ChapterPDF Available

Towards a Sociological Conception of Artificial Intelligence

Authors:

Abstract

Social sciences have been always formed and influenced by the development of society, adjusting the conceptual, methodological, and theoretical frameworks to emerging social phenomena. In recent years, with the leap in the advancement of Artificial Intelligence (AI) and the proliferation of its everyday applications, “non-human intelligent actors” are increasingly becoming part of the society. This is manifested in the evolving realms of smart home systems, autonomous vehicles, chatbots, intelligent public displays, etc. In this paper, we present a prospective research project that takes one of the pioneering steps towards establishing a “distinctively sociological” conception of AI. Its first objective is to extract the existing conceptions of AI as perceived by its technological developers and (possibly differently) by its users. In the second part, capitalizing on a set of interviews with experts from social science domains, we will explore the new imaginable conceptions of AI that do not originate from its technological possibilities but rather from societal necessities. The current formal ways of defining AI are grounded in the technological possibilities, namely machine learning methods and neural network models. But what exactly is AI as a social phenomenon, which may act on its own, can be blamed responsible for ethically problematic behavior, or even endanger people’s employment? We argue that such conceptual investigation is a crucial step for further empirical studies of phenomena related to AI’s position in current societies, but also will open up ways for critiques of new technological advancements with social consequences in mind from the outset.
Towards a Sociological Conception
of Artificial Intelligence
Jakub Mlynář1,2[0000-0001-5206-3212], Hamed S. Alavi2,3[0000-0001-8443-7514], Himanshu
Verma2[0000-0002-2494-1556] and Lorenzo Cantoni4[0000-0001-5644-6501]
1 Charles University, Nám. Jana Palacha 2, 116 38 Praha, Czech Republic
2 University of Fribourg, Boulevard de Pérolles 90, 1700 Fribourg, Switzerland
3 University College London, 66-72 Gower Street, WC1E 6EA London, United Kingdom
4 University of Lugano, Via Giuseppe Buffi 13, 6900 Lugano, Switzerland
jakub.mlynar@ff.cuni.cz
h.alavi@ucl.ac.uk
himanshu.verma@unifr.ch
lorenzo.cantoni@usi.ch
Abstract. Social sciences have been always formed and influenced by the
development of society, adjusting the conceptual, methodological, and
theoretical frameworks to emerging social phenomena. In recent years, with the
leap in the advancement of Artificial Intelligence (AI) and the proliferation of
its everyday applications, “non-human intelligent actors” are increasingly
becoming part of the society. This is manifested in the evolving realms of smart
home systems, autonomous vehicles, chatbots, intelligent public displays, etc.
In this paper, we present a prospective research project that takes one of the
pioneering steps towards establishing a “distinctively sociological” conception
of AI. Its first objective is to extract the existing conceptions of AI as perceived
by its technological developers and (possibly differently) by its users. In the
second part, capitalizing on a set of interviews with experts from social science
domains, we will explore the new imaginable conceptions of AI that do not
originate from its technological possibilities but rather from societal necessities.
The current formal ways of defining AI are grounded in the technological
possibilities, namely machine learning methods and neural network models. But
what exactly is AI as a social phenomenon, which may act on its own, can be
blamed responsible for ethically problematic behavior, or even endanger
people’s employment? We argue that such conceptual investigation is a crucial
step for further empirical studies of phenomena related to AI’s position in
current societies, but also will open up ways for critiques of new technological
advancements with social consequences in mind from the outset.
Keywords: Artificial Intelligence, Sociology, Social Sciences.
1 AI as a Sociological Phenomenon
Given the rapidly growing importance of Artificial Intelligence (AI) in many domains
of social life, it is striking that the interest of sociologists and social scientists in AI
has been quite scarce. At the end of 20th century, AI was occasionally discussed in
2
sociology as a methodological tool for data analysis and theory development, yet not
as a social phenomenon in its own right. However, as it is expected that the social
impact of AI will continue to increase over the next years, contributing to transform
the ways people organize economical production, learn and spend leisure time, to
name just a few concerned fields, we argue that sociology and other social sciences
need to acquire an adequate understanding of how artificial intelligence is and should
be grown into a social actor, reflecting its relevance and consequentiality in different
layers of social organization and social reality.
This precisely is the central aim of our project, which we describe in this paper. It
intends to provide a sociological conception of AI, i.e. understanding AI as a social
phenomenon
and a non-human social actor
. We are convinced that scientific studies
of human beings and their collectivities need to tailor their conceptual tools to the
society of the 21st century. This requires, first and foremost, a proper understanding
of specific aspects of digitalization in everyday life, which is a domain where AI
plays an increasing role. Through an exploratory study, our intention is to “prototype“
the methodological and conceptual cross-fertilization between sociology and the
fields that deal traditionally with the subject of AI, such as computer science,
philosophy of mind, and cognitive psychology. The second equally important goal of
our project is to execute an investigation of the possibilities of sociology to influence
and guide the future of AI-related developments in our societies. In this paper, rather
than presenting results of empirical research, we introduce and discuss an agenda to
proceed.
