How often is new content published?

Areas of Study

Social and Embodied Intelligence

AI systems that learn and use “world models” to enable context-sensitive social and embodied reasoning
Computational Models of Human Behavior

Realistic simulators of human behavior to stress-test AI systems before deployment
Context and Trust in Human-AI Interaction

How norms and trust emerge, evolve, and vary across the lifetime of a human-AI interaction
Stakeholder Collaboration for Mental and Behavioral Health

Engaging users, patients, clinicians, policy makers and civil society groups to guide our research and products