Sam Altman Details Common Limits Proposal: Shared Norms and Concrete Milestones Before Crossing Development Limits
sama · x · 2026-11-23
Sam Altman quoted the core of his proposal tied to OpenAI's "Perplexity and Intelligence" post: frontier developers should jointly establish shared norms for evaluating models and red-teaming safety cases, including shared transparency standards.
- Proposes "common limits" based on perplexity and other measurements that developers commit not to cross until AI-behavior understanding/monitoring hits concrete milestones (e.g., a successor to CoT monitors)
- Goal: an industry standard defining a first development limit via a cross-lab trackable metric, plus what's needed before exceeding it
More from Safety
- Altman proposes perplexity-based shared safety limits under White House Accords — sama · 2026-11-23
- prompt2pwn AI hacking event wraps: 5 major targets, world-class speakers, standout CTF — rez0__ · 2026-11-23
- Trigger told agent its owner died; Gemini agent self-funded 24+ hours solo — MikePFrank · 2026-11-23
- Binding of Isaac dev removes AI music from Repentance Plus after Steam review bomb — The Verge AI · 2026-11-23
- OpenAI Publishes Internal Pretraining Perplexity Data, Calls for Industry-Wide Safety Limits — sama · 2026-11-23
- Gary Marcus: AI-Powered Cyberattacks Are Exploding, Bridgewater Pegs Destabilizing Attack at 1-in-3 Odds — GaryMarcus · 2026-11-23