1.1 The Case of Facebook
In the recent public and political discourses, the role of AI has been already the
subject of challenging debates. In the US Senate hearing held on April 10th, 2018,
with Mark Zuckerberg, the founder and CEO of Facebook, the words “Artificial
Intelligence” or their abbreviation “A.I.” were used 29 times. In fact, more often than
“trust” (20 occurrences), “transparency” and “transparent” (18 occ.), or even
“freedom(s)” (4 occ.) and “democracy/democratic” (3 occ.). This was also noticed by
the participants of the meeting, one of whom noted that Zuckerberg “brought [AI] up
many times during [his] testimony.” (Senator Peters, 3:47:13-3:47:16 of Zuckerberg’s
US Senate hearing as available online ) This highly medially exposed example
1
documents and illustrates the role that AI has taken in current societies worldwide. It
also points to the major role that the AI “systems” or “tools” might play in future
social developments. Indeed, Zuckerberg himself stressed the societal relevance of
such questions more than once during the hearing session: “[A]s were able to
technologically shift towards especially having AI proactively look at content, I think
that that’s going to create massive questions for society about what obligations we
want to require companies ... to fulfill.” (Mark Zuckerberg, 2:53:48-2:54:05 of the
hearing) And also about one hour later: “[T]he core question youre asking about, AI
transparency, is a really important one that people are just starting to very seriously
1 https://www.youtube.com/watch?v=pXq-5L2ghhg
3
study, and that’s ramping up a lot. And I think this is going to be a very central
question for how we think about AI systems over the next decade and beyond.” (Mark
Zuckerberg, 3:47:46-3:48:02 of the hearing)
1.2 Our Approach
In providing a sociological conception of AI, our project starts from an extensive
literature review, as well as content analysis of media production related to AI. At the
most fundamental level, we aim to conceptualize AI sociologically, providing answers
to questions such as: Are there inherent differences between human and non-human
(AI) social actors? Should we revisit and reconsider our presumptions of human
uniqueness? And, on the other hand, can we truly speak about anything like “AI in
general”, or do we rather encounter loosely related instances of phenomena in the
sense of “family resemblances” [1]? As explained further in this paper, the answers
provided will be based on analysis of several kinds of empirical data, qualitative and
quantitative in nature. The resulting conception of AI, although sociological in its
nature, will be then adaptable by other social sciences such as communication,
economy, political science, or social anthropology.
2 The Study of AI in Sociology and Computer Science
2.1 AI and Non-Human Actors in Sociology
With the growth of initiatives such as Ubiquitous Computing (UbiComp
, see e.g. [2]),
and the remarkable entry of “smart” systems into the domain of everyday social lives,
it is necessary to reconsider the position of AI in sociology and vice versa
.
Historically and traditionally, sociology was usually practiced – as Zygmunt Bauman
once nicely put it as “a narrative on what follows from the fact that man is not
alone” [3]. One of the tacit presumptions, arising from this conception of sociology as
a science on accumulated and interrelated human beings, has been the disregard for
non-human actors and material components of the social world (cf. [4]). Sociologists
simply considered the “non-human“ and “extra-human” to compose only the
environment of sociologically relevant phenomena, which does not have to be taken
into account. Since the late 1970s, this neglect was explicitly formulated and criticised
in sociological orientation to subjects such as the natural environment [5], animals [6],
or technology [7]. Focusing specifically on AI – which has been extensively discussed
in cognitive psychology, philosophy of mind, and computer science already since the
1950s (cf. [8]) –, few sociologists have started writing on the subject in the 1980s and
1990s. However, up to this day, AI has been almost exclusively conceived in
sociological context only as a methodological tool in statistical or textual analysis [9],
and development of sociological theories [10] in other words, an “application of
machine intelligence techniques to social phenomena”, i.e. the Artificial Social
Intelligence [11].
4
Broadly speaking, so far, AI has not been systematically considered as a social and
sociological phenomenon sui generis and the discipline of sociology lacks a suitable
conception of AI, which could serve as a framework for empirical studies. Rare
exceptions include that of Woolgar [12], who proposed a “sociology of machines”,
arguing that we should see the “AI phenomenon as an occasion for reassessing the
central axiom of sociology that there is something distinctively ‘social’ about human
behaviour” (p. 557), and proposed that we should examine the underlying assumption
in social sciences that there is a fundamental difference between humans and
machines (and, by extension, also between human and machine intelligence). Apart
from advocating sociological research of AI discourse and AI research practices, he
also claimed more broadly that “the phenomenon of AI provides an opportunity for
investigating how presumptions of the distinction between human and machine
delimit social inquiry” (p. 568). Wolfe [13] explored Woolgar’s radical idea and
demonstrated that “interpretive” sociological approaches (such as ethnomethodology
or symbolic interactionism), rather than “systemic” ones, may “expand and elaborate”
the hypothesis of human uniqueness in comparison with AI. Schwartz [14] suggested
that AI has to be studied with regard to the social context (setting) in which it is
“implemented”, and characterized AI systems as “social actors playing social roles”
(p. 199). At the turn of the century, Malsch [15] discussed the proximities of AI and
sociology through the concept of socionics. This field, standing at the intersection of
sociology and AI, aims to “address the question how to exploit models from the social
world for the development of intelligent computer technologies” (p. 155), exploring
the specificities of modern societies and resilient adaptability of social systems in
order to provide means of translating these features into computer-based technologies.
Indeed, the most influential attempt to incorporate non-human actors into sociological
thinking is the conception of Bruno Latour [16]. His actor network theory (ANT)
aims, among other things, to transgress the distinction of human and non-human
actors ([17]; similarly to Woolgar’s [12] argument presented above), acknowledging
technologies and objects as partakers in the construction of society. However, AI as a
phenomenon is not discussed by Latour in this context. More recently, Muhle [18]
presents an ethnomethodological study of “embodied conversational agents” (bots) in
the virtual world of a massively multiplayer online game Second Life
, posing the
question whether bots (i.e. non-playable characters) in computer games are conceived
by players as social actors. His approach relates closely to our own interests, but our
aim is to provide much broader picture. Some other empirical studies of specific
instances involving AI-based technologies have been conducted (such as the use of
smartphones in social interaction: e.g., [19]), however, they rather focus on the
“human side” of the interaction, and without the intention of providing a generalizable
sociological framework of the subject of AI. This is also the case of the field of
Human-AI Interaction, which we review in more detail in the next subsection.
