Knowing Ourselves with AI: A Modified Tool-Assisted Approach

• In Event: Artificial Intelligence, Learning, or Teaching

Abstract

This paper addresses that question by examining how AI systems increasingly play a role in shaping individuals’ self-understanding, and by analyzing the epistemic and formative tensions that arise when AI-generated insights into the self are treated as authoritative. Drawing on philosophical debates around self-knowledge, particularly the contrast between agential and inferential models, this paper proposes a modified tool-assisted approach that clarifies the conditions under which AI can support rather than undermine epistemic agency. Ultimately, this paper offers conceptual and normative guidance for how personalized AI tools might be designed and deployed in educational contexts to support, rather than supplant, learners’ autonomy in self-knowledge.

Author

• Soomin Nam, University of Illinois at Urbana-Champaign