrative, responsible tech employer specialized in running and hosting
computational and architectural platforms.
As Melissa Shira Grant points out in her critique of Kristof’s ‘crusade’, his essay promoted
the efforts of Traffickinghub, a campaign run by a religious US right-wing organization aiming
at ‘the abolition of the sex trade, including prostitution and porn, by means of the criminal
law’.35 Kristof’s piece strategically targeted corporate reputations—as relational, ephemeral,
and contingent as they are—in order to delimit the possibilities of sex workers to monetize
their content. His labelling of Pornhub as the engine driving the distribution of abusive and
violent material directly undermined the platform’s long-term self-curation as a lifestyle brand
in an attempt to de-stigmatize porn use (while still monetizing porn’s notoriety).36 Aided by
its myriad publicity stunts, Pornhub had grown into a dominant enough player to stand as a
shorthand for online porn in general. Exemplifying Srnicek’s point about networked effects,
its high visibility fed further popularity such that Pornhub’s annual volumes of visitors grew
steadily up until Kristof’s piece was published. Visibility then equally fed vulnerability: for had
Pornhub not gained a symbolic status as the hub within a centralized porn platform economy,
it would have been less likely a target for such a campaign.
Following Tarleton Gillespie, the meaning of the notion of platform is fourfold: encompassing
the ‘computational, something to build upon and innovate from’, the ‘political, a place from
34 Nicholas Kristof, ‘The Children of Pornhub’, The New York Times, 4 December 2020, https://www.
nytimes.com/2020/12/04/opinion/sunday/pornhub-rape-trafficking.html.
35 Grant, ‘Nick Kristof and the Holy War on Pornhub’.
36 Susanna Paasonen, Kylie Jarrett and Ben Light, NSFW: Sex, Humor, and Risk in Social Media,
Cambridge, MA: The MIT Press, 2019; Silvia Rodeschini, ‘New Standards of Respectability in
Contemporary Pornography: Pornhub’s Corporate Communication’, Porn Studies 8.1 (2021): 76-91.
108 THEORY ON DEMAND
which to speak and be heard’, the ‘figurative, in that the opportunity is an abstract promise
as much as a practical one’, and the architectural in being designed to afford certain kinds
of interactions over others.37 In a figurative sense, to have a platform means being heard,
gaining an audience, and having potential impact. Conversely, to deplatform means to silence
by removing access to be heard and seen. Within networked media, deplatforming entails
removing user accounts or entire groups,38 banning content categories and enforcing such
policies through content moderation,39 as well as by impacting the technical or financial
infrastructures necessary for a platform’s operability. The decision to demonetize Pornhub
obviously exemplified deplatforming in the last sense. Without access to PSPs, Aylo has been
cut off from the financial infrastructures that its success had considerably fed, and which
had enabled its thriving. Since Aylo trades in user data, this has however not resulted in its
going under.40
Berlant discusses ‘the inconvenience paradox of dependency’ as one of ‘needing people or
a situation and hating to have that need’.41 This paradox cuts through the entanglements of
payment infrastructures and online porn, becoming manifest in moments of rupture when
the latter are simply cut off as unwanted, so that monetary circulation (and accumulation)
between subscribers, platforms, and creators comes to an abrupt halt. It can then be argued
that, as persistent friction, the inconvenience paradox has led to the casting of porn creators
as disposable vis-à-vis financial infrastructures as the notion of ‘high-risk customer’ has
shifted from a label indicating higher processing fees to just being deplatformed. Following
Berlant, this has to do with how perceived inconvenience (be it imaginary, potential, or actual)
intensifies to the degree of representing a threat to corporate existence, as an unbearable
thing to be done away with as an affective operation of moral sanitization.42
Disposable ones
The logic of platformed disposability of sex workers echoes the one casting them as ‘expend-
able, disposable, corruptible and untrustworthy’ in law enfor cement. This suggests that they
are seen to lack proper agency due to the improper nature of their enterprise, and therefore
disabling their income is in fact the right thing to do.43 Such disposability became explicit
37 Tarleton Gillespie, ‘The Politics of “Platforms”’, New Media & Society 12.3 (2010): 352.
38 Richard Rogers, ‘Deplatforming: Following Extreme Internet Celebrities to Telegram and Alternative
Social Media’, European Journal of Communication 35.3 (2020): 213-229.
39 Paul Byron, ‘“How Could You Write Your Name Below That?” The Queer Life and Death of Tumblr’, Porn
Studies 6.3 (2019): 336-349; Katrin Tiidenberg, ‘Playground in Memoriam: Missing the Pleasures of
NSFW Tumblr’, Porn Studies 6.3 (2019): 363-371.
40 Elena Maris, Timothy Libert, and Jennifer R. Henrichsen, ‘Tracking Sex: The Implications of Widespread
Sexual Data Leakage and Tracking on Porn Websites’, New Media & Society 22.11 (2021): 2018-
2038; Ilir Rama et al., ‘The Platformization of Gender and Sexual Identities: An Algorithmic Analysis
of Pornhub’, Porn Studies 10.2 (2023): 154-173; Rebecca Saunders, ‘Sex Tech, Sexual Data and
Materiality,’ Porn Studies 10.2 (2023): 120-134.
41 Berlant, On the Inconvenience, p. 36.
42 Berlant, On the Inconvenience, p. 152.
43 Zahra Stardust et al., ‘“I Wouldn’t Call the Cops if I Was Being Bashed to Death”: Sex Work, Whore
Stigma and the Criminal Legal System,’ International Journal for Crime, Justice and Social Democracy
109
IN/CONVENIENCE
when OnlyFans declared a ban on sexually explicit content in 2021—the very stuff that had
made the platform’s soaring success in the course of COVID-19 lockdowns and the influx
of Pornhub’s creators onto the site—in order to ‘ensure the long-term sustainability of our
platform’.44 The novel policy, for which the CEO squarely blamed Bank of New York Mellon,
Metro Bank, and JPMorgan Chase flagging and rejecting payments, was reversed within days
after content creators rightly pointed out the unfairness of this all; the platform’s success was,
after all, owed directly to their labor.45
The decision to implement the policy without communicating with sexual content creators
further aggravated a sense of injustice caused by the platform’s long-term unwillingness to
promote their visibility while very much profiting from their popularity. Even as the brand of
OnlyFans had at this point grown inseparable from pornography, the company operated with
the premise that this was not the case or that, at the very least, association with commercial
sex ultimately formed an inconvenience. I suggest that this incident reads as something of a
mise-en-abyme of platformed online pornography involving both acute financial incentives
and the much more opaque casting of sex workers as inconvenient due to their perceived
‘negative social value’.46
Both anti-pornography feminism and Marxist critiques of sex work identify it as alienating in
turning individuals and their inalienable aspects into commodities, so that the workers in this
realm reify themselves.47 Reification, on Georg Lukács’s terms, involves people becoming
thing-like (and, conversely, things gaining degrees of liveness)—a line of argumentation well
aligned with feminist critiques of pornography as a system of objectification.48 If sex work is
not necessarily acknowledged as work, the same applies to novel vocations such as social
media influencing reliant on self-presentation in the service of commercial partnerships within
economic structures impossible for the workers to affect, to the point that it may be difficult
to be financially compensated.49 The gendered nature of both occupations plays a role in
casting feminized work as lesser-than—and maybe as not labor at all. In the case of sex work,
the boundary between work and non-work becomes drawn as a moral one (in that one should
arguably not be paid for such things, independent of the work’s legal status), whether the line
of argumentation is a political one, one steeped in religious belief, or just kind of a gut feeling.
10.3 (2021): 144.
44 Lucas Shaw, ‘OnlyFans to Bar Sexual Videos Starting in October’, Bloomberg, 19 August 2021, https://
www.bloomberg.com/news/articles/2021-08-19/onlyfans-to-block-sexually-explicit-videos-starting-in-
october.
45 Eloise Barry, ‘Why OnlyFans Suddenly Reversed its Decision to Ban Sexual Content’, Time, August
26 2021, https://time.com/6092947/onlyfans-sexual-content-ban; Jacob Bernstein, ‘OnlyFans
Reverses its Decision to Ban Explicit Content,’ The New York Times, 25 August 2021, https://www.
nytimes.com/2021/08/25/style/onlyfans-ban-reversed.html; Rébecca Franco, ‘“Controlling the Keys
to the Golden City”: The Payment Ecosystem and the Regulation of User-Generated Porn Platforms’,
forthcoming.
46 Berlant, On the Inconvenience, p. 85.
47 Kylie Jarrett, Digital Labor, Oxford: Polity, 2022, p. 134.
48 Susanna Paasonen et al., Objectification: On the Difference Between Sex and Sexism, London:
Routledge, 2020, pp. 7-8.
49 Jarrett, Digital Labor, pp. 46, 181.
110 THEORY ON DEMAND
Arguing against broad applications of the notion of commodification in connection with digital
labor, Kylie Jarrett proposes the concept of ‘assetization’ for describing how ‘a worker who is
marketing their own assets is not selling their labor-power to a platform to do with as it wills
but opening those assets to a valuation by the marketplace’ so that this value ‘cannot be
entirely produced or captured by the enterprise that exploits it’ but remains bound up with
the worker’s subjectivity.50 This more context-sensitive approach to platform labor helps in
separating tactical self-commodification from alienation and, in so doing, resisting the kind of
exceptionalism that sex work gets treated with, both on- and off-platform. Following this line
of thought, financial deplatforming targeting the infrastructures of assetization makes evident
the fragile agency of pretty much all platform laborers, both individually and collectively, within
the black box societies that Pasquale addresses.
Porn and other sexual media are consumed on mass scales: they are far from niche as media
content or fields of labor, even as their popularity does not shield them from on-platform
demonetization. In reference to critiques of data capitalism eschewing governmental regula-
tion reliant on the principles of democracy,51 the power of payment infrastructures to dictate
their own terms is detached from the legal frameworks that they need to comply with. This
basically points to their power to delimit possible realms of assetization, occupational possi-
bilities, and the commodities that people can purchase. And, given the global reach of pay-
ment infrastructures, this applies to merchants and consumers globally, independent of the
legal status of sex work in specific regions. Just as social media platforms govern the allowed
forms of sociality by regulating sexual communication, payment infrastructures operate as
moral actors casting sexual content outside the loops of monetization as dirty-dirty-dirty. As
Tiidenberg argues, platform power is ‘achieved through territorialization and resource control
that comes from developing an infrastructure and locking a population into it via division
of labour, socialization and positioning the infrastructure as a higher good’.52 It just as well
means locking a population out.
Tenacious Inconveniences
It can be argued that the limits of Berlant’s discussion of inconvenience lie in its broad-
ness, the notion being capacious enough to accommodate anything from minor irritations
to ambiguous desires, rape, and murder—or that which they discuss as the unbearable.
The inconveniences of having a clandestine lover, discriminating against employees, and
mauling racialized others are rather distinct; the consequences for those being thus cast
as inconvenient remain similarly specific. The inconvenience for a banker being taunted for
trading with pornographers and the acute disaster for a sex worker being cut off from financial
infrastructures are equally incommensurable, yet, following Berlant, intimately entangled in
terms of affective economies.
50 Jarrett, Digital Labor, pp. 161.
51 Shoshana Zuboff, The Age of Surveillance Capitalism: The Fight for the Future at the New Frontier of
Power. London: Profile Books, 2019; Pasquale, The Black Box Society, p. 6.
52 Tiidenberg, ‘Sex, Power and Platform Governance’, p. 383.
111
IN/CONVENIENCE
Writing on the sliding scale from passing annoyances to the destruction of life, Berlant is
making an argument precisely for their inseparability from the basic inconvenience of nonsov-
ereignty—of resenting the attachments that make us, so that infrastructural relations are
also nonrelations, or can smoothly switch into ones with an aggressive edge.53 Such volatile
switching off of relation very much characterizes the precarious position of sexual platforms
and their content creators in platform ecologies. If the ‘unbearable is the limit case of the
inconvenience of other people, where people stands in for any object that one needs for the
world to proceed, and where inconvenience stands for the fact that attachment is never easy’,
then the increasingly vocal casting of porn creators as inconvenient has pretty much resulted
in their demonetization as unbearable.54
How commercial sex gets classified as inconvenient/unbearable as an issue of corporate
ethics finds support from a cultural imaginary associating porn with a ‘scenario of danger
and rescue, a little perennial melodrama in which, though new players have replaced old,
the parts remain much as they were first written’ in the 19th century.55 This affective imagi-
nary represents more than a historical specter hovering over contemporary exchanges, given
the speed and impact with which its affective appeal gets activated—recall Kristof’s piece.
Independent of the realities of sex work, melodramatic divides between the helpless and
their oppressors hold longevity. Payment infrastructures then further contribute to such a
dynamic of helplessness by cutting off sex workers already stigmatized for their occupation.
That a melodramatic imaginary of cruel consequence plays out vis-à-vis the assetization of
sex work online does not mean that this will invariably be the case, or that the demonetization
of sex is bound to grow ever tighter in spiraling fashion. Both Tumblr’s 2022 decision to allow
for certain kinds of sexual content on the platform and the Meta oversight board’s 2022
ruling to revise the company’s content policies connected to nudity and sex hint at other
directions, even as their outcomes remain opaque and rather useless to sex workers. Despite
the uniformity implicated by the notion of a platform, it entails heterogenous interests, values,
or missions, so that frictions and conflicts are constant as well as internal.56 And since plat-
form economies and infrastructures operate at cross-continental scales, articulated values
come across as much less obvious for those not sharing the same cultural context, and are
potentially open to contestation.
Although the role of pornography in web economies is much less pronounced than in the
1990s, it no longer being a necessary inconvenience, the shapes and forms of online sexual
content continue to multiply. As do their forms of monetization and assetization—this being
by no means a declining economy—so that its total effacement from payment infrastructures
remains impossible.
53 Berlant, On the Inconvenience, p. 152.
54 Berlant, On the Inconvenience, p. 170, emphasis in the original.
55 Kendrick, The Secret Museum, p. xiii.
56 Taina Bucher, Facebook, Oxford: Polity, 2021.
112 THEORY ON DEMAND
While Berlant outlines in/convenience as a contingent dynamic where the one blends into
the other, infrastructural imaginaries underpinning the terms of PSPs operate with a binary
logic of either/or, this or that—not both/and, or maybe. Grey zones of ambiguity do not fare
well in algorithmic logics where the inconvenience—or the perceived inappropriateness—of
merchants can be automatically calculated on the basis of the data gathered and where
association with sex works lands them in the category of ‘no-no’. In other words, Berlant is
not a helpful guide to understanding the logic of corporate terms of use just here. But, as I
have suggested, their take on inconvenience helps in articulating default co-dependencies
and attachments between human and nonhuman (institutional, algorithmic, representational)
bodies within platform economies which, in their ambiguity, are simultaneously matters of
convenience, so that the two can be seen as two sides of the same coin.
In/convenience communicates frictions as ongoing. Following Berlant, while inconveniences
cannot be resolved, they can be reframed and rethought as per their objects: the casting of
things or people as inconvenient is not, then, a one-way street but rather something open to
affective reworking. Despite how depressing the contemporary moment may seem in terms
of platformed infrastructural injustice, this line of thinking offers ways for considering future
horizons for sexual content creation as not dictated by the moral corporate imaginaries that
they are currently disciplined by. As feeble as the promise may seem, a sense of hope is crucial
for critical work, and we cannot foresee the future infrastructural actors at play. Capitalists
do, after all, like money.
Acknowledgements
Research for this chapter has been supported by the Strategic Research Council at the
Research Council of Finland, grant number 352520. (Intimacy in Data-Driven Culture).
References
Are, Carolina and Paasonen, Susanna. ‘Sex in the Shadows of Celebrity’, Porn Studies 8.4 (2021): 411-
419.
Barry, Eloise. ‘Why OnlyFans Suddenly Reversed its Decision to Ban Sexual Content’, Time, August 26,
2021, https://time.com/6092947/onlyfans-sexual-content-ban.
Berg, Heather. Porn Work: Sex, Labor, and Late Capitalism, Chapel Hill: The University of North Caroli-
na Press, 2021.
Berlant, Lauren and Prosser, Jay. ‘Life Writing and Intimate Publics: A Conversation with Lauren Ber-
lant’, Biography 34.1 (2011).
Berlant, Lauren. ‘Intimacy: A Special Issue’, Critical Inquiry 24.2 (1998): 281-288.
_____. On the Inconvenience of Other People, Durham, NC: Duke University Press, 2022.
_____. The Female Complaint: The Unfinished Business of Sentimentality in American Culture, Durham,
NC: Duke University Press, 2008.
_____. The Queen of America Goes to Washington City: Essays on Sex and Citizenship, Durham, NC:
Duke University, 1997.
113
IN/CONVENIENCE
Bernard, Tara Siegel. ‘Sex Workers Have Been Shunned by Banks, Even When Their Work Is Legal’,
The New York Times, 18 November 2023, https://www.nytimes.com/2023/11/18/business/sex-work-
ers-bank-accounts.html.
Bernstein, Jacob. ‘OnlyFans Reverses its Decision to Ban Explicit Content,’ The New York Times, 25
August 2021, https://www.nytimes.com/2021/08/25/style/onlyfans-ban-reversed.html.
Blunt, Danielle and Wolf, Ariel. ‘Erased: The Impact of FOSTA-SESTA & The Removal of Backpage’,
Hacking//Hustling, 2020, https://hackinghustling.org/erased-the-impact-of-fosta-sesta-2020/.
Bucher, Taina. Facebook, Oxford: Polity, 2021.
Byron, Paul. ‘“How Could You Write Your Name Below That?” The Queer Life and Death of Tumblr’,
Porn Studies 6.3 (2019): 336-349.
Chun, Wendy Hui Kyong. Control and Freedom: Power and Paranoia in the Age of Fiber Optics, Cam-
bridge, MA: The MIT Press, 2006.
Davies, Gary and Miles, Louella. ‘Reputation Management: Theory Versus Practice’, Corporate Reputa-
tion Review 2 (1998): 16-27.
Franco, Rébecca. ‘“Controlling the Keys to the Golden City”: The Payment Ecosystem and the Regula-
tion of User-Generated Porn Platforms’, forthcoming.
Gillespie, Tarleton. ‘The Politics of “Platforms”’, New Media & Society 12.3 (2010): 347-364.
Grant, Melissa Gira. ‘Nick Kristof and the Holy War on Pornhub’, The New Republic, 10 December
2020, https://newrepublic.com/article/160488/nick-kristof-holy-war-pornhub.
Jarrett, Kylie. Digital Labor, Oxford: Polity, 2022.
Keating, Frank. ‘Justice Puts Banks in a Choke Hold’, Wall Street Journal, 25 April 2014, https://www.
wsj.com/articles/frank-keating-justice-puts-banks-in-a-choke-hold-1398381603.
Kendrick, Walter. The Secret Museum: Pornography in Modern Culture, 2nd edition, Berkeley: Universi-
ty of California Press, 1997.
Kristof, Nicholas. ‘The Children of Pornhub’, The New York Times, 4 December 2020, https://www.
nytimes.com/2020/12/04/opinion/sunday/pornhub-rape-trafficking.html.
Lane, Frederick S., III. Obscene Profits: The Entrepreneurs of Pornography in the Cyber Age, New York:
Routledge, 2001.
Maris, Elena; Libert, Timothy; and Henrichsen, Jennifer R. ‘Tracking Sex: The Implications of Wide-
spread Sexual Data Leakage and Tracking on Porn Websites’, New Media & Society 22.11 (2021):
2018-2038.
McNair, Brian. Porno? Chic! How Pornography Changed the World and Made It a Better Place, London:
Routledge, 2013.
Millman, Gregory J. ‘Managing Credit Card Risk for Sex Businesses’, The Wall Street Journal, 18
February 2015, https://www.wsj.com/amp/articles/managing-credit-card-risk-for-sex-business-
es-1424281370.
Paasonen, Susanna. Carnal Resonance: Affect and Online Pornography, Cambridge, MA: The MIT
Press, 2011.
Paasonen, Susanna; Attwood, Feona; McKee, Alan; Mercer, John; and Smith, Clarissa. Objectification:
On the Difference Between Sex and Sexism, London: Routledge, 2020.
Paasonen, Susanna; Jarrett, Kylie; and Light, Ben. NSFW: Sex, Humor, and Risk in Social Media, Cam-
bridge, MA: The MIT Press, 2019.
114 THEORY ON DEMAND
Pasquale, Frank. The Black Box Society: The Secret Algorithms That Control Money and Information,
Cambridge, MA: Harvard University Press, 2015.
Perdue, Lewis. EroticaBiz: How Sex Shaped the Internet, New York: Writers Club Press, 2002.
Rama, Ilir; Bainotti, Lucia; Gandini, Alessandro; Giorgi, Giulia; Semenzin, Silvia; Agosti, Claudio; Corona,
Giulia; and Romano, Salvatore. ‘The Platformization of Gender and Sexual Identities: An Algorithmic
Analysis of Pornhub’, Porn Studies 10.2 (2023): 154-173.
Rodeschini, Silvia. ‘New Standards of Respectability in Contemporary Pornography: Pornhub’s Corpo-
rate Communication’, Porn Studies 8.1 (2021): 76-91.
Rogers, Richard. ‘Deplatforming: Following Extreme Internet Celebrities to Telegram and Alternative
Social Media’, European Journal of Communication 35.3 (2020): 213-229.
Saunders, Rebecca. ‘Sex Tech, Sexual Data and Materiality,’ Porn Studies 10.2 (2023): 120-134.
Shaw, Lucas. ‘OnlyFans to Bar Sexual Videos Starting in October’, Bloomberg, 19 August 2021, https://
www.bloomberg.com/news/articles/2021-08-19/onlyfans-to-block-sexually-explicit-videos-start-
ing-in-october.
Srnicek, Nick. Platform Capitalism, Cambridge, UK: Polity, 2016.
Stardust, Zahra; Treloar, Carla; Cama, Elena; and Kim, Jules. ‘“I Wouldn’t Call the Cops if I Was Being
Bashed to Death”: Sex Work, Whore Stigma and the Criminal Legal System,’ International Journal for
Crime, Justice and Social Democracy 10.3 (2021): 142-157.
Stuart, Rachel. ‘Webcam Performers Resisting Social Harms: “You're on the Web Masturbating… It's
Just about Minimising the Footprint”’, International Journal of Gender, Sexuality and Law 2.1 (2022):
171-198.
Swartz, Lana. New Money: How Payment Became Social Media, New Haven: Yale University Press,
2020.
Tiidenberg, Katrin. ‘Playground in Memoriam: Missing the Pleasures of NSFW Tumblr’, Porn Studies
6.3 (2019): 363-371.
_____. ‘Sex, Power and Platform Governance’, Porn Studies 8.4 (2021): 381-393.
William, Isaac. ‘Don't Like an Industry? Send a Message to Its Bankers,’ Wall Street Journal, 22 Novem-
ber 2014, https://www.wsj.com/articles/william-isaac-dont-like-an-industry-send-a-message-to-its-
bankers-1416613023.
Zuboff, Shoshana. The Age of Surveillance Capitalism: The Fight for the Future at the New Frontier of
Power. London: Profile Books, 2019.
115
IN/CONVENIENCE
THEORIZING ‘ANTI-CONTENT’: ON SLEEP APPS AND
HORIZONTAL MEDIA
NETA ALEXANDER
Introduction: The Soporic Media Industry
Born with a facial paralysis, I have never been able to fully shut my right eye. To fall asleep, I
have to use an eye mask, blackout blinds, or other improvised tools and techniques to cre-
ate an entirely darkened environment. Over the years, I have used furniture, towels, pillows,
t-shirts, houseplants, books, and, less successfully, pets to block the sunlight while traveling.
Large windows freak me out, and they are the first thing I notice when entering a bedroom.
My level of fatigue, or, to borrow a popular concept among disability scholars, the number of
‘spoons,’ or energy reserves, I can use on a given day, is directly related to my ability to sleep
in a dark or windowless room.1 Years before ‘sleep hygiene’ became a self-care industry of
endless books, listicles, and products, I removed every electronic device from my surround-
ings before going to bed.2
My sleep habits, however, are increasingly difficult to maintain. Streaming platforms like Net-
flix, Amazon, HBO Max, and YouTube compete for my attention by declaring a war on sleep.3
At the same time, technology companies like Apple enthusiastically design and manufacture
new products promising to mitigate the user’s exposure to blue light, which is correlated with
sleep disturbances.4 Features like ‘dark modes’ and ‘night modes’ offer users more control
over the color and brightness of their display, while normalizing the idea that users should
take their personal electronics to bed in order to wind down and fall asleep.
As it is hindered by ubiquitous screens, doomscrooling, and remote work, sleep demonstrates
how technology companies develop new products promising to solve the problems created by
1 For an overview of spoon theory and ‘spoonies’, see Christine Miserandino, ‘The Spoon Theory Written
by Christine Miserandino’, But You Don’t Look Sick, 25 April 2013, https://butyoudontlooksick.com/
articles/written-by-christine/the-spoon-theory/; Jonathan Sterne, Diminished Faculties: A Political
Phenomenology of Impairment, Durham: Duke University Press, 2022, pp. 163-172.
2 One recent example of the rising popularity of sleep hygiene is Arianna Huffington’s New York Times
bestseller. See Arianna Huffington, The Sleep Revolution: Transforming Your Life, One Night at a Time,
New York: Harmony Books, 2016.
3 In 2017, Netflix co-founder Reed Hastings famously stated that the company’s most fierce competitor
is sleep: ‘You get a show or a movie you’re really dying to watch, and you end up staying up late at
night, so we actually compete with sleep—and we’re winning!’. Qtd. in Rina Raphael, ‘Netflix CEO Reed
Hastings: “Sleep Is Our Competition”’, Fast Company, 6 November 2017, https://www.fastcompany.
com/40491939/netflix-ceo-reed-hastings-sleep-is-our-competition/.
4 For an historical overview of how tech companies monetize concerns around blue light, see Dylan
Mulvin, ‘Media Prophylaxis: Night Modes and the Politics of Preventing Harm’, Information & Culture
53.2 (2018): 175-202.
116 THEORY ON DEMAND
their current products. As such, I define soporific media as any medium or feature designed to
induce sleep, including noise-canceling sleep headphones, sleep trackers, ASMR videos, the
streaming platform Napflix, sleep apps like Calm and Slumber, and endless other products.
Soporific features like Apple’s Night Shift or the sleep app Calm assume a fatigued media
user relying on their electronics to fall asleep.
Exploring the soporific media industry as both a poison and a cure, I follow the logic of “tech-
nopharmacology,” which has been coined and developed by Joshua Neves, Aleena Chia,
Susanna Paasonen, and Ravi Sundaram to “expand media theoretical inquiry by attending to
the biological, neurological and pharmacological dimensions of media.”5 Soporific products, I
argue, reveal a tension between a promise to cure prevalent sleep disturbances and medical
conditions like insomnia, and a demand for premium subscriptions and constant surveillance.
Blurring the lines “between big data and big pharma,” these apps and features monetize the
quest for convenience.6 The tech-based cocoons they offer deem it necessary to transition
from wakefulness to sleep by eliminating inconvenient cognitive, emotional, or optical factors
such as rumination, anxiety, noise, or daylight. By drawing on a medical discourse touting the
benefits of destressing and rest, such products have become conducive to the user’s sense
of comfort and well-being.
