Altman: OpenAI sacrifices capability to preserve chain-of-thought monitoring
rohanpaul_ai · x · 2026-09-05
Speaking on Bloomberg Television, Sam Altman said OpenAI has discussed the importance of chain-of-thought monitoring "for well over a year" and has made decisions that help preserve chain-of-thought vulnerability monitoring — "even if it means we don't maximize the capabilities we could otherwise get." It's an unusually explicit statement that OpenAI prioritizes monitorability over raw model performance in its training decisions.
More from Models
- Simon Willison benchmarks GPT-6 Astra vs GPT-5.6 on SVG: max effort costs 63 cents per call — gaganghotra_ · 2026-09-05
- GPT-6 reportedly solves a SimpleBench question models have failed for years — ChrisGPT · 2026-09-05
- Unconfirmed GPT-6 Astra Benchmarks Leak, Traders Joke Everyone's Out of a Job — karlwaldman · 2026-09-05
- NVIDIA's fine-tuned Nemotron tops IOI 2026 with 535.4/600, beating top human — NVIDIAAI · 2026-09-05
- Astra on Medium Eats Tokens 2x Faster Than Sol, Users Report — Healthy-Nebula-3603 · 2026-09-05
- Astra model uses Apple Notes on Mac to draw a user's portrait — SIGKITTEN · 2026-09-05