According to CoinMeta, OpenAI claims that its secret AI model solved hundreds of mathematical problems in a single hint, which has raised doubts among mathematicians who have called for verification. On Tuesday, OpenAI released 722 mathematical manuscripts, all generated by an unpublished internal model. A spokesperson for OpenAI stated that nearly all of the results came from a single hint, although some may have involved multiple attempts. This statement sparked widespread discussion in the mathematical community. Andrew Sutherland, a mathematician from MIT, said that unless the model is made public and can be replicated, skepticism should be maintained towards these claims. OpenAI mentioned that they have posed about 4,000 questions to the model and have retained the outputs they consider to be of sufficient importance.