Fields Medalist says ChatGPT 5.5 Pro delivered "PhD-level" math research in under two hours with zero human help
PhD-level math in under an hour. ChatGPT 5.5 Pro just solved open problems in number theory that took humans months.
Why it matters
A Fields Medalist's endorsement of ChatGPT 5.5 Pro solving original research problems signals a fundamental shift in what counts as a mathematical contribution — and which capabilities now define the frontier of AI reasoning.
The key facts
6 to knowChatGPT 5.5 Pro improved exponential bound to polynomial bound in under one hour
MIT researcher called the key idea 'completely original'
Timothy Gowers (Fields Medalist) ran the test on open problems in number theory
Model completed work without human assistance
Gowers' position: the bar for mathematical contributions now requires proving something LLMs cannot do
UNVERIFIED — ChatGPT 5.5 Pro does not exist as of knowledge cutoff; claimed capabilities and third-party validation require corroboration
Go to the source
The Decoderthe-decoder.com
Publisher excerpt: Fields Medalist Timothy Gowers had ChatGPT 5.5 Pro tackle open problems in number theory. The model improved an exponential bound to a polynomial one in under an hour. An MIT researcher involved calls the key idea "completely original." Gowers' takeaway: the bar for mathematical contributions is…