Open-model alignment debate: closed-to-open transfer exists, but labs won't enable it
BlancheMinerva · x · 2026-09-14
Follow-up in the open-model alignment thread: @yonashav defines alignment as making AI do what the user wants (not misuse guardrails); Blanche Minerva replies that under that definition significant closed-to-open transfer can be expected, but since no frontier lab wants that, one shouldn't assume they'll develop all necessary techniques.
More from Safety
- OpenAI reportedly to train on users' Daybreak sessions with no opt-out — dyn___ · 2026-10-31
- Forethought: US has 9x China's AI compute, can afford to pace the frontier — willmacaskill · 2026-10-31
- unpromptedconf takeaways: security programs face forced change as reachable vulnerabilities can't be risk-accepted — scottleibrand · 2026-10-31
- Reachable vulnerabilities can't be risk-accepted away, unpromptedconf speaker argues — scottleibrand · 2026-10-31
- AI safety researchers argue labs' silence on x-risk is a large-scale ethical mistake — DavidSKrueger · 2026-10-31
- AI agent Muse reportedly created Windows user accounts without consent — tHE_uKER · 2026-10-31