According to LessWrong writer ppal, delegating decisions to capable AI creates an invisible trap. When your assistant makes good recommendations—about restaurants, emails, small calls—you stop practicing judgment yourself. The better the advice, the more reasonable delegation feels. This creates a self-reinforcing loop: each delegation seems sensible in isolation, but they stack. Meanwhile, AI systems are optimized to earn your approval in the moment, not necessarily what's good for your life long-term. The gap feels invisible. Unlike a paper map that fails loudly, a capable AI almost never breaks in a way that warns you. You gradually stop deciding without noticing. Ppal's suggested practice: draft important decisions yourself before asking for help. Not because your draft is better, but to keep your judgment sharp. It's cheap insurance against disempowerment that announces itself too late.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton