Open Challenges and Opportunities in Human-Agent Systems

Human-agent interaction requires modeling human perspectives through Theory of Mind, belief-desire-intention architectures, and recursive reasoning to manage coordination in multi-agent environments. Current computational approaches focus on goal and plan recognition to align agent decision authority and objectives with human participants across varying team scales.