#1A recent experience with ChatGPT 5.5 Pro
Fields Medal winner Tim Gowers tested ChatGPT 5.5 Pro on open problems in additive number theory and reports the model produced PhD-level research in hours, improving bounds from exponential to polynomial complexity. Gowers provided minimal mathematical guidance while the AI independently improved upon existing frameworks, raising profound questions about the future of mathematical research when LLMs can tackle previously open problems.