Teaching may be the best prerequisite for prompting agents, argues Geoffrey Litt
sh_reya · x · 2026-10-17
Geoffrey Litt suggests that teaching in a domain may be the best prerequisite for prompting agents: it trains you to see mediocre work and instantly explain why it's bad and how to improve it. He adds that the best case is having critiqued lots of work in that medium, especially junior or student work — feedback practice that maps directly to writing evaluation criteria for agents.
More from coding & agent
- AutoSep cuts agent reward hacking from 69.9% to 12.1% via weak-to-strong verifier hardening — Stanford · 2026-10-17
- Anthropic engineer: Claude Slides is now how I make slide decks — bcherny · 2026-10-17
- Claude Code Review goes GA on Team and Enterprise plans with Quick, Standard, and Deep modes — bcherny · 2026-10-17
- Dev built soft-body physics engine with Claude's help, claims it tops BeamNG — grandprixlegends · 2026-10-17
- tldraw ships Codex plugin for pair-programming on a live multiplayer canvas — max__drake · 2026-10-17
- Open Co-Scientist: open-source recreation of Google's AI Co-Scientist with Elo-ranked hypothesis evolution — Haunting-Stretch8069 · 2026-10-17