Content uploaded by Ilkka Tuomi
Author content
All content in this area was uploaded by Ilkka Tuomi on Jun 13, 2024
Content may be subject to copyright.
Eur J Educ. 2022;00:1–6. wileyonlinelibrary.com/journal/ejed
|
1
© 2022 John Wiley & Sons Ltd.
DOI: 10.1111/eje d.1253 0
EDITORIAL
Charting the futures of artificial intelligence in
education
1 | INTRODUCTION
In less than half a decade, Artificial Intelligence (AI) has become a major topic in the daily news flow and in policy
debates. Technological advance has been fast, and AI has been claimed to transform occupations and work tasks.
It has been argued that everyone should have at least a basic understanding of what AI is in order to make sense of
what is happening and to be able to survive in this new environment, and— to remain relevant— we have to realise
that “our education systems will need a radical change in their purpose, form and content ” (Šucha & Gammel, 2021, p.
38).
Sense- making, indeed, is necessary and useful. One starting point for the thematic first part of this journal
issue was an internal European Commission Joint Research Centre (EC- JRC) pilot project that in 2019 sketched a
process to create an AI Handbook with and for Teachers. A key assumption in that project was that educators, ad-
ministrators, and education policymakers will soon face strong pressures to adopt AI- based systems in education
practices. To be able to assess claims about Artificial Intelligence in education (AIED), avoid overloading teachers
with unnecessar y initiatives, and to find concrete oppor tunities for AIED, we thought it would be useful to engage
teachers in a joint sense- making process where the potential and challenges of AIED would be contextualised in
actual pedagogic settings. In somewhat impolite terms, at the time there was so much hyperbole and nonsense
about AI that we were afraid that both the potential advantages of using AI in education and the actual purposes
of education were being lost in much of the talk about AIED.
2 | IN THIS ISSUE
In this issue, we try to move beyond the hyperbole and make some sense of this potentially very important tech-
nology that, indeed, is already beginning to have a major impact on education. There have been massive invest-
ments in AI technology around the world, as well as high- profile policy statements about the need to promote and
regulate this emerging technology (see the articles in this issue by Blikstein et al., Holmes and Tuomi, and Selwyn).
All the Big Tech giants (Alphabet, Amazon, Apple, and Google) are heavily involved in AI- infused educational
technology. Meanwhile, there are now more than thirty multi- million- dollar funded AIED firms, and the market
is expected to become worth more than US$ 20 billion within five years (GMI, 2022). Commercial interests are
frequently translated into claims about the transformative power of Artificial Intelligence systems in education
(e.g., OECD, 2021). Yet, the technical complexity of AI and AIED systems can make it difficult for practitioners and
policymakers to question such claims or assess their relevance. An important motivation for making this issue was
to help educators to do this. In particular, the comprehensive introductory article by Holmes and Tuomi provides
a detailed overview of AIED- related concepts and the current state- of- the- art.
Beyond highly technical academic research, almost all claims about Artificial Intelligence and AIED are claims
about the future. Statements and opinions about how AI will change education and learning are now omnipresent,
2
|
EDITORIAL
but efforts to systematically study the futures of AI in education and learning, building on known futures studies
methods have been scarce. Expert discussions have generated some narrative fragments and examples about
possible AI- enabled futures (Pelletier, 2021; Roschelle et al., 2020; Vuorikari et al., 2020); more often, however,
AIED researchers have painted dualistic images of dystopian and utopian futures of education (e.g., Aiken &
Epstein, 2000; Pinkwart, 2016; Schiff, 2021), or focused on highlighting technological trends and the potential
of future AIED in transforming education or in solving existing problems in education (e.g., Baker, 2021; Woolf
et al., 2013). Ethical challenges and the impact of commercialisation have received increased attention in re-
cent years in critical AIED studies (e.g., Blikstein & Blikstein, 2021; Buckingham Shum & Luckin, 2019; Holmes &
Porayska- Pomsta, 2023; Nemorin et al., 2022; Perrotta et al., 2021; Selwyn, 2019; Williamson & Eynon, 2020),
but, beyond computer science teacher communities (e.g., AI4K12, 2022) and a few experiments (e.g., Pihlajamaa
& Rantapero- Laine, 2020), educators or policymakers have rarely been actively involved in AIED system develop-
ment, research, or technology articulation.