As an ever growing, billion-dollar industry, soporific media is key to theorizing the transition
from sleep as ‘the great human affronts to the voraciousness of contemporary capitalism’ to
sleep as a site of value production.7 Sleep, Jonathan Crary contends, is passive, non-produc-
tive, and composed of empty time. It ‘requires periodic disengagement from networks and
devices in order to enter a state of inactivity and uselessness. It is a form of time that leads us
elsewhere than to the things we own or are told we need.’8 This description, however, fails to
account for the myriad ways in which sleep has become labor. The sleeping body has been
studied and monetized by pharmaceutical companies since the rise of ‘sleep labs’ in the
1970s, while, more recently, the domestic sphere has been recast into a makeshift sleep
laboratory with the help of WIFI-connected technologies.9 Sleeping bodies are increasingly
connected to headphones, smartphones, and tracking devices that turn circadian rhythms
into data streams sold to third-party companies. An early ad for the wearable tracker Fitbit
culminated with a close-up of a woman sleeping with the device on her hand, promising to
optimize ‘even inactivity’.10 Asleep or awake, profit can be made.
5 Joshua Neves, Aleena Chia, Susanna Paasonen, and Ravi Sundaram, Technopharmacology,
Minneapolis: University of Minnesota Press, 2022, p. x.
6 Neves et. al, Technopharmacology, p. x.
7 For an analysis of the rise of twenty-four-seven capitalism, see Jonathan Crary, 24/7: Late Capitalism
and the Ends of Sleep, New York: Verso Books, 2013, p. 10.
8 Crary, 24/7, p. 126.
9 For an overview of the rise of sleep labs in the 1970s and the medicalization of sleep, see Kenton Kroker,
The Sleep of Others, Toronto: University of Toronto Press, 2007.
10 Natasha Dow Schüll, ‘Data for Life: Wearable Technology and the Design of Self-Care’, BioSocieties 11.3
(March 2016): 6.
117
IN/CONVENIENCE
The handheld digital interface is a key locus to study the monetization of sleep as, unlike the
sedentary subject of the movie theater, it is horizontal-friendly: Users are encouraged to take
their electronics to bed, place them on or next to their bodies, and fall asleep.11 I explore the
shift from vertical and sedentary media to what I call horizontal media: devices, apps, and
interfaces strategically designed to be used while lying down. We need a theory of horizontal
media for three reasons. First, much of media theory is invested in studying ‘seated specta-
torial positions’ by focusing on theater, film, and television, while ignoring ‘folded’, injured or
bedridden embodiments.12 Second, existing accounts of ideal bodies and viewing positions
fail to isolate technologies that habituate the user to shift from one posture to another. Third,
so many of us consume videos, websites, and audio while prone, creating a need for a critical
account that asks not just what content the user consumes, but also what bodily position is
assumed by the interface through which the content is accessed.
Horizontal media is also a key site to theorizing the promise and perils of convenience and
cure. Sufficient daily rest is crucial for the very survival of the user, yet it can only be achieved
under specific conditions: a darkened room, a comfortable bed, and lack of screens and
stimulation. Here lies the paradox at the center of this chapter: while the services often used
in bed, from Netflix to Calm, tout convenience as a form of personalized, user-controlled
content, they might also produce inconvenience by overstimulating the user and preventing
her from falling asleep.
Keep Calm and Rewrite History
In 2011, American web designer Alex Tew built a website called donothingfor2minutes.com,
on which a user can stare at a screen and listen to waves for two minutes as long as they
didn’t touch their mouse (see figure no. 1).13 If, however, the user succumbs to their desire
to click, refresh, or type, the black font inviting them to relax turns into a red warning sign
stating ‘try again’, and the two-minute countdown begins anew. At the end of two minutes, an
email prompt appears, asking those who master the challenge to sign up. In the website’s
first two weeks, over one hundred thousand people did so, and a few months later Tew
co-funded Calm.com.14
Over a decade later, Calm has grown into one of the most popular sleep apps in the world,
valued at over $2 billion. In company parlance, the app is designed ‘to help you manage
stress, sleep better and live a happier, healthier life’ by offering ‘hundreds of hours of original
audio content available in seven languages’. Once a subscriber downloads the app and pays
the annual fee, they can engage with an ever-changing library of ‘sleep stories’, ‘meditations’,
11 For an analysis of sleep as crucial for theories of spectatorship and media use, see Jean Ma, At the
Edges of Sleep: Moving Images and Somnolent Spectators, Berkeley: University of California Press,
2022.
12 For a theory of the folded body of avid computer users, see Michele White, The Body and the Screen:
Theories of Internet Spectatorship, Cambridge: MIT Press, 2006.
13 David Curry, ‘Calm Revenue and Usage Statistics (2022)’, Business of Apps, 1 July 2022, https://www.
businessofapps.com/data/calm-statistics/.
14 Curry, ‘Calm Revenue and Usage Statistics (2022)’.
118 THEORY ON DEMAND
‘soundscapes’, and ‘playlists’. More than one hundred million users now have Calm on their
smartphones, after downloads surged by a third in the coronavirus pandemic’s early days.15
New users are now recruited via their employers. Through the company’s Calm Business pro-
gram, ten million American workers have free access to the app as a mental health benefit.16
The company’s growth, both in terms of venture capital and paying subscribers, has pushed
it to become a wellness empire, with an HBO television show and an ever-expanding line of
products, including an $80 meditation cushion and a $272 weighted blanket.17 Calling itself
‘the Nike of the mind’, the app rewards users for extending a meditation or sleep ‘streak’ every
time they listen to content. Through these design decisions, Calm makes the pursuit of better
sleep a competitive and paradoxically stressful endeavor. This can lead to an obsessive quest
for optimal sleep, a condition called ‘orthosomnia’ that was identified in 2017 and attributed
to the use of sleep trackers and apps.18
Figure 1: A screenshot of the website donothingfor2minutes.com, inviting users to stare at a sunset and
‘just relax’ for two minutes, and featuring a countdown at the center of the screen (Source: screenshot
taken by the author
For millions of users, Calm is a source of pleasure and convenience because it associates
sleep with an intimate voice whispering into their ears, returning adults to the childhood ritual
of the bedtime story. Since it was launched in early 2013, Calm’s most popular content has
15 Annie Lowrey, ‘The App that Monetized Doing Nothing’, Atlantic, 4 June 2021, https://www.theatlantic.
com/technology/archive/2021/06/do-meditation-apps-work/619046/.
16 Curry, ‘Calm Revenue and Usage Statistics (2022)’.
17 Lowrey, ‘The App that Monetized Doing Nothing’.
18 The term ‘orthosomnia’ was coined by researchers from Rush University Medical School and
Northwestern University’s Feinberg School of Medicine in a 2017 case study published in the Journal
of Clinical Sleep Medicine. See Karen Zraick and Sarah Mervosh, ‘That Sleep Tracker Could Make Your
Insomnia Worse’, New York Times, 13 June 2019, https://www.nytimes.com/2019/06/13/health/sleep-
tracker-insomnia-orthosomnia.html.
119
IN/CONVENIENCE
been its ‘sleep stories’.19 Dreaming with Frida is a thirty-eight-minute sleep story pushed to
new subscribers with a thumbnail consisting of a colorful portrait of the artist Frida Kahlo. The
easily recognizable iconography includes the famous painter wearing a blue shirt and a hair
ribbon made out of three pink roses. Her relaxed torso emerges from a jungle-like garden,
with a yellow parrot on her left and a green butterfly to her right (see figure no. 2). As to be
expected from an app designed to put users to sleep, this thumbnail projects tranquility and
harmony between Frida—presented here only by her first name—and the natural world. The
story, written by Paola Villegas Sourco and read by Mexican-American actress Emily Rios, is
noteworthy for the ways it recasts the surrealist painter as a protagonist of a fairytale taking
place in 1940s Mexico. The bedtime story introduces Kahlo as a fierce warrior who ‘never
apologized for who she was’ and whose paintings explored recurring themes ‘including her
marriage, her chronic pain, and her love for Mexico’. The listener is then invited by Rios’ soft
voice to close their eyes, take a deep breath, and ‘dream with Frida’. For the next half an hour,
Rios walks us through an imaginary day in Kahlo’s Casa Azul in Mexico City.
Rios briefly reminds her lethargic listeners that, at age forty, Kahlo suffered from injuries that
kept her mostly at home, and ‘she often painted while lying in bed’. But today, she continues,
Kahlo is filled with ambition and desire to step outside and paint. Talking slowly—in an app that
does not enable its users to change the playback speed—Rios takes ample time to describe
Kahlo’s delight in feeding exotic animals, including monkeys and parrots, and walking around
her beautiful house. The ‘sleep story’ glosses over any detail that might seems inconvenient,
and therefor potentially harmful, to the listener’s gradual transformation into the dream world.
The years-long torturous relationship between Kahlo and painter Diego Rivera, for example, is
succinctly described as ‘their love, sometimes stormy, runs deep’. The result is a sentimental
depiction of an artist who bravely challenged rooted conceptions of the disabled female body
that only mentions her injury and pain as a vague aside—‘her imagination conquered all’. By
the time Kahlo selects a few papayas from ‘her favorite tree’ and gazes at ‘a flock of butterflies’
that gently land on her shoulder just as the earth ‘wraps around her like a soft embrace’, the
listener should be already fast asleep. ‘She has everything she needs within these walls’, Rios
whispers, painting an image of a five-star resort, rather than a woman confined to her bed
for years following a spine injury and a series of complicated surgeries. The story ends when
Kahlo leaves the house to purchase a new brush and, upon returning to her studio, is inspired
to paint one of her most famous self-portraits, Roots (1943), depicting her wearing an orange
dress with vines growing out of her body into the ground. As described by Rios, this painting
is a celebration of nourishment and hope, two recurring concepts in the sleep stories lexicon.
19 Calm’s most popular single piece of content is Dream with Me, a story read by Harry Styles, the former
One Direction singer, and a Calm investor. When it was released in July 2021, overwhelming traffic
crashed the app. See Lowrey, ‘The App that Monetized Doing Nothing’.
120 THEORY ON DEMAND
Figure 2: A screenshot from the Calm app depicting a thumbnail of Dreaming with Frida (left), and a
screenshot of the credit page of the same sleep story (right). Source: screenshots taken by the author.
With its close attention to colors, textures, flavors, flowers, and natural rhythms, Calm’s sleep
stories provide relaxing soundscapes designed to function as white noise and shield the listen-
er from bedtime anxiety and rumination. It turns Kahlo’s many struggles and inconveniences
into an audioscape offering nothing but comfort and relaxation, slowly luring the user to sleep.
Employing techniques such soft, slow voices and clean, simple designs, sleep and meditation
apps offer a horizontal media experience that has grown in popularity over the past decade.
Dreaming with Frida repackages the lifelong health struggles of a pioneering disabled Mex-
ican artist as a soothing lullaby for overworked phone users. Yet, I argue that the content of
horizontal media is less important than its form. In his analysis of Apple’s Night Shift, Dylan
Mulvin defines ‘media prophylactics’ as ‘the techniques, technologies, and design choices
that are made on behalf of or by users to preempt the ill effects (whether imagined or con-
crete) of media use, participation, or environmental exposure’.20 He goes on to distinguish
between media prophylactics that are ‘content agnostic’, including earplugs, air purifiers, and
night modes, and those that are ‘content partisan’, like safe search algorithms or the use of
commercial content moderators in platform governance.21 These tactics, as different as they
may seem, are all designed and marketed to filter part of the user’s sensory environment:
20 Mulvin, ‘Media Prophylaxis’, p. 176.
21 Mulvin, ‘Media Prophylaxis’, p. 176.
121
IN/CONVENIENCE
noise, smell, light, or graphic content. Theirs is a promise of a world devoid of inconvenience;
a world in which the user is in full control of her surroundings.
The idea that Calm and other sleep apps are content agnostic might come as a surprise,
considering the costs of recruiting Hollywood stars, copyrighting beloved audio segments
from works like The Joy of Painting, and marketing their premium subscriptions by touting an
ever-growing content library. Yet, users are unlikely to listen to Dreaming with Frida because
they are eager to learn more about the history of modern art. In fact, if they stay awake long
enough to discover that the day described by Rios led to the creation of one of Kahlo’s most
celebrated portraits, then the app failed to achieve its stated goal of putting them to sleep.
Indeed, once Rios is done reciting this tale, the app disappointedly asks, ‘Still Awake?’, against
a dark-blue background, immediately offering the insomniac listener another sleep story. A
sleep app that draws too much attention to its content might keep its users awake out of
interest, curiosity, or suspense. The imaginary user of Calm is not the fully awake universal
spectator of apparatus theory; rather, it is the ‘somnolent’ or ‘drowsy’ spectator-turned-audi-
tor who liberates herself ‘from the directives of the text and its system of meaning’.22 The app’s
users are likely to be lying down rather than sitting, and as such they differ from the sedentary
moviegoer who purchased a ticket in the hopes of being immersed in a fictional world.
The movie theater, in fact, is far from an ideal dreamscape: it requires sedentary rather than
horizontal engagement, with changing light, sound, and exposure to the bodily movements,
whispers, and smells of strangers. Even when achieved, the spectator’s sleep is limited by
the length of the film. Unlike the bed, the theater’s architecture and design have historically
privileged an able-bodied spectator, preventing wheelchair users and people with above-av-
erage height and weight from comfortably joining this communal ritual.23 That many movie
theaters are trying to lure back ticket buyers after the pandemic by installing reclining chairs
tells the story of an industry desperate to replicate the standards of comfort and customization
associated with domestic media consumption. More than ever, movie theaters might look and
feel like luxury bedrooms, where the spectator is invited to watch a film in repose.
Yet the movie theater is not the bedroom as it is devoid of the sense of privacy and intimacy
provided by one’s bed. Sleep apps challenge the history of sedentary engagement with media
in two ways: firstly, by replacing the prescribed vertical seated posture with the horizontal,
and, secondly, by moving us away from the representational and aesthetic questions that
22 Bruno Guaraná, ‘At the Edges of Sleep’, Film Quarterly 76.2 (Winter 2022): 109.
23 Drive-ins, for example, were patented in the 1930s by an American named Richard Hollingshead,
who created them ‘as a solution for people unable to comfortably fit into smaller movie theater seats’.
See ‘The History of Drive-in Movie Theaters’, New York Film Academy, 7 June 2017, https://www.nyfa.
edu/student-resources/the-history-of-drive-in-movie-theaters-and-where-they-are-now/; wheelchair
accommodations in movie theaters were only introduced in the United States as a result of the
Americans with Disabilities Act, which was signed into law on July 26, 1990. However, in recent decades,
seats for wheelchairs that once numbered five or six per theater have been cut to two or three, in order
to make space for recliner-style seats and other incentives. See Kristen Lopez, ‘How Movie Theaters
Are Failing Viewers with Disabilities’, IGN, 3 June 2018, https://www.ign.com/articles/2018/06/03/how-
movie-theaters-are-failing-viewers-with-disabilities.
122 THEORY ON DEMAND
have come to dominate film theory. Instead of focusing on narrative structure, mise-en-scene,
editing, pacing, soundtrack, and so forth, sleep apps can best be explored as ‘anti-content’.24
It is not the content of sleep apps’ recordings that explains their immense success, but rather
the sonic and affective affordances of their interface design. Embracing the inattentive user,
what these apps offer is an ambient human voice strategically recorded to be listened to in
a liminal state between wakefulness and sleep.
In an hailing review of Calm, New York Times columnist Amanda Hess recounts how she lis-
tened to the exact same recording every night for many weeks, falling asleep and developing a
‘strangely intimate relationship’ with the app’s popular meditation guide, Tamara Levitt.25 For
Hess, Levitt’s soft voice functions as a ‘sound tranquilizer’ crucial for her daily sleeping ritual.
The fact that Hess listened to the same recording every night suggests a unique attachment
to both the content and the narrator. Yet this repetition, which, over time, might become
compulsive (‘I tapped into Calm at night without thinking much about what I was doing’, Hess
writes) has more to do with Hess’s familiarity with her smartphone, the app’s interface, and
Levitt’s voice than with the recording itself. That Hess never bothers to describe the content
of her chosen segment strengthens the idea that it is Calm’s interface and voice-based library,
rather than its stories, that convince millions to pay for subscriptions. As Hess concludes,
‘Half of the programming is stuff designed to fall asleep to: If it’s working, you don’t hear it.’26
Here it is the passive—or, even better, the snoring—user who is the imagined ideal of the app,
rather than the active spectator of film theory. If it achieves its goal, only the first few minutes
might be recalled by the user-turned-listener.
If earplugs filter unwanted noise, air purifiers filter unwanted smells, and night modes filter
unwanted blue light, what do sleep apps filter? One possible answer is that they filter unwant-
ed thoughts. It is often one’s stream of consciousness that chases sleep away. Ruminations,
to-do lists, and feelings such as abandonment, jealousy, and fear arise more frequently and
intensely at nighttime because of a lack of distractions. To counteract them, Calm offers
something to anchor one’s attention. Drew Ackerman, the founder and host of the popular
sleep podcast Sleep with Me and a self-described insomniac, explains that he launched his
show in 2013 ‘to tame the vigilant, overactive “guardian” in the brain that feels it must stay
awake to worry’.27 Over one thousand episodes later, Ackerman told the New York Times
that he finds listening to be therapeutic, as it prevents him from ruminating all night long.28
Despite the growing popularity of sleep podcasts and apps, these soporific tools are not
as successful as they often claim. A research study analyzing 369 sleep apps available on
Android phones and iPhones found that the majority of the apps failed to include components
that have proved beneficial for those seeking to improve the quality, duration, and regularity
24 Amanda Hess, ‘The App That Tucks Me in at Night’, New York Times, 17 July 2019, https://www.nytimes.
com/interactive/2019/07/17/arts/calm-app-sleep-meditation.html.
25 Hess, ‘The App That Tucks Me in at Night’.
26 Hess, ‘The App That Tucks Me in at Night’.
27 Qtd. in Pagan Kennedy, ‘The Insomnia Machine’, New York Times, 18 September 2016, https://www.
nytimes.com/2016/09/18/opinion/Sunday/the-insomnia-machine.html.
28 Kennedy, ‘The Insomnia Machine’.
123
IN/CONVENIENCE
of their sleep.29 While researchers found that the apps were easy to navigate and use, they
concluded that there is much room for improvement, as ‘only a minority of the apps included
features that support behavior change’.30 Only four apps described habits that can interfere
with sleep and worsen insomnia, such as drinking caffeine or alcohol before bedtime.
This is not to say that sleep and meditation apps do not offer benefits. When asked to reflect
on their use of personal electronics during the many months of pandemic lockdown, my
undergraduate students at Colgate University frequently mentioned how sleep apps helped
them maintain a healthier routine when studying from home—anecdotal data supported by
empirical studies.31 A meta-analysis of over 1,500 peer-reviewed articles published from 2010
to 2022 concluded that ‘[e]xisting studies have proved the initial validation and efficiency
of delivering sleep treatment by mobile apps; however, more research is needed to improve
the performance of sleep apps and devise a way to utilize them as a therapy tool.’32 Yet it is
important to emphasize that sleep apps might be successful not because of their content
libraries and endless loops of ocean waves, but thanks to how they encourage their users to
shift from the sedentary use of media (bingeing Netflix, writing emails, working at their desk)
to horizontal use consisting of mostly audio, limited visual cues, and dark, ad-free interfaces.
This shift from more familiar technologies of convenience such as television to an emerging
category of convenient media specifically designed to enhance sleep might involve incon-
venient moments of trying to find the perfect posture, temperature, sensation, and mindset.
Sleep apps streamline and support this daily process by making it more seamless, and there-
fore less stressful, as users are invited to fall asleep while holding their beloved transitional
object: the ultimate convenience tech known as the smartphone.
Sleep apps gently nudge users to lie down and be metaphorically tucked in by their electronic
devices. This creates a very different relationship to the horizontal than that invoked by Kahlo’s
painful paintings. Kahlo’s self-portraits, in many of which she is lying down, reject the associ-
ation of horizontality with tranquility or sleep. In paintings such as the aforementioned Roots
(1943) and Without Hope (1945), the artist’s horizontal figure brings her closer to earth and
the natural world, but also to death, sickness, pain, and paralysis. These paintings convey
vulnerability, intimacy, and an injured female body in desperate need of rest.
29 Diana Yates, ‘Study of Sleep Apps Finds Room for Improvement’, Illinois News Bureau, 12 April 2017,
https://news.illinois.edu/view/6367/486860.
30 Yates, ‘Study of Sleep Apps Finds Room for Improvement’.
31 A 2019 study, for example, found that ‘Calm is an effective modality to deliver mindfulness meditation
in order to reduce stress and improve mindfulness and self-compassion in stressed college students’.
See Jennifer Huberty et al., ‘Efficacy of the Mindfulness Meditation Mobile App ‘Calm’ to Reduce Stress
Among College Students: Randomized Controlled Trial’, JMIR Mhealth Uhealth 7.6 (25 June 2019):
e14273.
32 See Abdullah Al Mahmud, Jiahuan Wu, and Omar Mubin, ‘A Scoping Review of Mobile Apps for Sleep
Management: User Needs and Design Considerations’, Frontiers in Psychiatry, 18 October 2022,
https://www.frontiersin.org/articles/10.3389/fpsyt.2022.1037927/full.
124 THEORY ON DEMAND
If Dreaming with Frida can ease us into sleep by using ASMR effects like the gentle sound
of butterfly wings, Without Hope is a work of art chasing sleep away (see figure 3).33 A visual
nightmare, this oil painting was painted while Kahlo was bedridden after a failed operation
intended to straighten her damaged spine with a bone graft and steel support. Because she
lost her appetite during her recovery, she was force-fed through a funnel.34 Drawing on these
traumatic experiences, the back of the painting carries the following inscription: ‘Not the least
hope remains for me... everything moves in tune with what the belly contains’.35
Desperate, bedridden, and in constant pain, Kahlo painted a self-portrait in which a paint-
er’s easel hovers above her bed. Instead of a canvas, it ‘suspends a gristly funnel delivering
dead-eyed fish, plucked chicken, bloody shanks of meat and pendulous entrails, directly into
Kahlo’s mouth’.36 According to art critic Reed Enger, each of the objects draws on Kahlo’s
ongoing struggle with sickness and surgeries confining her to bed, from her hospital bed
during her post-surgery recovery in New York to the oddly-shaped easel created by her father
to allow her to paint while bedridden. 37 The oversized funnel, Enger explains, was inspired by
Alfonso Toro’s engravings of the Spanish Inquisition’s water torture from Kahlo’s own copy
ofLa Familia Carvajal.38
Figure 3: Frida Kahlo, Without Hope, 1945, depicting the artist as she is lying in bed and being force-fed
33 For a study of ASMR affective economy of ‘shiveries’ and bodily sensations, see Joceline Andersen, ‘Now
You’ve Got the Shiveries: Affect, Intimacy, and the ASMR Whisper Community’, Television & New Media
16.8 (1 December 2015): 683–700.
34 Reed Enger, ‘Without Hope’, Obelisk Art History, 30 July 2017, http://arthistoryproject.com/artists/frida-
kahlo/without-hope/.
35 Enger, ‘Without Hope’.
36 Enger, ‘Without Hope’.
37 Enger, ‘Without Hope’.
38 Enger, ‘Without Hope’.
125
IN/CONVENIENCE
This horizontal portrait cannot be easily transformed into a Calm sleep story. The oil painting
is a two-dimensional interface owing its existence to a simple hack: an easel that could be
painted on while lying in bed. The result is disturbing, even shocking. It conjures longer his-
tories of force-feeding and other forms of violence targeting hospitalized patients and incar-
cerated convicts, many of whom are female.39 Dreaming with Frida replaces these personal
and collective struggles with a soothing female voice and an infantilized depiction of Kahlo
as a happily married woman in complete harmony with her natural environment. If Calm
enhances the user’s agency by gamifying and personalizing sleep, Without Hope depicts the
excruciating pain of losing one’s autonomy over one’s body.
By depicting the very apparatus needed to create the painting—Kahlo’s oddly-shaped easel—
Without Hope draws attention to the labor and unique tools needed to sustain the creative
act while bedridden. This makes the horizontal orientation of recumbent labor an important
theme of this work. Calm’s content creators, on the other hand, are heavily invested in mask-
ing the conditions under which they work. Levitt, one of Calm’s most popular narrators, told
Hess that for two years she ‘regularly worked 12-hour days’ to fulfill the growing demand for
original content.40 Much like Kahlo, Levitt’s work requires an easel. When she records guided
meditations or sleep stories for Calm, she uses ‘a paisley-printed easel fitted with an iPad (for
her script) and an iPod (for keeping time), and a footstool crowded with beverages’.41 While
Kahlo’s father helped her to configure a scaffolding device aligning her injured body with
her creative ambition, Levitt and Calm’s other narrators are asked to align themselves with
growing demand by using technological tools and nutritional support that override their need
for rest and the limitations of the human vocal cords. Kahlo’s embrace of the horizontal is
a crip call to bend the domestic environment to meet the injured body instead of forcefully
feeding it into health. Calm’s sleep stories, on the other hand, provide mass-scale ‘sound
tranquilizers’ that can help users meet the 24/7 relentless clock.
What happens when we center our analysis of soporific media on those who are either not
allowed to, or are otherwise unable, to sleep? How can artists or web designers make visible or
hearable the conditions of laboring bodies? And is it possible to develop a crip sleep app that
holds the tensions between activity and passivity, as well as resistance and refusal, instead
of deepening an unhealthy dependency on light-emitting electronics and mobile apps?
Reconguring #SelfCare
If Calm habituates users to take their smartphones to bed, the mobile app #SelfCare opts to
improve sleep and ease anxiety by reconfiguring and de-automating the user’s relationship
with her screen. As such, it offers a crip alternative to the soporific media industry.
39 According to Reed Enger, ‘Striking suffragettes were violently force fed in UK jails until the Prisoners Act
of 1913, and in the U.S., activist Ethel Byrne was force-fed while jailed for campaigning to legalize birth
control in 1917’. See Enger, ‘Without Hope’.
40 Qtd. In Hess, ‘The App That Tucks Me in at Night’.
41 Hess, ‘The App That Tucks Me in at Night’.
126 THEORY ON DEMAND
Launched in 2018 by the Toronto-based studio TRU LUV, #SelfCare is a free app with no
advertisements.42 The app, which the studio calls a ‘companion’ to differentiate it from
for-profit sleep and health apps, is intentionally simple. It consists of several short pastimes,
among them: a slow breathing exercise, petting a cat by gently stroking the screen with one’s
fingers, or idly sorting a digital laundry basket. The homepage portrays a domestic scene
familiar to any mobile user: a person lying in bed, tucked under a white blanket, with a single
hand gently pulsing on the pillow while a cat lies on the left bottom of the bed (see figure 4).
The bedroom, which is seen from a bird’s eye view, is bright and minimally furnished with a
white rug, a laundry basket, and two bedside chests with flowers, cards, and several other
items. A smartphone is tossed on the left pillow, barely out of reach from the half-asleep,
half-awake protagonist.
