Mohit Bansal's distinguished lectures at Columbia and Northwestern on trustworthy agents

mohitban47 · x · 2026-11-18

UNC professor Mohit Bansal gave Distinguished Lecture Series talks at Columbia and Northwestern on three major pitfalls of state-of-the-art AI agents and potential fixes: trustworthy/safe collaboration (speaker-listener confidence calibration, persuasion balance, compositional privacy and theory-of-mind belief steering), skills/world discovery, and long-horizon memory — plus implications for research, teaching, and work.

Related event: Mohit Bansal gives distinguished lectures at Columbia and Northwestern on AI challenges(2 posts)→

Original post →

More from Research

Research channel →