OpenAI claims it solved an 80-year-old math problem — for real this time

OpenAI says its reasoning model disproved a geometry conjecture that had remained unsolved since 1946, a notable validation of its math capabilities. The article also references prior scrutiny of an earlier claim, but says mathematicians who challenged that claim are now backing the new result. The news is positive for OpenAI’s credibility in AI reasoning, though the broader market impact is likely limited.

Analysis

OpenAI says its reasoning model disproved a geometry conjecture that had remained unsolved since 1946, a notable validation of its math capabilities. The article also references prior scrutiny of an earlier claim, but says mathematicians who challenged that claim are now backing the new result. The news is positive for OpenAI’s credibility in AI reasoning, though the broader market impact is likely limited.