Cambridge Paper: Reframe Education for AI — From Memorisation to Dialogic, Collaborative Learning

Education must be reframed to integrate AI, researchers argue

A conceptual paper from the University of Cambridge argues that education should be redesigned so artificial intelligence can help address the major challenges facing humanity — from ecological crises to the future of democratic societies.

The authors propose using AI not as a shortcut to answers but as a tool to support collaborative, dialogic learning: classrooms where students and teachers engage in conversation, jointly explore problems, and test ideas from multiple perspectives.

What dialogic learning looks like

To illustrate their proposal, the paper reimagines a basic science lesson on gravity. Rather than beginning with formulae to memorise, a dialogic lesson might open with a question such as "Why do objects fall to the ground?" Students would discuss ideas in groups and then probe those ideas with an AI chatbot that can adopt the voices of different thinkers — for example, Aristotle, Newton and Einstein.

According to the authors, this approach places learners inside scholarly conversations tied to the curriculum and helps them grasp core concepts by reasoning, arguing and refining ideas collectively.

"If ChatGPT can pass the exams we use to assess students, then at the very least we ought to be thinking deeply about what we are preparing them for," says co-author Rupert Wegerif, Professor of Education at the University of Cambridge.

Wegerif argues that technologies such as the internet, blackboards and writing have historically prompted major shifts in teaching — and AI may demand a similar rethink. He suggests emphasising conversation, collaboration and collective reasoning over rote reproduction of facts.

Risks: AI as a "cognitive poison" if systems don't change

The paper warns that, without changes to pedagogy and assessment, generative AI (GenAI) could function as a "cognitive poison." If schools remain anchored to traditional, print-based assumptions and assessments, students under pressure to produce individually authored essays may lean heavily on GenAI. That reliance risks diminishing their personal creativity, critical engagement and sense of agency.

Co-author Dr Imogen Casebourne (Hughes Hall, Cambridge) notes that GenAI has arrived while education systems are already under strain. The crucial choice, she says, is whether the technology will be adopted in ways that strengthen dialogue and critical thinking or in ways that undermine them.

The paper concludes that AI can be part of the remedy — but only if learning and assessment reward collaborative inquiry and collective reasoning rather than memorisation. The study appears in the British Journal of Educational Technology.