Researchers Advocate for Conversational Learning as AI Advances

Generative artificial intelligence is prompting educators to rethink traditional teaching methods, according to a new paper from researchers at the University of Cambridge. Published in the British Journal of Educational Technology on November 17, 2025, the paper argues for a shift towards more conversational and collaborative learning strategies as AI technologies, such as chatbots, become increasingly prevalent in educational settings.

The researchers, led by Rupert Wegerif, professor of education at the University of Cambridge, highlight that AI could challenge existing educational models that prioritize memorization and recall. They propose what they call “dialogic” learning, where students and teachers engage in discussions that explore problems and ideas collectively. This approach not only enhances critical thinking but also helps students connect with various sources of knowledge.

In a conventional science lesson on gravity, students typically memorize key principles and formulas. The authors suggest a more interactive method, beginning with a question like, “Why do objects fall to the ground?” Students would discuss their ideas in groups and then interact with an AI chatbot designed to represent the thoughts of historical figures like Aristotle, Newton, and Einstein. This method places students at the center of scholarly conversations, allowing them to grasp complex concepts through dialogue and reasoning.

Wegerif emphasized the historical significance of this moment, stating, “Every so often, a technology comes along that forces a rethink of how we teach.” He pointed to past innovations, such as the internet and writing, that have reshaped education. As he noted, if AI tools like ChatGPT can pass educational assessments, educators must reconsider the purpose of these evaluations and explore more engaging, conversational learning methods.

The paper raises concerns about the risks of integrating AI into education without thoughtful strategies. It warns that students struggling with assignments may resort to chatbots for assistance, thus undermining their learning. In this context, Wegerif describes AI as a potential “cognitive poison,” allowing students to avoid critical thinking and limiting their intellectual growth. To counter this, he advocates for educational practices that promote dialogue among students and scholarly ideas.

The authors propose a “double-dialogic pedagogy” that emphasizes two key components. First, it encourages teaching methods that prioritize conversation, allowing students to systematically explore different perspectives while AI serves as a guide. One example mentioned is “ModeratorBot,” currently under development at the University of Cambridge, which aims to facilitate group discussions by ensuring balanced participation and introducing open-ended questions.

Secondly, the authors argue that AI can help students engage with existing scholarly dialogues on various topics. By using AI tools to challenge their ideas, students can develop their reasoning skills. An example of this is QReframer, an AI tool that prompts students to critically reflect on their assumptions rather than simply providing answers.

The researchers stress that successful integration of generative AI into education requires a shift towards more conversational and collaborative approaches. Dr. Imogen Casebourne, a researcher at the Digital Education Futures Initiative, stated, “The question is whether it is adopted in ways that enable students to develop skills such as dialogue and critical thinking or ways that undermine this.”

By fostering learning environments that encourage dialogue, students can better navigate complex global challenges known as the “polycrisis,” which includes issues like climate change, rapid population growth, and threats to democracy. Wegerif concluded, “AI can be part of the remedy, but only with approaches to learning and assessment that reward collaborative inquiry and collective reasoning.”

The findings of this research underscore the need for educational systems to adapt in light of advancements in AI technology, ensuring that students are not merely passive recipients of information but active participants in their learning journeys.