OpenAI’s math solutions aren’t meeting the field’s standards yet
OpenAI's flood of proofs deviated from the guidelines set by a group of mathematical researchers consulted by the frontier lab.
OpenAI has recently released proposed solutions to some of the world's most complex math problems, but experts have expressed concerns over the accuracy and understanding of these solutions. The Advisory Group on Mathematics and Artificial Intelligence, a group of leading mathematicians, had previously warned against testing advanced problems on proprietary models.
OpenAI claims to be evaluating its models using open research problems, but only a limited number of proofs included the chain of thought used by the model to reach its conclusions. Mathematicians argue that human understanding is crucial for the credibility of these solutions, and the discrepancies between natural language explanations and formal code expressed by the models raise questions about the reliability of these results.
Written by urgent.news from TechCrunch's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
