Our position paper “AI Agents Push Humans Out of the Loop” (with Margaret Mitchell and Samir Passi) is getting press! WIRED drew on it in a piece on Meta’s Muse agent, and AI FrontPage covered it as “The Watchdog Problem.” The core argument is that current agent design undermines human oversight, and that long use of AI systems wears down the very skills overseers need. Our short version: oversight degrades the overseer.