Nate Retracts Overconfident Claims on LLM Introspection
joshua_saxe · x · 2026-09-01
Nate Witkin deleted a post making imprecise and overconfident claims about LLMs' capacity for introspection. While he believes most of it was correct, it did not meet his standards. He thanked peers for constructive pushback and has reposted a modified version on his Substack.
More from AGI Musings
- Ajeya Cotra on OpenAI Attack & Recursive Self-Improvement — Miles_Brundage · 2026-09-02
- Opinion: AI-generated spam floods social media, hurting both supply and demand — dotey · 2026-09-02
- Stanford HAI's Bernstein: X's For You algorithm should mend social divides, not inflame them — StanfordHAI · 2026-09-02
- New blog 'Proofs and Prompts' explores how AI is reshaping mathematics — giannis_daras · 2026-09-02
- Token doubling is AI's Moore's law; interactive clock puts AI at 1% of US GDP by 2028 — robleclerc · 2026-09-02
- Ask AI who wins from automation: corporations and shareholders, it says — Ok-Equivalent4080 · 2026-09-02