OpenAI's math drop sparks scientific skepticism
openai releases new math results from internal frontier model, but lacks transparency on methodology, sparking criticism from experts gary marcus and terence tao.
openai has released a broad range of new mathematical results produced by an internal frontier model, but the company has faced criticism for lacking transparency on the methodology used to achieve these results. experts gary marcus and terence tao have expressed skepticism on the scientific validity of the results, citing a lack of detail on the procedure, architecture, failure rate, and training process. this has raised questions about the generalizability of the results beyond math and the potential for these results to contribute to the development of agi.