scaling01.substack.com/p/accidental-scaling
1 correction found
hearing rumors that Anthropic had solved two Millennium Prize problems
OpenAI’s own writeup does not say Anthropic solved two Millennium Prize problems. It says OpenAI heard rumors that two Millennium Prize problems had been resolved, and later found the rumor related to Levent Alpöge and Tristan Buckmaster’s work on the forced Euler problem, which is not itself a Millennium Prize problem.
Full reasoning
OpenAI's Navier–Stokes post says: "On Tuesday, September 1, we heard rumors that two Millennium Prize problems had been resolved." It does not say Anthropic had solved two Millennium Prize problems. In the same post, OpenAI later clarifies what the rumor was actually about: it was "related to Levent Alpöge, an Anthropic employee, and Tristan Buckmaster, a math professor at NYU," and when OpenAI contacted them, it learned that they "had produced a resolution of the forced Euler problem." OpenAI further states: "In the Euler case, Alpöge and Buckmaster proved a result with external forcing." So the article overstates both who had supposedly solved the problems (not Anthropic as an organization) and what had supposedly been solved (the later-described work was on forced Euler, not two Millennium Prize problems).
2 sources
- On the Navier–Stokes Millennium Prize Problem | OpenAI
OpenAI writes: "On Tuesday, September 1, we heard rumors that two Millennium Prize problems had been resolved." Later, it says the rumor "was related to Levent Alpöge, an Anthropic employee, and Tristan Buckmaster, a math professor at NYU," and that they "had produced a resolution of the forced Euler problem."
- On the Navier–Stokes Millennium Prize Problem | OpenAI
In the same post, OpenAI adds: "In the Euler case, Alpöge and Buckmaster proved a result with external forcing," clarifying the work was on forced Euler rather than two solved Millennium Prize problems by Anthropic.