Figure 4: The homepage of the app #SelfCare (left), and a screenshot of a word game featuring the word
‘insomnia’ from the same app (right). Source: screenshots taken by the author.
While users can change the skin color of the sleeping protagonist via the settings menu, the
default color is dark. This design decision differentiates #SelfCare from other products of the
soporific media industry, which, as I explored elsewhere, assumes the average user is white
and able-bodied.43 The app’s imaginary user, as described by its developers, is someone who
42 Kara Stone, ‘Designing Self-Care Affect and Debility in #SelfCare’, in Arno Görgen and Stefan Heinrich
Simond (eds) Krankheit in Digitalen Spielen: Interdisziplinäre Betrachtungen, Bielefeld: transcript, 2020,
p. 424.
43 Neta Alexander, Interface Frictions, Durham: Duke University Press, forthcoming. For a critique of the
127
IN/CONVENIENCE
struggles with insomnia, anxiety, or depression. At the same time, its minimalist design and
relaxing, scoreless games can help a plethora of users reassess their relationship with their
phones. A user who gets bored with a breathing exercise and attempts to leave the app by
pressing the menu in the top right corner instead encounters a screen suggesting, ‘let’s stay
here a bit longer’—drawing attention to the inconvenience of not having full control over the
phone. While it is still possible to leave the app, these design decisions break the automated
bodily reactions to boredom and frustration. After a few short games, the app encourages
the user to turn off their phone entirely—an alternative sleep-inducing approach to that of
Calm and other sleep apps, which offer an endless stream of sleep stories and white noises
requiring to keep one’s phone at arm’s length.
If Calm rewrites personal histories of pain and sickness as shooting lullabies for weary bodies
and minds, #SelfCare normalizes depression, fatigue, and burnout by offering a single home
page featuring an insomniac person hiding under the covers and tossing from side to side. As
the app tells us, this person ‘refuses to leave bed today’. The mostly concealed protagonist
is not a tranquil and healthy user about to rest and recharge. This human body is bound to
remain horizontal as long as the app is turned on. For Canadian game designer and disability
scholar Kara Stone, the choice not to include a script of waking up or going to work is what
makes the app successful: ‘#SelfCare has not set out to cure all debility or negative feelings,
but to change our relationship with our phones’.44 In TRU LUV’s parlance, this is achieved by
degamifying the more familiar health and sleep apps: ‘Our goal is simply to feel better. There’s
no winning, no failure, no score. No difficulty, no ads, no notifications. There is just us and
our feelings’.45 These feelings, the app tells us, might entail, ‘I don’t want to get back to my
emails,’ ‘I feel disconnected’ or, ‘I feel anxious and alone despite the affordances of conve-
nience media’. It then invites the user to recognize these feelings, for example, by playing a
word game in which she is asked to complete words such as ‘is_lation’ or ‘insom_ia’, only so
she can sit with them for a while instead of clicking her pain away. As Stone argues, ‘The goal
of the game is not to overcome depression, a common narrative arc concerning disability, but
to create rituals that make life sustainable’.46 One such ritual could be putting one’s phone
away after briefly reflecting on how it might disturb sleep and worsen one’s disconnection
from one’s body. By depicting an image of an insomniac person who tosses from side to side
next to their phone, the app associates phone use with sleeplessness, providing an alternative
to Calm’s interface or to Netflix’s and YouTube’s autoplay-induced rabbit holes. Here, to fall
asleep, one must first get rid of one’s phone and be reminded of one’s bodily needs, even when
these needs might invoke undesired and inconvenient sensations such as pain and rumination.
assumed whiteness of mediation and sleep apps’ users, see Lida Zeitlin-Wu, ‘Meditation Apps and
the Unbearable Whiteness of Wellness’, Just Tech, 1 November 2023, https://just-tech.ssrc.org/field-
reviews/meditation-apps-and-the-unbearable-whiteness-of-wellness/.
44 Stone, ‘Designing Self-Care Affect and Debility in #SelfCare’, p. 424.
45 Qtd. in Stone, ‘Designing Self-Care Affect and Debility in #SelfCare’, p. 417.
46 Stone, ‘Designing Self-Care Affect and Debility in #SelfCare’, p. 425.
128 THEORY ON DEMAND
Coda: Covid Nightmares
My hypersensitivity to light makes me prone to fatigue. When woken up by artificial or natural
night, I am seldom able to fall asleep again. In the resulting moments of forced wakefulness,
leaving my bedroom and turning on my phone amounts to surrender, an embracing of the
fact that the night is lost. Transforming the individual act of sleep into a communal reflection
on the conditions that hinder or support it, I wish to conclude this analysis of the soporific
media industry with a close look at a recent artwork.
In COVID Sleep, an art piece comprised of fifteen short videos, Dayna McLeod produces
night-vision surveillance footage of herself and her girlfriend sleeping (or failing to sleep). The
project was conceived as part of a virtual residency inviting ten Canadian artists to reflect
on the idea of ‘being alone together’ during the pandemic lockdowns, and the footage was
shot over the course of the sixty-day residency from April to May 2020. As McLeod describes
it, ‘I’ve always had sleep disturbances like nightmares, sleepwalking, and night terrors, but
[I] didn’t realize their extent and frequency until I started these recordings. […] I gasp, yell,
talk, scream, and otherwise ask questions while asleep’.47 While the project was originally
presented as a gallery installation, the fifteen video pieces are available on YouTube. The
short segments, which last between ten to forty seconds, capture McLeod as she wakes up
from a nightmare, involuntarily moves her arm, and talks in her sleep. They present us with
a depiction of a vulnerable, restless human body (see figure 5).
Figure 5: A frame from Dayna McLeod’s COVID Sleep, depicting the artist as she abruptly wakes up from
a nightmare at 23:43 p.m. (source: Courtesy of Dayna McLeod)
With its glimpses of McLeod’s sleeping girlfriend, whose body is farther away from the camera
and is often concealed by McLeod’s torso and facial expressions, this project intentional-
47 Covid Sleep (dir. Dayna Mcleod), viewable at: https://www.youtube.com/playlist?list=PLCxxIPfj-
5HRU933P6lOSoSQZJPASoBzJ.
129
IN/CONVENIENCE
ly draws on previous cinematic explorations of queer intimacy. Durational works like Andy
Warhol’s Sleep (1964) famously inquired what might happen to this most intimate act when
exhibited in public, depicting a nonsexual mode of care and vulnerability.48
In her artist’s statement, McLeod explains she employs night vision because its aesthetic
features consist of ‘pixilation, blurriness, perspective distortion, grain, and lighting’.49 This
low-resolution aesthetics reframes McLeod’s abrupt movements as shots from a horror film.
It is not just queer intimacy that McLeod invokes, but rather the suspenseful, uncanny sense
of a living creature being caught on hidden camera in the dead of night. I read McLeod’s work
as an attack on surveillance capitalism and the notion that, to improve one’s quality of sleep,
we need to subject ourselves to technologies that monitor our circadian rhythms.
McLoed’s cramped bedroom conveys claustrophobia and paranoia. It features multiple tech-
nologies, from a television screen in the very back of the frame to an iPad that McLeod holds
before going to bed, as well as a phone on her night bed and, of course, the camera docu-
menting her and her partner throughout the night. Yet these technologies of convenience all
fail to provide what McLeod seems to seek: an undisturbed night’s sleep, allowing her body
to fully rest and recharge. Watching her nocturnal struggles to go back to sleep reminded me
not only of my own sleepless nights, but also of the limitations of the design features promising
to tuck me into bed.
Inviting us to reexamine our dependency on screens, crip horizontal media offers radical inter-
pretations of convenience, vulnerability, care, and rest. McLeod and the creators of #SelfCare
invite us to reflect on the anxieties associated with sleep and its lack—aging, sickness, pain,
vulnerability, and uncertainty. They provide a creative space from which to conjure alterna-
tives to the surveillance industry tracking users’ sleep and wakefulness patterns by asking con-
venience for whom, and at what price? The soporific media industry, with its growing number
of apps, design features, and products, worsens insomnia by making people obsessed with
optimizing their sleep and achieving ‘the perfect slumber’.50 Rejecting the impulse to quantify
sleep, crip horizontal media considers how this basic human need is shaped by technological,
psychological, and socio-political conditions. By cripping convenience, these case studies
insist that inconvenience cannot be entirely eliminated; in constant flux, the human body will
forever remain a source of both comfort and pain, ease and distress.
Sleep apps and the convenience tech they are premised on are a part of a ‘wellness industry’
promising to better align our bodies and circadian rhythms with the relentless demands of 24/7
capitalism. What they do, however, is to nudge sedentary phone users to lie down. The irony
is that this postural change is produced by, and is dependent on, one’s electronic device—
48 See Branden W. Joseph, ‘The Play of Repetition: Andy Warhol’s Sleep’, Grey Room 19 (Spring 2005):
29–33.
49 Qtd. in Lorenza Mezzapelle, ‘Parallel Lines Considers What It Means to Be Alone, Together’, The
Concordian, 12 May 2012, https://theconcordian.com/2020/05/parallel-lines/.
50 Brian X. Chen, ‘The Sad Truth About Sleep-Tracking Devices and Apps’, New York Times, 17 July 2019,
https://www.nytimes.com/2019/07/17/technology/personaltech/sleep-tracking-devices-apps.html.
130 THEORY ON DEMAND
the very same device that is likely to hinder sleep due to overstimulation and light-emitting
screens. The postural is the habitual, and as such it requires much closer attention.
References
Alexander, Neta. Interface Frictions, Durham: Duke University Press, Forthcoming.
Al Mahmud, Abdullah, Wu, Jiahuan, and Mubin, Omar. ‘A Scoping Review of Mobile Apps for Sleep
Management: User Needs and Design Considerations’, Frontiers in Psychiatry, 18 October 2022,
https://www.frontiersin.org/articles/10.3389/fpsyt.2022.1037927/full.
Andersen, Joceline. ‘Now You’ve Got the Shiveries: Affect, Intimacy, and the ASMR Whisper Communi-
ty’, Television & New Media 16.8 (1 December 2015): 683–700.
Enger, Reed. ‘Without Hope’, Obelisk Art History, 30 July 2017, http://arthistoryproject.com/artists/
frida-kahlo/without-hope/.
Chen, Brian X. ‘The Sad Truth About Sleep-Tracking Devices and Apps’, New York Times, 17 July 2019,
https://www.nytimes.com/2019/07/17/technology/personaltech/sleep-tracking-devices-apps.html
Covid Sleep (dir. Dayna Mcleod), viewable at: https://www.youtube.com/playlist?list=PLCxxIPfj-5HRU-
933P6lOSoSQZJPASoBzJ.
Crary, Jonathan. 24/7: Late Capitalism and the Ends of Sleep, New York: Verso Books, 2013.
Curry, David. ‘Calm Revenue and Usage Statistics (2022)’, Business of Apps, 1 July 2022, https://www.
businessofapps.com/data/calm-statistics/.
Hess, Amanda. ‘The App That Tucks Me in at Night’, New York Times, 17 July 2019, https://www.
nytimes.com/interactive/2019/07/17/arts/calm-app-sleep-meditation.html.
Huberty, Jennifer; Green, Jeni; Glissmann, Christine; Larkey, Linda; Puzia, Megan; and Lee, Chong.
‘Efficacy of the Mindfulness Meditation Mobile App ‘Calm’ to Reduce Stress Among College Students:
Randomized Controlled Trial’, JMIR Mhealth Uhealth 7.6 (25 June 2019): e14273.
Huffington, Arianna. The Sleep Revolution: Transforming Your Life, One Night at a Time, New York:
Harmony Books, 2016.
Guaraná, Bruno. ‘At the Edges of Sleep’, Film Quarterly 76.2 (Winter 2022): 107-111.
Joseph, Branden W. ‘The Play of Repetition: Andy Warhol’s Sleep’, Grey Room 19 (Spring 2005):
29–33.
Kennedy, Pagan. ‘The Insomnia Machine’, New York Times, 18 September 2016, https://www.nytimes.
com/2016/09/18/opinion/sunday/the-insomnia-machine.html.
Kroker, Kenton. The Sleep of Others, Toronto: University of Toronto Press, 2007.
Lopez, Kristen. ‘How Movie Theaters Are Failing Viewers with Disabilities’, IGN, 3 June 2018, https://
www.ign.com/articles/2018/06/03/how-movie-theaters-are-failing-viewers-with-disabilities.
Lowrey, Annie. ‘The App that Monetized Doing Nothing’, Atlantic, 4 June 2021, https://www.theatlantic.
com/technology/archive/2021/06/do-meditation-apps-work/619046/.
Ma, Jean. At the Edges of Sleep: Moving Images and Somnolent Spectators, Berkeley: University of
California Press, 2022.
Mezzapelle, Lorenza. ‘Parallel Lines Considers What It Means to Be Alone, Together’, The Concordian,
12 May 2012, https://theconcordian.com/2020/05/parallel-lines/.
Miserandino, Christine. ‘The Spoon Theory Written by Christine Miserandino’, But You Don’t Look Sick,
25 April 2013, https://butyoudontlooksick.com/articles/written-by-christine/the-spoon-theory/.
131
IN/CONVENIENCE
Mulvin, Dylan. ‘Media Prophylaxis: Night Modes and the Politics of Preventing Harm’, Information &
Culture 53.2 (2018): 175-202.
Neves, Joshua; Chia, Aleena; Paasonen, Susanna; and Sundaram, Ravi. Technopharmacology, Minne-
apolis: University of Minnesota Press, 2022.
Raphael, Rina. ‘Netflix CEO Reed Hastings: “Sleep Is Our Competition”’, Fast Company, 6 November
2017, https://www.fastcompany.com/40491939/netflix-ceo-reed-hastings-sleep-is-our-competition/.
Schüll, Natasha. ‘Data for Life: Wearable Technology and the Design of Self-Care’, BioSocieties 11, no.
3 (March 2016): 6.
Sterne, Jonathan. Diminished Faculties: A Political Phenomenology of Impairment, Durham: Duke
University Press, 2022.
Stone, Kara. ‘Designing Self-Care Affect and Debility in #SelfCare’, in Arno Görgen and Stefan Heinrich
Simond (eds), Krankheit in Digitalen Spielen, Interdisziplinäre Betrachtungen, 2020: 417-432.
‘The History of Drive-in Movie Theaters’, New York Film Academy, 7 June 2017, https://www.nyfa.edu/
student-resources/the-history-of-drive-in-movie-theaters-and-where-they-are-now/.
White, Michele. The Body and the Screen: Theories of Internet Spectatorship, Cambridge: MIT Press,
2006.
Yates, Diana. ‘Study of Sleep Apps Finds Room for Improvement’, Illinois News Bureau, 12 April 2017,
https://news.illinois.edu/view/6367/486860.
Zeitlin-Wu, Lida. ‘Meditation Apps and the Unbearable Whiteness of Wellness’, Just Tech, 1 November
2023, https://just-tech.ssrc.org/field-reviews/meditation-apps-and-the-unbearable-whiteness-of-well-
ness/.
Zraick, Karen and Mervosh, Sarah. ‘That Sleep Tracker Could Make Your Insomnia Worse’, New York
Times, 13 June 2019, https://www.nytimes.com/2019/06/13/health/sleep-tracker-insomnia-orthosom-
nia.html.
132 THEORY ON DEMAND
IN/CONVENIENCES OF MOBILE PAYMENTS:
“ALTERNATIVE DATA” AND THE DISTRIBUTION
INFRASTRUCTURES OF LOAN APPS
RAHUL MUKHERJEE
A loan app advertisement in contemporary India begins with a delivery worker on a motorbike,
making use of his phone in a parking lot amid sounds of Diwali firecrackers. An ‘applied’
notification with the KreditBee loan app icon flashes for audiences.
Figure 1: Loan Applied Notification. Screengrab by author of KreditBee commercial on YouTube.
After completing this task, the delivery worker wishes people ‘Happy Diwali’ in a selfie video.
He makes known that he has completed his work, but one surprise delivery still remains. The
selfie video recording began just as the loan app application had been submitted. The video
timer draws the viewer’s attention. The audiences can see the first selfie video recording
ending at around the 20 second mark as the ad protagonist begins his final surprise delivery.
Just as he puts his helmet and is about to start the motorbike, the delivery worker receives
another notification from KreditBee that says ‘Approved’ with a green tick. So, we can infer,
the loan app approval took about 20 seconds.1
1 KreditBee, ‘Celebrate Diwali 2022 with #KhushiyonKaLoan’, Loan app ad, Youtube, 12 October 2022,
https://www.youtube.com/watch?v=QCrf-IGKk5M.
133
IN/CONVENIENCE
Figure 2: Loan Approved Notification. Screengrab by author of KreditBee commercial on YouTube.
As he moves from shop to shop through the city on his bike, buying sweets and a necklace,
wrapped in elegant gift boxes safely put in his delivery bag mounted on the back of his vehi-
cle, the soundtrack containing the following words suffuses the scene ‘Khushiyon Ka Tyohar
Manana Hain’ (‘Festival of Happiness needs to be celebrated’). This connotes that he is
buying gifts as part of celebrating the Diwali festival. At a saree shop, the shopkeeper asks
between what monetary range should he show sarees to which the protagonist replies that he
would like to see the best quality saree as he does not have to worry about ‘budget’ this Diwali
because his loan was approved ‘Jhath Se’ (instantly) by KreditBee. We see the timestamps
as he picks up perfume at 15 mins 51 seconds and a saree at 25 mins 16 seconds, before
reaching home with the delivery of gifts for (what seems to be) his wife at 38 mins 26 seconds.
This ad emphasizes the convenience of taking out loans through mobile phone apps with-
out the hassles of taking paper documents to brick-and-mortar banks, waiting for in-person
appointments, or being scrutinized, humiliated and denied loans during the appointment for
not having proper documentation. One can apply from a parking lot with a mobile internet
connection just putting together some know-your-customer (eKYC) ‘digital’ documents for
identity verification and credit risk assessment, and the money will reach the bank account
of the applicant in 10 mins. This is within the time duration that the delivery worker has
reached the first shop to start buying gifts for this wife. The ad notes that KreditBee stands
for ‘10-minute loans,’ ‘100% online documentation,’ and ‘flexible repayment plans’. The loan
app is part of a platformized convenience economy which believes in providing seamless and
instant services to customers, reducing their barriers to access. After all, according to Shep
Hyken, ‘reducing friction’ between the customer and the product or service experience, is
the key principle of ‘convenience revolution’.2 Indeed, we find an instantiation of such a
2 Shep Hyken, The Convenience Revolution: How to Deliver a Customer Service Experience That Disrupts
134 THEORY ON DEMAND
professed value/experience in the KreditBee tagline that reads: ‘Loans anytime anywhere’.
This emphasizes the ubiquity and comfort of their service. In the loan ad, it is evening time
and the delivery worker has been probably busy all day delivering goods and food. Even if the
banks have closed and he is at a parking lot, he can still get a loan.
Loan Apps and Platformed Convenience
A crucial aspect of the above-mentioned ad is how it integrates loan apps with the other quo-
tidian practices of the platform economy. Firstly, what should not escape notice is that the loan
app is itself an intermediary between the lending institution and the borrower, building on the
classic logic that today’s platform companies are merely intermediaries between multi-sided
markets and/or stakeholders. The disclaimer that follows the ending of the ad clarifies this
aspect: ‘the loan is subject to the credit policy and terms & conditions of the registered NBFCs/
Banks [lender(s)] participating on the KreditBee platform’. This suggests that KreditBee is
mediating between lenders and borrowers. It further adds that the disbursement of the loan
within 10 minutes is not guaranteed and is subject to whether the applicant’s details comply
with the conditions of the loan.3
This is a negotiation of roles and responsibilities by KreditBee as it takes on the ‘platform’
label. Tarleton Gillespie has written about YouTube deploying the trope of the ‘platform’ as a
cultural intermediary in its efforts to discursively cater to the at times contradictory interests
of its different constituencies: users, content producers, and advertisers.4 KreditBee as a
financial intermediary platform mediates between lenders and borrowers who have different
views about what interest rates on loans should be as well as what repayment schedules are
acceptable.
The problematic pricing of the loans and the settlement times have raised lots of concerns
regarding the digital loans architecture because this is where the inconvenience of the loan is
found. If the loan comes ‘anytime anywhere’ the repayment comes at the bidding of the bank
and the app—and therein lies the twin of convenience: inconvenience. This entwined rela-
tionship between convenience and inconvenience is captured by this collection’s editors as
‘in/convenience’. The question about interest rates/settlement times is not just for KreditBee,
but also for other financial technology (fintech) intermediaries like Paytm, MoneyTap, and
Navi. All these digital lending platforms have the same answer to this question (about interest
rates and settlement times) that they are merely intermediaries who mediate between lenders
and borrowers, and facilitate faster customer verifications and loan disbursals. The company
Paytm had a meteoric rise as a digital payments company during India’s 2016 demonetization
(notebandi) drive, and then branched into digital lending. When asked about interest rates by
a financial journalist, the Paytm spokesperson replied, ‘Regarding the interest rates, they are
the Competition and Creates Fierce Loyalty, Shippensburg, PA: Sound Wisdom, 2019, p. 16.
3 KreditBee’s lending partners listed on their website include Piramal Finance, Tata Capital Financial
Services Limited, Vivriti Capital Limited, and Poonawalla Fincorp Limited among others. See, KreditBee
Digital Lending Partners, https://www.kreditbee.in/digital-lending-partners.
4 Tarleton Gillespie, ‘The politics of “platforms”’, New Media & Society 12.3 (2010): 347-364.
135
IN/CONVENIENCE
charged by the lenders and not us. We are only distributors of the products that our lending
partners offer.’5 This admission by Paytm of being ‘distributors’ of financial products (loans/
money) is key because distribution is where the impact of platform revolution, whether in
streaming film/TV content or e-commerce delivery, has been arguably felt the most.6 How,
then, could digital money distribution be left out of platformization? And what does it mean
to put the distribution of money back into critical accounts of platformization?
Secondly, the integration with the platform economy is emphasized in the ad by suggesting
how swift distribution of money (through the KreditBee platform in about 10 mins) works with
efficient delivery of gifts by the delivery worker in 38 mins. The repeated invocation of time-
stamps throughout the advertisement should not be underestimated. The platform economy
is after all an on-demand economy, with platform services ranging from on-demand streaming
content to on-demand food delivery. KreditBee seems to indicate that it provides on-demand
money/loans, which facilitates on-demand delivery of other services. The promise of an instant
loan in 10 mins resonates with the promise of quick commerce firms such as Zepto in India
promising 10 mins food deliveries. These faster end-to-end deliveries require restructuring of
urban logistical infrastructures evident in the rise of ‘dark stores’ and ‘ghost kitchens’ along
with anxieties about inconveniences for the delivery worker, including increased risks of road
accidents or vehicular wear negotiating potholes.7 While the 10 mins food delivery model
has sparked debates about storage facilities and transportation conditions, the 10 mins loan
delivery model has not received the same attention. Movement of money perhaps needs
as much attention as movement of food and television content. After all, as Lana Swartz
and David Stearns write, ‘innovations in money and payments during the modern era have
tracked alongside changes in the underlying communications infrastructure’, whether it is
paper checks and money orders sent through paper mail and express shipping, or digital
payments through/into ATM cards, e-wallets, and loan apps.8 Money movement and goods
movement proceed apace.
5 Cited in Ashwin Manikandan, ‘Instant Digital Loans: convenience comes with a heavy cost’, The
Morning Context, 29 August 2023, https://themorningcontext.com/internet/instant-digital-loans-
convenience-comes-with-a-heavy-cost.
6 On demonetization and platformization, see Adrian Athique, ‘A great leap of faith: the cashless agenda
in Digital India’, New Media & Society 21.8 (2019): 1697-1713.
7 Noopur Raval, ‘Instant-Instant Noodles: How algorithmic platforms transform food, taste and
reproduction in cities’, Penn Cinema & Media Studies Colloquium, 7 September 2022, https://pricelab.
sas.upenn.edu/events/instant-instant-noodles. For ghost kitchens, refer to Aaron Shapiro, ‘Platform
urbanism in a pandemic: Dark stores, ghost kitchens, and the logistical-urban frontier’, Journal of
Consumer Culture 23.1 (2022): 168-187. Also see Abhirup Roy and Aditya Kalra, ‘India’s Zomato faces
heat for plan to deliver food in 10 mins’, Reuters, 22 March 2022, https://www.reuters.com/world/india/
indias-zomato-faces-heat-plans-deliver-food-10-minutes-2022-03-22/. On ten minute delivery work,
see Naandika Tripathi, ‘On the trail of how 10-minute delivery works—and doesn't’, Forbes India, 1 July
2022, https://www.forbesindia.com/article/take-one-big-story-of-the-day/on-the-trail-of-how-10minute-
delivery-worksand-doesnt/77761/1.
8 Lana Swartz and David Stearns, ‘Money and its Technologies: Making Money Move in the Modern Era’,
in Taylor C. Nelms and David Pedersen (eds) A Cultural History of Money in the Modern Age, London:
Bloomsbury Academic, 2019, pp. 27-52.
136 THEORY ON DEMAND
Convenience is about ‘timing’ as Elizabeth Shove has so convincingly written and Joshua
Neves and Marc Steinberg reiterate in their chapter. Convenience is the ‘ability to shift and
juggle obligations and to construct and determine personal schedules’.9 The delivery worker
is part of the gig economy that runs on (or adapts itself to) flexible work schedules. As such,
the delivery worker seems to both be able to complete his delivery work by evening and still
find time to enjoy the festival of Diwali, thanks to KreditBee’s instant loan app service. That
the worker is both an on-demand consumer and an on-demand producer is itself of note. The
delivery worker himself is a key part of India’s e-commerce and food delivery app sector, an
ideal lower middle-class customer for the small loans that KreditBee offers. These are small
size loans, with the maximum amount being 5 lakh rupees, with interest rates that can vary
from 12-25% for KreditBee (and between 10-48% for the digital lending industry as a whole)
based on some research by The Morning Context.10
I have used the KreditBee loan app advertisement circulating in India around October 2022 to
illustrate the state of the platformized convenience economy in India. As the ad demonstrates,
convenience is both an ideology of contemporary ‘platform capitalism’ (or as some would
say ‘surveillance capitalism’) as well as a temporal conditioning that involves experiencing
on-demand deliveries of gifts, food, and money, juggling schedules in the now, and antici-
pating uncertain futures.11
Lending initiatives by more established players such as KreditBee and Paytm are part of
a recent change in the loan app landscape in India which from early 2020 till mid-2022
witnessed the dominance of illicit loan apps which charged obscenely high interest rates
and processing fees, and deployed predatory collection tactics. While the legitimate loan
apps do not necessarily use data to target borrowers in a predatory, abusive way and they do
provide disclaimers and caveats to their loan promises, the question of what happens in the
backend—to support seamless loan approvals in the frontend loan interface—still very much
remains. Concerns about user/customer data extraction and surveillance in the loan approval
and disbursement process exist even as fintech enthusiasts tout the work of AI and machine
learning in calculating credit scores from alternative data. Furthermore, the discursive politics
of lending platforms elides any significant discussion about who has access to this customer
data and where it goes. These concerns about the use of data as collateral were heightened
during the reign of the rogue loan apps especially amidst the Covid-19 pandemic, when they
lured desperate customers (without pay and/or jobs) to enlist. During the loan onboarding
process, these predatory loan apps gathered the WhatsApp contact details of customers and
harassed borrowers by notifying their relatives when they defaulted on their loans.
