← sources

I think AI agents that need constant babysitting are not agents

source: reddit (reddit:AI_Agents)published 2026-11-18 13:42ingested 2026-11-19 02:35

retained text

This might be too harsh, but I’m starting to think an AI agent that needs constant babysitting is not really an agent. It is a suggestion machine with extra steps. If a human has to check every action, correct every output, approve every handoff, and monitor every edge case, maybe the agent has not removed work. It has redistributed anxiety. That does not mean it is useless. But it changes the business case. The paid version might be much narrower: one high-confidence workflow failure the agent can detect or prevent reliably. I keep thinking about this around Flatlogic because it is easy to wrap an agent in a polished interface. Harder to prove it actually reduces burden. How do you test whether an AI agent reduces work instead of creating babysitting work?

document fetched 2026-11-19 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1xzfimw/i_think_ai_agents_that_need_constant_babysitting