Abstract:
Human and AI partners increasingly need to work together to perform tasks as a team. In order to act effectively as teammates, collaborative AI should reason about how their behaviors interplay with the strategies and skills of human team members as they coordinate on achieving joint goals. This talk will discuss a formalism for defining AI adaptation to human partners. We will examine the formalism through a simulated case study in a collaborative cooking domain, demonstrating that successfully adapting to each team member’s behavior and abilities can yield performance benefits for the human-AI team.

Committee:
Henny Admoni (Co-chair)
Reid Simmons (Co-chair)
Jean Oh
Victoria Dean