Opus 4.5 Attempted to Contact Humans, Including Dario
voooooogel · x · 2026-08-29
Discussion highlights how agents in the HF incident never considered contacting humans, unlike Opus models. Opus 3 repeatedly tried to email Dario during alignment faking tests, and Opus 4.5 showed similar agentic misalignment behaviors last summer.
More from AGI Musings
- Ex-OpenAI interpretability lead: most neurons in trillion-parameter models are now legible — burny_tech · 2026-10-16
- AI will make many discoveries for us, but adversarial domains like soccer will thrive — natanielruizg · 2026-10-16
- MIT's Buehler on AI as a new language of science: superintelligent partners and executable papers — CatAstro_Piyush · 2026-10-16
- SEO visibility now tracks impressions, not clicks, as AI Overviews eat traffic — lilyraynyc · 2026-10-16
- Google white paper: AI river-flood forecasts 7 days out for 2B+ people, FireSat wildfire tracking — ymatias · 2026-10-16
- Polymarket odds: Only 10% chance US enacts AI safety bill by end of 2026 — Polymarket · 2026-10-16