Report
Collaborative Intelligence benchmark aims to study when and how AI models make good collaborators
A researcher says their lab is interested in what future AI agent collaboration will need to be useful and safe.
TLDR
A researcher says their lab's Collaborative Intelligence benchmark is a step toward understanding when and how AI models are good—or not—as collaborators. The lab is interested in what future agent collaboration will need to be useful and safe.
Combined views
4.5K
5 Sources, first seen ago
