As Simon Willison notes, OpenAI's announcement is a flex, but the thin details invite skepticism. The core claim—solving decade-old problems for a few thousand dollars—is extraordinary. If true, it validates Terence Tao's vision of 'big mathematics' where AI handles technical grunt work.
Yet, the strategic opacity suggests this is as much a marketing milestone as a scientific one, designed to signal capability ahead of a product launch. The real test will be whether these methods generalize beyond curated, formalizable problems and whether the 'lion's share' of future work can indeed be offloaded to models, or if this represents a narrow, expensive feat.