OpenAI Solves Navier-Stokes; Buckmaster Cites Opacity — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI discovered a solution to the Navier-Stokes equations — one of six Clay Mathematics Institute Millennium Prize Problems — showing the equations can 'blow up' under certain conditions; the work has not been peer-reviewed but has been formalised (turned into computer-checkable code).
- Tristan Buckmaster (NYU) and Levent Alpöge (Anthropic) rushed out their own solution to the closely related Euler equations after hearing rumors of OpenAI's announcement; Buckmaster said OpenAI gave 'opaque responses' and that his work was stored on OpenAI's servers as a customer, not a research partner.
- OpenAI explicitly denied any wrongdoing, stating no person or AI from the company had seen Buckmaster and Alpöge's work and that its Navier-Stokes solution was different from theirs.
- Terence Tao (UCLA) warned that AI's acceleration of mathematics means new findings will arrive faster than the field can absorb them, saying AI companies are 'dumping carcasses of raw meat onto our communal village table' and leaving cleanup to mathematicians.
- Clay Mathematics Institute president Martin Bridson said the evaluation process 'is deliberately unhurried' and 'absolutely rigorous,' with the $1 million prize still undecided and potentially going to either OpenAI or Buckmaster and Alpöge.
- The cost of solving the problem using OpenAI as a customer would run to $15 million — many times the prize money — but the headline could make OpenAI's upcoming IPO 'a much more attractive proposition for investors,' the source notes.
Why it matters: If the $1M Millennium Prize goes to OpenAI, the company gains a marquee credential for its upcoming IPO — the source notes the solution cost ~$15M as a customer but could recoup many times that in investor appeal. Buckmaster and Alpöge risk being eclipsed despite their related Euler breakthrough, while mathematicians like Tao warn the field cannot keep pace with AI-generated results.
Ask SkimNews
