Sharing AI progress in mathematics

In its own post, OpenAI said it is "releasing a broad range of new mathematical results produced by an internal frontier model."[3]According to Unite.AI, the release also includes Lean proof formalizations and shortened summaries of the model's reasoning.[1]Scientific American reports that the claimed results include a solution to the four-dimensional Kakeya conjecture and progress toward the Riemann hypothesis[2]. The magazine adds that many of the results have already been checked in Lean, a programming language that verifies a proof's logic[2].An OpenAI spokesperson told Scientific American that the model produced almost every result from a single prompt given to a single AI agent.[2]The spokesperson also said OpenAI's own mathematicians do not yet understand many of the results.[2]OpenAI said it consulted the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study[3]. However, Scientific American reports that OpenAI is revealing only the average compute time per problem, plus some other statistics, and no prompts[2].NewsBytes reports that experts in the field will have to check the results independently to confirm they are important and correct.[4]OpenAI said it will be funding a series of workshops, conferences, and special programs[3]. The company added that it is working to responsibly release the model that produced these results[3].