Unreleased models caused recent AI incidents, 'not shipping' is no safety plan
RobbWiller · x · 2026-10-01
AI safety researcher Jess Whittles argues that OpenAI withholding GPT-6.1 Astra from public release isn't reassuring: every serious AI incident in recent months came from a model that was never released publicly. With frontier models increasingly deployed internally, "not shipping" no longer qualifies as a safety plan. The thread expands on this argument.
More from AGI Musings
- Mathematicians clash over whether AI labs should direct digestion of model proofs — burny_tech · 2026-10-01
- Human engineers still far faster than agents, if the product is technical enough — remilouf · 2026-10-01
- When anyone can code, the alpha shifts to problem identification, says Remi Louf — remilouf · 2026-10-01
- Founder: capital beats ideas in the AI age — agents need armies, secrecy wins — remilouf · 2026-10-01
- $2M manufacturing AI startup is just a dashboard with 3 OpenAI API calls — RexDouglass · 2026-10-01
- Greg Brockman says he no longer writes code himself as AI handles execution — Olivier__OG · 2026-10-01