Aarhus University Seal

Socratic Ignorance in working with generative AI chatbots

Instead of seeing  the  GAIs as threats  we want to explore GAIs a learning tools to enhance digital citizenship with a new Socratic Ignorance approach. Just like the Socratic Dialogues in Ancient Greece, ChatGTP works by answering questions, many of them relating to knowledge and learning – and the aim of this project is to lift our understanding of the moral and educational issues of ChatGPT. The learning potential in working with GAIs would be build up around giving students Socratic Ignorance, by letting them give tasks to the GAI: 1. Write an essay about Bourdieu’s concepts of ‘capitals’ or 2. Explain what is meant by ‘interpretative anthropology’ as examples. Then they read the essay, and (supervised by us as teachers) develop and use their Socratic Ignorance. What was new to them in the text? What do they maybe question in the GAIs essay? What has spurred their curiosity? And then we evaluate their essays based on 1. Their ability to formulate questions to GAI, 2. The GAI text, and 3. a written essay from the students where they explain how their Socratic Ignorance meets the GAI text. The students essays are about reflecting about what they have learned they did not know, what it means to know something, and how to be certain of knowledge (their own or the one presented by the GAI).

Maja Hojer Bruun

Associate Professor Danish School of Education - Educational Anthropology, Aarhus

Cathrine Hasse

Professor Danish School of Education - Educational Anthropology, Emdrup

Christian Christrup Kjeldsen

Associate Professor Danish School of Education - Educational Sociology, Aarhus