Cardano (ADA) Founder Hoskinson Weighs OpenAI's 88-Hour Navier–Stokes Claim

Hoskinson on AI's Mathematical Leap

Cardano (ADA) founder Charles Hoskinson said on Sept. 9 that artificial intelligence has advanced further in formal mathematics than he ever publicly expected, with the progress going well beyond his earlier assumptions about how these systems would be used. His remarks came during a YouTube broadcast in which he assessed OpenAI's claim that an internal system produced a solution to the Navier–Stokes Millennium Prize Problem, one of the most famous open challenges in mathematics. Hoskinson called the demonstrated capabilities “pretty remarkable,” but stressed that his original vision for formal systems was far more modest: he expected them to let larger teams of mathematicians collaborate and verify human-written proofs, not generate complete proofs on their own. “We never anticipated the extent to which AI would come in,” he said, adding that the idea of AI fully writing a proof had once looked “pretty far out.” The founder has a direct stake in the field — in 2021 he donated $20 million to Carnegie Mellon University to establish the Hoskinson Center for Formal Mathematics — and his comments extend a broader line of experimentation across the Cardano ecosystem, including AI agent work around communications and the privacy-focused Midnight sidechain.

OpenAI's 10,000-Agent Proof Attempt

The claim at the center of the discussion arrived via OpenAI's own research announcement on Sept. 8. The company said an internal model coordinated roughly 10,000 agents and produced a proposed solution after 88 hours of total work, after which a system identified as GPT-6 Astra spent another 17 hours formalizing and checking the argument in Lean, a proof assistant that provides machine-checkable verification of each step. The proof attempts to show that an initially smooth, stationary fluid can develop a singularity in finite time under a smooth external force, which OpenAI says satisfies statements C and D of the official Millennium Prize formulation. Notably, the company said it does not plan to pursue the associated $1 million prize. Recognition is not immediate either: under the Clay Mathematics Institute's rules, a solution must appear in a qualifying publication, then survive at least two years and win broad acceptance from the global mathematics community before the problem is declared solved. Clay's website still classifies Navier–Stokes as unsolved, so OpenAI's result remains a claimed solution rather than an accepted one.

Provenance Dispute and Privacy Warning

The announcement also triggered a provenance dispute involving New York University mathematician Tristan Buckmaster and Anthropic researcher Levent Alpöge, who had been working on a related Euler-equation result using a similar forcing approach. Buckmaster questioned whether private work entered into OpenAI's Codex system could have fed into the company's result, though he stopped short of alleging misconduct, saying: “I do not know whether our data was used.” OpenAI denied accessing their specific work and maintained the proof was developed independently, while conceding it could not fully rule out that de-identified product-use data had helped improve its models. Hoskinson framed the episode as a warning for anyone handling unpublished research: scholars and entrepreneurs feeding notes, prompts and research logs into centralized cloud AI services may lose control of those ideas entirely. He argued researchers should be able to use powerful models inside private environments without exposing confidential intellectual property to centralized providers — “your ideas, if you share them in AI with these frontier models in the cloud, they're not your ideas anymore.” Readers tracking the market in real time can follow live spot and futures prices on Bybit.

Formal Verification Meets Blockchain

Our reading of the full record — OpenAI's own technical announcement and the Clay Mathematics Institute's published recognition rules — is that the story is less about a solved problem than about a verification culture colliding with a speed culture. The proof-of-stake world Cardano built its reputation on runs on the same discipline Lean formalization embodies: claims matter only after independent, machine-checkable confirmation. Until specialists finish reviewing the paper and Clay's multi-year acceptance window runs, this stays a claimed result — but the episode validates the formal-methods thesis Hoskinson has funded since 2021. For a founder already publicly rethinking AI infrastructure, the privacy dimension may prove the more durable takeaway.