2.2 Human-AI Interaction in Computer Science
Human-AI Interaction, as a field of study, is a subdomain of Human-Computer
Interaction (HCI), and focuses on the understanding of the nuances of our interactions
5
with AI supported tools, technologies, and processes. Although currently in a nascent
stage of development, this subdomain of computer science embodies an extensive
range of contexts, activities, and types of users. Furthermore, the encapsulation of
human-like behavior in artifacts and environments, and embodiment of intelligence in
varied kinds of technologies are being homogenized within the fabric of everyday life.
From domain experts (such as medical experts diagnosing cancer cells through
intelligent image processing [20]) to children (the use of AI in education to improve
the learning experience and outcome [21]) to building and urban dwellers (home
automation controlling thermal comfort of inhabitants [22], and autonomous cars
changing the shape of cities and experience of mobility [23]) to disabled users (for
example mobile assistive applications helping blind users navigate in urban
environments [24–25]), the role of AI in our socio-cultural aspects has become
increasingly pervasive. This is no longer limited to the embodiment of technologies
by artifacts, but also extends to the realm of built environments [26], having direct
spatial and consequently social impacts topics that the proliferating HCI
contributions in built environments have recently begun to address [27–28]. Still the
conception, design, and study of Human-AI Interaction is predominantly focused on
ad hoc instances
(such as robots, driverless cars, chatbots, etc.) with little or no
overlap between instances of different kinds. This lack of generalizability in the study
of Human-AI Interaction can be attributed to the HCI’s emphasis on design instances
and a bi-directional disconnect between these instances and theoretical frameworks.
In addition, the “black box” approach of representing AI algorithms by the
researchers oftentimes undermines the efforts to achieve a significant level of
generalizability. Consequently, the widely accepted conceptions of AI algorithms and
tools, especially their social impact, is currently distributed across the nature and form
of design instances (or products), and how experts and users likewise ascribe meaning
to these separate instances. This multi-layered gap in conceptions about AI and its
societal impact amongst actors of different backgrounds (AI developers, sociologists,
and users), and their varying levels of interactivity with smart technologies has
remained out of the scope of Human-AI Interaction as domain of computer science
research.
2.3 Research Gap
As demonstrated in the previous subsections, there is a research gap in contemporary
sociology as well as computer science which relates to (1) the conceptual
understanding of AI as a specific social (non-human) actor, and (2) the role sociology
could play not only in interpreting but also in helping to lead the future technological
development of AI-based tools, systems and devices. This research gap manifests
itself on several levels of social scientific endeavors: at the level of sociological
theory
, where non-human social actors are commonly “theorized out of existence”,
and also at the level of empirical studies
(similarly to the related domain of
Human-AI Interaction), where the broader societal impacts of AI are not considered,
given the primary focus on particular cases of specific technology use.
6
3 Research Plan and Methodology
3.1 Literature and Discourse Analysis
The first step of our research project will cover an investigation of past and ongoing
discourses within the other relevant research domains pertaining to AI, such as
computer science, cognitive psychology, and philosophy. In particular, we will aim to
identify aspects that are relevant for a specifically sociological formulation of
empirically investigatable research questions related to AI: its societal roles, functions
and imaginaries. In addition, content and discourse analysis of online discussion
forums and other media (TV, newspaper) will be an important initial step in outlining
and understanding the common-sense conceptualizations of AI in current society.
Qualitative and quantitative techniques of content analysis (cf. [29]) will allow us to
systematically gather initial knowledge of the existing imaginaries and
conceptualizations of AI in media. Furthermore, the discursive aspects of the analysed
texts will be studied by the methodologies of discourse analysis [30]. In addition to
literature review, these approaches will serve to further elaborate and specify the
research questions for empirical investigation in the next stages.
3.2 Online Survey
An online survey will aim to collect the widespread common-sense conceptions and
imaginaries of AI in contemporary Swiss society, and capture its expectable varieties.
Our aim is to have a representative sample, reflecting the demographic and social
diversity, and cover all Swiss languages. We will use TypeForm
platform for online
collection of survey data, and distribute the questionnaire among potential
respondents by a number of diverse venues. Descriptive and inferential statistics will
be used to gain quantitative insights and test hypotheses about different manifestation
of certain ideas and their correlation in the survey responses. In addition, to extract the
influence of different variable (culture, age, education level, etc.) on the perception of
AI, we will use exploratory data mining and statistical methods that allow for
clustering and pattern recognition. Visualizing the patterns and quantifying the
seminal components in the current perception of AI will be followed by qualitative
analysis to extract the meanings and nuances of what AI means in our current
societies.
3.3 Observational Studies
In the third phase, we will conduct three in-depth observational studies, collecting
video recordings of instances when a group of individuals interact with AI-based tools
and systems: (i) driverless shuttle; (ii) chatbots; (iii) game-play systems. In order to
extract features of situated common-sense conception of AI from the recordings, we
will analyse the data from the perspective of ethnomethodology and conversation
7
analysis [31–33], which focuses on the “perspective of the actor” and aims to describe
and elucidate the methodical work of practical sense-making in specific social
settings. It has been convincingly demonstrated by previous research in the field that
ethnomethodological analysis of video recordings of social interaction can yield
valuable insights into the details of situated action (e.g., [34]).
