Mollick Finds a Multi-Agent Interface in the New Claude Projects

Ethan Mollick has a new favorite surface for multi-agent work.

The new Claude Projects is “a nice interface for multi-agent work,” Mollick says. It does a good job with a user-facing coordinator model and independent agents running threads that pass messages around. “Takes some fiddling but quite powerful.”

He adds a footnote: same name, but no relation to the old Projects.

The group chat is the wrong shape for multiplayer AI.

Earlier on Monday, Mollick argued that multiplayer AI needs to move past the group chat phase. Chats are chaotic. Everyone sees everything, which creates social and management issues, and they are bad for long projects.

Worse, he says, they keep AI trapped inside existing teams, when its real promise is working and helping across them.

There are many more options for multiplayer, Mollick says. They require serious experimentation with organizational form, which takes more effort than adding AI to an existing chat. In his view, that effort also makes them a better strategic moat than a generic capability.