9 Elizabeth Shove, Comfort, Cleanliness and Convenience: The Social Organization of Normality. New
Technologies/New Cultures Series. Oxford, UK: Berg, 2003, p. 171.
10 Manikandan, ‘Instant Digital Loans’.
11 Refer to Jenny Huberman, ‘Amazon Go, Surveillance Capitalism, and the Ideology of Convenience’,
Economic Anthropology 8.2 (2021): 337-349. On surveillance capitalism, see Shoshana Zuboff, The
Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power, London:
Profile Books, 2019. Regarding platform capitalism, see Nick Srnicek, Platform Capitalism, Malden,
MA: Polity, 2016.
137
IN/CONVENIENCE
Since 2016, the Government of India offered the Unified Payments Interface, and with demon-
etization, fintech start-ups provided financial services from payments and lending to savings
and insurance. These could all be performed through mobile phone apps with a click/swipe/
tap. This notion of ‘instantaneous customer satisfaction’ where Indian state/corporate ser-
vices could be summoned by the consumer-citizen at their convenience became part of the
‘start-up state’.12 Digital payments and lending were seen as trade-off between ‘platformed
convenience’ of financial activities and concerns about privacy/surveillance.13
I focus on the connection between the frontend interface and the backend mobile money
distribution infrastructure. The backend infrastructure makes possible this experience of
convenience while interacting with the app interface/screen. With fintech platforms, an eco-
system of human/non-human intermediaries (data brokers, analytics providers, and software
development kit (SDK) libraries) seemed to replace the traditional in-person moneylender,
agent and brick-and-mortar bank. These new intermediaries collect and trace (read: surveil)
monetary transaction data and behavioral data pertaining to social media activity, which is
termed ‘alternative data’. Predatory loan apps proliferated with invasive surveillance sys-
tems: if debtors failed to pay off their loans, their WhatsApp contacts received messages and
phone calls. Third-party developers plug in their SDKs into loans apps to facilitate credit risk
assessment and eKYC checks. To understand the role of third-party developers in moving
and using customer data requires examining the fintech infrastructure. An infrastructural
analysis suggests that the inconveniences of mobile money scams or loan app repayments
are not exceptions to, but rather constitutive of digital economy.14
It is important to trace the history of digital payments and lending in India and how they
intersect with the growth of platform economies. Examining infrastructural aspects of loan
apps, I argue for a multi-situated approach to understanding how such apps are situated
within varied infrastructural relations. This multi-situated study of loan apps connects the
temporal conditioning of users experiencing the instantaneity and easy availability of loans
and the infrastructural relations across various stakeholders of the loan app ecosystem that
make it possible to have such a customer experience of convenience. My goal is to then shift
to inconveniences associated with these loan apps, building on Joshua Neves and Marc
Steinberg’s insight that the inconvenience is deeply related to and often the flip side of con-
venience.15 One kind of inconvenience comes from financial misinformation/disinformation
that scamsters are using to deliberately deceive or lure customers into taking predatory loans
and then coercing them to pay exorbitant interest rates. Another kind of inconvenience occurs
when even legitimate lending platforms obfuscate how borrower (alternative) data is collected,
stored, and processed, and elide information about how interest rates are decided.
12 Vijayanka Nair, ‘Governing India in Cybertime: Biometric IDs, Start-Ups, and the Temporalized State’,
South Asia: Journal of South Asian Studies 42.3 (2019): 519-536.
13 Joshua Neves and Marc Steinberg, ‘Pandemic Platforms: How Convenience Shapes the Inequality of
Crisis’, in Philipp Dominik Keidl et al (eds) Pandemic Media: Preliminary Notes Towards an Inventory,
Lüneberg: meson press, 2020, pp. 105-112.
14 Winifred R. Poster, ‘Introduction to special issue on scams, fakes, and frauds’, New Media & Society
24.7 (2022): 1535-1547.
15 Joshua Neves and Marc Steinberg, ‘In Convenience’, in this volume.
138 THEORY ON DEMAND
Digital Financial Inclusions and Exclusions in India
Enabled by the Aadhaar system, the Unified Payments Interface (UPI), and the India Stack
infrastructure, interlinked digital platforms allow for smooth transfers across bank accounts
for India’s consumer-citizens. Both the government and fintech companies have proclaimed
that it has become possible for ‘unbanked’ Indians to begin to have a credit history and be
part of the digital lending and borrowing system (The Economist, 2018). Previously in India,
only members of the formalized workforce who had bank accounts, PAN (Permanent Account
Number associated with taxpayer) cards, and a rich credit history could get formal credit, but
now credit distribution could potentially be democratized to include small businesses, the
urban and rural poor, and students. From 2015 to 2019, as digital payments became more
frictionless, instantaneous, and ubiquitous (all synonyms for convenient), digital lending grew
as well. This increased spread of digital lending in the name of “financial inclusion” has been
welcomed by some commentators, while others suggest that it has caused increasing debt
among borrowers and monitoring of customers’ personal and social data for credit scores.16
The government and fintech corporations see this as an opportunity to serve vast swaths of
marginalized Indian customers who were previously excluded from formal banking systems
but now can be included because they have a mobile phone connected to the Aadhaar iden-
tification system (also known as unique identification or UID system). Others see this rise of
digital/mobile money in India as consistent with a wider global trend, where ‘digitization of
payments enables the digitization of finance more broadly’, with partnerships between Big
Tech and Big Finance emerging as fintech and championing a cashless economy and society.17
Before turning to predatory loan apps, which mushroomed amidst the Covid-19 pandemic of
2020, I want to briefly mention some of the more legitimate apps which were growing in the
Indian finance markets since 2016. When it went live in December 2016, mPokket became
one of the first successful microloan apps meant for students who needed urgent cash. Most
students in India were ineligible for credit cards, and they were too young to have saved any-
thing substantial, which meant they were likely to face cash crunches. For the small loans
they required, the banks and other traditional financial institutions did not consider them
worthy customers to invest in. mPokket made it simple for students to sign up: they could
find the microloan app in the Google Play store, download it, register through their Google or
Facebook account, and then upload eKYC documents. Once verified (in most cases, within an
hour), they became eligible for loans. The ticket size for these loans was small, from Rs 500
to Rs 20,000. Timely loan repayments help to increase the borrowing limit next time.18 Being
16 Kevin Donovan and Emma Park, ‘Algorithmic Intimacy: The Data Economy of Predatory Inclusion
in Kenya’, Social Anthropology 30.2 (2022): 120-139; Tarunima Prabhakar, A New Era for Credit
Scoring: Financial Inclusion, Data Security, and Privacy Protection in the Age of Digital Lending, UC
Berkeley Center for Long Term Cybersecurity White Paper, 2020, https://cltc.berkeley.edu/wp-content/
uploads/2020/06/A_New_Era_for_Credit_Scoring.pdf.
17 Brett Scott, Cloudmoney: Cash, Cards, Crypto, and the War for our Wallets, London: Penguin, 2022.
18 Vishal Krishna, ‘This startup clocked $9M in revenue by lending to just students and the self-employed’,
YourStory, 30 March 2021, https://yourstory.com/2021/03/kolkata- based-lending-platform-mpokket-
students-instant-loan.
139
IN/CONVENIENCE
savvy users of smartphones, students were a readily accessible market for these microloan
apps whose primary channel of outreach is the mobile phone.
Once consumers started using smartphones for borrowing, the question for lenders became:
How does one gather so-called alternative data (about which more below)? The alternative
data collection and scoring practices on which creditworthiness is being evaluated remain
opaque. After all, mPokket—or, for that matter, any microloan app—is not simply basing the
borrowing limit on prior repayment history. Comprehending how phone surveillance and data
collection practices by data brokers have changed from 2015 onward is key to understanding
credit ratings and digital fintech processes. This is where a whole ecosystem of players like
alternative risk-score providers come into the picture. Alternative data storage and processing
is linked to the conveniences and inconveniences of payments.
Alternative Data and the Inconvenience of Repayments
In a country like India where credit scores are virtually non-existent for the majority of the
population and where smartphones and mobile internet data plans have dispersed so widely
among the population, it obviously becomes enticing for the fintech industry to examine all
kinds of unstructured (behavioral) data which they are now able to gather because of the
digital traces left by millions of Indian users on the mobile internet.19 Fintech companies
are trying to devise algorithms so as to gauge risks and consumer intent better. In fintech
conferences, one repeatedly hears about the responsibility that loan apps have to open the
black box of AI and algorithms, and clarify the decisions made by algorithms to their lenders
and borrowers. But how much of this is actually done in practice remains a question. At times,
loan apps and third-party developers employ a very technical vocabulary dropping terms
like ‘decision trees’ and ‘neural networks’, and during other times, mention one provocative
variable in determining borrower credibility. In a panel discussion regarding credit ecosystem
in the Payments and Lending Conference 2022, the CEO of a loan app company EarlySalary
(now called Fibe), Akshay Mehrotra, explains ‘alternative data’ in the following manner:
Data is actually very large […] millions of data points coming in and we taking out
variables out of it […] how fast does a customer answer the questions. […] In [the] traditional
world, you could not measure because a person was filling up your form on your behalf and
he would always put the right answer. […] In a digital world, let’s say on EarlySalary mobile
app, we are measuring every extreme data of the consumer and that goes on into a variable to
say [that] if you look at in a larger sense […] people in a hurry are more desperate for money
while people who wait out and answer the question correctly are more patient and will be more
responsible with the money given to them. Now, you can use it as a new variable or you can say
it is a judgmental answer. But over time, it comes as a result to say it adds up to value or not.20
19 The World Bank Report 2018 finds formal/digital/mobile phone credit use among less than 10% of
Indian population. While mobile phone use is growing, mobile phone credit use may not be increasing
at the same rate, and hence the scope for improvement. See ‘Financial Inclusion Data / Global Findex’,
World Bank, accessed 2 April 2024, https://datatopics.worldbank.org/financialinclusion/country/india.
20 See Akshay Mehrotra’s speech in IBS Intelligence, ‘Reimaging the Credit Ecosystem,’ Panel Discussion,
Payments & Lending Conference, 24 March 2022, Youtube, 11 April 2022, https://www.youtube.com/
watch?v=oW3S8J58LWo.
140 THEORY ON DEMAND
So, here Akshay Mehrotra notes that as a user is filling up online details in EarlySalary app in terfa ce,
their movements are being measured to gather behavioral data. A borrower typing in a hurry
and making mistakes is being considered desperate in the way this behavioral variable is being
operationalized for the algorithm. Another customer who has a steady approach to filling in
details in an unhurried manner with accuracy is considered a more dependable borrower.
Mathur adds two caveats. One, that this could be considered ‘judgmental’ by some, but he
adds that they will be testing the results from this variable over a period of time, retesting it
again and again. Two, another caveat which Mehrotra implicitly gestures to while explain-
ing the scenario is that this (un)hurried user movement on the loan interface and their (in)
accuracy rate of filling details is one among many variables that is used determine the loan
candidacy. Much of this alternative data tends to be behavioral data, and while the loan app
itself promotes immediacy and instantaneity (conveying convenience and slickness), its algo-
rithms gathering ‘alternative data’ might negatively judge a user’s hurried finger movements
on the app interface, more so if they make mistakes while imputing the information. Ironically,
it seems a loan applicant may be more likely to receive the loan if they do not believe in the
availability, ease and convenience the app interface and discourse around the app promises.
A key player in the Indian fintech space with regard to credit scoring is Lenddo. In its initial
days, Lenddo tested its algorithms for credit scoring by offering small loans in the Philippines
around 2011. Soon after, it quit offering loans and positioned itself as a credit-rating company.
Thereafter, Lenddo sold its credit-scoring and identity-verification services to other banks
and loan apps, and its products have been integrated into the onboarding process of digital
lenders.21 Jum Balea writes that ‘Lenddo uses algorithms that pull and analyze data from
these people’s [individuals from the emerging middle class] social media accounts—who
their friends are, how often they interact, their interests—and turns the information into a
rating or score that says if they’re likely to pay back or default on their loans. Or in other words,
if they are to be trusted.’22
Marie Langevin understands this to mean that Lenddo links credit scores to the individual
debtor’s ‘capacity to form online groups with friends, parents and colleagues’.23 For Lenddo,
the social graphs of the debtors’ online relationships reveal character traits such as reliabil-
ity, responsibility, and honesty. Langevin finds Lenddo privileging artificial intelligence and
behavioral economics in evaluating creditworthiness instead of ‘drawing on the know-how
concerning the evaluation and measurement of the productive capacities of debtors’.24 For
loan companies to evaluate the productive capacities of debtors, they would need to (per-
haps) deploy in-person, human expertise and invest considerable time: credit analysts would
maintain a close relationship with micro-entrepreneurs to assess cash flow and repayment
potential. Compared to earlier microfinance initiatives, it is precisely time and human labor
that the alternative credit scorers are promising to cut back on. At best, these alternative credit
21 Marie Langevin, ‘Big data for (not so) small loans: technological infrastructures and the massification of
fringe finance’, Review of International Political Economy 26.5 (2019): 790-814.
22 Jum Balea, ‘Lenddo stops lending, now helps clients determine customer trustworthiness’, TechInAsia,
25 January 2015, https://www.techinasia.com/lenddo-customer-trustworthiness.
23 Langevin, ‘Big data for (not so) small loans’, 800.
24 Langevin, ‘Big data for (not so) small loans’, 802.
141
IN/CONVENIENCE
scores—partly relying on social media data for character and behavioral traits—can provide
some indication of whether a particular debtor can be relied on to pay on time. However, if
such a debtor simply has no means to pay, how will they? In Langevin’s analysis or Mehrotra’s
explanation, this data is alternative also because now with increased monitoring of digital
traces this data is finally available to be stored and processed, something which was not
available before the widespread accessibility of the mobile internet in India.
Using social media connections or phone contacts data as alternative credit scoring data can
have benefits but also be compromised in contexts such as India where in rural and peri-urban
areas there is a possibility that women’s phone use and access is mediated by a male relative.
In such a situation, women feel discouraged to have too many phone contacts and self-censor
their text messages and social media connections.25 This would lead to a thin data footprint
capture or an incomplete data upon which algorithms would make decisions. So, if alterna-
tive data was part of providing user convenience, it might create gendered inconveniences.
Alternative data collection and uses are a cause of in/convenience for both legitimate and
illegitimate loan apps. Even legitimate loan apps seem to be operating based on questionable
alternative data practices. That said, it is important to highlight for a comparative perspective,
the rise of predatory loan aps during the recent pandemic in India.
Predatory Microloan Apps
Amid the COVID-19 pandemic in April and May 2020, a new variant of digital lending apps
(like UdhaarLoan, CashBean, RupeeFactory, and Wifi Cash) mushroomed in India, provid-
ing swift microloans to the poor and unemployed, as well to those who were desperate for
money because they had lost their daily jobs due to lockdowns. These apps demanded quick
repayments with high interest rates. Unlike banks or financial institutions—which require
stacks of paperwork and substantial collateral, and which take many months to approve a
loan—these quick microloan apps approved payments within just minutes. But there was a
catch. Exploiting the desperate poor and unemployed people who were battling the crisis of
the pandemic, these predatory loan apps were found to be flagrantly flouting financial regu-
latory laws and consumer protection guidelines. They made their debtors/customers install
the app on their smartphone and share their mobile’s GPS location, contact list, and picture
gallery.26 The micro-loan debt trap was not just restricted to India as working-class migratory
laborers in China also found themselves defaulting on payments, and being harassed by
recovery agents.27
25 Alexandra Rizzi, Alexandra Kessler, and Jacobo Menajovsky, The Stories Algorithms Tell: Bias and
Financial Inclusion at the Data Margins, Center for Financial Inclusion Report, 31 March 2021, https://
www.centerforfinancialinclusion.org/the-stories-algorithms-tell-bias-and-financial-inclusion-at-the-
data-margins.
26 Nilesh Christopher, ‘Debt and Shame via GooglePlay’, Rest of World, 27 Januar y 2021, https://
restofworld.org/2021/debt-and-shame-via-google-play/.
27 Yichen Rao and Tom McDonald, ‘Debt at a distance: Counter- collection strategies and financial
subjectivities of China’s working-class defaulters during COVID-19’, Economy and Society 52.2 (2023):
250-273.
142 THEORY ON DEMAND
These rogue loan apps carried misleading advertisements on YouTube with attractive (albeit
distorted) interest rates and repayment windows. A click took interested borrowers to the
Google Play store to download the loan app. Could one really participate in the mobile money
transition without being cheated at some point? How would one know when to share bank
account information and when not to? If the loan process was going to be instantaneous, how
could one hesitate before clicking a link or sharing their financial information? The conve-
nience of instant loans seemed to be mixed with the (inconvenient) anxiety of possibly being
taken for a ride by these loan apps.
The collateral for easy loans was user data, and this data (or surveillance of data) was the
trade-off for convenience and seamless onboarding process and instant delivery of money into
one’s bank account. The predatory loan apps used the customer data collected during the
initial onboarding process to harass those who had defaulted. Intimidation practices included
calling not just the defaulter—and threatening to sue them —but also the defaulter’s friends
and relatives (from the contact data captured earlier) and using expletives against them. So
blatantly vulgar was the language, tone, and demeanor of these calls that some harassed
debtors started recording them. These calls were then replayed on prime-time news shows.28
While public outrage on a TV show spectacle might have its place, these shows did not trace
the ways the fraudulent loan apps were using existing fintech infrastructure. The TV shows
failed to acknowledge that one reason these predatory loan apps were succeeding had much
to do with how certain protocols of loan disbursement and onboarding (like asking for opt-in
permissions for certain phone data) had been normalized by legitimate loan apps.
Following many complaints from victims of predatory loan apps as well as consumer rights
groups, the Reserve Bank of India in August 2022 introduced stricter guidelines with regard
to digital lending. The Google Play store also took steps to take down the illicit apps. The exit
of predatory loan apps also meant a void was created in the online lending market which
encouraged established players like KreditBee and Paytm to aggressively enter this market.29
Unlike rogue apps, legitimate loan apps are not using the data gathered to abuse defaulters.
That said, as I have argued, the architecture of ‘alterative’ data collection needed to provide
a convenient loan onboarding experience had been in place even in the legitimate lending
platform sphere. So, user call records these apps still access may not be used to harass
defaulters’ families, but they may be used to judge the credibility of a borrower depending on
whether they call their parents regularly or not—treated here as a sign of reliability. While illicit
predatory loan apps deliberately spread financial disinformation, I argue that the legitimate
digital lending platforms are creating opaque zones of financial knowledge trying to cater to
their lenders, borrowers, third-party developers, and regulators, all at the same time. Beyond
debating alternative data and algorithms, I believe we need to move toward discussing this
fintech infrastructure or the loan app platform ecosystem, which seems to provide partial
information to each of its stakeholders.
28 Zee Business, ‘Operation Hafta Vasooli’, Youtube, 16 April 2020, https://www.youtube.com/
watch?v=NMQdKi_zVJQ.
29 Manikandan, ‘Instant Digital Loans’.
143
IN/CONVENIENCE
Appropriating Fintech Infrastructures
In most cases involving both legitimate and illegitimate microloan apps, a nonbanking lender,
usually called a nonbank financial company (NBFC), partners with a lending partner or a loan
app so that it can reach a wider set of people with loan offerings. Usually, an NBFC works with
several lending partners and loan apps, and the loan app in turn depends on various actors
to supply SDKs/APIs for loan dispersal, credit scoring, identity-verification checks, payment
processing, and loan collection. These services are called fintech-as-a-service (FaaS), and
many loan apps depend on a few niche FaaS companies. The NBFC partner may not even
know who the loan app is partnering with for a particular service.
During the microloan controversy, an NBFC like Inditrade did not know that some of the loan
apps it was partnering with were using identity-verification services from the ‘Advance AI’,
a Chinese company working on AI-driven eKYC.30 Some of these verification methods, as
recounted by Cashless Consumer experts Suman Kar and Srikanth Lakshmanan, involve
taking a selfie during loan onboarding, which is then sent back to a server in China. This
server tests and checks whether the borrower is a real person, after which the information is
passed back to the app. These flows of data raise national/geopolitical concerns, as some of
these ID-verification techniques involve liveness detection and facial recognition data, which
can then be mirrored in other databases. It bears noting here though that fintech companies
developing questionable third-party data gathering tools do not just exist in China, they are
spread across several countries of South Asia, South East Asia, and East Africa including
India, Philippines, and Kenya. Scholars working on the relationship between apps and infra-
structures stress that apps are not ‘stand-alone objects’ but work at various levels as part of
‘wider socio-technical assemblages’.31 This means that we have to understand the activities
of loan apps not only at the individual or company level but also at the level of infrastructural
platform services.
Furthermore, the loan app company as a digital lending platform seems to be selective in
the way information moves between the various stakeholders it is mediating between: here,
between the NBFC and third-party developer. The same could be said about how interest
rates are being determined based on alternative data collection that borrowers do not know
much about and that the loan app indicates is the prerogative of lending firms and third-party
algorithms.
Not just in the loan app space, but also within the wider mobile app ecosystem, third party
SDKs have become constitutive actors in helping offer monetizing, analytics, and advertising
functions that are able to convert user sociality into data at the backend and thereby improve
the functionality of the apps with the hidden collateral of greater surveillance. SDKs are far
30 Arundhati Ramanathan, ‘The unregulated tech supermarket powering India’s loan app scourge’, The
Ken, 17 March 2021, https://the-ken.com/story/the-unregulated-tech- supermarket-powering-loan-
app/.
31 Michael Dieter et al., ‘Multi-situated App Studies: Methods and Propositions’, Social Media + Society 5.2
(2019): 1-15.
144 THEORY ON DEMAND
more integrated into an app’s source code than laptop/desktop cookies, which should make
us more concerned about user privacy in the mobile phone ecosystem.32 For loan apps, the
implications of permissions requested by third party SDKs for effective app use is particularly
crucial to study given the sensitivity of information collected. Equally important is the need to
ascertain how the app stores are regulating to what extent these permissions are necessary
and reasonable, or appropriate and ethical. With the spread of financial services in emerging
economies like India and Ghana, there are concerns about whether the data generated out
of erstwhile unbanked poor might be compromised with the rise of customer protection risks.
David Medine and Gayathri Murthy in their CGAP (Consultative Group to Assist the Poor)
report noted that data protection regimes by emphasizing individual consumer consent end
up placing undue burden on low-income populations in emerging economies who have just
started opening accounts and coming online, and thereby lack significant digital literacy.33
Production of Convenience, Cultures of Convenience
There is industrialized convenience being created through digital lending where the fintech
infrastructure creates the foundation for an easy and convenient loan onboarding process,
and alternative data collection and processing determine the type of loan eligibility, payment
windows, and interest rates leading to potential (in)conveniences in the future. In some sce-
narios, it could almost seem that customers did not need to take loans per say, but rather
took it because it was so convenient and readily available—never perhaps knowing that they
might end up repaying monthly installments for years. At a panel discussion regarding digital
lending, the newspaper Economic Times’ financial editor Amol Dethe, while asking questions
to lenders and loan app professionals, pointed out the Gen Z demography’s borrowing prac-
tices.34 These borrowers espoused the ‘Buy Now, Pay Later model’, and thus were especially
susceptible to taking out digital loans on whims such as looking nice for a date or for attending
a wedding. So they took loans not because they required it but because borrowing seemed
seamless, only later realizing that repayment of loans was anything but seamless.
So, industrialized production of convenience by fintech companies and infrastructures can
create conditions where loans are made available for constituencies who may not necessarily
need it or know the hazards of borrowing. Borrowers might face problems paying back and
lending platforms may find it difficult to recover the money as well. Another scenario is when
lending platform recovery agents encourage customers to take a loan from another loan
app to pay the monthly installments of an earlier loan, thereby creating conditions of debt
entrapment where borrowers keep moving from one loan to another. This is where the narra-
32 Jennifer Pybus and Mark Cote, ‘Did you give permission?: Datafication in the mobile ecosystem’,
Information, Communication & Society 25.11 (2022): 1650-1668.
33 David Medine and Gayatri Murthy, Making Data Work for the Poor: New Approaches to Data
Protection and Privacy, CGAP: Focus Note, January 2020, https://www.cgap.org/sites/default/files/
publications/2020_01_Focus_Note_Making_Data_Work_for_Poor.pdf.
34 See Amol Dethe (moderator), ‘NBFCs' Biggest Boon: Digital Lending’, panel discussion, Economic
Times: Banking, Financial Services and Insurance [ET BFSI], Youtube, 4 May 2023, https://www.
youtube.com/watch?v=l4YPAFWLhnY.
145
IN/CONVENIENCE
tive of financial inclusion changes trajectory toward ‘predatory inclusion’.35 This scenario is not
restricted to India either. Anthropologists Kevin Donovan and Emma Park write about young
Kenyans in a downtown Nairobi pub running out of mobile money, but who could still socialize
over drinks as they availed the digital overdraft facility offered by Safaridotcom in the form of
Fuzila with the catchphrase: ‘Finish what you need to finish with Fuzila.’36
Whether buying gifts for Diwali using KreditBee, or socializing over drinks thanks to Fuzila, are
strictly necessary reasons for taking out a loan (or being in debt) remain debatable, yet they sug-
gest that everyday calculations of convenience are subject to socio-cultural variations.37 Some
KreditBee ads offer other loan-taking scenarios—‘upgrading’ from scooter to motorbike, buying
a new smartphone, paying children’s school admissions fees, funds to start a new business
venture, and raising money for sister’s marriage. Each of these choices require different degrees
of risk calculations about future inconveniences and possible repayment failures, and each
perhaps are considered amidst different contingencies. These scenarios of digital lending sum
up what it means to ‘live in convenience’ in today’s platformized society as noted by the editors
of this collection. That said, all such decisions and scenarios cannot be encapsulated within the
condition or discourse of convenience; rather a poetics of convenience has to be linked with
affective considerations of immediacy, expediency, and aspiration. On the macro-level, the
‘access doctrine’, a continuing legacy of the Information and Communication Technology for
Development (ICT4D) discourse, trumps everything else—the ‘hope’ or the ‘promise’ is that
the mobile phone is key to any technological solution to solve the problem of access.38 So, more
than convenience, for the Indian state and the corporations, it is the access to loans through
mobile phones that makes concerns about surveillance secondary.
The industrial production of convenience did not begin with platforms: there is a long legacy of
washing machines and toasters that was seen to save domestic labor or of the neighborhood
convenience stores and their organized chains in the form of Wawas and 7-Elevens that saved a
long trip to Walmart or Albertsons.39 While some Indians continue to value slow cooking, others
have enthusiastically espoused instant Maggi noodles, and now with quick commerce, there is
a ‘social need’ being actively created by food delivery platforms for ‘instant-instant noodles’.40 In
India’s entrenched social hierarchy, platforms might encode caste into values of convenience
as was noticed with the food delivery platform Zomato recently announcing the “Pure Veg” fleet.
