AI companies rumored to have solved longstanding theoretical computer science problems

On September 15, 2026, a user said Scott Aaronson had written about hearing rumors that AI companies had solved longstanding open problems in theoretical computer science and were delaying major announcements because of a Navier-Stokes controversy. The same user cited Greg Brockman for a claim that OpenAI had made “significant progress” on a second Millennium problem. They also shared a rumor that OpenAI was close to solving two remaining Millennium problems—or already had—without identifying which two.