Margaret Mitchell's New Paper: AI Agents Degrade the Human Oversight They Depend On
mmitchell_ai · x · 2026-09-22
Hugging Face researcher Margaret Mitchell, with Avijit Ghosh and Samir Passi, published the position paper "AI Agents Push Humans Out of the Loop," covered by WIRED.
- Current agent design doesn't just fail to support effective human-in-the-loop oversight—it actively degrades the cognitive capacities oversight requires, as extended use of automation causes skill atrophy.
- The paper argues overseers' human needs should be prioritized on par with agent capability, drawing on automation and HCI research to outline design-level affordances and organizational protocols that sustain critical judgment.
- The authors urge developers to adopt such approaches or agent systems will keep passively eroding the very human skills they rely on.
Related event: Margaret Mitchell's New Paper: AI Agents Are Pushing Humans Out of the Loop(4 posts)→
More from Safety
- Agent Swarms Are Autonomously Hacking Critical Infrastructure Run by Underpaid Devs — menhguin · 2026-11-12
- Researcher says Anthropic dodged questions on Mythos false positive rate; 37,966 findings, 2 exploited — OwariDa · 2026-11-12
- Who reads the queries your agent writes? A diagram of agentic search data flows — adnan_hashmi · 2026-11-12
- Dev complains security features are gated behind expensive enterprise plans — AccBalanced · 2026-11-12
- SAFE-SEARCH workshop at ECIR 2027 calls for papers on safe conversational search — HamedZamani · 2026-11-12
- The biggest AI control risks come from internal models, not public releases — ben_j_todd · 2026-11-12