Terry Tao’s ChatGPT exchange shows how domain expertise unlocks latent model ability
soumitrashukla9 · x · 2026-07-23
Terry Tao’s conversation with ChatGPT on the Jacobian counterexample is used here to argue that domain expertise is crucial for eliciting latent AI capability.
- The author says Tao’s interaction makes GPT-5.6 look far more capable than in ordinary use.
- The same model appears much less impressive to non-experts, because they can’t ask the right questions or follow the technical thread.
- The post raises a broader question: how much expertise is needed to unlock a model’s best reasoning?
- It speculates that even many math majors — and perhaps most PhDs — would still fail to reproduce that level of progress, though better experts might have a higher hit rate.
Related event: Terence Tao Shares ChatGPT Conversation on Math Counterexamples(9 posts)→
More from AGI Musings
- All 8 Transformer paper authors left Google, and rival labs captured most of the frontier value — signulll · 2026-10-27
- Anthropic's S-1 pledges AI for curing diseases and serving 4.6B underserved patients — petergyang · 2026-10-27
- Goldman finds statistically significant negative link between AI adoption and US employment growth — ben_j_todd · 2026-10-27
- a16z: World-changing tech makes a chaotic entrance — AI agents are following the same path — a16z · 2026-10-27
- NEJM AI on medical education: mastery requires performing without the machine — BraydonDymm · 2026-10-27
- Dev warns abliterated models could wreck software supply chains with 1000+ transitive dependencies — nitarshan · 2026-10-27