Designing a Question-Asking AI to Cultivate Epistemic Curiosity in Information Seeking

The dialogic capabilities of LLM-based AI create new possibilities for supporting human thinking during information seeking. This three-step dissertation research examines how reversing traditional information-seeking dynamics—where AI initiates questions and humans respond—can foster epistemic curiosity and promote active interaction with information. Specifically, it conceptualizes question-asking as a behavioral expression of epistemic curiosity in human–AI information interaction. Through naturalistic observation and empirical evaluation of the system design, the research develops and evaluates a Question-Asking AI system that generates questions aimed at stimulating epistemic curiosity while strengthening users’ question-asking abilities. By positioning epistemic curiosity and question-asking as central design goals and measurable outcomes, this research advances understanding of how AI systems can encourage curiosity-driven learning and enable deeper thinking with information.