Latenode

OpenAI’s math release sparks backlash from Tao and others

For teams shipping high-stakes agents, the bottleneck is shifting from generation to verification. Provenance, expert review queues, and retraction paths need to be part of the product.

OpenAI’s math release sparks backlash from Tao and others

OpenAI’s mathematics release drew backlash after The Decoder reported that the company published more than 700 manuscripts at once, including claimed solutions to hundreds of open problems, and that more than 100 researchers reacted afterward.

Terence Tao’s response was mixed. The report says he suggested some proofs may contain genuinely useful ideas, but also described deep frustration with the lack of human exposition and discussion needed for mathematicians to evaluate and build on the results. Tao also argued that AI companies’ goals are severely misaligned with those of the math community. The same report quotes Fields Medalist Hugo Duminil-Copin saying: "The shock is immense. I am paralysed."

In its mathematics release, OpenAI said many proofs include Lean formalizations and that it would add more material over time. The criticism from mathematicians points to a different problem: systems can now generate technical claims faster than experts can verify, explain, or contest them.

When models and APIs shift keep the workflow running

Connect the models and apps from stories like this in one scenario. If a vendor changes routing, pricing, or availability, you update the workflow — you don't start over.

Start Free

Free forever plan. No credit card required.