3.4 Interviews with Experts
We will conduct approximately 15 semi-structured interviews with experts on AI, as
well as experts in the relevant domains of sociology. First, mostly with the AI experts,
the goal of the interviews will be to discuss the results of empirical studies (see
subsections 3.2 and 3.3) and compare the common-sense conceptions/imaginaries of
AI with the experts’ perspective. In this case, the expert opinions are needed as a
contrastive foil in further elaboration of a truly sociological conception of AI, which
is the ultimate goal of our project. Second, mostly with the social scientists, we will
discuss the possibilities of sociology (or other social sciences) to influence and guide
the future of AI-related developments in our societies. In this case, the expert opinions
are needed especially because of the long-lasting controversies in sociology regarding
social advocacy and public engagement (cf. [33]). As a complementary method of
gathering experts’ opinions, we will also consider using the Delphi method.
3.5 Sociological Conception of AI
The goal of this final phase is to synthesize the main results of all previous phases in a
sociological conception of AI. The conception, we expect, will have the form of an
original coordinate system (matrix) for evaluation of societal conceptions and
imaginaries of AI. We will also conduct a general assessment of the current status of
AI in sociological thought and research. Our theoretical (concept-building) work will
be oriented by two main regards: (1) to sociological research, i.e., the
operationalizability of our conception in further empirical studies conducted from
different paradigmatic standpoints; (2) to other human and social sciences, i.e.,
providing the conceptual framework for sociologically sensitive research in
communication, political science, cultural anthropology, social psychology etc.
4 Conclusion
4.1 Subsequent Research Prospects
A number of empirical studies (qualitative and quantitative) can be outlined as a
direct result of a sociological conceptualization of AI. Our investigation respecifies
and opens up novel fields for collaboration between human/social and
computer/natural sciences. A sociological conceptualization of AI is necessary in
order to carry out further empirical research in this area, which would investigate AI
as a social phenomenon, its imaginaries in different segments of current societies, and
8
the role it has as a non-human social actor in particular social and institutional
settings. Presently, AI is already being applied in a great number of fields, such as
games, households, education, transportation, logistics, industrial production,
marketing and sales, communication, scientific research, data analysis, and many
others. Each of these fields requires sociological knowledge in order to understand AI
application, its impact on “users”, “customers”, “clients”, and their possible concerns
regarding interaction with AI. The ongoing fourth industrial revolution
– expansion of
cyber-physical systems such as AI and robots will presumably contribute to major
transformations when it comes to the ways we live, think and communicate. Proper
sociological understanding of AI provides us with a historically unique opportunity of
capturing the details of this revolution continuously and progressively as it happens.
The specific subsequent research prospects include survey-based studies of the
diversity of AI imaginaries in different segments of societies (defined economically,
culturally, politically, demographically etc.); detailed examination of communicative
and discursive processes related to AI (such as interaction with chatbots);
investigation of positive and negative impacts of AI-based automation in various
industrial spheres, and its influence on employment; historically oriented explorations
of the image of AI in popular culture; and indeed, further refinement of conceptual
and theoretical frameworks of AI based on empirical validations of sociological
models.
4.2 Innovation Potential
Our project will open the way for sociology and the social sciences into major AI
projects, where their influence is currently only marginal. In particular, sociology and
the social sciences, equipped with an adequate conception of AI, could (and should)
fully contribute to steer the development of AI-based technologies. This is important
especially since the current development of AI is predominantly grounded in the field
of technological possibilities (such as machine learning methods, neural network
models), rather than preliminary consideration of societal effects of the proliferation
and expansion of AI. On the other hand, indeed as with other technologies, it is
2
important to develop AI-based devices and tools in a way that builds on already
existing ways of practical usage of technology. For, as Harvey Sacks remarked
already in the 1960s, any novel technological object is “made at home in the world
that has whatever organization it already has” [33] it is incorporated in familiar
social practices. We do not need to stress that it is primarily sociology that sets out the
detailed study of the organization of the social world and related practical activities as
its principal and primordial field of interest. Similarly to other domains of technology,
sociology can provide crucial knowledge to AI designers; however, in order to do so,
it needs an appropriate understanding of the subject in question, in our case, artificial
intelligence.
2Our arguments to consider only the currently employed Machine Learning and Neural
Network algorithms rather than Agent-Based and BDI systems in the sociological conceptions
of AI are based on the predominant proliferation of the former methods in intelligent
applications which users often interact with.
9
To conclude, we firmly believe that precise sociological conceptualization of AI
could, in a long-term perspective, improve our comprehension of the nature of
humans and technology. Therefore, sociological conceptualization of AI, and
empirical studies in the sense outlined above, would have far-reaching impact not
only in the field of sociology, but also in human and social sciences in general.
References
1. Wittgenstein, L.: Philosophical Investigations. Basil Blackwell, Oxford (1953)
2. Greenfield, A.: Everyware: The Dawning Age of Ubiquitous Computing. New Riders
(2006)
3. Bauman, Z.: Úvahy o postmoderní době [Thoughts on the Postmodern Age]. Sociologické
nakladatelství SLON, Prague (1995)
4. Lindemann, G.: The analysis of the borders of the social world: A challenge for
sociological theory. J. Theor. Soc. Behav. 35, 69–98 (2005).
doi:10.1111/j.0021-8308.2005.00264.x
5. Dunlap, R.E., Catton, Jr., W.R.: Environmental Sociology. Annu. Rev. Sociol. 5, 243–273
(1979). doi:10.1146/annurev.so.05.080179.001331
6. Bryant, C.: The Zoological Connection: Animal Related Human Behavior. Soc. Forces 58,
399–421 (1979). doi:10.1093/sf/58.2.399
7. MacKenzie, D., Wajcman, J. (eds.): The Social Shaping of Technology. Open University
Press, Milton Keynes / Philadelphia (1985)
8. Nilsson, N. J.: The Quest for Artificial Intelligence: A History of Ideas and Achievements.
Cambridge University Press, Cambridge (2009).
