Both John Dewey and Paulo Freire theorized about education by contrasting two approaches.
Dewey made a distinction between curriculum-focused education and child-focused education. The former heavily emphasizes logically organized, often symbolic knowledge transmitted through a pre-arranged curriculum, which he likened to a map. On the other hand, child-focused education is based on the psychological journey of the learner, starting directly from their lived experience.
Freire then contrasted ‘banking education’ characterized by one-way transmission of knowledge (which he likened to making ‘deposits’ into the minds of learners with the goal of adapting them to the current societal system) and ‘problem-posing’ education in which the learner’s reflection and action are combined and a dialogical approach is used toward learner liberation.
What is implicit in both theorists’ views is that while ‘curriculum-focused’ and ‘banking education’ are systematically regarded as the more efficient options because they are less constrained by the sometimes slow developmental journey of students, in both approaches the teacher still has choice. In this paper I want to argue that recent advances in AI tools do not take away the teacher’s choice but rather emerge as a new variable affecting their authority and method of achieving educational goals. Both perspectives on AI – as a tool in the hands of a teacher and as a structural force reshaping teacher authority – are considered.
Having previously relied heavily on a state-approved curriculum and planning lessons according to its pre-defined logical structure of knowledge, a teacher might use AI to generate conceptual or visual maps of a topic, retaining the bird’s-eye perspective characteristic of a city map, a strong metaphor used by Dewey. Equally, the same teacher may combine their direct personal knowledge of their students with AI tools to help “psychologize” the subject in a Deweyan sense and create materials that meet students in their lived experience, helping them see logical connections and provoking learner thinking. AI can help translate “symbols detached from life” into relatable content which the student can connect with.
Looking at the capabilities of AI tools from the Freireian perspective, again we might have a teacher whose main goal is to impart specific knowledge and even values to students to prepare them for the current society (Freire might even say, shape them to accept current ‘structures of oppression’). Such a teacher would use AI to strengthen their control – over the subject matter as ‘banking deposits’ and over students themselves—reducing them to objects of algorithmic processing. AI tools could be used to generate a coherent and persuasive video celebrating a historical national hero. On the other hand, a different educator – more inclined to see education as liberation through reflection and action – would use AI to create activities helping students achieve reasoned dialogical understanding. The same educator might generate a less celebratory video on the same historical person inviting a more nuanced and contextualized exploration.
Putting Dewey and Freire side by side, a foundational difference is clear – and it has to do with what kind of power a teacher holds.
Dewey ascribes epistemological power to the curriculum-focused teacher, with the teacher expected to have traveled the territory and be in control of the map. Here, AI complicates the picture because users attribute increasing informational authority to AI. Students can challenge the accuracy of the teacher’s presentation when ChatGPT contradicts it, questioning the teacher (or in fact the approved curriculum) as the primary source of knowledge.
Freire then ascribes political power to the proponent of banking education, effectively shaping students in a specific way so they ingest and accept knowledge serving the interests of current political structures. Here, AI – if used directly by students – can either reinforce or challenge the dominant narrative constructed by power structures by either confirming narrow politically-motivated banking-education content or showing its limitations and encouraging alternative opinions.
In both approaches, AI is no longer the power-neutral element in education that many expected it to be as an educational technology. Despite having been shown to be biased and suffering from hallucinations, it is increasingly being implemented in schools with a particular set of objectives and often promoted for its novelty and justified by the fear of falling behind. It starts to actively influence and even transform the epistemological power of teachers (Dewey) and their political power (Freire).
If an educational system in a particular country pushes teachers into ‘banking education’ by setting strict requirements on curriculum coverage, teachers will be incentivized to use AI toward the goals of banking education – generating presentations and exercises and tests exactly mirroring the prescribed curriculum. Yet, teachers who value a liberal arts approach to education and critical thinking and individual students’ self-development, might still purposefully choose the “less effective” option as well – the problem-posing approach (Freire) or child-centered ‘psychologizing’ activities (Dewey). AI will give them previously inaccessible creative opportunities because it makes preparation faster and teaching more customizable.
The practice and adoption of AI will then be determined partly by individual values of teachers but with significant influence from official educational policy requirements. In my country, Slovakia, the new educational reform shifts the focus from static knowledge reproduction to functional competences, from performing well on subject-based assessments to demonstrating that children can apply knowledge in practice. These shifts in emphasis are intended to make education in Slovakia more child-centered (Dewey) and more problem-posing (Freire) but experts warn that structural incentives have not changed and political actors retain the power to tighten or loosen school autonomy with each legislative cycle, open to reversing this trend. The new capabilities of AI are also likely to influence what teachers optimize for.
In conclusion: neither Dewey’s nor Freire’s theories lose relevance today but AI has emerged as a new factor influencing which educational content (‘map’ or ‘deposits’) appears in classrooms and how it shapes the new generation (to become agents in their own lives or recipients made to fit existing societal structures).
References
Dewey, J. (1990). The child and the curriculum. In The school and society and the child and the curriculum (pp. 181–209). University of Chicago Press. (Original work published 1902).
Freire, P. (2000). Pedagogy of the oppressed (M. B. Ramos, Trans.; 30th anniversary ed.). Continuum. (Original work published 1968).