The thematic first part of this issue opens with a perspective paper by Riel Miller and Ilkka Tuomi, Making
the futures of AI in education: Why and how imagining the futures matters. Miller and Tuomi reflect on the future
of AIED from the point of view of futures studies. As noted also by Selwyn in the fifth article in this issue, there
are two very different kinds of AIED. One is the actual AI in use. The other— much more common— is the AIED of
imagined futures. Paradoxically, both rest on the foundation of anticipatory models that we use to make sense
of the present and the future. In general, anticipatory models— as explained in the perspective paper— shape
the futures we can imagine. These anticipations, therefore, largely determine what AI is and can be in those
imagined futures. To understand AIED, therefore, requires that we take a closer look at different articulations of
futures. This brings us to the field of futures studies. Building on earlier work on futures literacy and theory of
anticipation (Fuller, 2017; Miller, 2007, 2018; Poli, 2017; Tuomi, 2019), Miller and Tuomi reflect on how different
futures of AIED emerge as potential futures are used in different ways. AI, as a meaningful technology with real
social, economic, and cultural consequences, can only be understood as a product of our imagined futures. Our
capabilities to imagine futures and understand different ways of using futures, therefore, determine in important
ways what we can make of AI.
The first article in the thematic first part of this issue on The state of the art and practice in AI in education is by
Wayne Holmes and Ilkka Tuomi. Holmes and Tuomi identify some of the key concepts for Artificial Intelligence
in education and provide an overview of the AIED field. The article outlines some of the historical background
and introduces different ways in which AIED is currently being used, noting that different practices need to be
considered both collectively and separately. To organise the different types of AIED, an updated typology of
AIED applications is elaborated (see also Holmes et al., 2019). An earlier version of the typology has previously
informed a number of reports (e.g., Holmes et al., 2022; Miao & Holmes, 2021; Vuorikari et al., 2020). The article
also introduces some key roadblocks, these are obstacles that need to be addressed to integrate AIED in education
practices, ranging from lacking engagement with the ethics of AIED to evidence of impact, commercialisation of
education, and the use of AI as a tool for colonising.
The second article, Ceci n'est pas une école: Discourses of artificial intelligence in education through the lens of se-
miotic analytics, by Paulo Blikstein, Yipu Zheng and Karen Zhuqian Zhou, is an empirical study that focuses on the
commercial discourses of Artificial Intelligence in education. Using text mining to analyse AIED vendor websites,
the article shows how the words used by AIED vendors shape the understanding not only of AI in education but
of education itself. Through rhetorical moves that juxtapose ‘old- fashioned’ teachers and ‘advanced technology’
that works at the speed of light and is able to ‘bring students into the 21st century’, a comparison is constructed
in which technology is portrayed as unequivocally superior. In a highly thought- provoking way, the article points
out that the realisation of AIED comes with a new discourse, and that the narrative part of AIED has important
consequences for education practices and policy.
The third article, Still w(AI)ting for the automation of teaching: An exploration of machine learning in
Swedish primary education using Actor- Network Theory, is by Katarina Sperling, Linnéa Stenliden, Jörgen
|
3
EDITORIAL
Nissen and Fredrik Heintz. The article documents an attempt to use a data- driven AIED system to support
mathematics teaching in primary classrooms in Sweden. Following the tradition of science and technology
studies, it provides an ethnography- informed description of the successes and failures in appropriating
new technological functionality in actual teaching practices. Although the realisation of the planned em-
pirical study was in part limited by the Covid pandemic, the article provides a useful insider view of what
using AIED can be like in practice. To describe the complex dynamics of AIED adoption, the article frames
the process using actor- network theory, and shows the importance of thick descriptions of AIED in real
educational settings.
The fourth article, Artificial intelligence, 21st century competences, and socio- emotional learning in education:
More than high- risk?, by Ilkka Tuomi, focuses on the possible consequences of a collision of three highly influential
education policy discourses: namely, (1) the use of AI in education, (2) suppor ting the development of 21st century
competences, and (3) measuring social and emotional learning. Reviewing existing research on ‘soft skills’ and their
relation to personality traits, abilities, and interest structures, the article shows that the use of AI to analyse and
support the development of 21st century competences and non- epistemic learning may have important social
consequences that require careful consideration. Machine learning models developed using data on 21st century
competences may have fundamental implications for the ways in which societies organise themselves. The article
therefore suggests a moratorium on the use of data on non- epistemic competence components until researchers
and policymakers better understand these implications.
