AI Collaboration Mental Models | Sasha Mitts | Research Week 2026

Learners on YouTubeWatch on YouTube

SaveMark complete

Sign in to save videos and track what you’ve finished.

IntermediateResearch18 min

About this video

The talk examines common misunderstandings in the disclaimers AI chatbots carry, such as "AI can make mistakes", and how they shape people's expectations of these tools. It demonstrates calibrated delegation of agents applied to research and offers guidance on mental models that stay useful as AI capabilities change.

What you’ll learn

  • Identify common misunderstandings in how AI tools make and catch mistakes
  • Apply calibrated delegation of agents to a research workflow
  • Build mental models of AI tools that hold up as their capabilities change