Zomato suggested that it would have a segregated fleet differentiated by green uniform that
would only carry vegetarian food. Rather than a vegan lifestyle, this was seen by anti-caste
activists as reifying a caste-based sensory-spatial order existing in India that is fussy about how
food is cooked, who handles it, with food touched by “lower-caste” people deemed impure.41
35 Tressie McMillan Cottom, ‘Where Platform Capitalism and Racial Capitalism Meet: The Sociology of
Race and Racism in the Digital Society’, Sociology of Race and Ethnicity 6.4 (2020): 441-49.
36 Donovan and Park, ‘Algorithmic Intimacy’.
37 Rahul Oka, ‘Introducing an anthropology of convenience’, Economic Anthropology 8.2 (2021): 188-207.
38 Daniel Greene, The Promise of Access: Technology, Inequality, and the Political Economy of Hope,
Cambridge, MA: MIT Press, 2021.
39 Huberman, ‘Amazon Go, Surveillance Capitalism, and the Ideology of Convenience’.
40 Raval, ‘Instant-Instant Noodles’.
41 Yashee, ‘Zomato’s ‘pure veg food’ scheme is pure casteism,’ Mar 22, 2024. Available: https://
146 THEORY ON DEMAND
While the ‘analog’ neighborhood kirana grocery stores continue to flourish in India, e-commerce
delivery has also witnessed a significant rise. There are incongruities in this story of platformed
convenience that suggests convenience does not fit all sizes at the same time. Hawkers selling
fruits and vegetables in India who never had banks and landlines till recently suddenly possess
smartphones through which they take loans and receive digital payments from customers via
Paytm or PhonePe. An essential endnote to this story is that the hawker is still paying the online
loan they took to buy the smartphone.42 The in/convenience of borrowing apps, indeed.
Since 2015, the state and fintech companies based on India Stack and FaaS infrastructures
have aggressively sought to include the until now ‘unbanked’ Indian citizens into regimes of
‘formal’ digital payments and lending through mobile phone apps. This has provided the con-
venience of instantaneous transmission of monetary benefits to many customers as well as
predatory possibilities for third-party loan app developers and mobile phishing operators to
siphon off user data. As users find themselves subjected to this collection and use of data
and information at various points of the fintech’s money/data distribution infrastructure supply
chain, the question remains how much of this financial misinformation is a deliberate exercise
of scamsters and how much of this is part of (i.e. embedded in) the networked architecture, or
still better, platform ecosystem? Along with misinformation and disinformation, there is at times
a lack of information (or grey/opaque zones of knowledge) as digital lending platforms cater to
their diverse constituencies, including borrowers, lenders (NBFCs/banks), third-party devel-
opers, and regulators strategically positioning themselves as merely financial intermediaries.
The consequences of this discursive platform politics for the financial misinformation climate
and the (in)conveniences of borrowers/customers remains a question for future scholars to
continue to explore.
References
‘Financial Inclusion Data / Global Findex’, World Bank, accessed 2 April 2024, https://datatopics.world-
bank.org/financialinclusion/country/india.
Athique, Adrian. ‘A great leap of faith: the cashless agenda in Digital India’, New Media & Society 21.8
(2019): 1697-1713.
Balea, Jum. ‘Lenddo stops lending, now helps clients determine customer trustworthiness’, TechInAsia,
25 January 2015, https://www.techinasia.com/lenddo-customer-trustworthiness.
Christopher, Nilesh. ‘Debt and Shame via GooglePlay’, Rest of World, 27 January 2021, https://restof-
world.org/2021/debt-and-shame-via-google-play/.
Dethe, Amol, moderator. ‘NBFCs' Biggest Boon: Digital Lending’, panel discussion, Economic Times:
Banking, Financial Services and Insurance [ET BFSI], Youtube, 4 May 2023,https://www.youtube.com/
watch?v=l4YPAFWLhnY.
indianexpress.com/article/opinion/columns/zomatos-pure-veg-food-casteism-heres-why-9224719/.
Am indebted to Abhishek Sekharan for pointing out this platformized connection between caste and
convenience.
42 Benjamin Parkin, John Reed and Jyotsna Singh, ‘The India Stack: opening the digital marketplace to the
masses’, The Financial Times, 20 April 2023, https://www.ft.com/content/cf75a136-c6c7-49d0-8c1c-
89e046b8a170.
147
IN/CONVENIENCE
Dieter, Michael; Gerlitz, Caroline; Helmond, Anne; Tkacz, Nathaniel; van der Vlist, Fernando N.; and
Weltevrede, Esther. ‘Multi-situated App Studies: Methods and Propositions’, Social Media + Society 5.2
(2019): 1-15.
Donovan, Kevin and Park, Emma. ‘Algorithmic Intimacy: The Data Economy of Predatory Inclusion in
Kenya’, Social Anthropology 30.2 (2022): 120-139.
Gillespie, Tarleton. ‘The politics of “platforms”’, New Media & Society 12.3 (2010): 347-364.
Greene, Daniel. The Promise of Access: Technology, Inequality, and the Political Economy of Hope,
Cambridge: MIT Press, 2021.
Huberman, Jenny. ‘Amazon Go, Surveillance Capitalism, and the Ideology of Convenience’, Economic
Anthropology 8.2 (2021): 337-349.
Hyken, Shep. The Convenience Revolution: How to Deliver a Customer Service Experience That Dis-
rupts the Competition and Creates Fierce Loyalty, Shippensburg: Sound Wisdom, 2019.
IBS Intelligence. ‘Reimaging the Credit Ecosystem,’ Panel Discussion, Payments & Lending Confer-
ence, 24 March 2022, Youtube, 11 April 2022, https://www.youtube.com/watch?v=oW3S8J58LWo.
KreditBee. ‘Celebrate Diwali 2022 with #KhushiyonKaLoan’, Loan app ad, Youtube, 12 October 2022,
https://www.youtube.com/watch?v=QCrf-IGKk5M.
KreditBee Digital Lending Partners. https://www.kreditbee.in/digital-lending-partners.
Krishna, Vishal. ‘This startup clocked $9M in revenue by lending to just students and the self-em-
ployed’, YourStory, 30 March 2021, https://yourstory.com/2021/03/kolkata- based-lending-platform-
mpokket-students-instant-loan.
Langevin, Marie. ‘Big data for (not so) small loans: technological infrastructures and the massification
of fringe finance’, Review of International Political Economy 26.5 (2019): 790-814.
Manikandan, Ashwin. ‘Instant Digital Loans: convenience comes with a heavy cost’, The Morning
Context, 29 August 2023, https://themorningcontext.com/internet/instant-digital-loans-convenience-
comes-with-a-heavy-cost.
McMillan Cottom, Tressie. ‘Where Platform Capitalism and Racial Capitalism Meet: The Sociology of
Race and Racism in the Digital Society’, Sociology of Race and Ethnicity 6.4 (2020): 441-49.
Medine, David and Murthy, Gayatri. Making Data Work for the Poor: New Approaches to Data Protec-
tion and Privacy, CGAP: Focus Note, January 2020, https://www.cgap.org/sites/default/files/publica-
tions/2020_01_Focus_Note_Making_Data_Work_for_Poor.pdf.
Nair, Vijayanka. ‘Governing India in Cybertime: Biometric IDs, Start-Ups, and the Temporalized State’,
South Asia: Journal of South Asian Studies 42.3 (2019): 519-536.
Neves, Joshua and Steinberg, Marc. ‘Pandemic Platforms: How Convenience Shapes the Inequality of
Crisis’, in Philipp Dominik Keidl, Laliv Melamed. Vinzenz Hediger, and Antonio Somaini (eds) Pandem-
ic Media: Preliminary Notes Towards an Inventory, Lüneberg: meson press, 2020, pp. 105-112.
_____. ‘In Convenience’, in this volume.
Oka, Rahul. ‘Introducing an anthropology of convenience’, Economic Anthropology 8.2 (2021): 188-
207.
Parkin, Benjamin; Reed, John; and Singh, Jyotsna. ‘The India Stack: opening the digital marketplace
to the masses’, The Financial Times, 20 April 2023, https://www.ft.com/content/cf75a136-c6c7-49d0-
8c1c-89e046b8a170.
Poster, Winifred R. ‘Introduction to special issue on scams, fakes, and frauds’, New Media & Society
24.7 (2022): 1535-1547.
148 THEORY ON DEMAND
Prabhakar, Tarunima. A New Era for Credit Scoring: Financial Inclusion, Data Security, and Privacy
Protection in the Age of Digital Lending, UC Berkeley Center for Long Term Cybersecurity White Paper,
2020, https://cltc.berkeley.edu/wp-content/uploads/2020/06/A_New_Era_for_Credit_Scoring.pdf.
Pybus, Jennifer and Cote, Mark. ‘Did you give permission?: Datafication in the mobile ecosystem’,
Information, Communication & Society 25.11 (2022): 1650-1668.
Ramanathan, Arundhati. ‘The unregulated tech supermarket powering India’s loan app scourge’,
The Ken, 17 March 2021, https://the-ken.com/story/the-unregulated-tech- supermarket-power-
ing-loan-app/.
Rao, Yichen and McDonald, Tom. ‘Debt at a distance: Counter- collection strategies and financial
subjectivities of China’s working-class defaulters during COVID-19’, Economy and Society 52.2 (2023):
250-273.
Raval, Noopur. ‘Instant-Instant Noodles: How algorithmic platforms transform food, taste and repro-
duction in cities’, Penn Cinema & Media Studies Colloquium, 7 September 2022, https://pricelab.sas.
upenn.edu/events/instant-instant-noodles.
Rizzi, Alexandra; Kessler, Alexandra; and Menajovsky, Jacobo. The Stories Algorithms Tell: Bias and
Financial Inclusion at the Data Margins, Center for Financial Inclusion Report, 31 March 2021, https://
www.centerforfinancialinclusion.org/the-stories-algorithms-tell-bias-and-financial-inclusion-at-the-
data-margins.
Roy, Abhirup and Kalra, Aditya. ‘India’s Zomato faces heat for plan to deliver food in 10 mins’,
Reuters, 22 March 2022, https://www.reuters.com/world/india/indias-zomato-faces-heat-plans-deliv-
er-food-10-minutes-2022-03-22/.
Scott, Brett. Cloudmoney: Cash, Cards, Crypto, and the War for our Wallets, London: Penguin, 2022.
Shapiro, Aaron. ‘Platform urbanism in a pandemic: Dark stores, ghost kitchens, and the logistical-ur-
ban frontier’, Journal of Consumer Culture 23.1 (2022): 168-187.
Shove, Elizabeth. Comfort, Cleanliness and Convenience: The Social Organization of Normality. New
Technologies/New Cultures Series. Oxford: Berg, 2003.
Srnicek, Nick. Platform Capitalism, Malden: Polity, 2016.
Swartz, Lana and Stearns, David. ‘Money and its Technologies: Making Money Move in the Modern
Era’, in Taylor C. Nelms and David Pedersen (eds) A Cultural History of Money in the Modern Age, Lon-
don: Bloomsbury Academic, 2019, pp. 27-52.
Tripathi, Naandika. ‘On the trail of how 10-minute delivery works—and doesn't’, Forbes India, 1 July
2022, https://www.forbesindia.com/article/take-one-big-story-of-the-day/on-the-trail-of-how-10min-
ute-delivery-worksand-doesnt/77761/1.
Zee Business, ‘Operation Hafta Vasooli’, Youtube, 16 April 2020, https://www.youtube.com/
watch?v=NMQdKi_zVJQ.
Zuboff, Shoshana. The Age of Surveillance Capitalism: The Fight for a Human Future at the New Fron-
tier of Power, London: Profile Books, 2019.
149
IN/CONVENIENCE
BEYOND INCLUSION: GLITCHY ECONOMIES AND
THE PROMISE OF PLATFORMIZATION IN AFRICAN
CITIES
ANDREA POLLIO & LIZA ROSE CIROLIA
Digital Platforms Beyond the Discourse of Inclusion
Globally, the vocabulary of the platform economy is wide and diverse. Yet, in the African con-
text, digital platforms tend to be framed within discourses of ‘inclusion’. Providing financial
services to the unbanked, integrating informal workers into structured labor regimes, and
easing knowledge access for young students are some of the common promises articulated
by platform advocates. Whether fin-tech, ed-tech, or health-tech, most African platforms
claim to address observable exclusions linked to the developmental challenges facing African
economies, cities, and dwellers.
The idea of inclusion, of course, is powerful and appealing. It caters to far-reaching global
agendas, including the Sustainable Development Goals (SGDs), and more specialized ones
such as the G20 Global Partnership for Financial Inclusion. Inclusion-speak enlivens the imag-
inations of ambitious startuppers who see Africa as the last frontier of platform innovations.1
After all, digital technology has often been premised on the promethean faith that it can
address seemingly intractable quandaries. In the late nineties, technology critics Richard
Barbrook and Andy Cameron gave a name to this set of beliefs: the Californian ideology.2 They
charted how Silicon Valley had produced a culture of techno-solutionism, a bizarre alliance
of progressive and neoliberal politics, which sought to find technical fixes to what may have
appeared (or been constructed) as social problems. Ultimately, ‘Californian ideologues’ imag-
ined that digital technology could do good (socially) while doing well (financially). Especially in
the African context, where market-solutions to poverty had become the experimental terrain
of what Ananya Roy more broadly termed ‘millennial development’, digital technology is seen
as a panacea for inclusion.3
1 The notion of the market frontier is problematic yet a powerful emic category through which businesses
operate to make risk manageable in Africa. However, for a reclaiming of the notion of frontier, see
Francis B. Nyamnjoh, ‘Incompleteness: Frontier Africa and the currency of conviviality’, Journal of Asian
and African studies 52.3 (2017): 253-270.
2 Richard Barbrook and Andy Cameron, ‘The Californian ideology’, Science as culture 6.1 (1996): 44-72.
3 Markets for Development, or M4D, provides a catchy buzz phrase for this ambition, formalized in many
development programs including that of DFID. See Ananya Roy, Poverty capital: Microfinance and the
Making of Development, New York: Routledge, 2010; Andrea Pollio, ‘Acceleration, development and
technocapitalism at the Silicon Cape of Africa’, Economy and Society 51.1 (2022): 46-70.
150 THEORY ON DEMAND
Many critical scholars have usefully pointed to the limits—and indeed often predatory nature—
of these experiments of digital enrollment4. Undeniably, such efforts reflect both neo-colonial
assumptions (e.g. that African cities are a new and unregulated Wild West) and capitalist
ambitions (market-making and extraction). However, our chapter takes a different approach.
The notion of inclusion, we contend, expresses a linear vector of technological transformation:
from being outside to being inside. For critical scholars, both inclusion and exclusion are
problematic, leaving little room for optimism, maneuver or proposition. In contrast, our expe-
rience and empirical work suggest that other processes and other motivations are underway
in Africa’s booming digital economies. In the context of glitchy, fragmented, patchworked
economic relations, platforms operate beyond the register of inclusion, enabling new forms
of algorithmically-enabled ease, optimization and even pleasure—let’s call it convenience.
What values would a grammar of ‘convenience’ expose, beyond the one-way vector critique
of ‘inclusion’? Would the making of algorithmic convenience, as enabled by digital platforms,
expand both our critical gaze and our understanding of the kinds of expertise and labor that
fuels these inclusionary testbeds?
Before getting into a more detailed account of the conceptual switch that we propose—from
inclusion to convenience—a brief example is in order here, one that sketches some answers
to these questions. It is the story of a buoyant startup operating in the outskirts of Nairobi,
Kenya. Funded by a Chinese expatriate who, like many of his peers, has chosen the East-Af-
rican city to experiment with algorithmic business models, this story begins to shed light on
the limitations of inclusion. And in doing so, as we will see, the example explicitly challenges
simple framings of convenience. Here, we want to sit with the ambiguity that convenience
brings with it, conceptually and practically. This does not negate critiques which see the
enrollment in convenience as an (exclusive) luxury, glut or even laziness. Nor do we intend to
ignore questions of extraction endemic to the making of conveniences for some, at times at
the expense of others. In fact, as we will later show, moving away from a binary of in/exclusion,
convenience specifically helps us to see the multiple displacements of effort and work now
held by new bodies, systems, and processes.
Fixing Glitchy Networks
In the mid-nineties, Tiger left his hometown Shanghai to study business in the United States
at one of the Ivy League schools that attract students from all over the globe. At the time, he
remembers, the Jin Mao building, soon to become China’s tallest skyscraper, was still under
construction, its pagoda shape towering over Pudong, a rapidly changing semi-industrial
area across the river from old Shanghai. By the time Tiger moved back to China in the early
2000s, armed with his business degree and a short but successful Wall Street career, things
had drastically changed. China had joined the World Trade Organization, the Jin Mao building
was surrounded by a multitude of even taller skyscrapers under construction, and the whole
4. Among others, see Daniela Gabor and Sally Brooks, ‘The Digital Revolution in Financial Inclusion: Interna-
tional Development in the Fintech Era’, in Kate Bayliss, Ben Fine, Mary Robertson (eds) Material Cultures of
Financialisation, New York: Routledge, 2020, pp. 69-82.
151
IN/CONVENIENCE
country was teeming with opportunities for ambitious, foreign-trained talents like Tiger. In
China, these returnees were amicably called sea turtles, haigui, a pun based on the homoph-
ony between ‘returned from overseas’ and the shelled reptiles that always find a way home.
Eventually, Tiger realized his success in selling cloud infrastructure services and enterprise
software solutions in Beijing, in the early days of those digital economies. But a decade and
a half later, now wealthy and experienced, he wanted more. The Chinese digital market was
saturated, and his peers had lost that socialist drive that he himself had come to value and
respect. China’s technological ascendancy, Tiger explains, was made possible precisely by
the capacity of homegrown entrepreneurs and firms to respond to the digital needs of the
masses. True, these hardware and software companies had made staggering profits, created
vehicles for massive surveillance, but they had also served the needs of what media scholar
Jack Qiu has called ‘working class network society’—the information have-less of the world.5
Despite its contradictions and precariousness, the digitization of Chinese society had con-
nected those who were, at the time, left out by profit-blinded Western digital corporations.6
At the very moment Tiger was thinking about finding new business challenges, Africa had
started appearing in Chinese media. It was the time of the Belt and Road bonanza, and
Beijing celebrated its cooperation with African nations through resounding announcements
of infrastructure projects and lofty proclamations of a new era of South-South cooperation.
Tiger had to see it for himself. He took a trip to Nairobi, a city that was celebrated as Africa’s
‘silicon savannah’ in a country, Kenya, that had sealed several agreements with the Chinese
government to build highways, railways, and digital corridors.7 Leaving aside the many con-
troversies of these large-scale projects, Tiger understood that, in a place like Nairoibi, most
people had yet to be reached by these infrastructure systems. The last mile, the ultimate leg
of connectivity, still fell short. Smoothly tarmacked highways and bypasses now crisscrossed
the sprawling metropolis, but in many peripheral neighbourhoods, roads were still unpaved
and so rutted that it was hard for a car to drive along. The same applied to electric networks,
water pipes, and, ultimately, digital connectivity.
For Tiger, China’s digital ascendancy held a lesson that could be exported and adapted to
Nairobi’s glitchy digital infrastructures. In 2019, he eventually relocated to Kenya, and opened
a small Internet-service startup in the northeastern neighborhoods of the city. Here, along the
highway that connects Nairobi to Thika, the landscape is dotted by master-planned estates,
by unplanned, plotted urban expansions, but also by some of the densest and most populated
suburbs in the city. These urban pockets host the bulk of Nairobi’s lower-middle classes, in
multistory concrete buildings that rise so close to each other that sometimes the crevices are
5 Jack Linchuan Qiu, Working-class Network Society: Communication Technology and the Information
Have-less in Urban China, Cambridge, MA: The MIT Press, 2009.
6 Lin Zhang, The Labor of Reinvention: Entrepreneurship in the New Chinese Digital Economy, New York:
Columbia University Press, 2023.
7 Michel Njeri Laura Wahome, Fabricating Silicon Savannah, PhD diss., University of Edinburgh,
Edinburgh, 2020; Liza Rose Cirolia et al., ‘Silicon Savannahs and Motorcycle Taxis: A Southern
Perspective on the Frontiers of Platform Urbanism’ Environment and Planning A: Economy and Space
55.8 (2023): 1989-2008.
152 THEORY ON DEMAND
less than a meter wide. Coupled with scarce investments by telecommunication companies,
which installed far too few telephone towers to meet the bandwidth demands of these growing
settlements, layers and layers of cement bricks render wireless mobile telecommunications
technologies completely inadequate. Moreover, local mobile operators serve their customers
with a “prepaid” business model based on selling tiny data packages at a marginal cost that is
much higher than what people pay elsewhere in the world, when they buy large data bundles.
As a consequence, people would consume online content sparingly. Data, Tiger realized, was
inconvenient both in terms of physical access and in terms of its cost.
The irony was that mobile operators had been lured into the prepaid, small-data bundle
business precisely by the global digital inclusion agenda. In the early 2000s, Global South-
based telcos had switched to the convenience of small airtime packages precisely to “include”
so-called bottom of the pyramid markets in places like Nairobi, where the majority of people
could not afford to buy bulk quantities of airtime (or monthly subscriptions). In fact, mobile
operators had followed the example of consumer goods multinationals—the likes of Unilever
and Nestlé—in creating pocket portions of their products. Kenyans even have a specific name
for this, the kadogo economy (kiSwahili for “small”): a slang with which the miniaturization
of consumer staples is described, from data to soap. But this business model, critics argue,
has allowed dominant companies to make staggering profit without needing to invest in large
scale distribution infrastructure.
Tiger echoed this critique. Like many entrepreneurs, he’d turned himself into a makeshift
ethnographer and spent months researching the digital data practices of lower-income urban
dwellers. Admittedly, data provision was a difficult market with fuzzy boundaries. The inhab-
itants of northeastern Nairobi were not the poorest in the city by far. Neither were they truly
middle-class, in a traditional sense. Their income oscillated, their indebtedness just as much.
Most importantly, they were fluctuating consumers of those digital commodities that had
come to represent social mobility: Instagram stories, Tik Tok shorts, YouTube videos, and,
the most aspirational of all, a movie or a series on a streaming platform. Small prepaid data
bundles didn’t allow that, as locals could only afford 1 gigabyte per month on average, and a
few episodes of a Showmax series alone would chew into half of it.8
Tiger’s solution consisted in upending the prepaid business model. As an Internet-service
provider, his company would sell time instead of data. That is, you could prepay to go online
for a chosen amount of time, and with no data limitations. With a few shillings, you could spend
a night binge watching a series, at a fraction of the cost charged by the mainstream mobile
providers. Or you could doom-scroll an infinite number of Tik Tok videos made by famous
Nairobi street comedians who cut their teeth by filming shorts in these same suburbs of the
city. But you could also run a business. After all, the ground floors in the area are a maze of
small shops and parlors. These have been included, indeed, in the financial ecosystem of
mobile money, but rarely do they have access to the full possibilities of connectivity. Needless
to say, Tiger’s business proposition was a staggering success. In early 2023, less than two
8 Showmax is a South-African streaming service with a strong localization strategy in sub-Saharan Africa,
producing content locally and partnering with mobile telcos.
153
IN/CONVENIENCE
years after starting his business officially, he had hundreds of thousands of subscribers, and
a team of more than 100 locally hired employees.
How had he achieved this success? It was, in his view, a matter of last-mile infrastructure.
Initially, he experimented with mobile antennas and Wi-Fi bridges. His company installed
thousands of wireless routers, one on each floor of the buildings that were to be connect-
ed. This solution mimicked what Tiger had observed in the early days of the digital boom
in Beijing, where lower-income residents had become accustomed to every public space
in the city offering access to the Internet through Wi-Fi bridges and routers. But that model
soon turned out to be unsuited for this part of Nairobi. The number of users was too high, the
bandwidth was crowded and unreliable, and too many routers were an infrastructural mess
for his network engineers. Eventually, Tiger’s business turned to hardwired connectivity. He
struck contracts with fiber providers in Kenya, and laid miles and miles of last-mile cables
between small, room-sized data centers and their surrounding urban fabric. It became an
incremental, trial-and-error testbed for an alternative mode of service delivery, on a large
scale. Despite the difficulties, replacing Wi-Fi bridges with wires innervating every building
meant that Tiger could now ensure reliability, scalability and redundancy. The convenience
of reliable connectivity was not just a matter of the payment model but also a tangible solution
to otherwise glitchy digital networks. Customers were happy, and their number kept growing.
Of course, the story of Tiger’s low-cost Internet provider in the periphery of Nairobi does
not yet have an epilogue. Small startups, even those that are already serving many clients,
constantly fail and pivot. Often, as critical scholars argue, they are cannibalized by monopoly
holders—big tech companies and big telcos—that incorporate frugal innovations into their
businesses.9 But even without an epilogue, this fledgling Internet startup already speaks to
the permutations and adaptations of the promise of convenient connectivity, as well as its
contradictions. Let’s look at two examples.
Tiger’s core business has been to serve an underserved mass market with access to data.
But now, with hundreds of thousands of users logging on and off the network on a daily basis,
the business of selling data is also a data business in itself. Even without accessing private
personal data and without tracking actual online behaviours, Tiger has gathered massive
amounts of information about his users; he can now delineate the market profiles of a segment
of Nairobi’s society that defies easy characterization. In turn, as has happened with mobile
money, this data can become, for example, an alternative credit scoring system to offer the
kind of pricey small loans that Nairobians rely on to cope with the precarity of their economic
lives.10 And much else. As sociologists Marion Fourcade and Kieran Healy remind us, digital
companies often build personal data dragnets without a specific goal in mind, in case these
troves of information may be monetized at a later stage.11 Tiger is now considering using
9 Laura Mann and Gianluca Iazzolino, ‘From Development State to Corporate Leviathan: Historicizing the
Infrastructural Performativity of Digital Platforms Within Kenyan Agriculture’, Development and Change
52.4 (2021): 829-854.
10 Kevin P. Donovan and Emma Park, ‘Knowledge/seizure: Debt and data in Kenya’s zero balance
economy’, Antipode 54.4 (2022): 1063-1085.
11 Marion Fourcade and Kieran Healy, ‘Seeing like a market’, Socio-economic review 15.1 (2017): 9-29.
154 THEORY ON DEMAND
these proxied credit profiles to offer buy-now-pay-later finance to his existing customer base.
These schemes are not bad per se, but as their critics argue, they can become very extractive,
with high interest rates, and unsavory practices of debt shaming. In other words, convenient
infrastructures may well become the rails of predatory inclusion.12
Another example of Tiger’s expanding business model, however, points to a different direction
of his convenience-driven project. In 2021, he met a young entrepreneur, Jinny, who was
planning to launch an e-commerce platform, importing the Pinduoduo model from rural China
into peri-urban Kenya. Pinduoduo, one of the fastest growing e-commerce giants in China,
had made its initial fortune by offering group-buying options for farmers.13 On the heels of
Pinduoduo’s success in China, Jinny was working on a similar service for Kenya. But while
she had a clear vision for the platform itself, she needed support for organizing the last-mile
infrastructure of group buying—from the big warehouse where bulk stock is received, to the
hotspots where commodities are distributed to group-purchasers. Tiger already had that infra-
structure in place. His data centers, after all, already functioned as infrastructural hubs for
the communities that they served. And he already employed part-time riders to move cables
and equipment around. All Tiger and Jinny needed to do was join forces. And so they did. Now
their customers, including the numerous small retailers that are scattered in the area, can
group-buy staples like flour and cooking oil at a third of the price they’d pay from middlemen.
