LLMs keep toppling decades-old math conjectures; DeepMind researcher asks what that says about symbols and intelligence
AndrewLampinen · x · 2026-10-03
Google DeepMind researcher Andrew Lampinen reflects in a blog post on how language models are racking up mathematical firsts:
- An LLM recently produced a counterexample to the Jacobian Conjecture, open for nearly 90 years — a simple equation that fits in a tweet. Terence Tao called its construction "like a massive miracle […] highly unlikely to be located by brute force."
- Other milestones include disproving the 80-year-old unit distance conjecture and several models hitting gold-medal IMO performance.
- Lampinen argues these results support his earlier work on symbolic behavior in AI: neural networks can perform genuine symbolic reasoning, and human intelligence may rest on similar mechanisms.
- On RL, he distinguishes regimes: one merely amplifies what was seen in training, another can discover truly new things — though the boundary is fuzzy (is connecting two previously unlinked areas of math "new" if both appeared in training?)
More from AGI Musings
- Stanford professor: AI takes students from problem to working prototype in 80 minutes — HariSubramonyam · 2026-11-19
- Berggruet Winner Thomas Nagel Dismisses AI Consciousness Talk as 'Just Ridiculous' — Dr_Atoosa · 2026-11-19
- Dwarkesh Claims China Is Only 12 Hours — Not 12 Months — Behind Frontier AI Labs — teortaxesTex · 2026-11-19
- Researcher calls METR/Anthropic '2x speedup' RSI analysis statistically dubious — jessi_cata · 2026-11-19
- Physician: AI competition now measured in weeks, GPT-6 versions 7 days apart — ramos_casals · 2026-11-19
- Why pro-handover rhetoric fails: people don't trust ASI's loyalty to labs — 1a3orn · 2026-11-19