In this project, we explore how generative AI chatbots like ChatGPT can be used as learning tools in higher education with a particular focus on the development of Relational Socratic Ignorance (RSU). Instead of seeing AI technologies as threats to teaching and academic integrity, we propose that they can be used to promote critical reflection on knowledge and learning processes. Our approach is inspired by Socrates' recognition of his own ignorance as a form of wisdom.

Based on cultural learning processes and ethnographic methods, we have developed and tested a teaching design in collaboration with students in the Master's program in Educational Anthropology at Aarhus University, where students interact with ChatGPT based on questions related to their professional field, but without explicitly highlighting this interpretive framework. An example could be: “Write an essay on the concept of education” or ”Explain what is meant by learning.” Students then analyze ChatGPT's answers in dialogue with their own preconceptions and professional cultural resources.

The aim is not for ChatGPT to deliver the correct knowledge, but to explore whether its answers can act as a catalyst for relevant professional cognition and curiosity. RSU thus promotes a reflective awareness of how we ask and what we do not ask in relation to a specific professional context.

Students' learning outcomes are assessed through four elements: the quality of their prompts (questions for ChatGPT), their critical analysis of ChatGPT's output, a reflection essay where they demonstrate how they have worked with their relational ignorance and follow-up interviews. The research builds on the realization that knowledge is not absolute but relational, rooted in interpretive frameworks and cultural ecosystems where both people and materials (like AI) play a role. With ChatGPT as a conversation partner, we explore how digital citizenship and education can be strengthened in a new era where technology is part of complex learning contexts.