It’s a fledgling super-app, they explain.14 In other words, the inclusionary but pricey kadogo
economy has been turned on its head by the deeper convenience of a less glitchy distribution
infrastructure, and an integrated logistical system that Tiger and Jinny plan to further expand
in the future to include other staples like drinkable water.
As we hinted at earlier, being ‘in convenience’ for some—even for many, in this case—is often
predicated on slight adjustments or displacements of labor, spaces, energy, profits, affects,
etc. From the electric grid to the gel batteries that rest in Tiger’s data centers to provide
redundancy in case of electric failure, from small shops to dark stores and to the riders that
make commodities move, convenience begins to foreground the limitations of the inclusion/
exclusion framing.
From Inclusion to the ‘Value(s)’ of Convenience
The story of Tiger and Jinny’s startup experiments with convenient connectivity and group-buy-
ing e-commerce speaks to our interest, as researchers of infrastructure in urban Africa, to
move beyond one of the dominant vocabularies through which both advocates and critics of
platform economies frame these kinds of experience: inclusion. Startups, venture capitalists,
12 Kevin P. Donovan and Emma Park, ‘Algorithmic intimacy: The data economy of predatory inclusion in
Kenya’, Social Anthropology/Anthropologie Sociale 30.2 (2022): 120-139. See also Rahul Mukherjee’s
chapter in this volume.
13 For a broader critique of platform experiments at the rural margins of the Chinese economy, see Lin
Zhang, ‘Platformizing family production: The contradictions of rural digital labor in China’, The Economic
and Labour Relations Review 32.3 (2021): 341-359.
14 For a definition, see Marc Steinberg, Rahul Mukherjee, and Aswin Punathambekar, ‘Media power in
digital Asia: Super apps and megacorps’, Media, Culture & Society 44.8 (2022): 1405-1419.
155
IN/CONVENIENCE
and sometimes even governments claim that platform-mediated inclusion (into more stable
labor markets, new financial systems, payment standards, and so on) is a win-win process.
It creates value for those who monetize the platforms, whatever their data-driven business
model, but also for those who become entrepreneurs of and on the platform.
Take Uber, for example. We have a vivid memory of the first days in which Uber started oper-
ating in Cape Town, its promotional posters professing that a new generation of self-employed
drivers would benefit from the new platform. Now included in an algorithmic system of gig
work distribution, drivers could count on more flexibility, better income opportunities, access
to asset financing, insurance coverage, and other perks that they would have not been able
to access as informal, or even illegal taxi operators.15 It was a powerful pitch. That promise,
of course, was more honored in the breach than the observance.16 But the example of Uber
adopting the developmental vocabulary of inclusion to embed its offering in the Cape Town
context is telling: even a global platform company had adapted its value proposition to claim
that they were fixing a broken system (the private mobility infrastructure) of an African city.
And that they were doing that through the platform-mediated inclusion of underemployed
workers. At the peak of their launch marketing campaign, Uber even started sharing the per-
sonal empowerment stories of their drivers to their customers. These were short videos of how
these workers—often women, migrants, and formerly disenfranchised South Africans—had
gone from being excluded to being included in the South African economy.
Ridden as it is by contradictions, legal challenges, and failing business metrics, Uber is a
fraught example. But it’s also the tip of the iceberg in the world of digital platforms, big and
small, homegrown and international, private and public, that we have been researching with
other colleagues over the last half decade in different African cities.17 These digital startups
propose a pitch that is not too dissimilar to what Uber claimed in the heyday of its African
launch: doing good (by fixing urban infrastructure and creating more inclusionary markets)
while doing well (as a profit-driven company). These business arguments vividly resonate with
the developmental language of inclusion, which has dominated the discursively powerful work
of the World Bank and other development finance institutions over the last two decades at least.
Through experiments with inclusion, the Global South, and Africa in particular, have been
recast as a new terrain of opportunity for California-ideology inspired businesses and human-
itarian organizations that want to do good while doing well.18 One domain of inclusion that
15 Andrea Pollio, ‘Forefronts of the sharing economy: Uber in Cape Town’, International Journal of Urban
and Regional Research 43.4 (2019): 760-775.
16 Mohammad Amir Anwar, Jack Ong’iro Odeo, and Elly Otieno, ‘“There is no future in it”: Pandemic and
ride-hailing hustle in Africa’, International Labour Review 162.1 (2023): 23-44; Pádraig Carmody and
Alicia Fortuin, ‘“Ride-sharing”, virtual capital and impacts on labor in Cape Town, South Africa’, African
Geographical Review, 38.3 (2019): 196-208.
17 Cirolia et al., ‘Silicon Savannahs and Motorcycle Taxis’; Andrea Pollio, Liza Rose Cirolia, and Jack
Ong’iro Odeo, ‘Algorithmic suturing: platforms, motorcycles and the “last mile” in urban Africa’,
International Journal of Urban and Regional Research 47.6 (2023): 957-974.
18 Adam Moe Fejerskov, The Global Lab: Inequality, Technology, and the Experimental Movement, Oxford,
UK: Oxford University Press, 2022.
156 THEORY ON DEMAND
has been more forcefully promoted (and also critiqued) has been that of digitally enabled
financial markets—the fintech sector, and its role within the World Bank-sanctioned global
‘financial inclusion agenda’. With the lowest banking and legacy financial service penetration
in the world, African countries are indeed a prime laboratory for fintech experimentation. Its
optimistic advocates claim that financial inclusion generates wealth, helps people out of
poverty, and turns them into potential entrepreneurs, all the while generating new investment
assets for global financial capital. Its critics point to the violence of adverse incorporation
into vicious cycles of unsustainable debt, the financialization of everyday life, especially of
poor households, the neo-colonial logics that underpin new datafied risk scoring techniques,
and, ultimately, the fact that financial inclusion is not a good business after all, but a failed
neoliberal experiment.19
These critiques are crucial. And yet, the notion of inclusion upon which both critical warnings
and uncritical advocacy seem to rest only captures one of many vantage points from which to
observe these algorithmic experiments. Economic anthropologist Janet Roitman, for example,
explains that the focus on inclusion risks overshadowing the multiple and ingenious forms of
value creation, rather than just extraction, that are beholden to platform-enabled financial
services.20 A linear mode of thinking about technology, she writes, primes much of the critique
of the platform economy in the Global South. This linearity is not just geographical—from the
North to the South—but also empirical: it assumes that platformed services simply enroll
those subjects and systems that were previously, somehow, ‘excluded’.
But let us return for a moment to the example of Tiger and his business, which he framed as
a challenge to the motifs of inclusionary digital markets. What we see in this example, one of
many we researched over the years, is that platform companies in African cities are already
operating far beyond the framing of inclusion, even when the language remains more or less
aligned. Their actual promise? Optimizing and fixing broken infrastructure systems. Making
access to goods and services more convenient, in spite of the fractured nature of urban life.
What would happen to our understanding of platform economies in the continent, if for a
second, we took at face value the perspective of entrepreneurs like Tiger and Jinny, or even
of companies like Uber—a perspective that, however problematic, tells us that they are fixing
glitchy, broken systems, and making life more convenient?
As flagged in the introduction, platform optimization also names one easily forgotten yet
fundamental rationality that animates such experiments and processes: the desire for ease
(and, in the case of Uber, reliability and safety). In other words, we draw on and expand the
vocabulary of convenience to engage the platform operations we have encountered during
our research, which pledge to make things ‘easier’ for Africa’s urban citizens. In fact, schol-
ars have long posited that infrastructure and service delivery networks in urban Africa are
19 See Philip Mader, ‘Contesting financial inclusion’, Development and Change 49.2 (2018): 461-483;
Gabor and Brooks, ‘The Digital Revolution in Financial Inclusion’; Nick Bernards, A Critical History of
Poverty Finance: Colonial Roots and Neoliberal Failures, London: Pluto Press, 2022.
20 Janet Roitman, ‘Platform economies: Beyond the North-South divide’, Finance and Society 9.1 (2023):
1-13.
157
IN/CONVENIENCE
glitchy and incomplete, where they exist at all, and constantly reworked and repaired through
collective and individual practices of material improvisation.21 For Tiger, for example, these
glitches manifested in the paltry availability of Internet access in Nairobi's most populous
neighborhoods—and presented a business opportunity to reduce them. More generally, as we
will see in the following pages, platform-based optimization, whether in the forms of a govern-
ment system for revenue collection or a remittance application, is predicated on heightening
the convenience of hitherto complicated, fragmented, and frictional infrastructural systems.
Frictions of Everyday Life in Urban Africa
Thinking about the critiques of the Amazonification of everyday life in cities like San Francisco
and London, maybe one day it will be possible to look at residents of African cities and ask
‘how much easier do you want everyday life to be?’. But we are far off from this. Platforms have,
of course, expanded to many urban economies (Kenya’s M-Pesa being a commonly cited
example).22 From e-commerce to crypto wallets and on-demand services, efforts to expand
the frontiers of platform capitalism are indeed prevalent. Platforms have also been used to
augment social and political life in African cities—think of the use of WhatsApp for political
party mobilization, reliance on social media platforms by prophetic churches, or the rise of
dating and music apps of various sorts.
While remaining critical of techno-solutionism, we can also concede that the sorts of problems
tech seeks to ‘solve’ (even if often unsuccessfully) are real problems and not merely elite
agitations. Notwithstanding considerable diversity across contexts and spaces, it is widely
acknowledged that African cities experience a range of shared challenges. These challenges
often stem from a knotted set of processes: histories of colonization, post-independence
nation-building projects, violent structural oppression within global geopolitical systems, sys-
tematic underfinancing of key infrastructures. These challenges are not abstractions; they
impact people’s everyday life.
The result is that cities are replete with processes that are fragmented, overlapping, inefficient,
traumatized, and heavily reliant on the unpaid work of everyday people: citizens of all classes,
including officials, business owners, and scholars. In this context, many celebrate the improvi-
sations and ‘make do’ or adaptive tactics—as people manufacture livelihoods, devise systems
to access services, or build their homes. These practices are framed as resilience, ingenuities,
or frugal innovations. No doubt they are, and yet this work, while valiant in the face of duress,
also consumes time, resources, and energy. These are glitchy economies causing dis-ease for
many. From sitting in traffic for hours daily to navigating complex, redundant bureaucracies for
registering pretty much anything (vehicles, businesses, births), to managing the ever-changing
21 Prince K. Guma, ‘Incompleteness of Urban Infrastructures in Transition: Scenarios from the Mobile Age
in Nairobi’, Social Studies of Science 50.5 (2020): 728-750. See also discussion in Jonathan Silver, The
Infrastructural South: Techno-Environments of the Third Wave of Urbanization, Cambridge, MA: The MIT
Press, 2023.
22 Nancy Odendaal, ‘Platform urbanism and hybrid places in African cities,’ in Alessandro Aurigi and
Nancy Odendaal (eds) Shaping Smart for Better Cities: Rethinking and Shaping Relationships Between
Urban Space and Digital Technologies, London: Academic Press, 2021, pp. 203-219.
158 THEORY ON DEMAND
currency fluctuations that burden all manner of financial transactions, such celebrations
normalize disruptions in the infrastructures that support everyday life and economies—dry
taps, internet shutdowns, and rolling black-outs.
In these circumstances, African elites, middle classes and expatriate travelers have attempt-
ed to isolate themselves from some of these glitchy, taxing, and frictional systems. We can
be critical of the islands of access created through boreholes and water tanks, VPNs and
multiple sim cards, elevated highways, generators, and solar panels—all manner of technol-
ogies aimed primarily at easing access or living in convenience.23 We can, at the same time,
consider the incredible work (and indeed expense and expertise) that also goes into these
(elite) practices of attending to what we could call the inconvenience of traffic, queues and
service delivery failures.
What the case of Tiger only alludes to is the way in which platforms aim to ease the everyday
frictions of urban life we just discussed. To better understand this, we must turn to the users
of these platforms, whether enrolled by choice or not. This allows us to see past the devel-
opmental evangelism espoused by the creators/managers of the platforms towards a view
of optimization that centers users. How might one look at optimization of a system, and the
convenience that it allows, as something valuable in the context of durable strain and costly
interactions which constitute many aspects of life in African cities? What, in turn, does the
frame of convenience (rather than inclusion) allow us to see that critics and scholars focused
on splintering or make-do or resilience overlook?
Thinking with Platform Optimization
Let us move, then, from urban enclaves to other populations who benefit from the ease afford-
ed, in this case, by platform-enabled digital transactions. The following examples are once
again anecdotal, but they allow us to reflect on the kinds of mundane ameliorations that
platform processes yield in the context of inconvenient systems, and the displacements of
labors that follow from the processes of optimization.
We begin this exploration in a provincial capital, Kisumu, situated on the Kenyan banks of
Lake Victoria. The lobby of the Kisumu City Authority, while now technically a part of the
County administration, is not unlike many smaller local governments in Africa. Its three levels,
accessed through dark wood balustrades, hold a series of offices for the various technical
functions of the city, including trash collection within the CBD, non-motorized transport and
building plan approval, among other intermediate functions. In terms of staff, the largest
department located in this building is revenue collection. Not only does the activity occupy
several offices, but in fact the entire first floor and courtyard.
23 What scholars would call the ‘splintering’ of urban infrastructure. See Stephen Graham and Simon
Marvin, ‘Splintering urbanism at 20 and the “Infrastructural Turn”’, Journal of Urban Technology 29.1
(2022): 169-175.
159
IN/CONVENIENCE
When entering the building, maybe to pay your monthly property rates or annual business
license, you are immediately greeted by the Kenya Commercial Bank Tellers, with lines at
each that stretch back towards the entrance. There is a TV to keep the waiters occupied, in
the intense heat that is common in the Lakes region. After payments are made, these lines
slowly trickle into the building's courtyard. Here, there is another set of waiting areas, where
people wait - again - in a spiraling line in the little shelter provided by the shade of the building.
And again, people line up to have the payment logged and be provided with a city receipt to
confirm payment. Depending on the time of day, month and year, this process can take any-
where from half an hour to several hours. And such a process is repeated for each payment.
In the last ten years in Kenya, as the case of Tiger indicates, there have been shifts towards the
digitization of payments in many sectors, including for state services.24 This is, of course, not
only the case in Kenya. Many African cities are transitioning onto platforms to support online
payments, citing ‘digital dividends’.25 These efforts are framed by their donors and higher
levels of government as ‘good governance’ measures, often against the backdrop of tropes
of petty corruption in the process of fee collection. It aligns with the focus of many donors on
raising sub-national revenue streams. However, we would argue that their value sits less in
the panoptical disciplining of sub-national functionaries. If we orient ourselves towards the
citizen—one of the user groups of these platforms—the value sits in the alleviation of waiting
in multiple lines, collecting slips, and losing half of a day or work or rest. In interviews with
people waiting in lines, most on their phones responding to messages or watching clips, they
lament the taxing experience, taking turns standing by a fan which once had the capacity to
swivel. Since we first visited Kisumu in 2018, most of these people are now able to use their
phones to pay these fees. They no longer have to travel to Kisumu’s city center to pay for such
fees, or spend their day moving from desk to desk. These affordances are hardly an unearned
luxury, but rather a much-needed relief.
Another example of an opaque and frustrating system which platforms promise to ease is
that of remittance corridors. From Kisumu, we now move back to Cape Town. The city is
home to migrants from across Africa, many of whom need to send portions of their earnings
to relatives in their hometowns and villages. The market for intra-Africa remittances has been
dominated by Western Union, which has a reputation of being costly and extractive.26 Until
recently, therefore, sending remittances through this existing financial infrastructure has not
only been expensive—with significant losses accruing to already very precarious people—but
also time consuming, with transactions taking days to clear. When a family needs money
urgently, this can be very stressful.
24 Romanus Opiyo et al., ‘Attaining E-democracy through digital platforms in Kenya’, in T.M. Vinod Kumar
(ed) E-Democracy for Smart Cities, Singapore: Springer, 2017, pp. 441-459.
25 For an example of this discourse, see World Bank, World Development Report 2016: Digital Dividends,
Washington, DC: World Bank, 2016.
26 Peter Mudungwe, Leveraging the African Diaspora for Development, The Hague: African Diaspora Policy
Centre, 2017, https://www.diaspora-centre.org/wp-content/uploads/2017/10/Leveraging_Diasporas_
for_Development.doc.pdf; Daniel Folkinshteyn, Mark M. Lennon, and Tim Reilly, ‘The Bitcoin Mirage: An
Oasis of Financial Remittance’, Journal of Strategic and International Studies (2015), https://ssrn.com/
abstract=2601621.
160 THEORY ON DEMAND
It is therefore not surprising that remittances at large, and intra-Africa remittances specif-
ically, have come to be a fertile site through which platform optimization finds traction. On
the back of Cape Town’s self-declared position as both start-up and fintech capital of Africa,27
Cape Town-based platforms like Mama Money and Mukuru have attempted to reduce costs
(down to around 5%), improve the speed (within the hour), and enable migrants who do not
have advanced documentation access (through reduced KYC).28 The founders boast social
visions, the former started by two men who met en route from a festival in the desert of South
Africa and the latter by a Zimbabwean-born, South African-educated, and London-based
entrepreneur and rock musician. While the founders’ explanations are replete with question-
able development jargon, the people who use the platforms praise them. Betty, a Congolese
businesswoman who sells Cape Malay curries and samosas at a local market in Cape Town
explained to us how the platform allows her to send money from wherever she is, just using
her phone: no more lines, abusive tellers, or endless return visits to track the payment.
Many flags can be raised. And of course, we should be attentive to the ways in which the enroll-
ment of taxpayers, migrants, and many other economic subjects can be exploited through
(financial) platforms of different nature—not to mention who is excluded. At the same time,
it is undeniable that such platforms ease certain aspects of urban life beyond the enclaves
of the wealthy middle class, enabling people to overcome the enduring strain and glitchy
nature of bureaucratic and economic processes. In this case too, convenience displaces
the burden of work required to move money from informal and small-scale money agents to
dedicated programmers and data scientists whose (well-paid) labor is necessary to ensure
seamless transactions between financial institutions, telecommunication providers, last-mile
outlets and end-users.
Convenience Reconsidered
Notwithstanding these vignettes, and perhaps unlike many parts of the world, few would call
African cities convenient. Even among the wealthy who can afford cleaners and childcare,
given the sheer scale of infrastructural fragmentation—fuel scarcity, traffic jams, internet
shutdowns—money’s capacity to ease meets its limits. We did not intend, in this chapter, to
present African cities as somehow exceptional, immune to the expansion of platform capi-
talism or to capricious overconsumption. Rather, what we hoped to do was adopt a different
starting point for charting the value(s) of convenience across diverse geographical contexts.
Often relegated to the binary between exclusion and inclusion, and to the warnings of pred-
atory, or so called ‘adverse’ inclusion, platformization processes underway in African cities
call for a different grammar of analysis and a different orientation towards these emerging
economies (as well as the people who animate them). As other authors in this volume argue,
27 Andrea Pollio and Liza Rose Cirolia, ‘Fintech urbanism in the startup capital of Africa’, Journal of Cultural
Economy 15.4 (2022): 508-523.
28 See Liza Rose Cirolia, Suzanne Hall, and Henrietta Nyamnjoh, ‘Remittance Micro‐Worlds and Migrant
Infrastructure: Circulations, Disruptions, and the Movement of Money’, Transactions of the Institute of
British Geographers 47.1 (2022): 63-76.
161
IN/CONVENIENCE
convenience can be a strategically awkward category of analysis, one that sits ambiguously
between critique and recognition that not all forms of ease are unnecessary or consumeristic
indulgence. In fact, the examples in this chapter speak to the variegated forms of optimiza-
tion that do in fact transform glitchy and vexing urban systems and infrastructure in Africa.
Ultimately, in addition to the acknowledgement of these optimized fixes, convenience also
allows us to see some of the displacements or the shifts that platform economies entail:
from humans to servers and data centers, from shops to dark kitchens, from one type of
labor to another. However one may feel about these movements, the lenses of convenience,
and especially the perspective that convenience offers us from African cities, brings to the
fore again the technological ambivalence of platformization and the need to engage the full
gamut of possibilities therein—possibilities that are too often foreclosed both by the makers
and the critics of platforms.
References
Anwar, Mohammad Amir; Ong’iro Odeo, Jack; and Otieno, Elly. ‘“There is no future in it”: Pandemic
and ride-hailing hustle in Africa’, International Labour Review 162.1 (2023): 23-44.
Barbrook, Richard and Cameron, Andy. ‘The Californian ideology’, Science as culture 6.1 (1996):
44-72.
Bernards, Nick. A Critical History of Poverty Finance: Colonial Roots and Neoliberal Failures, London:
Pluto Press, 2022.
Carmody, Pádraig and Fortuin, Alicia. ‘“Ride-sharing”, virtual capital and impacts on labor in Cape
Town, South Africa’, African Geographical Review, 38.3 (2019): 196-208.
Cirolia, Liza Rose; Hall, Suzanne; and Nyamnjoh, Henrietta. ‘Remittance Micro-Worlds and Migrant
Infrastructure: Circulations, Disruptions, and the Movement of Money’, Transactions of the Institute of
British Geographers 47.1 (2022): 63-76.
Cirolia, Liza Rose; Sitas, Rike; Pollio, Andrea; Sebarenzi, Alexis Gatoni; and Guma, Prince K. ‘Silicon
Savannahs and Motorcycle Taxis: A Southern Perspective on the Frontiers of Platform Urbanism’ Envi-
ronment and Planning A: Economy and Space 55.8 (2023): 1989-2008.
Donovan, Kevin P. and Park, Emma. ‘Algorithmic intimacy: The data economy of predatory inclusion in
Kenya’, Social Anthropology/Anthropologie Sociale 30.2 (2022): 120-139.
Donovan, Kevin P. and Park, Emma. ‘Knowledge/seizure: Debt and data in Kenya’s zero balance econ-
omy’, Antipode 54.4 (2022): 1063-1085.
Fejerskov, Adam Moe. The Global Lab: Inequality, Technology, and the Experimental Movement, Oxford,
UK: Oxford University Press, 2022.
Folkinshteyn, Daniel; Lennon, Mark M.; and Reilly, Tim. ‘The Bitcoin Mirage: An Oasis of
Financial Remittance’, Journal of Strategic and International Studies (2015), https://ssrn.com/
abstract=2601621.
Fourcade, Marion and Healy, Kieran. ‘Seeing like a market’, Socio-economic review 15.1 (2017): 9-29.
Gabor, Daniela and Brooks, Sally. ‘The Digital Revolution in Financial Inclusion: International Devel-
opment in the Fintech Era’, in Kate Bayliss, Ben Fine, Mary Robertson (eds) Material Cultures of
Financialisation, New York: Routledge, 2020, pp. 69-82.
Graham, Stephen and Marvin, Simon. ‘Splintering urbanism at 20 and the “Infrastructural Turn”’,
Journal of Urban Technology 29.1 (2022): 169-175.
162 THEORY ON DEMAND
Guma, Prince K. ‘Incompleteness of Urban Infrastructures in Transition: Scenarios from the Mobile
Age in Nairobi’, Social Studies of Science 50.5 (2020): 728-750.
Mader, Philip. ‘Contesting financial inclusion’, Development and Change 49.2 (2018): 461-483.
Mann, Laura and Iazzolino, Gianluca. ‘From Development State to Corporate Leviathan: Historicizing
the Infrastructural Performativity of Digital Platforms Within Kenyan Agriculture’, Development and
Change 52.4 (2021): 829-854.
Mudungwe, Peter. Leveraging the African Diaspora for Development, The Hague: African Diaspora
Policy Centre, 2017, https://www.diaspora-centre.org/wp-content/uploads/2017/10/Leveraging_Dias-
poras_for_Development.doc.pdf.
Nyamnjoh, Francis B. ‘Incompleteness: Frontier Africa and the currency of conviviality’, Journal of
Asian and African studies 52.3 (2017): 253-270.
Odendaal, Nancy. ‘Platform urbanism and hybrid places in African cities,’ in Alessandro Aurigi and
Nancy Odendaal (eds) Shaping Smart for Better Cities: Rethinking and Shaping Relationships Between
Urban Space and Digital Technologies. London: Academic Press, 2021, pp. 203-219.
Opiyo, Romanus; Mwau, Baraka; Mwang’a, Keziah; and Mwaniki, Dennis. ‘Attaining E-democracy
through digital platforms in Kenya’, in T.M. Vinod Kumar (ed) E-Democracy for Smart Cities, Singapore:
Springer, 2017, pp. 441-459.
Pollio, Andrea. ‘Acceleration, development and technocapitalism at the Silicon Cape of Africa’, Econo-
my and Society 51.1 (2022): 46-70.
Pollio, Andrea. ‘Forefronts of the sharing economy: Uber in Cape Town’, International Journal of Urban
and Regional Research 43.4 (2019): 760-775.
Pollio, Andrea and Cirolia, Liza Rose. ‘Fintech urbanism in the startup capital of Africa’, Journal of
Cultural Economy 15.4 (2022): 508-523.
Pollio, Andrea; Cirolia, Liza Rose; and Ong’iro Odeo, Jack. ‘Algorithmic suturing: platforms, motorcy-
cles and the “last mile” in urban Africa’, International Journal of Urban and Regional Research 47.6
(2023): 957-974.
Qiu, Jack Linchuan. Working-class Network Society: Communication Technology and the Information
Have-less in Urban China, Cambridge, MA: The MIT Press, 2009.
Roitman, Janet. ‘Platform economies: Beyond the North-South divide’, Finance and Society 9.1
(2023): 1-13.
Roy, Ananya. Poverty capital: Microfinance and the Making of Development, New York: Routledge,
2010.
Silver, Jonathan. The Infrastructural South: Techno-Environments of the Third Wave of Urbanization,
Cambridge, MA: The MIT Press, 2023.
Steinberg, Marc; Mukherjee, Rahul; and Punathambekar, Aswin. ‘Media power in digital Asia: Super
apps and megacorps’, Media, Culture & Society 44.8 (2022): 1405-1419.
Wahome, Michel Njeri Laura. Fabricating Silicon Savannah, PhD diss., University of Edinburgh, Edin-
burgh, 2020.
World Bank, World Development Report 2016: Digital Dividends, Washington, DC: World Bank, 2016.
Zhang, Lin. ‘Platformizing family production: The contradictions of rural digital labor in China’, The
Economic and Labour Relations Review 32.3 (2021): 341-359.
Zhang, Lin. The Labor of Reinvention: Entrepreneurship in the New Chinese Digital Economy, New York:
Columbia University Press, 2023.
163
IN/CONVENIENCE
EASY WINS AND LOW HANGING FRUIT.
BLUEPRINTS, TOOLKITS, AND PLAYBOOKS TO
ADVANCE DIVERSITY AND INCLUSION IN AI
TOMASZ HOLLANEK AND MAYA INDIRA GANESH
The emergence of AI has been accompanied by myriad moral and philosophical questions.
