Geoffrey Irving: Internal Misjudgment on Why OpenAI Disabled CoT Monitoring
geoffreyirving · x · 2026-08-31
Geoffrey Irving discusses the meme that OpenAI 'forgot' to look at Chain of Thought (CoT). He argues that existing CoT monitoring fails out-of-distribution, causing high false positives, necessitating its disabling to train models with new capabilities. Irving notes an employee pushback claiming this is inaccurate, which he finds frightening as it suggests a fixed internal explanation, even when adjusting thresholds for low false positives at the expense of negatives.
More from Companies & People
- US PhD ROI declines as visa tightens and China's AI ecosystem grows — DrDatta_AIIMS · 2026-08-31
- User predicts Tesla and SpaceX will eventually merge with Grok as CEO — beffjezos · 2026-08-31
- Anthropic Hiring Spree: Sales and Compute Roles Surge — FinanceYF5 · 2026-08-31
- AI automation builders: how did you define and sell your first offer? — xtraai · 2026-08-31
- Report: Tim Cook to step down as Apple CEO on August 31 — AccBalanced · 2026-08-31
- Suno leak reveals codebase including v5, no weights or data — realmrfakename · 2026-08-31