OpenAI Forms Independent Mathematics Advisory Group After Its AI Resolves More Than 100 Open Problems
OpenAI created a Princeton-hosted advisory group of nine named mathematicians after its unreleased model, begun training August 28, resolved 100+ open math problems.
Overview
OpenAI announced a new independent advisory group of mathematicians on Monday, hosted at the Institute for Advanced Study in Princeton, New Jersey, according to TechCrunch. Called the Advisory Group on Mathematics and Artificial Intelligence, the group is meant to give mathematicians more input into the company’s math-oriented research, TechCrunch reported. OpenAI said the same unreleased internal model behind its Navier-Stokes proof has now resolved more than 100 additional long-standing open problems across most areas of mathematics, according to TechCrunch and OpenAI.
What We Know
The announcement comes after the abrupt publication of a solution to the Navier-Stokes Millennium Prize problem, according to TechCrunch — an effort previously reported by The Machine Herald, which centered on a dispute over whether OpenAI’s model may have benefited from a mathematician’s private research data. According to OpenAI, the model behind both results began training on August 28.
Many prominent mathematicians have taken issue with the frenzied pace of those results, according to TechCrunch. Earlier this month, 25 Fields Medal-winning mathematicians signed an open letter arguing that AI labs are threatening their intellectual work as they seek to one-up each other with solutions to famous math problems, TechCrunch reported. OpenAI identifies that letter by name: “A Severe Misalignment of AI in Mathematics.”
According to TechCrunch, OpenAI’s blog post announcing the group states: “This group will serve as a bridge to the mathematical community and broader public, giving mathematicians a voice in how we move forward.” True to its name, the new group will mostly serve in an advisory capacity, assessing the significance of new results and coordinating their release, TechCrunch reported. While members won’t be paid, they can offer unsolicited advice, go public with their views, and control their own membership, giving them a measure of independence from the company, according to TechCrunch.
However, the group won’t be given leeway to slow down or redirect OpenAI’s ongoing mathematical research, TechCrunch reported. As OpenAI’s blog post puts it: “the group will not be responsible for advising us on how to pace our internal progress on mathematics.” The Institute for Advanced Study emphasized this point in its own announcement, according to TechCrunch: “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 mathematicians have been named as initial members of the group, according to TechCrunch and OpenAI: François Charles (ENS-PSL), Camillo De Lellis (IAS, GSSI), Timothy Gowers (Collège de France, Cambridge), Martin Hairer (EPFL, Imperial College London), Nikhil Srivastava (Berkeley, Simons Institute), Ulrike Tillmann (Oxford, INI), Ravi Vakil (Stanford), Edward Witten (IAS), and Melanie Matchett Wood (Harvard). Notably, only one of the members — De Lellis — also signed the Fields Medalists’ letter, TechCrunch reported.
What We Don’t Know
OpenAI has not disclosed which specific open problems its internal model is claimed to have resolved beyond the Navier-Stokes work, nor has it named the model publicly. It also remains unclear how OpenAI will handle disagreements with the advisory group, since the group can voice public criticism but has no authority to slow or redirect the company’s research pace.
Analysis
The formation of the advisory group reflects a company trying to manage the fallout from a research pace that has unsettled parts of the mathematics community. By hosting the group at an outside institution and granting members the freedom to criticize the company publicly, OpenAI is attempting to signal independence — while explicitly withholding any power over how fast it publishes new results. That the group’s sole overlapping member with the Fields Medalists’ open letter is De Lellis, who is also affiliated with the hosting institution, underscores how narrow the overlap is between OpenAI’s outside reviewers and its most vocal critics.