Mathematician Tony Feng: current-paradigm LLMs lack the diversity of thought research needs
littmath · x · 2026-11-02
Asked if he still stands by his earlier take, mathematician Tony Feng doubles down: even oracle-level LLMs in the current paradigm can't substitute for research overall, because they lack the "diversity of thought" fundamental to collective human research.
His argument:
- For most of his career he felt he was the only person in the world working on his particular questions — math research is sparse exploration of an infinite universe, not many people attacking the same problem.
- The questions themselves are continuously discovered during exploration, which isn't "solved" by oracle-level LLMs; they can only accelerate the exploration.
- Discussed by Daniel Litt and others; whether AI can do core research remains open.
More from AGI Musings
- vikhyat predicts alternating compute oversupply and undersupply on the road to AGI — TheZachMueller · 2026-11-02
- Sam Altman: AI has surpassed top scientists, and 2026 will be a turning point year — kimmonismus · 2026-11-02
- DeepSeek design doc: a person can only follow 3-5 agent sessions at once, attention is the real bottleneck — heyshrutimishra · 2026-11-02
- AI researcher: we have no reliable way to tell whether an AI system is conscious — dioscuri · 2026-11-02
- Toby Ord: track both what models can and can't do to see real capability gaps — tobyordoxford · 2026-11-02
- Forward Deployed Verifier will soon be as valuable as FDE, argues poster — vaibhavbetter · 2026-11-02