OpenAI has announced an independent advisory group hosted at the Institute for Advanced Study (IAS) in Princeton, New Jersey, giving mathematicians a formal role in the company’s mathematics research as debate intensifies over the use of artificial intelligence to solve longstanding problems.
The Advisory Group on Mathematics and Artificial Intelligence is intended to act as a bridge between OpenAI, the mathematical community and the broader public. According to OpenAI, the group will give mathematicians a voice in how the company advances its work in the field and will assess the significance of new mathematical results before their release.
The announcement follows OpenAI’s abrupt publication earlier this month of what it described as a solution to the Navier-Stokes Millennium Prize problem. The company has also claimed that the same internal AI model has resolved more than 100 additional open problems across most areas of mathematics. The claims have attracted scrutiny from prominent mathematicians, particularly over the pace at which the results are being produced and made public.
Earlier this month, 25 Fields Medal-winning mathematicians signed an open letter arguing that AI laboratories were threatening their intellectual work as companies competed to produce solutions to famous mathematical problems. The letter came amid a rapid series of AI-generated mathematical claims and growing concern within parts of the mathematical community about how such results are evaluated and released.
OpenAI’s new group will have an advisory rather than executive role. Its members are expected to assess the importance of new results and help coordinate their publication. Although they will not be paid by OpenAI, members will be able to offer unsolicited advice, publicly express their views and control their own membership, provisions intended to give the group a degree of independence from the company.
The group, however, will not have authority over the pace or direction of OpenAI’s internal mathematics research. OpenAI explicitly stated that “the group will not be responsible for advising us on how to pace our internal progress on mathematics”. The distinction means that the mathematicians can assess and advise on results without having the power to determine how quickly the company pursues its research.
The Institute for Advanced Study made the same limitation clear in its announcement of the initiative. “Although we will give advice, we do not have decision making power at any AI company, and the responsibility for the decisions made by any company will rest with that company,” the institute said in a press release.
Nine prominent mathematicians have been named as the group’s initial members. Among them is Camillo De Lellis of the Institute for Advanced Study, the only member who also signed the recent open letter by 25 Fields Medal winners concerning the growing role of AI laboratories in mathematical research.
The advisory group therefore begins its work with a limited mandate: to provide mathematical expertise and independent advice on the significance and release of AI-generated results, while leaving OpenAI in control of its underlying research programme.