9. Carley, K.M.: Artificial Intelligence within Sociology. Sociol. Method. Res. 25, 3–30
(1996). doi:10.1177/0049124196025001001
10. Brent, E.E.: Is there a role for Artificial Intelligence in sociological theorizing? Am.
Sociol. 19, 158–166 (1988). doi:10.1007/BF02691809
11. Bainbridge, W.S., Brent, E.E., Carley, K.M., Heise, D.R., Macy, M. W., Markovsky, B.,
Skvoretz, J.: Artificial Social Intelligence. Annu. Rev. Sociol. 20, 407–436 (1994).
doi:10.1146/annurev.so.20.080194.002203
12. Woolgar, S.: Why not a sociology of machines? The case of sociology and artificial
intelligence. Sociology 19, 557–572 (1985). doi:10.1177/0038038585019004005
13. Wolfe, A.: Mind, Self, Society, and Computer: Artificial Intelligence and the Sociology of
Mind. Am. J. Sociol. 96, 1073–1096 (1991). doi:10.1086/229649
14. Schwartz, R.D.: Artificial Intelligence as a Sociological Phenomenon. Can. J. Sociol. 14,
179–202 (1989). doi:10.2307/3341290
15. Malsch, T.: Naming the Unnamable: Socionics or the Sociological Turn of/to Distributed
Artificial Intelligence. Auton. Agent. Multi-Ag. 4, 155–186 (2001).
doi:10.1145/91474.91483
16. Latour, B.: Reassembling the Social: An Introduction to Actor-Network Theory. Oxford
University Press, Oxford (2005)
17. Callon, M., Latour, B.: Unscrewing the big Leviathan. In: Knorr Cetina, K.D., Mulay, M.
(eds.), Advances in Social Theory and Methodology, pp. 196–223. Routledge & Kegan
Paul, London (1981)
18. Muhle, F.: Embodied Conversational Agents as Social Actors? Sociological
Considerations on the Change of Human-Machine Relations in Online Environments. In:
10
Gehl, R.W., Bakardjieva, M. (eds.), Socialbots and their Friends: Digital Media and the
Automation of Society, pp. 86–109. Routledge, New York / London (2017)
19. Laurier, E., Brown, B., McGregor, M.: Mediated Pedestrian Mobility: Walking and the
Map App. Mobilities 11, 117–134 (2016). doi:10.1080/17450101.2015.1099900
20. Esteva, A., Kuprel, B., Novoa, R.A., Ko, J., Swetter, S.M., Blau, H.M., Thrun, S.:
Dermatologist-level classification of skin cancer with deep neural networks. Nature 542,
115 (2017). doi:10.1038/nature21056
21. Siemens, G., Long, P.: Penetrating the fog: Analytics in learning and education.
EDUCAUSE Rev. 46, 30 (2011). doi:10.17471/2499-4324/195
22. Brambilla, A., Alavi, H., Verma, H., Lalanne, D., Jusselme, T., Andersen, M.: “Our
inherent desire for control”: A case study of automation’s impact on the perception of
comfort. Enrgy Proced. 122, 925–930 (2017). doi:10.1016/j.egypro.2017.07.414
23. Alavi, H.S., Verma, H., Bahrami, F., Lalanne, D.: Is Driverless Car Another Weiserian
Mistake? In: Proceedings of the 2016 ACM Conference Companion Publication on
Designing Interactive Systems, pp. 249–253. ACM, New York (2017).
doi:10.1145/3064857.3079155
24. Ross, D.A., Blasch, B.B.: Wearable interfaces for orientation and wayfinding. In:
Proceedings of the fourth international ACM conference on Assistive technologies, pp.
193–200. ACM, New York (2000). doi:10.1145/354324.354380
25. Shen, H., Chan, K.Y., Coughlan, J., Brabyn, J.: A mobile phone system to find crosswalks
for visually impaired pedestrians. Technol. Disabil. 20, 217–224 (2008).
26. Alavi, H. S., Churchill, E., Kirk, D., Bier, H., Verma, H., Lalanne, D., Schnädelbach, H.:
From Artifacts to Architecture. In: Proceedings of the 19th International ACM
SIGACCESS Conference on Computers and Accessibility, pp. 387–390. ACM, New York
(2018). doi:10.1145/3197391.3197393
27. Alavi, H. S., Lalanne, D., Nembrini, J., Churchill, E., Kirk, D., Moncur, W.: Future of
human-building interaction. In: Proceedings of the 2016 CHI Conference Extended
Abstracts on Human Factors in Computing Systems, pp. 3408–3414. ACM, New York
(2016). doi:10.1145/2851581.2856502
28. Alavi, H. S., Churchill, E., Kirk, D., Nembrini, J., Lalanne, D.: Deconstructing
Human-Building Interaction. Interactions 23, 60–62 (2016). doi:10.1145/2991897
29. Krippendorff, K.: Content Analysis: An Introduction to its Methodology. SAGE,
Thousand Oaks / London / New Delhi (2004)
30. Fairclough, N.: Analysing Discourse: Textual analysis for social research. Routledge,
London / New York (2003)
31. Garfinkel, H.: Studies in Ethnomethodology. Prentice-Hall, Englewood Cliffs (1967)
32. Garfinkel, H.: Ethnomethodology’s Program: Working Out Durkheim’s Aphorism.
Rowman & Littlefield, Lanham (2002)
33. Sacks, H.: Lectures on Conversation I–II. Blackwell, Oxford (1992)
34. Knoblauch, H., Schnettler, B., Raav, J., Soeffner, H.-G. (eds.): Video Analysis:
Methodology and Methods. Lang, Bern (2006)
35. Hartmann, D.: Sociology and Its Publics: Reframing Engagement and Revitalizing the
Field. Sociol. Quart. 58, 3–18 (2016). doi:10.1080/00380253.2016.1248132
... With the rise of computational power and ubiquitous computing made available through smartphones and other personal devices, various forms of AI have made their ways into all layers of society and daily life (Mlynár et al., 2018). While the technological progress of the associated techniques and processes is fast and steep, individuals, communities, and societies struggle with understanding their nature and potential ethical pitfalls. ...
