David Duvenaud rebuts "LLMs don't reason": all the stated reasons apply to humans too
DavidDuvenaud · x · 2026-10-03
In a MIT Tech Review piece, DeepMind's Thore Graepel argues — using AlphaGo's Move 37 as a starting point — that today's most advanced AI systems still lack something fundamental: generating longer chains of thought is not the same as reasoning.
David Duvenaud, professor at the University of Edinburgh, publicly pushed back, noting that every cited reason LLMs "don't reason" applies equally to humans: no explicit, persistent, and inspectable epistemic state; no clean separation between what a system knows and how it manipulates that knowledge; and explanations concocted after the fact. "Can I reason?" he asks. The exchange reopens the debate over what counts as reasoning in LLMs.
Related event: AlphaGo Veteran Questions LLM Reasoning, Sparking Researcher Debate(6 posts)→
More from AGI Musings
- Dev challenges "AI will solve humanity's problems": where exactly in the next 5 years? — aronchick · 2026-11-17
- Top researchers are quietly stopping paper submissions, observer notes — avt_im · 2026-11-17
- "Abundant intelligence should terrify businesses built on keeping it scarce" — _AustinCalvert_ · 2026-11-17
- 130-page paper introduces "bottleneck pressure": work AI can't do gains value as the rest speeds up — soumitrashukla9 · 2026-11-17
- AI conference authors and reviewers complain unprecedentedly about AI slop — khademinori · 2026-11-17
- Sam Altman: AI already speeds up OpenAI's research, but a self-improving button? 'Definitely not' — ChrisGPT · 2026-11-17