OpenAI recently claimed to solve a Millennium Prize problem, a mathematical challenge worth one million dollars, without providing a complete proof. The company announced this achievement alongside other major breakthroughs in the discipline, prompting immediate concern from the academic community. Mathematicians argue that the system likely relied on training data containing existing solutions rather than genuine discovery. This approach mirrors earlier controversies where AI models generated plausible but incorrect code or proofs. The backlash centres on the lack of transparency regarding how the model reached its conclusions.
The core issue is that solving these problems requires rigorous human verification, which current AI systems cannot yet perform independently. Without a verifiable proof, the solution holds no scientific value and risks misleading future research. OpenAI has since formed an advisory group to consult with elite mathematicians on improving its methods. The organisation aims to avoid repeating past mistakes by prioritising accuracy over speed.
* The Navier-Stokes existence and smoothness problem was the specific challenge addressed.
* Experts demand full, step-by-step proof before accepting any AI-derived solution.
* OpenAI now seeks guidance from mathematicians to validate its future outputs.
