Conference Agenda

This paper examines the relationship between artificial intelligence and democratic education. AI and other digital technologies are currently being touted for their potential to “democratise” education, even if it is not clear what this would entail (see, e.g., Adel et al., 2024; Kamalov et al., 2023; Kucirkova & Leaton Gray, 2023). By analysing the discourse surrounding educational AI, I distinguish four distinct but interrelated meanings of democratic education: equal access to quality learning, education for living in a democracy, education through democratic practice, and democratic governance of education. I argue that none of these four meanings can render education democratic on its own, and present Dewey’s (1956; 2016) notion of democratic education as integrating these distinct conceptualisations. Dewey emphasises that education needs to provide children with skills and dispositions necessary for democratic living, experience in communication and cooperation, opportunities to codetermine the shape of democratic institutions and education itself, and equal opportunities to participate in learning. By examining today’s commercial AI tools (Holmes & Tuomi, 2022; Khan, 2024) and the information-centric models of learning underlying them (focusing in particular on Individual Tutoring Systems and educational chatbots such as the GPT-4-based Khanmigo), I argue that their emphasis on individualisation of learning, their narrow focus on the mastery of the curriculum, and the drive to automate teachers’ tasks are obstacles to democratic education. I demonstrate that: 1) AI deprives children from opportunities to gain experience in democratic living by reducing quality education to efficient transmission of information and divorcing knowledge from practical engagement; 2) AI makes it difficult for children to acquire communicative and collaborative skills and dispositions by substituting engagement with peers and teachers with conversation with always agreeable and patient machines; the increased corporate influence over education systems habituates students to an environment over which they have little or no control, potentially impacting how they will aproach shared problems as democratic citizens. I conclude by outlining some suggestions for aligning educational AI with a pragmatist notion of democracy and democratic education, and by connecting the contemporary trends in educational AI to wider, historical debates surrounding educational technology.

References

Adel, Amr, Ali Ahsan, and Claire Davison. ‘ChatGPT Promises and Challenges in Education: Computational and Ethical Perspectives’. Education Sciences 14, no. 8 (August 2024): 814. https://doi.org/10.3390/educsci14080814.

Bergviken Rensfeldt, Annika, and Lina Rahm. ‘Automating Teacher Work? A History of the Politics of Automation and Artificial Intelligence in Education’. Postdigital Science and Education, 2023. https://doi.org/10.1007/s42438-022-00344-x.

Dewey, John. The Child and the Curriculum: And The School and Society. University of Chicago Press, 1956.

Dewey, John. Democracy and Education. Gorham, Me: Myers Education Press, 2018.

Holmes, Wayne, and Ilkka Tuomi. ‘State of the Art and Practice in AI in Education’. European Journal of Education 57, no. 4 (2022): 542–70. https://doi.org/10.1111/ejed.12533.

Kamalov, Firuz, David Santandreu Calonge, and Ikhlaas Gurrib. ‘New Era of Artificial Intelligence in Education: Towards a Sustainable Multifaceted Revolution’. Sustainability 15, no. 16 (January 2023): 12451. https://doi.org/10.3390/su151612451.

Khan, Salman. Brave New Words: How AI Will Revolutionize Education (and Why That’s a Good Thing). New York: Viking, 2024.

Kucirkova, Natalia, and Sandra Leaton Gray. ‘Beyond Personalization: Embracing Democratic Learning Within Artificially Intelligent Systems’. Educational Theory 73, no. 4 (2023): 469–89. https://doi.org/10.1111/edth.12590

Watters, Audrey. Teaching Machines: The History of Personalized Learning. The MIT Press, 2021. https://doi.org/10.7551/mitpress/12262.001.0001.

AI is expected to play a prominent role in primary and secondary education. Both the promises to solve persistent educational problems as well as the potential for risks are widely acknowledged (Holmes 2022; UNESCO 2023). In this presentation we provide results of a scoping review of European and Dutch policy documents for AI in K-12 education. The aim of this research was two-fold: Firstly, to identify the educational values and practices most affected by AI-technology. Secondly, to provide insight into the types of value-disruptions (Swierstra and Vermaas 2022) that take place in education.

Using the sphere transgression framework (Sharon 2021a; 2021b; Stevens, Kraajieveld, and Sharon 2024) we argue that the introduction of AI into education brings with it values from outside the educational sphere – such as efficiency, control and personalization, which may potentially disrupt the long-standing norms, values and practices foundational to education – such as personal development, pedagogical autonomy and inclusion – thus potentially reshaping the education sphere. Using this framework, we were able to identify three types of value-disruptions triggered by the introduction of AI in the educational sphere. Firstly, explicit value conflicts: conflicts identified by the authors of the policy documents: efficiency vs. pedagogical autonomy, personalization vs. socialization, and standardization vs. inclusivity. Secondly, tacit value conflicts: conflicts that are not explicitly identified as such in the documents but that can be expected to emerge according to the documents: standardization vs personal development and personalization vs. democracy. Thirdly, value redefinitions, i.e., the reinterpretation of thick educational values into thinner versions that align better with the capacities of AI: a redefinition of “inclusivity” as “digital accessibility” and “personalized learning” without “personal development”.

Our findings provide a novel insight into how primary and secondary education are influenced through AI. We conclude by arguing that the value-disruptions we have identified are not exhaustive and further research into the broad landscape of value-disruptions through a spere-transgression theory as well as in-depth analyses of particular value-disruptions and their consequences is urgently warranted to develop AI that support educational values and practices.