DeepMind claims its AI performs better than International Mathematical Olympiad gold medalists
DeepMind claims its AI performs better than International Mathematical Olympiad gold medalists
DeepMind claims its AI performs better than International Mathematical Olympiad gold medalists
DeepMind claims its AI performs better than International Mathematical Olympiad gold medalists
6 Comments
“The system, called AlphaGeometry2, is an improved version of a system, AlphaGeometry, [that DeepMind released last January](https://techcrunch.com/2024/01/17/deepminds-latest-ai-can-solve-geometry-problems/). In a [newly published study](https://arxiv.org/pdf/2502.03544), the DeepMind researchers behind AlphaGeometry2 claim their AI can solve 84% of all geometry problems over the last 25 years in the International Mathematical Olympiad (IMO)
Proving mathematical theorems, or logically explaining why a theorem (e.g. the Pythagorean theorem) is true, requires both reasoning and the ability to choose from a range of possible steps toward a solution. These problem-solving skills could — if DeepMind’s right — turn out to be a useful component of future general-purpose AI models.”
AI solves cookie-cutter problems with known solutions, what a shock!
My dictionary knows more words than any scholar throughout history…
Whoa, it can do math?
Next they’ll claim it could win a spelling bee.
Calculators are better than international mathematical gold medalists too…
Looking at the paper: it was given hundreds to thousands of attempts per problem.
Section 8 – Results https://arxiv.org/pdf/2502.03544
Far less impressive when you read between the lines.
OpenAI fudges stats with math in the same ways, which made me suspicious enough to go look at the paper.