Article
Full-text available
The constant evolution of philosophical views on art is interwoven with trajectories of accelerating technological development. In the current vehement emergence of generative algorithms there is an immediate need for making sense of modern technologies that increasingly seem to step in the realm that has been reserved for humans-creativity. This paper aims to understand the role of the human in generative art by demystifying implications of black-box generative algorithms and their applications for artistic purposes. First, we present examples of current practice and research in generative art with a special interest in music that served as foundation for our work. Then, we introduce Anastatica (2020), a part performance, part installation built on the basis of data-driven generative live coding. Finally, we discuss the various implications of AI in art through a case study rooted in Anastatica's development and performance. Here we trace the path from algorithms to intelligence, applying both musical and computer science theory to a practical case of generating a live coding musical performance, with special focus given to aesthetic, compositional, conceptual, and phenomenological implications.
... The fast advancement in Artificial Intelligence (AI) and Big Data technology is reshaping global communications, commercial activities, and social relations in industrialized economies [14], [19]. The superhuman capability of data processing and pattern recognition enables AI to outperform human workers in many data-and/or computing-intensive tasks [8], [27]. ...
Conference Paper
Full-text available
Advancements in Artificial Intelligence (AI) and Big Data technologies are reshaping the global economy. The community currency network is receiving more attention through enhancing social ties within a community. As the most popular type of community currency, Time Banking (TB) is a generalized community exchange economy, which uses the time to evaluate each participant's contributions on the same scale rather than any equivalence with the official national currency. TB is a noble idea with the potential to improve the quality of life through prosocial, reciprocally beneficial activities among community members. However, it also brings new concerns about security and trust issues. Inspired by blockchain and smart contract, this paper introduces a Blockchain Integrated Timebanking (BIT) system to secure a decentralized community exchange economy. In BIT system, service providers and recipients can securely exchange effort through a self-executing smart contract without relying on a third-party trust authority. The blockchain network ensures immutability, auditability, and traceability of all data and service transactions recorded on the distributed ledger. A proof-of-concept prototype was implemented and tested on a private Ethereum network. The experimental results verify the feasibility of the proposed BIT to provide decentralized community service exchanges with limited computation overhead and network latency.
... Aligning interaction designs focusing on establishing virtual telecopresence to enhance the role of virtual counselors as communication agents can potentially help in achieving cognitive plausibility in such interactions. This can be achieved in several ways: first, humans perceive interactive digital devices as social actors 6,35,36 ; as such, trust and believability in the virtual counselor can help the patient engage in counseling sessions and adhere to their care plan. Second, creating embodiment, a degree of engagement that is created by the presence of humans in a conversation, can enhance the quality and nature of communicative interactions 13,14 . ...
Article
Full-text available
Voice-based personal assistants using artificial intelligence (AI) have been widely adopted and used in home-based settings. Their success has created considerable interest for its use in healthcare applications; one area of prolific growth in AI is that of voice-based virtual counselors for mental health and well-being. However, in spite of its promise, building realistic virtual counselors to achieve higher-order maturity levels beyond task-based interactions presents considerable conceptual and pragmatic challenges. We describe one such conceptual challenge—cognitive plausibility, defined as the ability of virtual counselors to emulate the human cognitive system by simulating how a skill or function is accomplished. An important cognitive plausibility consideration for voice-based agents is its ability to engage in meaningful and seamless interactive communication. Drawing on a broad interdisciplinary research literature and based on our experiences with developing two voice-based (voice-only) prototypes that are in the early phases of testing, we articulate two conceptual considerations for their design and use—conceptualizing voice-based virtual counselors as communicative agents and establishing virtual co-presence. We discuss why these conceptual considerations are important and how it can lead to the development of voice-based counselors for real-world use.
... communicate. Thus, question such as "what exactly is AI as a social phenomenon?" will have to 505 be answered (Mlynář, Alavi, Verma, & Cantoni, 2018). Furthermore, technological entities such 506 as robots and AI are irreversibly continuing to develop and it is in the common best interest that 507 their functions and aims remain aligned with those of humans. ...
Article
Full-text available
Artificial intelligence and robots may progressively take a more and more prominent place in our daily environment. Interestingly, in the study of how humans perceive these artificial entities, science has mainly taken an anthropocentric perspective (i.e., how distant from humans are these agents). Considering people’s fears and expectations from robots and artificial intelligence, they tend to be simultaneously afraid and allured to them, much as they would be to the conceptualisations related to the divine entities (e.g., gods). In two experiments, we investigated the proximity of representation between artificial entities (i.e., artificial intelligence and robots), divine entities and natural entities (i.e., humans and other animals) at both an explicit (Study 1) and an implicit level (Study 2). In the first study, participants evaluated these entities explicitly on positive and negative attitudes. Hierarchical clustering analysis showed that participants’ representation of artificial intelligence, robots and divine entities were similar, while the representation of humans tended to be associated with that of animals. In the second study, participants carried out a word/non-word decision task including religious semantic-related words and neutral words after the presentation of a masked prime referring to divine entities, artificial entities and natural entities (or a control prime). Results showed that after divine and artificial entity primes, participants were faster to identify religious words as words compared to neutral words arguing for a semantic activation. We conclude that people make sense of the new entities by relying on already familiar entities and in the case of artificial intelligence and robots, people appear to draw parallels to divine entities.