The fifth article, The future of AI and education: some cautionary notes, by Neil Selwyn, reflects on how to avoid
some major caveats in future discussions about AIED. Aligned with the empirical study by Blikstein et al. in this
issue, it highlights the importance of avoiding exaggerated claims about AIED and suggests a more critical stance
on the potential and challenges of AIED. For example, it asks for a clear distinction between ‘actually existing AI’
and ‘speculative AI.’ It also calls for a more refined understanding about what aspects of education can actually
be quantified and represented as data. For current policy debates about digital education, this is a fundamental,
theoretically and practically important question. Addressing it would require rethinking some common beliefs
about the nature of data and digital computation (Rosen, 1978, 1987; Tuomi, 2000). The article argues that, in-
stead of optimistic hyperbole, a balanced view that acknowledges the potential negative impacts, including social
and environmental harms, and the ideological and political nature of AIED is necessary. This is what Facer and
Selwyn (2021) have called non- stupid optimism.
The sixth and final article of the thematic part of this issue, Towards hybrid human- AI learning technologies, by
Inge Molenaar, builds on her earlier research (2021), and suggests elements of a new conceptual language for
thinking about AIED. Suggesting a detect- diagnose- act framework for understanding student- facing AIED and the
six levels of automation model as a way to classify AIED systems, the article shows how different AIED systems re-
quire different ways of dividing control and tasks among teachers and technology. A starting point for the ar ticle is
that AIED needs to be understood as a tool that facilit ates learning and augments teacher and student capabilities.
This is in contrast to the automation perspective that has commonly been associated with AI. The augmentation
perspective on AI leads to a view where human cognition and AI form a hybrid, which Molenaar explores in her
article using self- regulated learning as an example. Molenaar develops the point that shared language is necessary
for the effective involvement of key stakeholders in AIED development and adoption. The article contributes by
articulating concepts and language for this purpose.
AIED is often viewed as a technological speciality that can be competently addressed only by computer sci-
entists specialised in machine learning and AI. However, as Holmes and Tuomi explain, we need to acknowledge
that AIED comprises a complex variety of technologies that therefore require multiple understandings. Further,
as the articles by Blikstein et al., Sperling et al., and Selwyn in this issue illustrate, narrative elements are key parts
of technology, and enable us to make sense of it. Part of the process of technology development, therefore, is the
creation of new languages— an effort Molenaar engages with in her article. The Miller and Tuomi article, in turn,
turns towards the foundations of the meaningful reality that organises our societies and policy debates, showing
4
|
EDITORIAL
that anticipatory models make expectations, explanations, justifications— the key elements of stories we tell— and
action possible. Futures and AIED, therefore, are deeply linked conceptually, and an improved capability to use
futures also makes possible new forms of AI, AIED, and education.
3 | PART II ARTICLES
Part II opens with an article on Third space workers in higher education in times of dislocated complexity, by
Kay Livingston and Lorraine Ling, that examines the changing nature of the higher education workforce.
Specifically, with reference to the increasing influence and importance of third space workers such as e-
learning developers, partnerships managers and learning technology specialists. Livingston and Ling deploy
Giddens' Theory of Structuration to analyse two cases of how the higher education workforce is changing,
one drawing from a study in Scotland and the other from Australia. Third space workers are forging new iden-
tities, crossing traditional boundaries and facilitating change internally within the university and externally
through partnerships. The conclusion identifies complex features of ongoing changes in higher education and
highlights the need for structural and policy changes; in particular, a need to recognise and legitimise the role
of third space workers.
The second article on The transition from higher education to first employment in Spain, is by Encarnación
Cordón- Lagares, Félix García- Ordaz and Juan José García del Hoyo. Cordón- Lagares and colleagues report on
findings from secondary analysis of survey data on the time it takes higher education graduates to obtain their
first job in Spain. The statistical analysis draws on parametric and nonparametric analysis of duration models to
estimate the exit rate to employment of university graduates. The results show that graduates with prior work ex-
perience, graduates from private universities and men have a comparative advantage in the transition to employ-
ment. Additional factors discussed include the subject field studied, Information and Communication Technology
(ICT) skills, international experience and the timing of job searches.
