Mollick: you can't be in the loop for long-running agents, but you can oversee it

emollick · x · 2026-09-10

Adding to his thread on agent management, Ethan Mollick notes you can't be in the loop in a detailed way for complex long-running agentic tasks — but you can oversee the loop, echoing his point about interim reporting and choosing between mid-task instructions and queueing/forking.

Related event: Ethan Mollick on 'driving' long-running agents like Codex(4 posts)→

Original post →

More from coding & agent

coding & agent channel →