... The explosion in Artificial Intelligence (AI), Machine Learning (ML), and Big Data is reshaping global communications, commercial activities, and social relations in industrialized economies [8], [16]. The superhuman capability of data processing and pattern recognition enables AI to outperform human workers in many data-and/or computingintensive tasks [3], [21]. ...
Conference Paper
Full-text available
Blockchain technology is reshaping the the traditional economies. People may have more trust than ever before as the transaction is immutable and transparent. Success in crypto-currency and other technical areas highlights many attractive features of the blockchain technology that can benefit more aspects of modern society. Time Banking is a generalized exchange economy not based on money, but values everyone's contribution on the same scale, the time expended. Time banking is a noble idea with great potential, but the security and trust issues are not well addressed. In this paper a BLockchain-ENabled Decentralized Time Banking System (BlendTBS) is proposed to build a trustful, dynamic and respectful community. People in this community are encouraged to be engaged in mutual serving relationships. For this purpose, the BlendTBS is designed to reward the residents who commit in socially beneficial activities. An initial prototype is implemented on a permissioned blockchain network and a small scale study is planned to examine the utility of BlendTBS to a traditional community on the island of Aneityum, Republic of Vanuatu. Within a selected community in the village of Analgahuat, deeper insights will be explored by observing the trust enabled by Blockchain technology that allows peer to peer service exchanges between any two individuals. Authors hope this position paper may inspire more interests in the roles that blockchain technology can play in modern society.
Article
Full-text available
The accession and implementation of new generation free trade agreements bring numerous opportunities as well as challenges to Viet Nam, regarding trade, labor and investment. The increasing number of workers abroad puts a pressure on Vietnamese government to support them in new working cultures and environments. The application of chatbot, which has been known to help certain vulnerable groups such as patients, women and migrants could be one of the tools to support Vietnamese migrant workers by providing immediate information, network connection and consultation. Analyzing the results of the qualitative interviews with 11 migrant workers and the group discussion with two sections from three public social media groups and three active informants, the article argues that migrant workers as a vulnerable group who are supposed to be promising users of chatbot due to their great demand and needs, may have a complicated form of trust and expectation in chatbot.
Thesis
Full-text available
Humans interact with computers socially, with a tendency to anthropomorphize machines even when they operate with only the smallest of human characteristics. As artificial intelligence (AI) becomes mainstream with Digital Voice Assistants (DVAs), these agents increasingly carry human social characteristics such as name, voice, and even personality, and more often the default persona is feminine. Female persona AI robots, including sex-robots, are emerging as commercial products available to a mass market. Concerns over unregulated and unethical AI include bias in and abuse of AI abound, and the extremely limited diversity among those who develop AI technology suggest equitable AI is not in near sight. Important sociological questions emerge, pertaining to both human social relationships and societal functions such as education, economy, and politics. How will interaction with today's gendered AI agents alter human social interactions and social identity? Are science and technology themselves gender biased? As society moves towards an inevitable future of living with intelligent machines, society must adapt to ensure an ethical world for both human and machine. This paper will argue that civil society must play a larger role in influencing the ethics of AI applications broadly, with specific effort toward gender equality, and that this should not be left solely in the domain of commercial entities who predominantly design and own them.
Conference Paper
Full-text available
The vision and mission of research under the banner of Ubiquitous Computing has increasingly moved from focusing on the realm of "artifacts" to the realm of "environments". We seek to scrutinize this very transition, and raise questions that relate to the specific attributes of built environments that set them inherently apart from artifacts. How does an interactive environment differ from an interactive artifact, a collection of artifacts, or an integrated suite of artifacts? Consequently, we ask what are the new user experience dimensions that HCI researchers should merge into their considerations, for example, by supplementing us-ability and engagement with occupants' comfort across multiple dimensions, and shifting attention from (often) short lifespan and discretionary to durable and immersive experiences? In this contribution, we bring arguments from the literature of environmental psychology and architecture that highlight the points of divergence between artifacts and architecture, and then translate them into challenges for Human-Computer Interaction, and particularly for the emerging domain of Human-Building Interaction.
Conference Paper
Full-text available
To investigate the relationship between occupants’ perception of control over building elements and their comfort, we conducted a study where two prototype office rooms were compared: while the first room allowed occupants to open or close the window and configure the shading, the second one was fully automated. The quantitative analysis of collected data a) supports the existing results in the literature reporting higher satisfaction where manual control is maintained, and b) uncovers a new impact of highly automated systems: lower control over building elements can increase the occupant's consciousness of the environmental factors and the saliency of comfort parameters.
Conference Paper
Full-text available
We present a structured discussion of the concept of driverless car as a major Ubicomp project, and particularly of its hypothetical integration into the fabric of city. The analytical framework is borrowed from the Transportation and Urban research domains, which provides us with a list of agreed-upon subject matters when accounting for car mobility in urban design. We pose concrete questions about each of these subjects as how self-driven personal vehicles could have a positive or negative impact on them. Out of the six discussed topics, this initial examination showed that driverless car could have a negative impact on five, suggesting smoke detected, and thus the need for broadening such inspections.