The final article, Partnership of schools and civil society organisations to support education of students of varied
linguistic backgrounds— The situation in the Czech Republic, Italy and Spain, is by Janet Wolf, Raquel Casado- Muñoz
and Francesca Pedone. The article reports on a study in which the authors surveyed 34 non- profit organisations
in Czech Republic, Spain, and Italy and interviewed fifteen teachers. Specifically, the study examined how non-
profit organisations and teachers viewed collaboration between schools and non- profits as a potential resource
for teachers for supporting students with a different mother tongue (L2 students). The authors highlight the
potential of non- profit organisations as partners in education for covering themes from intercultural education.
Obstacles identified by non- profit organisations included lacking communication, funding and coordination by
public administration authorities.
Ilkka Tuomi1
Wayne Holmes2
Riel Miller3
1Meaning Processing Ltd., Helsinki, Finland
2Knowledge Lab, Institute of Education (IOE), Faculty of Education and Sociology, University College London
(UCL), London, UK
3J. Herbert Smith Centre, University of New Brunswick, Canada
Correspondence
Ilkka Tuomi, Meaning Processing Ltd., Arkadiankatu 20 A 20, 00100 Helsinki, Finland.
Email: ilkka.tuomi@meaningprocessing.com
|
5
EDITORIAL
ORCID
Ilkka Tuomi https://orcid.org/0000-0002-4179-7103
Wayne Holmes https://orcid.org/0000-0002-8352-1594
Riel Miller https://orcid.org/0000-0001-6329-5983
REFERENCES
AI4K12. (2022). AI4K12.org. https://ai4k12.org
Aiken, R. M., & Epstein, R. G. (2000). Ethical guidelines for AI in education: Starting a conversation. International Journal
of Artificial Intelligence in Education, 11(2) , 1 6 3– 176.
Baker, R. S. (2021). Artificial Intelligence in education: Bringing it all together. In S. Vincent- Lancrin (Ed.), Digital ed-
ucation outlook: Pushing the frontiers with AI, blockchain, and robots (pp. 43– 54). OECD. https://doi.org/10.1787/
f54ea 644- en
Blikstein, P., & Blikstein, I. (2021). Do educational technologies have politics? A semiotic analysis of the discourse of ed-
ucational technologies and artificial intelligence in education. In Algorithmic rights and protections for children. https://
doi.org/10.1162/ba67f 642.646d0673
Buckingham Shum, S., & Luckin, R. (2019). Learning analytics and AI: Politics, pedagogy and practices. British Journal of
Educational Technology, 50(6), 2785– 2793. htt ps://doi.org/10.1111/bjet.1288 0
Facer, K., & Selwyn, N. (2021). Digital technology and the futures of education— Towards ‘non- stupid’ optimism. ED- 2020/
FoE- BP/27. UNESCO. https://unesd oc.unesco.org/ark:/48223/ pf000 0377071
Fuller, T. (2017). Anxious relationships: The unmarked futures for post- normal scenarios in anticipatory sys-
tems. Technological Forecasting and Social Change, 124 (November), 41– 50. https://doi.org /10.1016/j.te chf
ore.2016.07.045
GMI. (2022). AI in education market size & share, growth forecast 2022- 2030. Report. Global Market Insights Inc. h t t p s ://
www.gmins ights.com/indus try- analy sis/artif icial - intel ligen ce- ai- in- educa tion- market
Holmes , W., Bialik , M., & Fadel, C. (2019). Artificial intelligence in education: Promises and implications for teaching & learning.
The Center for Curriculum Redesign.
Holmes, W., Persson, J., Chounta, I.- A., Wasson, B., & Dimitrova, V. (2022). Artificial intelligence and education. A critical
view through the lens of human rights, democracy, and the rule of law. Council of Europe.
Holmes, W., & Porayska- Pomsta, K. (Eds.). (2023). Ethics of artificial intelligence in education. Taylor & Francis.
Miao, F., & Holmes, W. (2021). AI and education: Guidance for policy- makers. UNESCO https://unesd oc.unesco.org/
ark:/48223/ pf000 0376709
Miller, R. (2007). Futures literacy: A hybrid strategic scenario method. Future s, 39(4), 341– 362. https://doi.org/10.1016/j.
futur es.2006.12.001
Miller, R. (Ed.). (2018). Transforming the future: Anticipation in the 21st centur y. Routledge.
