- OpenAI Says Unreleased AI Model Solved 100 Math Problems in Under a Month
- People who revealed victim’s face are equally guilty: Sanjay Saraogi on Jamui’s molestation horror
- Bangladesh: Fuel price hikes raise fears of fresh inflation, higher household costs
- Delhi: ATM fraudster arrested in Mayur Vihar; three cards recovered
- Puneet Issar says J.P. Dutta’s ‘Border’ was iconic: The original is the original
- 48 Years Of Megastar Chiranjeevi: A Journey Of Stardom, Success & Love
- Anjali Arora says she landed Mata Sita role as she visits Hanuman temple every Tuesday: Maybe this is a blessing
- Fact check: Punjab Police denies issuing advisory asking personnel to avoid commuting in uniform
OpenAI Says Unreleased AI Model Solved 100 Math Problems in Under a Month
In Short
After solving the Navier-Stokes problem-one of the Millennium Prize Problems posed 90 years ago-OpenAI asserts that its unreleased AI model has solved over 100 long-standing mathematical problems within a month of beginning its training.
Share :
OpenAI Says Unreleased AI Model Solved 100 Math Problems in Under a Month
AI models are improving rapidly. Earlier this month, OpenAI revealed that an internal AI model-not yet released to the public-had solved the Navier-Stokes problem, a 90-year-old mathematical challenge linked to the Millennium Prize Problems. Now, the company claims that the same model successfully solved more than 100 long-unsolved mathematical problems. According to OpenAI, the model in question began training less than a month ago.
In a blog post, OpenAI stated that this AI model began training on August 28. Reportedly, in less than a month, it solved over 100 open problems spanning most areas of mathematics. The company notes that the speed of this progress surprised OpenAI's mathematicians and sparked internal discussions on how to inform the broader mathematics community.
OpenAI creates a new mathematics advisory group
This has led OpenAI to announce the creation of a new independent advisory group. This group, named the Mathematics and Artificial Intelligence Advisory Group, will be based at the Institute for Advanced Study in Princeton, New Jersey.
OpenAI stated that the group was formed after several mathematicians expressed concern-in a recent open letter titled "A Severe Misalignment of AI in Mathematics"-regarding the effects of using open problems as benchmarks to evaluate new AI systems. "This group will serve as a bridge to the mathematics community and the general public, giving mathematicians a voice in how we move forward," the post noted. According to the company, the advisory group will help review and communicate emerging results while evaluating their relevance. It will also advise OpenAI on academic and professional standards in mathematical research. OpenAI stated that the new advisory group will operate independently; its members will be free to offer advice not solicited by the company, comment publicly on OpenAI’s impact on mathematics, and make their recommendations public.
Members will not receive remuneration from OpenAI and may modify the group's composition as they see fit. However, the group will not have decision-making power regarding any AI company. Initial members of the group include Nikhil Srivastava (UC Berkeley), Ravi Vakil (Stanford University), and Melanie Matchett Wood (Harvard), among others.
It is worth noting that OpenAI’s claim to have solved part of the Navier-Stokes problem-a challenge that has remained unsolved for decades-has drawn some criticism. At the time, the company announced that over 10,000 artificial intelligence agents had arrived at a solution in 88 hours, after exchanging nearly 3 million messages and utilising 130 billion output tokens.
However, Tristan Buckmaster of New York University stated that he and Anthropic researcher Levent Alpoge had been working on the same problem using Codex, OpenAI’s tool. Buckmaster claimed to have learned that "information regarding our progress had been passed on to OpenAI" and questioned both the company's timing and its methods. Although OpenAI clarified that it had not seen Buckmaster and Alpoge’s work before they made it public, the company acknowledged it could not rule out the possibility that anonymised product data had contributed to improving its models.
All this is taking place amidst a growing debate regarding the need to regulate artificial intelligence, given fears that it could pose an existential threat to humanity. OpenAI is a leading advocate for slowing the pace of technological development.