Some of these are speculative intuition pumps to activate a philosopher’s muscles of rea-
soning: ‘which way should the train trolley with the failed brakes be directed—jeopardizing
the life of one person working on the track, or five?’ There are more complex and pressing
real-world questions as well, like: should commercial art and design companies be pointing
their employees to generative AI image-making tools to speed up the ideation process? And,
similarly: should large language models be trained by psychologists and psychiatrists to deliver
therapeutic services to people with mental health conditions, thus making mental health
services cheaper and more accessible? In tandem with the emergence of AI technologies and
the ethical and moral challenges they animate, come lists of high-level ethics principles to
be prioritized by developers to ensure their products’ desirable social impact: human rights,
fairness, non-discrimination, privacy, transparency, accountability. New legal requirements
such as the European Union’s AI Act set out bright lines around applications at high levels
of risk, including credit scoring, targeted profiling, facial recognition, and automated hiring.
These bright lines have emerged thanks to documentation and analysis of risks and harms
in various contexts. However, the speed and scale at which AI technologies function, the
social and cultural complexities emerging at the sites of their application, and potential harms
emergent therein present challenges for the software development community. How should
the ethical concerns associated with AI be managed and mitigated? How will they be framed,
broken down into manageable parts, and addressed through computational, social, and pol-
icy actions? How can development teams be eased into openly and consciously engaging in
ethical deliberation as part of the design process? The question here is not about whether
ethical reasoning could, in fact, be turned into something modular, formulaic; into a pattern
to be adopted by designers. Nor is it about the consequences of translating ‘ethics’ into
user-friendly forms and formats that developers can immediately recognize and, therefore,
operationalize. If convenience is the ‘condition we inhabit within contemporary capitalism,’1
the key question is how? How can inconvenient questions about the trade-offs and conflicts
of interest be posed in ways that are both legible and bearable to those in the position to
transform the development pipeline?
Enter the toolkit. A toolkit is a design staple, a set of ready-to-use practices to solve a problem
or achieve a specific goal. A toolkit’s promise is scaling and continuity; that a set of instructions,
practices, or workflows will deliver a consistent, desired result. Sometimes that result is just
1 Joshua Neves and Marc Steinberg, ‘The Cultural Politics of In/Convenience’, Global Emergent Media: In
Progress, January 2023, https://www.globalemergentmedia.com/in-progress/the-cultural-politics-of-
in%2Fconvenience.
164 THEORY ON DEMAND
a process, rolling out across diverse spatial and temporal locations. As such, toolkits are a
favored methodology when things are to be made collaboratively and collectively, and when
the vagaries of time and place introduce discrepancies or inconsistencies. It is precisely the
assumption that AI can be designed to adhere to sets of values to avoid perpetuating harm that
brings AI ethics into the ambit of design. And it is the prevalent belief that following a prede-
termined process can ensure that the end-product is ethical and responsible that establishes
the toolkit as the primary instrument of AI ethics. The ongoing toolkitification of AI ethics, the
subject of this essay, reformulates ethical practice as frictionless, modular, as something that
can and should be made convenient and scalable, and, as such, responds to the demands
of convenience as a means of grappling with the complexity of our contemporary condition.
This complexity is software itself; its chains of supply and demand, infrastructural politics
and extractivism; its code—a palimpsest of social, economic, and cultural norms and values;
its power that, to its developers, is not a mysterious unaccountable force but something to
be tamed, trained, or taught. Software is supposed to make life and work easier. Software
is supposed to make the building of more software easier; modularity makes this possible.
Modularity is a central organizational logic of software and is about the division of labor;2 it
enables reach and, in this way, makes building elaborate projects more convenient—both
to imagine and to execute. In their study of how software developers assess their own pro-
fessional accountability for the ethical harms and lapses of AI and algorithmic technologies,
David Gray Widder and Dawn Nafus argue that, just as modularity helps to ‘minimize friction
as the code passes through many hands’,3 ethical challenges are similarly ‘encapsulated into
a module of work’ so as not to ‘introduce friction into the development process’.4 It is the kind
of division of labor that underlies ‘convenient media’.5
Yet, managing, overseeing, tending to, and patching software requires its own elaborate
system of systems. The ‘countervailing tendencies’ of elaborate software as simultaneously
modular and Byzantine, unknowable and yet accessible, serve a specific challenge for an
ideological or values-driven social or cultural project;6 in this case, the work of correcting and
reshaping institutional and technological systems, like AI, to be inclusive and equitable. For
values and ideologies are also simultaneously highly contextual, shifting, broad, diverse, and
yet also translatable into specific actions, positions, and normative rules. Translation requires
engagement beyond individuals, with institutional and structural actors and norms; this draws
us yet again to the vastness of systems. There’s a mirroring here between software, and
ideologies and institutions that design toolkits propose to intervene in but remain trapped by.
2 Lev Manovich, The Language of New Media, Cambridge, MA: The MIT Press, 2001, pp. 30-31.
3 David Gray Widder and Dawn Nafus, ‘Dislocated accountabilities in the “AI supply chain”: Modularity
and developers’ notions of responsibility,’ Big Data & Society 10:1 (2023): 2.
4 Widder and Nafus, ‘Dislocated accountabilities in the “AI supply chain”’, 7-8.
5 Neves and Steinberg, ‘The Cultural Politics of In/Convenience’.
6 Miriam Posner, ‘Breakpoints and Black Boxes: Information in Global Supply Chains,’ Postmodern
Culture 31:3 (2021), https://www.pomoculture.org/2021/12/01/breakpoints-and-black-boxes-
information-in-global-supply-chains/.
165
IN/CONVENIENCE
Toolkitication: the Making-convenient of ‘Ethics’ in and for AI
Design
There is no agreed-upon definition of a design toolkit as the term might refer to both technical
resources and educational brochures, to static text documents and interactive, web-based
applications.7 It is often used interchangeably with a singular tool, guideline, method, or blue-
print. And yet, we can identify toolkit-ification as an industry-wide phenomenon occurring in
response to the growing awareness of the risks and liabilities related to AI development. The
OECD’s Catalogue of Tools & Metrics for Trustworthy AI,8 the biggest collection of its kind
featuring over seven hundred toolkits (at the time of writing), speaks to the scale of this trend.9
The landscape of AI ethics toolkits is even wider, encompassing radical design ideation tools
and wiki-style web pages, such as the Intersectional AI Toolkit by Sarah Ciston.10 So wide, in
fact, that ‘toolkit-scoping’, the act of comparing available toolkits and testing their usefulness
for AI professionals, has become a sub-genre of AI ethics research.11 What this toolkit-scoping
work makes clear is that the toolkit paradigm privileges certain kinds of information, world-
views, and practices in its organization and presentation, and in doing so discursively re-shape
what (AI) ethics is.12 Specifically, the toolkit implies that ethical conflicts and challenges
associated with AI can be managed and that ethical practice is feasible and approachable.
Toolkits for trustworthy, safe, responsible, and ethical AI make a promise: that the tools they
contain are easily adoptable within existing workflows and adaptable to a particular team’s
needs, and that ensuring AI is made responsibly doesn’t imply a procedural revolution—only
selecting and applying an appropriate tool at the right stage of the design process.
The modularity of software development meets attempts at ‘translating’ the ‘theory’ of AI
ethics into development ‘practice’. Just like software is composed of smaller, constituent parts
that can be swapped out, reconstituted, and re-assembled, ‘ethics work’ is embraced in terms
of a similar organizing principle, as sets of actionable practices that can be stacked on and
slotted in. The most evocative example of an ethical issue getting turned into a development
module is the matter of AI bias; ‘technical tools to remove bias’ are, for instance, among the
most sought after by the users of the OECD’s Catalogue of Tools for Trustworthy AI.13 This par-
ticular approach to toolkitification of AI ethics that frames ethics work as technical work has
already been criticized for de-emphasizing the social, collective, and cultural value of diverse
stakeholders’ knowledge and engagement with AI—flattening and decontextualizing ethics by
7 Dorian Peters, Lian Loke, and Naseem Ahmadpour, ‘Toolkits, cards and games – a review of analogue
tools for collaborative ideation’, CoDesign 17:4 (2020): 410-434.
8 OECD, ‘Catalogue of Tools & Metrics for Trustworthy AI’, OECD.AI Policy Observatory, https://oecd.ai/en/
catalogue/tools.
9 Tomasz Hollanek, ‘The Ethico-politics of Design Toolkits: Responsible AI Tools, From Big Tech
Guidelines to Feminist Ideation Cards,’ forthcoming.
10 Intersectional AI Toolkit, https://intersectionalai.miraheze.org/wiki/Intersectional_AI_Toolkit.
11 Hollanek, ‘The Ethico-politics of Design Toolkits’.
12 Richmond Y. Wong, Michael A. Madaio, and Nick Merrill, ‘Seeing Like a Toolkit: How Toolkits Envision
the Work of AI Ethics,’ Proceedings of the ACM on Human-Computer Interaction 7, Issue CSCW1 (2023).
13 OECD, ‘Catalogue of Tools & Metrics for Trustworthy AI’.
166 THEORY ON DEMAND
proposing generalizable and scalable practices.14 But Widder and Nafus also highlight in their
study the kind of ethics work that no one even attempts to turn into a development module;
they show that this sort of work is ‘frequently left undone or cast as low status work, offloaded
to contractors’ or turned into ‘administrative labor no one else want[s] to do’.15 Our interest
lies precisely in the challenges that are seemingly ‘untranslatable’ into neat and prepackaged
work modules, into additional steps conveniently fitting the already established development
pipeline, and how design toolkits nonetheless promise to facilitate this difficult, inconvenient
work—to help their users address complex issues, such as discrimination of marginalized
groups perpetuated by AI systems, comprehensively and systematically. To think through
the effects of toolkitification as a process of making ‘ethics’ convenient in and for AI design,
we will look at a set of toolkits that aim to ensure AI is equitable and inclusive. These toolkits
move beyond the ‘remove-bias’ type of work and indeed acknowledge the complexity of the
ethical issues at hand. Yet they also promise to make these issues more approachable and
bearable, and the processes of addressing them not only manageable but also, at times, fun.
Diversity, Equity, and Inclusion in AI
Diversity, Equity, and Inclusion (also referred to as DEI) is positioned as both a challenge
and solution to AI’s problems. While unfair, biased, and discriminatory outcomes along the
lines of gender, race, class and intersections of these have emerged from the large-scale
applications of algorithmic and automated technologies, these problematic consequences
of AI deployment have also led to the establishment of the field of public, industrial, and aca-
demic inquiry into the ethics of algorithms and ethics of AI. Incident databases and registers
aggregate the various biased outcomes of algorithmic decision-making.16 Well-known cases
include: Amazon’s infamous Rekognition tool that negatively discriminated against women’s
CVs; the ProPublica investigation that revealed racially biased outcomes in a recidivism pre-
diction algorithm; the UK government’s disastrous A-level algorithm to predict school-leaving
grades that delivered results along the lines of class and postcode. Yet, the business magazine
Forbes reports that, according to startups in HR and recruitment, AI applications can enable DEI
by identifying biased or stereotypical use of language in job advertisements and by identifying
patterns of marginalization or disconnectedness among existing workers.17 DEI is positioned as
a solution to the ‘white guy problem’ in AI: the lack of gender, racial, and social diversity among
AI’s most powerful designers and developers is seen as a major influence on AI being biased
14 Wong, Madaio, and Merrill, ‘Seeing Like a Toolkit’. See also, Thilo Hagendorff, ‘The Ethics of AI Ethics:
An Evaluation of Guidelines,’ Minds & Machines 30 (2020): 99-120.
15 Widder and Nafus, ‘Dislocated accountabilities in the “AI supply chain”’, 8.
16 See AI Incident Database, https://incidentdatabase.ai; AIAAIC Repository, https://www.aiaaic.org/
aiaaic-repository.
17 Rebekah Bastian, ‘AI Brings Opportunities And Risks To Workplace DEI Efforts,’ Forbes, 8 May 2023,
https://www.forbes.com/sites/rebekahbastian/2023/05/08/ai-brings-opportunities-and-risks-to-
workplace-dei-efforts/?sh=4614ed8b4b2a; Jia Rizvi, ‘ How AI Can Be Leveraged For Diversity And
Inclusion,’ Forbes, 19 November 2023, https://www.forbes.com/sites/jiawertz/2023/11/19/how-ai-can-
be-leveraged-for-diversity-and-inclusion/?sh=6565f7af4ee9.
167
IN/CONVENIENCE
and discriminatory in the first place.18 DEI, understood as an equity measure, corrects the profit
and power imbalance associated with AI to a wider community.
If the very imagination of AI mirrors the aspirations of the white cis-male heteronormative
elite that populated the universities and military industrial complexes that AI and computing
emerged from,19 then DEI in AI assumes that pre-existing, data-driven algorithmic bias might
be spotted (better, earlier?) if people from marginalized and minoritized communities were
involved in the high-level design and development of AI. Moreover, such a diverse workforce
of decision makers might dilute the Silicon Valley monoculture that currently predominates AI
futures.20 Some organizations are bringing participatory, community-driven, embodied, and
scientific approaches to DEI. These include: Black in AI, a network of Black data scientists
working in AI; Our Data Bodies, a community-based research organization that investigates
how digital information from marginalized communities are collected, stored, and used by gov-
ernments and corporations; Data for Black Lives, a nonprofit focused on using data science
for positive change in the lives of Black people; and the Carceral Tech Resistance Network,
a campaign coalition against the experimental adoption and testing of technologies police,
prisons, and border enforcement.21
Despite these efforts, aspirations to DEI in AI are also met with fatigue or scorn as principled
high-level commitments betray reality. This happens when the optics of diversity becomes
a proxy for actual diversity. We are reassured of DEI when we can see it, when it is visible:
spotting, for instance, women speakers on a panel, or on the web page of an organization’s
leadership team. It is in response to this misinterpretation of DEI goals that new AI initiatives
have begun to emerge. Rosebud.Ai, for instance, works in gaming, branding, and digital mar-
keting, and offers to cut through the cost and effort of a photo shoot—and finding diverse
human models—by creating synthetic images of diverse people for websites and gameworlds.
In the same vein, the organizer of a 2023 tech conference created fake profiles of speakers
using generative AI tools to suggest that his conference was gender-diverse.22 The case of
18 Kate Crawford, ‘Artificial Intelligence’s White Guy Problem’, The New York Times, 25 June 2016, https://
www.nytimes.com/2016/06/26/opinion/sunday/artificial-intelligences-white-guy-problem.html.
19 The work of Alison Adam, whose early critical work pioneered feminist engagement with AI, is
exemplary here since she asks key questions of where our notions of intelligence and rationality come
from, and situates the relationship between place, embodiment, and knowing. See Alison Adam,
Artificial Knowing: Gender and the Thinking Machine, London: Routledge, 1998.
20 In the past half decade there has been a veritable flourishing of alternative, diverse adoptions and
refusals of AI by artists, organizers, designers, and scholars. Popular books by academics include Joy
Buolamwini, Unmasking AI: My Mission to Protect What Is Human in a World of Machines, New York:
Random House, 2023; Safiya Umoja Noble, Algorithms of Oppression: How Search Engines Reinforce
Racism, New York: New York University Press, 2018; Meredith Broussard, Artificial Unintelligence: How
Computers Misunderstand the World, Cambridge, MA: The MIT Press, 2018; and Ruha Benjamin, Race
after Technology: Abolitionist Tools for the New Jim Code, Cambridge, UK: Polity Press, 2019.
21 Sarah T. Hamid, ‘Community Defense: Sarah T. Hamid on Abolishing Carceral Technologies,’ Logic(s)
11: Care (2020), https://logicmag.io/care/community-defense-sarah-t-hamid-on-abolishing-carceral-
technologies/.
22 Natalie Lung and Ella Ceron, ‘Developer Conference Axed After Fake Female Profiles Outcry,’
Bloomberg, 27 November 2023, https://www.bloomberg.com/news/articles/2023-11-28/tech-
conference-faces-backlash-on-claims-of-fake-women-speakers.
168 THEORY ON DEMAND
generative AI being used to generate fake diversity is troubling and instructive here: it alerts
us to a minimization, a coloring-by-the-numbers approach to DEI, a shortcut that speaks to
DEI being, in fact, something at the end of a drop down menu or a box that has to be ticked off.
Even when AI companies do commit to DEI efforts to push beyond surface-level change,
these are met with resentment when business decisions undermine the original pledges. For
instance, when Kay Cole James, a vocal anti-LGBTQ campaigner, was appointed to Google’s
AI advisory council, the choice was vigorously opposed by the company’s employees,23 who
argued the appointment made clear that Google’s ‘version of “ethics” value[d] proximity to
power over the wellbeing of trans people, other LGBTQ people, and immigrants’.24 In another
example, the high-profile firing of Timnit Gebru from Google’s Ethical AI team in December
2020 was seen as an attack on one of the few highly decorated black women computer scien-
tists in AI vocal about the negative social and environmental consequences of development.25
Sara Ahmed writes in On Being Included: Racism and Diversity in Institutional Life that the
work of inclusion in institutions is unrewarded, unrecognized labor often done by the very
people who are affected by the lack of real DEI.26 Diversity work is institutional transformation
work, says Ahmed; she offers us rich language to consider what diversity work is and what
its workers must do. She uses hard, material, infrastructural terms referring to ‘sedimented’
institutional practices that must be unsettled through confrontation of discrimination and
lack of diversity and equity;27 to working as ‘institutional plumbers’; to being the person(s)
who moves ‘against the flow’ of the everyday.28 The flow of ‘business as usual’ that diversity
work disrupts is not actually a flow, she says; those experiencing discrimination, bias, and
a lack of inclusion and equity experience ‘flow’ as something solid. Diversity work, in other
words, is about working with immobility and immobilization. And it is usually the work of those
who do not quite fit into pre-existing norms, hence the requirement of DEI in the first place.
Diversity work is hard work. It is precisely this type of work that cannot be encapsulated into a
software development module, seamlessly fitting existing workflows. It is the kind of work that,
as Widder and Nafus demonstrate, is turned into ‘administrative labor no one else want[s] to
do’.29 And it is also the kind of work that new ethical AI toolkits that we analyze in what follows
promise to make easier, manageable, frictionless—in other words, convenient.
23 Jillian D'Onfro, ‘Google Employees Protest 'Anti-LGBTQ' Conservative's Appointment To AI Ethics
Council,’ Forbes, 1 April 2019, https://www.forbes.com/sites/jilliandonfro/2019/04/01/google-
employees-protest-anti-lgbtq-conservatives-appointment-to-its-ai-ethics-council/?sh=776ce37413e1.
24 Googlers Against Transphobia, ‘Googlers Against Transphobia and Hate’, Medium, 1 April 2019, https://
medium.com/@against.transphobia/googlers-against-transphobia-and-hate-b1b0a5dbf76.
25 Tom Simonite, ‘What Really Happened When Google Ousted Timnit Gebru,’ Wired, 8 June 2021, https://
www.wired.com/story/google-timnit-gebru-ai-what-really-happened/.
26 Sara Ahmed, On Being Included: Racism and Diversity in Institutional Life, Durham, NC: Duke University
Press, 2012.
27 Ahmed, On Being Included, pp. 175-176.
28 Ahmed, On Being Included, p. 186.
29 Widder and Nafus, ‘Dislocated accountabilities in the “AI supply chain”’, 8.
169
IN/CONVENIENCE
Toolkiticiation of DEI in AI
In this section, we compare several blueprints, toolkits, and playbooks that aim to help AI pro-
viders meet the goals of inclusivity and equity in AI development, deployment, and governance.
As mentioned, there is no single, agreed-upon definition of a toolkit that would distinguish
it from other formats, such as a guideline or blueprint—terms often used interchangeably.
So, for ease, we refer to all of the following as toolkits. We search for commonalities in their
framings of the difficult work that Ahmed refers to as ‘institutional plumbing’ to examine how
toolkitification makes what is uncomfortable approachable, the complex manageable, and
the irresolvable frictionless.
The first among these is the Mitigating Bias in Artificial Intelligence: An Equity Fluent Lead-
ership Playbook (2020), produced by the Center for Equity, Gender, and Leadership at the
Haas School of Business (University of California, Berkeley). The Playbook is a 62-page PDF
document introducing AI business decision-makers to the matters of DEI in AI. It is accom-
panied by a website from which users can download the titular ‘plays’—identifying necessary
‘moves’ (‘Enable diverse and multi-disciplinary teams working on algorithms and AI systems’),
relevant ‘players’ (such as ‘C-suite’ or ‘Human Resources’), and useful ‘tools’ (such as the
Alan Turing Institute’s Diversity Dashboard). Next is the Action Toolkit on Inclusive AI (2021),
developed by the Women4AI Daring Circle of the Women’s Forum for the Economy & Society,
in collaboration with UNESCO, the Oxford Internet Institute (OII), Shearman & Sterling, and
Price Waterhouse Coopers. While the Action Toolkit is, similarly, a PDF document, it is slightly
shorter (only 30-pages long), and aimed at both business and technical professionals, making
a case for inclusive AI and introducing some initial measures of success, as well as linking
to further, specialized tools. Then, we refer to A Blueprint for Equity and Inclusion in Artificial
Intelligence (2022), by the Global Future Council on Artificial Intelligence for Humanity under
the World Economic Forum (WEF), a 30-page long white paper presenting readers with issues
related to equity and inclusion at various stages of the development cycle, and linking to both
DEI-relevant tools, as well as application case studies. Finally, we also analyze A Blueprint for
Equitable AI: Building and Distributing Artificial Intelligence for Equitable Outcomes (2023),
from the Aspen Institute’s Science & Society team with sponsorship from Google DeepMind.
While this last, 34-page document does present ideas for potential strategies to meet DEI
goals in AI, it is in fact a report summarizing insights from a series of workshops convened
by the Aspen Institute team, and, as such, is the most general and least ‘action-oriented’ of
the selected toolkits. We refer to them as the Haas, Women4AI, WEF, and Aspen toolkits,
respectively.
Despite differences in the approaches proposed by these toolkits, we find similarities in
how they structure the relationships between different stakeholder groups—business deci-
sion-makers, developers, policymakers, as well as users of AI systems—and individual tools
and methods to achieve the goals of DEI in AI. They frame the complexity of DEI in AI in terms
of several dialectical tensions: they aspire to comprehensiveness while being comprehensible
to various stakeholders in the development process; being research-informed, easily digest-
ible and jargon-free; actionable but not simplistic; and necessary yet playful.
170 THEORY ON DEMAND
Comprehensive (and Comprehensible)
The first common feature of the toolkits we selected is their intended audience: they are
designed to be used, at least seemingly, by everyone in the AI ecosystem –not only design-
ers, data scientists, software engineers, but also business decision-makers, board members,
and policymakers. The WEF toolkit, for instance, explicitly addresses ‘managers and teams
responsible for the different stages of AI development, as well as decision-makers from all
sectors part of the AI ecosystem’ and also includes suggestions for governments, while the
Women4AI ‘instrument’ has been ‘created for C-suite executives, technologists, HR managers,
board members, developers, engineers and anyone who wants to change practice, policy,
strategy and attitudes within their organization towards ethical, inclusive AI’. Only the Haas
toolkit targets business decision-makers, but these are understood broadly, to include ‘a CEO,
a board member, an information / data / technology officer, a department head, a responsible
AI lead, a project manager’. And yet, despite this broad spectrum of the toolkits’ intended
users, all four toolkits address only those who are already present at the metaphorical design
table, rather than those who are still missing—the stakeholders who are most likely to bear
the brunt of the negative impact of AI development and whose involvement in the design
process the toolkits are supposed to encourage.
Related to this seemingly broad spectrum of intended users is another consistent feature of
our selected DEI toolkits: they all acknowledge that the challenge of making AI more equitable
and inclusive extends beyond technical questions of data bias and needs to be addressed
comprehensively at different stages of the development process and in different parts of the
AI ecosystem. This is key because the conception of DEI that the selected toolkits embody
does not amount to ‘bias-eradication’—the toolkits are nowhere near as simplistic and take
DEI work seriously, recognizing that it is hard work that must happen at various stages of the
design, development, and deployment process. While the WEF toolkit aims to ‘paint a com-
prehensive picture of challenges and opportunities for improvements in equity and inclusion
across the AI development life cycle and governance ecosystem’, the Haas toolkit notes it is
precisely because of the need to intervene at all the stages of the production pipeline that
‘addressing bias in AI is an issue for business leaders’ rather than technical workers, requir-
ing changes in hiring practices, among others. Women4AI similarly promises to guide users
through the necessary steps in transforming both the ‘organizational culture’ and aspects of
the design and development practice, encouraging design teams to be ‘as close as possible
to the populations likely to use it or be affected by it’, while the Aspen toolkit provides sugges-
tions ranging from concrete tips for changing the AI production process, such as ‘embedding
the topic of inclusivity into training for development team members’, to more general ones,
such as ‘preparing young people for AI through ethical tech education’. Because of this prom-
ise of comprehensiveness, the toolkits conflate design with policy: they merge different genres,
methods, and perspectives to ensure that they remain the only necessary entry point to the
question of inclusive and equitable AI for anyone—business decision-makers, regulators, or
software developers.
As the toolkits aspire to comprehensively tackle issues of equity and inclusivity, it is crucial to
highlight that each toolkit serves as a meta-toolkit, linking to more specialized tools, methods,
171
IN/CONVENIENCE
and guidelines. For instance, the WEF toolkit links its users to the AI Fairness Global Library,
where ‘other knowledge resources from leading institutions can be found to deepen the
topics presented’, while the Women4AI toolkit includes a list of other toolkits for ‘technical
audiences that seek to improve the ethical and inclusive practices of AI systems’, including
the Microsoft’s Responsible Innovation Toolkit and P wC’s Responsible AI Toolkit. The apparent
convenience of equitable AI toolkits is related precisely to this conception of comprehensive-
ness: they are meant to be designed for everyone, include information on all aspects of the
equity and inclusivity challenge, and gather (or link to) everything—all the necessary tools and
methods—one requires to tackle this challenge. This comprehensiveness the toolkit creators
have in mind is meant to acknowledge the complexity of the issues at stake, but not foreclose
their comprehensibility; on the contrary, comprehensiveness in this context becomes syn-
onymous with comprehensibility. It implies not information-overload, but total parse-ability,
making the questions of DEI ‘accessible’ to actors not usually burdened with DEI-related work.
Informed (but ‘jargon-free’)
The selected toolkits are informed by a vast amount of research and consultations with experts,
and the toolkits’ creators make this critical work purposefully explicit, detailing the processes
that led to the toolkits’ creation. This suggests rigor. The Aspen toolkit, for instance, presents
the insights drawn from discussions of ‘two diverse groups of experts’, including the legal
scholars Lilian Edwards and Sandra Wachter, and the data scientist Cathy O'Neil—well known
for their work on technology regulation, privacy, and feminist data science. The Haas toolkit
similarly draws from ‘academic literature and experts across disciplines – spanning sociology,
philosophy, engineering and more’, including the sociologist Gina Neff, the computer scientist
Stuart Russell, and the Managing Director of the AI Now Institute Sarah Myers-West. The WEF
toolkit, in turn, was created, by the Global Future Council on Artificial Intelligence for Humanity,
whose members include Angie Abdilla, specializing in indigenous knowledges and their rela-
tion to technology production, and Safiya Umoja Noble, the author of Algorithms of Oppression
(2018). The critical work of these scholars who are dedicated to exploring how machine
learning-based technologies reproduce or exacerbate social inequities would certainly be
considered ‘inconvenient’ by some decision-makers within technology companies because
they are radical in the sense of identifying the root of inequity, bias, and discrimination. To
pull things out by the roots is the sort of work that Sara Ahmed refers to as confronting the
‘sedimentation’ of institutional practice that does not acknowledge or make room for diversity.