Molenaar, I. (2021). Personalisation of learning: Towards hybrid AI- human learning technologies. In S. Vincent- L ancrin
(Ed.), Digital education outlook: Pushing the frontiers with AI, blockchain, and robots (pp. 57– 77). OECD. h t t p s : //d o i .
org /10.1787/589b2 83f- en
Nemorin, S., Vlachidis, A., Ayerakwa, H. M., & Andriotis, P. (2022). AI hyped? A horizon scan of discourse on artifi-
cial intelligence in education (AIED) and development. Learning, Media and Technology (ahead of print). ht t p s : //d o i .
org/10.1080/17439 884.2022.2095568
OECD. (2021). OECD digital education outlook 2021: Pushing the frontiers with ar tificial intelligence, blockchain and robots.
OECD. https://doi.org/10.1787/589b2 83f- en
Pelletier, K. (2021). 2021 EDUCAUSE Horizon Report: Teaching and Learning Edition. EDUCAUSE. https://libra ry.educa use.
edu/- /media/ files/ libra ry/2021/4/2021h rteac hingl earni ng.pdf
Perrotta, C., Gulson, K. N., Williamson, B., & Witzenberger, K. (2021). Automation, APIs and the distributed labour of
platform pedagogies in Google Classroom. Critical Studies in Education, 62(1), 97– 113. https://doi.org/10.10 80/17508
487.2020.1855597
Pihlajamaa, J., & Rantapero- Laine, A. (2020). School as an innovation platform— A unique model for co- creation. The Finnish
Smart Learning Environments for the Future projec t. European EdTech Network. https://eetn.eu/case- study/ detai l/
Schoo l- as- an- innov ation - platf orm- - - a- uniqu e- model - for- co- creat ion.- The- Finni sh- Smart - Learn ing- Envir onmen ts-
for- the- Futur e- project
Pinkwart, N. (2016). Another 25 years of AIED? Challenges and opportunities for intelligent educational technologies of
the future. International Journal of Artificial Intelligence in Education, 26(2), 771– 783. https://doi.org /10.1007/s4059
3- 016- 0099- 7
Poli, R. (2017). Introduction to anticipation studies. Springer International Publishing.
6
|
EDITORIAL
Roschelle, J., Lester, J., & Fusco, J. (2020). AI and the future of learning: exper t panel report. CIRCLS, Center for Integrative
Research in Computing and Learning Sciences. https://circls.org/wp- conte nt/uploa ds/2020/11/CIRCL S- AI- Repor t-
Nov20 20.pdf
Rosen, R. (1978). Fundamentals of Measurement and Representation of Natural Systems. North- Holland.
Rosen, R. (1987). On the scope of syntactics in mathematics and science: The machine metaphor. In J. L. Casti & A.
Karlqvist (Eds.), Real brains, artificial minds (pp. 1– 23). North- Holland.
Schiff, D. (2021). Out of the laboratory and into the classroom: The future of artificial intelligence in education. AI &
SOCIETY, 36(1), 331– 348. https://doi.org/10.1007/s0014 6- 020- 01033 - 8
Selwyn, N. (2019). Should robots replace teachers?: AI and the future of education. Pol it y.
Šucha, V., & Gammel, J.- P. (2021). Humans and societies in the age of artificial intelligence. European Commission,
Directorate- General for Education, Youth, Sport and Culture.
Tuomi, I. (2000). Data is more than knowledge: Implications of the reversed knowledge hierarchy to knowledge man-
agement and organizational memory. Journal of Management Information Systems, 6(3), 103– 117. h t t p s : //d o i .
org/10.1080/07421 222.1999.11518258
Tuomi, I. (2019). Chronotopes of foresight: Models of time- space in probabilistic, possibilistic and constructivist futures.
Futures & Foresight Science, 1(2), 1– 15. https://doi.org /10.10 02/ffo2.11
Vuorikari, R., Punie, Y., & Cabrera, M. (2020). Emerging technologies and the teaching profession: Ethical and pedagogical con-
siderations based on near- future scenarios. Publications Office of the European Union. https://doi.org/10.2760/46933
Williamson, B., & Eynon, R. (2020). Historical threads, missing links, and future directions in AI in education. Learning,
Media and Technology, 45(3), 223– 235. https://doi.org/10.1080/17439 884.2020.1798995
Woolf, B. P., Lane, H. C ., Chaudhri, V. K., & Kolodner, J. L. (2013). AI grand challenges for education. AI Magazine, 34(4),
66– 84. https://doi.org/10.1609/aimag.v34i4.2490
