OpenAI claims it has solved an 80-year-old math problem — for real this time
OPENAI'S BREAKTHROUGH IN DISPROVING AN 80-YEAR-OLD MATH CONJECTURE
The breakthrough is attributed to OpenAI's latest reasoning model, which has been designed to tackle complex mathematical problems. According to OpenAI, this model has generated an entirely new family of constructions that outperform previously accepted solutions, which mathematicians believed were optimal. The company stated that for nearly 80 years, the prevailing belief was that the best solutions resembled square grids, but their model has now challenged this notion, suggesting that there are better configurations yet to be discovered.
THE SIGNIFICANCE OF OPENAI'S CLAIM IN THE MATHEMATICAL COMMUNITY
The implications of OpenAI's claim are profound for the mathematical community. If validated, this proof could not only reshape the understanding of the specific problem at hand but also encourage further exploration into the capabilities of AI in mathematical research. OpenAI has emphasized that this achievement represents the first instance where AI has autonomously solved a prominent open problem in mathematics, a milestone that could pave the way for future AI applications in various scientific domains.
COMPARING OPENAI'S LATEST CLAIM TO PAST MISTAKES IN MATH SOLUTIONS
This announcement comes on the heels of a previous misstep by OpenAI, where the company claimed that its earlier model, GPT-5, had solved multiple Erdős problems. However, it was later revealed that GPT-5 had merely identified existing solutions rather than providing new insights. This past experience has led to skepticism from some quarters, particularly from rivals in the AI field, such as Yann LeCun and Google DeepMind's CEO Demis Hassabis, who have publicly criticized OpenAI's previous claims. OpenAI's current assertion, however, appears to be backed by a more rigorous approach, as the company has included supportive remarks from recognized mathematicians to bolster its credibility.
SUPPORT FROM MATHEMATICIANS VALIDATES OPENAI'S LATEST DISCOVERY
In a bid to lend credibility to its claim, OpenAI has published remarks from several mathematicians, including Noga Alon, Melanie Wood, and Thomas Bloom, who have expressed support for the disproof presented by OpenAI's model. Their endorsement is particularly significant given the scrutiny that OpenAI faced after its previous claims. These mathematicians have acknowledged the potential impact of OpenAI's findings on the field and have highlighted the importance of this advancement in mathematical reasoning. This support from experts in the field could help to assuage doubts and foster a more constructive dialogue regarding the role of AI in mathematical discovery.