- A Fields Medalist, Timothy Gowers, tested the capabilities of a state-of-the-art AI language model (ChatGPT 5.5 Pro) and found it capable of delivering “PhD-level” math research in under two hours without any human assistance.
- The AI improved an exponential bound to a polynomial one for an open problem in number theory within an hour, with the key idea considered “completely original.”
- This finding suggests that the bar for mathematical contributions may have shifted as LLMs can now solve problems beyond what was previously believed possible without human intervention.
Originally published at the-decoder.com. Curated by AI Maestro.
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.