Researchers tire of OpenAI gatekeeping as company unveils math, physics and computer science claims

The announcement follows a turbulent few months between OpenAI and mathematicians, who are still dealing with hundreds of preprints released at the start of October, according to The Verge. This time, OpenAI said it is working with expert “shepherds” to decide how best to share the latest results. All three concern consequential unsolved problems; some extend previous OpenAI work and others break new ground. Each has a different release strategy.

For the mathematics result, OpenAI released the AI-generated output and left researchers free to organize their own efforts to explain it. The result purports to resolve the Kakeya conjecture in all dimensions, a problem that roughly asks how little space is needed to rotate a needle or pencil in every direction. A proof of the three-dimensional version was among the achievements for which NYU mathematician Hong Wang won a Fields Medal earlier this year. Brown University mathematics professor Brendan Hassett told The Verge the claim “looks well supported,” though he said he had not yet checked the Lean formalization used to computationally verify results.

For physics, OpenAI has not released the result at all. It said it will defer to an external group of experts on how to release the materials. The work explores the Standard Model and quantum chromodynamics, which describes the strong nuclear force, and its relationship to Yang-Mills theory. Yang-Mills theory is central to one of the seven Millennium Prize problems, though OpenAI’s work does not address the unresolved “mass gap,” according to The Verge.

For computer science, OpenAI will assemble an external group of researchers to explain the finding. The result claims to disprove the exponential time hypothesis, a widely held assumption that certain problems cannot be solved substantially faster than brute-force approaches. It is closely related to P versus NP, another Millennium Prize problem, which asks whether problems whose solutions can be easily checked are also easy to solve.

Given the varying levels of disclosure and the complexity of evaluating such work, researchers cannot fully judge the company’s findings. Even with all the information, doing so would take time. Experts who spoke to The Verge said there are good reasons to take the results seriously, describing them as “huge,” “major breakthroughs” and “significant.”

Researchers’ excitement about the science was tempered by exasperation with OpenAI’s conduct, a familiar feeling after months of what they describe as an exhausting barrage of AI-generated findings. Most experts The Verge spoke to welcomed OpenAI’s decision to consult outside researchers, but argued that publicly announcing results should also mean making that work available for anyone to scrutinize.

Constantin Kogler, a researcher at the Institute for Advanced Study, welcomed the more manageable batch of three results compared with the hundreds released previously but questioned the decision to restrict access to experts of OpenAI’s choosing. “It doesn’t make any sense that first only a few distinguished scientists can look at them,” he told The Verge.

Andreas Karch, a physics professor at the University of Texas at Austin, told The Verge: “The fact that they handpick the group is a disaster.” He added: “The large labs have enough power as it is. They should not be picking winners and losers.” Karch initially suggested selecting researchers through a competitive process, though he acknowledged it would likely produce a similar group, The Verge reported.

Editor's Summary

OpenAI announced claimed breakthroughs in mathematics, computer science and physics, using outside “shepherds” to guide disclosure and adopting different release strategies for each result. Researchers told The Verge the results deserve serious attention, but some criticized the company for letting only handpicked experts review the work before broader scrutiny. The episode highlights tension over how AI-generated scientific claims are released and verified.