Soojin Lee

Human–AI Interaction for Evidence-Informed Decision-Making

Areas of Interest: Human–AI Interaction · Applied AI · Social Computing

Most real decisions are made with incomplete information. People piece together scattered facts, try to work out whether those facts apply to their own situation, and then act on what they have. I’m interested in how AI can support that process without overstating what it knows. In particular, I want to design interactions that make it clear where the evidence ends and where the AI’s own inference begins.

I’m also curious about the social side of this. What leads people to turn to AI for advice or support in the first place, and how do they decide how much to rely on what it tells them?