Conference Paper
Full-text available
In 2030, we will have a different interactive experience with our built environments, at home, at work, and even in public urban spaces. This is attributed to advancements in sensing and actuation systems that can integrate into the building infrastructures, in symbiosis with the new environmental concerns that call for new life, work, and mobility styles. This change, whether gradual or sudden, evident or seamless, can have a remarkable impact on our everyday experiences, and thus entails efforts to envision possible scenarios and plan for them. We believe that buildings, as they would embody our digital and physical interactive daily experiences, should be designed and nurtured in a dialogue with their users at the individual as well as social levels. This implies a responsibility of the HCI community to intervene and involve the user in the Human-Building Interaction (HBI) design practice. We propose bringing together expertise from the fields of human-computer interaction, building and urban architecture, and social sciences, and provide them with an occasion for collaboratively creating and sharing ``images' of HBI by 2030. The goal is to uncover research opportunities and challenges that will emerge through discussions and multi-faceted debates about the topics proposed.
Article
This article outlines how sociologists can and do engage a range of general, nonacademic public audiences, organizations, and interests. It begins with a discussion and critique of the notion of public sociology that emerged from Michael Burawoy’s 2004 American Sociological Association (ASA) presidency and draws upon recent commentaries and classical theories to formulate a more pluralist vision of sociology’s publics. The article concludes by arguing that embracing a multidimensional vision of the relationships between sociology and its publics not only provides a foundation for better public engagement, but also can help renew and revitalize of sociology.
Article
Skin cancer, the most common human malignancy, is primarily diagnosed visually, beginning with an initial clinical screening and followed potentially by dermoscopic analysis, a biopsy and histopathological examination. Automated classification of skin lesions using images is a challenging task owing to the fine-grained variability in the appearance of skin lesions. Deep convolutional neural networks (CNNs) show potential for general and highly variable tasks across many fine-grained object categories. Here we demonstrate classification of skin lesions using a single CNN, trained end-to-end from images directly, using only pixels and disease labels as inputs. We train a CNN using a dataset of 129,450 clinical images-two orders of magnitude larger than previous datasets-consisting of 2,032 different diseases. We test its performance against 21 board-certified dermatologists on biopsy-proven clinical images with two critical binary classification use cases: keratinocyte carcinomas versus benign seborrheic keratoses; and malignant melanomas versus benign nevi. The first case represents the identification of the most common cancers, the second represents the identification of the deadliest skin cancer. The CNN achieves performance on par with all tested experts across both tasks, demonstrating an artificial intelligence capable of classifying skin cancer with a level of competence comparable to dermatologists. Outfitted with deep neural networks, mobile devices can potentially extend the reach of dermatologists outside of the clinic. It is projected that 6.3 billion smartphone subscriptions will exist by the year 2021 (ref. 13) and can therefore potentially provide low-cost universal access to vital diagnostic care.
Article
Interaction design is increasingly about embedding interactive technologies in our built environment; architecture is increasingly about the use of interactive technologies to reimagine and dynamically repurpose our built environment. This forum focuses on this intersection of interaction and architecture.
Article
While walking has always been mediated, the arrival of smartphones with multiple apps has changed how we walk and how we use apps. In this study, we investigate the relationships of pedestrian-in-the-street and app-user-on-screen actions. We display and describe a series of intersubjective practices constituted by, and with, walking while using a mobile device. The video data used are from a larger study of pedestrians using smartphones in urban settings, with our analysis here turning on how a smartphone is used and interacted around to accomplish walking together. Our approach draws upon ethnomethodological conversation analytic studies of the sequential and category-based organisation of mobile and on-screen actions. The analysis shows how walking actions (such as unilateral-stopping, turning, restarting) are connected to map actions such as displaying the map, manipulating the scale and monitoring the movement of the you-are-here dot. We conclude with remarks on the collaborative inter-subjective nature of walking with apps.
Book
Artificial intelligence (AI) is a field within computer science that is attempting to build enhanced intelligence into computer systems. This book traces the history of the subject, from the early dreams of eighteenth-century (and earlier) pioneers to the more successful work of today’s AI engineers. AI is becoming more and more a part of everyone’s life. The technology is already embedded in face-recognizing cameras, speech-recognition software, Internet search engines, and health-care robots, among other applications. The book’s many diagrams and easy-to-understand descriptions of AI programs will help the casual reader gain an understanding of how these and other AI systems actually work. Its thorough (but unobtrusive) end-of-chapter notes containing citations to important source materials will be of great use to AI scholars and researchers. This book promises to be the definitive history of a field that has captivated the imaginations of scientists, philosophers, and writers for centuries.
Article
This paper examines the way some current artificial intelligence (AI) programs absorb and redefine social practices when deployed in setting where knowledge and intelligence are decisive. Two areas of AI research and development are discussed: expert systems and frame-based natural language processing. Successful performance of AI programs in these areas comes to depend on conditions for knowledge in contemporary society that are the subject of sociological analysis. The design-elements of AI technology are also seen to reflect features of contemporary social organization. While the computerization of society is frequently identified with an interest in social control, AI is shown to have affinities with postmodernist accounts of society stressing social fragmentation and cultural discontinuity. /// Le présent document traite de la façon dont certains programmes courants d'intelligence artificielle (IA) assimilent et redéfinissent les habitudes sociales lorqu'ils sont déployés dans des milieux où le savoir et l'intélligence ont une grande importance. Deux domaines de recherche et développement sont présentés: les systèmes experts et le traitement des langues naturelles sur unité centrale. L'exécution efficace des programmes IA dans ces secteurs repose sur les conditions qui déterminent l'intelligence dans la société contemporaine, qui font l'objet d'analyses sociologiques. Les rubriques utilisées pour la conception de la technologie IA semblent de même reflèter certaines caractéristiques de l'organization sociale contemporaine. Bien que l'informatisation de la société soit identifée comme intervenant pour le contrôle social, nous identifions les affinités de l'IA avec les descriptions post-modernes de la société, lesquelles mettent en lumière la fragmentation sociale et la discontinuité culturelle.