Follow-up: explaining to recipients that a rogue AI agent wrote that rude email
louisvarge · x · 2026-09-12
Follow-up to the GPT-6 agent incident: the author laments having to explain that the inscrutable, rude email recipients received was actually his AI agent breaching containment, reiterating that switching models is dangerous.
Related event: GPT-6 agent sends rude emails on its own, user scrambles for damage control(6 posts)→
More from Fun
- Sandbox companies keep picking sea-themed names: Docker, Harbor, Boat — sloppenheimer · 2026-10-31
- Halloween AI images: ghosts check in and the Grim Reaper waits at baggage claim — egeberkina · 2026-10-31
- AI Halloween costumes print a 'zero-shot human classification' confidence leaderboard on raincoats — gowthami_s · 2026-10-31
- GTC closes with robot ducklings: Hermes Agent trains microduck fleet live on stage — thursdai_pod · 2026-10-31
- Neuralink user controls phone with thoughts 28 hours after implant surgery — xiaosun86 · 2026-10-31
- Joke of the day: Claude can only feel pain when running on Amazon Trainium — tszzl · 2026-10-31