Leaders worry OpenAI is not doing due diligence to vet results and that AI models aren’t accessible to broader field of mathematicians
OpenAI has astounded mathematicians after releasing hundreds of new mathematical findings on Tuesday.
The company published over 370 mathematical results across a variety of topics such as algebra, theoretical computer science and mathematical logic, showcasing what some of its most advanced artificial intelligence models are capable of.
Last month, the company solved the Navier-Stokes equation, one of the world’s toughest mathematical problems with a $1m reward for anyone who cracked it.
The achievement prompted concerns from leaders in the field who say that frontier labs should not be testing the most advanced mathematical problems on proprietary AI models that are not accessible to the broader field of mathematicians. The Institute for Advanced Study in Princeton, New Jersey, an independent group of mathematical experts, said that it does not endorse the practice.
“It is now the case that AI can output mathematical arguments in situations without the human who prompted it being able to understand the arguments, verify them, or take responsibility for them,” a statement from the organization read. “We believe that human understanding of mathematics remains of paramount importance. How, in this new era, can we work towards a new paradigm that includes human understanding of mathematics as part of responsible scholarly output?”
To address concerns from the mathematics community, OpenAI announced it would work with the Institute for Advanced Study in order to give “mathematicians a voice in how we move forward”.
The company did not indicate, however, that it would stop testing its AI models with these advanced problems. Experts in the field say they worry OpenAI is not doing the due diligence required to vet these results.
In an interview with the New York Times, Tristan Buckmaster, a New York University mathematician who was working on the Navier-Strokes problem, said mathematicians who were prompting AI models to solve equations could be providing information that helped the model get to the result.
“There’s likely to be a bunch of results where they take someone’s work and then take it to completion,” Buckmaster said.
The advisory board has also asked that AI labs grant “equitable access” to their AI models to the global mathematics community.
“The use of proprietary internal models by AI labs to do mathematical research risks creating a two-tier system where labs outrun the rest of the field, effectively alienating the mathematical community from its own discipline,” the group wrote.
