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: Mollick: success with long-running AI agents means supervising the loop(4 posts)→
More from coding & agent
- He benchmarks new models by having them build their own copilot — $28 of compute — bucolucas · 2026-10-27
- Should agents be allowed to change plans mid-task? Reddit debates the drift boundary — wistfulcracker84 · 2026-10-27
- HeyGen + Canva MCP turns slide speaker notes into an avatar presentation video — HeyGen · 2026-10-27
- Local beats cloud: Qwen 3.8 FN handles private student data as well as frontier models — dr_lm · 2026-10-27
- LLM Tails cuts agent input tokens 80% by making agents read before they write — savadks1818 · 2026-10-27
- 10 practical ways to run Astra computer use on Premiere Pro for video editing — beechinour · 2026-10-27