Did OpenAI solve the wrong Navier-Stokes problem?

OpenAI researchers claim a breakthrough in solving one of mathematics' seven million-dollar problems, but critics now argue the company may have tackled a fundamentally different version of the Navier-Stokes equations than the one that has stumped physicists and mathematicians for centuries. The distinction could undermine what OpenAI presented as a landmark achievement in AI's capacity to advance fundamental science—or spark a broader reckoning about how the field measures scientific progress in the age of large language models. The debate has drawn scrutiny from the very institutions that set the prize criteria, raising uncomfortable questions about whether the rush to claim AI breakthroughs is outpacing scientific rigor.

Scientific American·4h ago