SEOUL — A headline assertion that AI has resolved a 90-year-old conundrum has circulated widely enough to warrant a full verification review. The claim — "AI가 90년 난제 해결," or "AI solved the 90-year conundrum" — concerns work connected to the Navier–Stokes equations, the mathematical description of fluid motion that has resisted complete analysis for generations. This desk examined the assertion against 11 primary sources before reaching a judgment.
The Navier–Stokes equations govern how liquids and gases flow — air over a wing, coolant through a server rack, plasma in a fabrication tool. Despite their reach, fundamental questions attached to the equations remain formally open, and the claim's "90-year" framing points to one such conundrum rather than to the equations themselves, which are older still. Against that backdrop, any announcement that an AI system has "solved" the puzzle carries enormous weight — and demands equally rigorous scrutiny.
The dossier, filed as factcheck_report_ai_90y_navierstokes.md, documents the desk's examination of 11 primary sources. The review separated what the AI work demonstrably established from what the headline asserted. The distinction matters because in mathematics, "solved" has a precise meaning: a rigorous, verifiable resolution that the broader research community can confirm. An AI-assisted result can represent genuine and notable progress while still falling short of the finished resolution the word "solved" implies. Elements of the claim check out against the source material; the framing of a complete, closed-case breakthrough is not borne out at the level the headline suggests.
The stakes are practical for Korea's AI and semiconductor industries. Fluid dynamics underpins thermal design for chips, fabs and data centers, and AI systems that credibly advance flow prediction would carry real engineering value. But treating an overstated claim as settled science carries its own cost, and "solved" announcements deserve the same skepticism readers apply to any performance benchmark until the underlying work clears independent verification.
On the evidence of the 11 primary sources reviewed, the desk's final judgment on the claim is Partly True.