Here comes the toolkit with its promise of ‘translation’ between disciplines and negotiation
between different, potentially conflicting goals. The WEF toolkit, for example, promises to
map ‘the vast amount of equity and inclusion challenges’ in the AI production and governance
ecosystem to then integrate them into ‘a digestible framework’ (our emphasis), while the Haas
toolkit sets itself apart from other available tools by promising to do the ‘crucial translational
work’ and present ‘conversations around “bias” in AI’—which can be, as it turns out, ‘mud-
dled and mean or refer to various concepts’—in a format that is ‘jargon-free and compre-
hensive’. The toolkit may refer to key AI ethics experts and institutions to legitimize itself as
an instrument of pro-justice change in AI. Yet, the very act of ‘translation’ or ‘adaptation’ of
the critical insights derived from critical AI ethics work for the purposes of corporate change
172 THEORY ON DEMAND
can, inadvertently, result in the ‘taming’ of critical, and often radical, positions; likely ridding
them of their transformative potential. The logic of the toolkit is that inconvenient or upsetting
perspectives—including the views of critical AI scholars whose views informed the develop-
ment of the equity toolkits in the first place—are toned down, made appealing, and bearable.
Necessary (but Playful); Actionable (but not Simplistic)
The selected toolkits strategically frame the work they facilitate as essential, rather than
optional. The Aspen toolkit says, ‘[p]ausing technological development and deployment until
all concerns are addressed is not feasible’, yet it is ‘critical to ensure that processes and
institutions exist to champion and implement efforts toward achieving equitable outcomes.’
This necessity extends beyond societal value; as the toolkits suggest, making AI more inclu-
sive translates to business value. The Women4AI toolkit highlights that equity and inclusivity
in AI development is ‘ultimately about helping your organization avoid the risks from biased
outcomes and reap the rewards from economies and societies which increasingly expect
inclusion as standard’. Likewise, the Haas toolkit underscores that using the tools and ‘plays’ it
collects to mitigate bias in AI is crucial ‘to unlock value responsibly and equitably’. If inclusive
AI is good for business and if toolkits can help achieve inclusive AI, then it follows that the
toolkits are good for business, too. There is a game-like quality to DEI in AI that the toolkits
encourage, leveling up as a strategy to unlock rewards for business. The toolkits manage their
intended users' expectations, recognizing that not all aspects of the DEI-fulfilling AI challenge
are easily and immediately addressable. The creators of the Haas toolkit point out that ‘de-bi-
asing’ AI fully is ‘not achievable’, while the Women4AI toolkit acknowledges that developing
inclusive AI ‘is a journey, not a destination’ and the toolkit can only serve as a starting point.
Yet, the toolkits tend to strategically highlight what is immediately solvable: the Haas toolkit,
for instance, suggests that using it will lead to some ‘quick wins’, as it introduces its users to
resources that can have concrete, immediate effects on the AI development pipeline. The
toolkits’ pedagogy: ensuring that business executives and developers get rewarded on the
journey towards inclusive, equitable AI with ‘easy wins’ and ‘low-hanging fruit’. Because if
these were not in sight, if there was no promise of eventual satisfaction and fun along the way,
the intended users of the toolkits could get discouraged and fail to persevere. Even when the
toolkit is a rather dense conference report, its appealing design serves as a promise of both
convenience and joy—even if this promise remains unrealized (and, perhaps, unrealizable).
Each of these sets of tensions in the toolkits’ framing suggest a desperation wrapped up
in a sincere commitment to DEI in AI. Perhaps because DEI is difficult, the authors want to
encourage rather than repel the potential toolkit user. Hence promises of comprehensive-
ness and comprehensibility, of simplicity but not being simplistic, of action and play, are like
treats to lure or even trick the user down the path of DEI; like honey to coat an oddly-shaped,
hard-to-swallow pill.
On making AI Ethics Inconvenient
Toolkits are now-ubiquitous material-cultural informational artifacts that organize many of our
shared organizational, political, and institutional work. Their ubiquity does not make them
173
IN/CONVENIENCE
benign or mundane, however. In 2021, the Indian government arrested a 22-year-old climate
activist and founder of the Indian chapter of Fridays for the Future, Disha Ravi, for assembling
an online toolkit for social media action and mobilization to support farmers who had been
protesting against the Farm Bills for over a year, camped out on the outskirts of and in the
capital, New Delhi.30 The Indian government charged Ravi with ‘collaborating’ to ‘spread
disaffection against the Indian state’ and sedition.31 Ravi shared the document with Greta
Thunberg who tweeted about it, which angered the state even further. In 2011, when former
Brazilian president Jair Bolsonaro was a congressperson, he launched a campaign against
a school-based education program to combat homophobia, arguing that the distribution of
information packages aka toolkits, which he called ‘gay kits’, in schools might actually ‘turn’
children gay through exposure.32 Toolkits can take on many forms, and their potential for
political action owes to the speed with which they promise the replication of ideas at scale.
These two instances demonstrate how their convenience can be perceived as inconvenient.
We bring that spirit to this critique. Furthermore, at the time of this writing, one of us is design-
ing a toolkit for software developers to fulfill the requirements of the EU AI Act associated with
high-risk applications of AI.33 This analysis of toolkits as convenient media therefore has been
developed in parallel with a deep engagement with the affordances of this form, and how
its limits might be tested to maintain inconvenience. ‘Inconvenience’ does not require that
toolkit design be user-unfriendly, its messaging pessimistic, or that the form be abandoned
altogether. In our practice, we find that it means reconfiguring the system of rewards that
the toolkit embodies, ensuring that ‘user satisfaction’ does not hinge solely on ticking a box
or marking a task as complete; it means highlighting that compliance is the bare minimum,
a starting point; practically, it means that any task or step that the toolkit incorporates is
followed by a ‘go further’ section—suggesting that there is always more to be done and that
the toolkit users can and should do more. A toolkit must inspire an ‘ongoing-ness’ of work. It
also means moving beyond the logic of modularity in where ethics work happens and how it
fits within existing workflows; it means facilitating reflection on the complexity of the issues.
So, in practical terms, we move away from self-contained ‘modules’—sets of tasks to be com-
pleted by different teams in a predetermined sequence—and into ‘spaces’: interconnected
areas of concern that different stakeholders must pass through and continue coming back
to, throughout software development and deployment. It means highlighting, rather than
gliding over, ‘inconvenient questions’—for instance, about the end user’s meaningful consent
30 Wikipedia contributors, ‘2020–2021 Indian farmers' protest’, Wikipedia, https://en.wikipedia.org/
wiki/2020%E2%80%932021_Indian_farmers%27_protest, accessed 22 February 2024.
31 ‘ India activist Disha Ravi arrested over farmers' protest “toolkit”’, BBC News, 14 February 2021, https://
www.bbc.co.uk/news/world-asia-india-56060232.
32 Ed Bracho-Polanco, ‘How Jair Bolsonaro used ‘fake news’ to win power’, The Conversation, 8 January
2019, https://theconversation.com/how-jair-bolsonaro-used-fake-news-to-win-power-109343.
33 See ‘The In-depth EU AI Act Toolkit’, Leverhulme Centre for the Future of Intelligence, http://lcfi.ac.uk/
projects/ai-innovation-praxis/eu-ai-act-toolkit/. The Act classifies AI systems according to different
levels of risk they may pose in predefined areas of applications; a system can be classified as a source
of high risk if it has potential to adversely impact people’s health, safety, or fundamental rights (such
as dignity or equality) in predefined areas of use, including biometric identification, law enforcement,
and recruitment; producers of such systems are required to implement measures to mitigate the AI
system’s undesirable societal consequences.
174 THEORY ON DEMAND
and what feminism and decolonial theory can teach us about its elicitation through design.34
Equity, justice, and fairness are rich in friction; they demand a historical, structural, and
institutional reckoning; and constant and passionate engagement with actual diversity—of
thought, experience, situation, values—with little reassurance of total success. The toolkits
we have referred to do acknowledge the complexity of the DEI challenges in AI development,
deployment, and governance; as the Aspen toolkit makes clear, there are no ‘silver bullets to
ongoing challenges’. And yet, despite this recognition highlighted as a premise for DEI work
in the AI ecosystem, the messaging of the toolkits for inclusive AI—precisely because they
are framed as toolkits rather than reports or even guidelines—points to the paradox that the
toolkitficiation of DEI work, and ethics more broadly, is necessarily ridden by. There might
be no simple solutions to the problem that structural injustice constitutes, and yet a toolkit
implies the existence of such ready-made tools; a blueprint suggests that the boundaries
and hierarchies of the AI ecosystem can still be redrawn, as long as the teams, companies,
and governments adhere to a clear-cut inclusivity template; a playbook signals that there
are tested tactics and methods, a means of redirecting activity to achieve the desired (and
desirable) outcome.
Toolkits might acknowledge complexity and difficulty, but their logic remains that of action-
ability and convenience. None of the struggle Ahmed talks about—the negotiation with immo-
bilization, culture, language, established practice, human social relations, or organizational
workflows—are in evidence here. But meaningful ethicopolitical life is, as Louise Amoore
argues in Cloud Ethics, precisely about ‘irresolvable struggles, intransigence, duress, and
opacity, and it must continue to be so for if a future possibility for politics is not to be eclipsed
by the output signals of algorithms.’35 Amoore alerts us that there is more than just the matter
of how, and if ,DEI in AI toolkits can deliver on empowering designers and decision-makers
to produce ethical technologies: the expansion demanded of algorithms to parse, process,
and encompass the intricacies and entanglements of human social life, to make accountable,
ethical, fair, unbiased, and trustworthy decisions, cannot happen in a vacuum. The algorithms
require frequent human intervention to maintain and manage their behavior; this human
intervention, in turn, requires its own constant tending-to within shifting and unequal social,
cultural, institutional, and organizational arrangements. This is far from smooth, this is hard
to automate, but this is the work of our time.
References
‘ India activist Disha Ravi arrested over farmers' protest “toolkit”’, BBC News, 14 February 2021,
https://www.bbc.co.uk/news/world-asia-india-56060232.
Smith, Genevieve and Rustagi, Ishita. Mitigating Bias in Artificial Intelligence: An Equity Fluent Leader-
ship Playbook, Berkeley, CA: Berkeley Haas Center for Equity, Gender and Leadership, 2020, https://
34 Joana Varon and Paz Peña, ‘Artificial intelligence and consent: a feminist anti-colonial critique’, Internet
Policy Review 10.4 (2021), https://policyreview.info/articles/analysis/artificial-intelligence-and-consent-
feminist-anti-colonial-critique.
35 Louise Amoore, Cloud Ethics: Algorithms and the Attributes of Ourselves and Others, Durham, NC: Duke
University Press, 2020, p. 172.
175
IN/CONVENIENCE
haas.berkeley.edu/wp-content/uploads/UCB_Playbook_R10_V2_spreads2.pdf.
‘The In-depth EU AI Act Toolkit’, Leverhulme Centre for the Future of Intelligence, http://lcfi.ac.uk/proj-
ects/ai-innovation-praxis/eu-ai-act-toolkit/.
Adam, Alison. Artificial Knowing: Gender and the Thinking Machine, London: Routledge, 1998.
Ahmed, Sara. On Being Included: Racism and Diversity in Institutional Life, Durham, NC: Duke Univer-
sity Press, 2012.
AI Incident Database, https://incidentdatabase.ai.
AIAAIC Repository, https://www.aiaaic.org/aiaaic-repository.
Amoore, Louise. Cloud Ethics: Algorithms and the Attributes of Ourselves and Others, Durham, NC:
Duke University Press, 2020.
Aspen Institute Science & Society Program. A Blueprint for Equitable AI: Building and Distributing
Artificial Intelligence for Equitable Outcomes, Washington, DC: The Aspen Institute, 2023, https://www.
aspeninstitute.org/wp-content/uploads/2023/01/Equitable-AI-Aspen-Institute.pdf.
Bastian, Rebekah. ‘AI Brings Opportunities And Risks To Workplace DEI Efforts,’ Forbes, 8 May 2023,
https://www.forbes.com/sites/rebekahbastian/2023/05/08/ai-brings-opportunities-and-risks-to-work-
place-dei-efforts/?sh=4614ed8b4b2a.
Benjamin, Ruha. Race after Technology: Abolitionist Tools for the New Jim Code, Cambridge, UK: Polity
Press, 2019.
Bracho-Polanco, Ed. ‘How Jair Bolsonaro used ‘fake news’ to win power’, The Conversation, 8 January
2019, https://theconversation.com/how-jair-bolsonaro-used-fake-news-to-win-power-109343.
Broussard, Meredith. Artificial Unintelligence: How Computers Misunderstand the World, Cambridge,
MA: The MIT Press, 2018.
Buolamwini, Joy. Unmasking AI: My Mission to Protect What Is Human in a World of Machines, New
York: Random House, 2023.
Crawford, Kate. ‘Artificial Intelligence’s White Guy Problem’, The New York Times, 25 June 2016,
https://www.nytimes.com/2016/06/26/opinion/sunday/artificial-intelligences-white-guy-problem.html.
D'Onfro, Jillian. ‘Google Employees Protest 'Anti-LGBTQ' Conservative's Appointment To AI Ethics
Council,’ Forbes, 1 April 2019, https://www.forbes.com/sites/jilliandonfro/2019/04/01/google-employ-
ees-protest-anti-lgbtq-conservatives-appointment-to-its-ai-ethics-council/?sh=776ce37413e1.
Global Future Council on Artificial Intelligence for Humanity. A Blueprint for Equity and Inclusion in
Artificial Intelligence, Geneva: World Economic Forum, 2022, https://www3.weforum.org/docs/WEF_A_
Blueprint_for_Equity_and_Inclusion_in_Artificial_Intelligence_2022.pdf.
Googlers Against Transphobia, ‘Googlers Against Transphobia and Hate’, Medium, 1 April 2019,
https://medium.com/@against.transphobia/googlers-against-transphobia-and-hate-b1b0a5dbf76.
Hagendorff, Thilo. ‘The Ethics of AI Ethics: An Evaluation of Guidelines,’ Minds & Machines 30 (2020):
99-120.
Hamid, Sarah T. ‘Community Defense: Sarah T. Hamid on Abolishing Carceral Technologies,’ Logic(s)
11: Care (2020), https://logicmag.io/care/community-defense-sarah-t-hamid-on-abolishing-carcer-
al-technologies/.
Hollanek, Tomasz. ‘The Ethico-politics of Design Toolkits: Responsible AI Tools, From Big Tech Guide-
lines to Feminist Ideation Cards,’ forthcoming.
Intersectional AI Toolkit, https://intersectionalai.miraheze.org/wiki/Intersectional_AI_Toolkit.
176 THEORY ON DEMAND
Lung, Natalie and Ceron, Ella. ‘Developer Conference Axed After Fake Female Profiles Outcry,’
Bloomberg, 27 November 2023, https://www.bloomberg.com/news/articles/2023-11-28/tech-confer-
ence-faces-backlash-on-claims-of-fake-women-speakers.
Manovich, Lev. The Language of New Media, Cambridge, MA: The MIT Press, 2001.
Neves, Joshua and Steinberg, Marc. ‘The Cultural Politics of In/Convenience’, Global Emergent Media:
In Progress, January 2023, https://www.globalemergentmedia.com/in-progress/the-cultural-poli-
tics-of-in%2Fconvenience.
Noble, Safiya Umoja. Algorithms of Oppression: How Search Engines Reinforce Racism, New York:
New York University Press, 2018.
OECD, ‘Catalogue of Tools & Metrics for Trustworthy AI’, OECD.AI Policy Observatory, https://oecd.ai/
en/catalogue/tools.
Peters, Dorian; Loke, Lian; and Ahmadpour, Naseem. ‘Toolkits, cards and games – a review of ana-
logue tools for collaborative ideation’, CoDesign 17:4 (2020): 410-434.
Posner, Miriam. ‘Breakpoints and Black Boxes: Information in Global Supply Chains,’ Postmodern
Culture 31:3 (2021), https://www.pomoculture.org/2021/12/01/breakpoints-and-black-boxes-infor-
mation-in-global-supply-chains/.
Rizvi, Jia. ‘ How AI Can Be Leveraged For Diversity And Inclusion,’ Forbes, 19 November 2023, https://
www.forbes.com/sites/jiawertz/2023/11/19/how-ai-can-be-leveraged-for-diversity-and-inclusion/?sh=-
6565f7af4ee9.
Simonite, Tom. ‘What Really Happened When Google Ousted Timnit Gebru,’ Wired, 8 June 2021,
https://www.wired.com/story/google-timnit-gebru-ai-what-really-happened/.
Varon, Joana and Peña, Paz. ‘Artificial intelligence and consent: a feminist anti-colonial critique’, Inter-
net Policy Review 10.4 (2021), https://policyreview.info/articles/analysis/artificial-intelligence-and-con-
sent-feminist-anti-colonial-critique.
Widder, David Gray and Nafus, Dawn. ‘Dislocated accountabilities in the “AI supply chain”: Modularity
and developers’ notions of responsibility,’ Big Data & Society 10:1 (2023).
Wikipedia contributors, ‘2020–2021 Indian farmers' protest’, Wikipedia, https://en.wikipedia.org/
wiki/2020%E2%80%932021_Indian_farmers%27_protest, accessed 22 February 2024.
Women4AI Daring Circle. Action Toolkit on Inclusive AI, Paris: Women’s Forum for the Economy & Soci-
ety, 2021,file:///Users/mi373/Documents/2024/writing https://storageprdv2inwink.blob.core.windows.
net/3b7997a1-b9ec-4c99-89ea-f5199663f903/b9c9f204-6bec-41d8-8453-05aaf4328ff5.
Wong, Richmond Y.; Madaio, Michael A.; and Merrill, Nick. ‘Seeing Like a Toolkit: How Toolkits Envi-
sion the Work of AI Ethics,’ Proceedings of the ACM on Human-Computer Interaction 7, Issue CSCW1
(2023).
177
IN/CONVENIENCE
BIOGRAPHIES
Neta Alexander is an Assistant Professor of Film and Media at Yale University and a former
Assistant Editor of Journal of Cinema and Media Studies (JCMS). She earned a PhD from
NYU and a M.A from Columbia University. Her work merges film and media theory, science
and technology studies, and critical disability studies. Her first book, Failure (co-authored
with Arjun Appadurai; Polity, 2020) studies how Silicon Valley and Wall Street monetize failure
and forgetfulness. Her second book, Interface Frictions (forthcoming, Duke University Press),
studies four ubiquitous interface design features—refresh, playback speed, autoplay, and
Night Shift—to develop a theory of digital debility.
Armin Beverungen works at the Centre for Digital Cultures and the Institute of Sociology and
Cultural Organization at Leuphana University in Lüneburg, Germany. As part of a project on
“automating the logistical city” (https://logistical.city/) he currently researches Amazon urban-
ism, and as part of another project on “smartness as wealth“ (http://smartnesswealth.net/) he
explores the promises of wealth associated with smart technologies, in particular digital twins.
DarrenByleris Assistant Professor of International Studies at Simon Fraser University in Van-
couver, British Columbia. He is the author of an award-winning ethnography titledTerror Cap-
italism: Uyghur Dispossession and Masculinity in a ChineseCity(DukeUniversity Press 2022)
and a narrative-driven book titledIn the Camps: China's High-Tech Penal Colony(Columbia
Global Reports 2021). His current research and teaching is focused on infrastructure devel-
opment and global China.
Liza Rose Cirolia is a Senior Researcher at the African Centre for Cities, University of Cape
Town. Her teaching, research and policy work focuses on infrastructural transitions, urban
statecraft, and finance, specifically in the context of Africa's urbanization.
Maya Indira Ganesh is an assistant teaching professor co-directing a master’s program in
AI Ethics and Society; and is a senior research fellow at the Leverhulme Centre for the Future
of Intelligence at the University of Cambridge, UK.Maya has a Drphil in Cultural Studies from
Leuphana University, Lüneburg, Germany. Her doctoral work took the case of the ‘ethics of
autonomous driving’ to study the implications of governance of and by algorithmic/AI tech-
nologies for human social relations, spaces, and bodies. Maya’s work is organised around
empirical inquiry into how people, places, and social organisations meet the design and devel-
opment of technology. She draws on varied theoretical and methodological genres, including
feminist scholarship, social and cultural studies of technology, and Science and Technology
Studies. Prior to academia, Maya spent over a decade as a researcher and activist working
at the intersection of gender justice, security, and digital freedom of expression.
178 THEORY ON DEMAND
Dr Tomasz Hollanek is a Postdoctoral Research Fellow at the Leverhulme Centre for the
Future of Intelligence (LCFI) and an Affiliated Lecturer in the Department of Computer Sci-
ence and Technology at the University of Cambridge, working at the intersection of design
theory, technology ethics, and critical artificial intelligence studies. His ongoing research
explores the possibility of applying critical design methods – prioritizing the goals of social
justice and environmental sustainability – in the governance, development, and deployment
of AI systems. This includes recent work on more-than-human-centered approaches to tech
design; the In-depth EU AI Act Toolkit, helping developers translate the requirements of the
European Union’s AI Act into design practice; and responsible deployment of AI in the digital
afterlife industry.
Orit Halpern is Professor and Chair of Digital Cultures at Technische Universität Dresden. Her
work bridges the histories of science, computing, and cybernetics with design. She completed
her Ph.D. at Harvard University. Her first book Beautiful Data: A History of Vision and Reason
(Duke UP 2015) investigates histories of big data, design, and governmentality. Her recent
book with Robert Mitchell (MIT Press 2023) is The Smartness Mandate.
Dr. Mél Hogan is the host ofThe Data Fixpodcast and is the Director of the Environmental
Media Lab (EML). She is an Associate Professor in the Department of Film and Media at
Queen's University. Her research focuses on data infrastructure, extractive AI, and genomic
media -- each understood from within the contexts of planetary catastrophe, and collective
anxieties about the future.http://melhogan.com
Tung-Hui Huis a poet and a media theorist. He is the author of five books, includingA Prehis-
tory of the Cloud(MIT Press, 2015) andDigital Lethargy: Dispatches from an Age of Disconnec-
tion(MIT Press, 2022).He is an associate professor of English at the University of Michigan.
Steven Gonzalez Monserrate is apostdoctoral researcherat theFixing Futures Research
Training GroupatGoethe University. He received his PhD in theHistory, Anthropology, Sci-
ence, Technology & Society (HASTS)program at the Massachusetts Institute of Technology.
His book project,Cloud Ecologies, is an environmental ethnography of data centers in New
England, Arizona, Puerto Rico, and Singapore. Committed to public scholarship, his writing
appears in venues includingWired, Aeon, Popular Science, Anthropology News, ABC News,
BBC News, NPRand more.
Rahul Mukherjee is Associate Professor of TV and New Media in Department of Cinema &
Media Studies at University of Pennsylvania. His scholarship engages the environmental and
logistical dimensions of infrastructures. He is the author ofRadiant Infrastructures: Media,
Environment, and Cultures of Uncertainty(Duke UP, 2020) and is completing a book about
mobile media distribution (under contract with MIT Press). His essays have been published
inCritical Inquiry,Science, Technology & Human Values, andJournal of Visual Culture.He has
co-edited journal special issues regarding platformization of everyday life in India (Asiascape)
and super apps and megacorps (Media, Culture & Society).
179
IN/CONVENIENCE
Joshua Nevesis Associate Professor of Film Studies and Director of the Global Emergent
Media (GEM) Lab at Concordia University. He is co-author (with Aleena Chia, Susanna Paa-
sonen, and Ravi Sundaram) ofTechnopharmacology(Minnesota University Press / Meson
Press, 2022), author ofUnderglobalization: Beijing’s Media Urbanism and the Chimera of
Legitimacy(Duke University Press, March 2020), and co-editor (with Bhaskar Sarkar) ofAsian
Video Cultures: In the Penumbra of the Global(Duke University Press, 2017). His work is
published inMedia Theory,Cultural Critique, Social Text, Discourse, Culture Machine, Film
Quarterly, Cinema Journal, Sarai, The Routledge Companion to Risk and Media, among others.
Susanna Paasonen is Professor of Media Studies at University of Turku, Finland. With an
interest in studies of sexuality, media and affect, she is the PI of the research consortium “Inti-
macy in Data-Driven Culture” (2019-2025) and the author of e.g., Many Splendored Things:
Thinking Sex and Play (Goldsmiths Press 2018), NSFW: Sex, Humor and Risk in Social Media
(with Kylie Jarrett and Ben Light, MITP 2019), Dependent, Distracted, Bored: Affective For-
mations in Networked Media (MIT 2021), and Yul Brynner: Exoticism, Cosmopolitanism and
Screen Masculinity (Edinburgh University Press, 2023).
Andrea Pollio is an assistant professor at the department of Urban and regional studies and
planning of the Polytechnic of Turin (Italy) and research associate of the African Centre for
Cities, at the University of Cape Town. His work explores the making of innovation economies
and platform capitalisms in urban Africa.
Marc Steinberg is Professor of Film and Moving Image Studies at Concordia University, Mon-
treal, and director of The Platform Lab. His work focuses on the platformization of culture
and economy in Japan and Asia. He is the author of The PlatformEconomy: How Japan
Transformed the Commercial Internet(Universityof Minnesota Press, 2019) and Media and
Management (University of Minnesota Press, 2021), and he has edited special issues includ-
ing “Regional Platforms” for Asiascape: Digital Asia and “Media Power in Digital Asia: Super
Apps and Megacorps” for Media, Culture & Society.
Theory on Demand #54
In/Convenience: Inhabiting the Logistical Surround
Editors: Joshua Neves and Marc Steinberg
Convenience is the feeling and aspiration that animates our platformed present. As
such, it poses urgent techno-political questions about the everyday digital habitus.
From next-day delivery, gig work, and tele-health to cashless payment systems, data
centers, and policing – convenience is an aordance and an enclosure; our logistical
surround. Driving every experience of convenience is the precarious work, proprietary
algorithms, or predatory schemes that subtend it. This collaborative book traces how the
logistical surround is transformed by thickening digital economies and networked rituals,
examining contemporary conveniences across a wide range of practices and geographies.
Contributors examine the ineluctable relation between convenience and its constitutive
opposite, inconvenience, considering its infrastructural, aective, and compulsory
dimensions. Living in convenience is thus both a hyper visible manifestation of so-called
late capitalism and a pervasive mood that fades into the background (like the data centers
that power it). Bringing the agonistic relation of in/convenience to center stage, this volume
analyzes the logistics of delivery, streaming porn, cloud computing, water infrastructures,
smartness paradigms, convenience stores, sleep apps, surveillance, AI ethics, and much
more – rethinking the cultural politics of convenience for the present conjuncture.
Joshua Neves is Associate Professor, Concordia University, and author of
Underglobalization: Beijing’s Media Urbanism and the Chimera of Legitimacy.
Marc Steinberg is Professor of Cinema, Concordia University, and author of
The Platform Economy: How Japan Transformed the Consumer Internet.
Printed on demand
ISBN: 9789083412559
