From today artificial intelligence Artificial intelligence systems possess many capabilities, but they often fail when it comes to solving complex math problems. That’s why Google is excited that two of its DeepMind AI systems have been able to solve several challenging problems posed in a prestigious math competition.
in a New post published on ThursdayGoogle touted artificial intelligence and the achievements of its DeepMind AlphaProof and AlphaGeometry 2 AI models. As we enter the year 2024 International Mathematical Olympiad (In my opinion) the two systems solved four out of six problems. That effort earned Google’s AI the same status as a first-time silver medalist in this contest, which is typically aimed at young mathematicians.
Also: OpenAI releases SearchGPT: Here’s what it can do and how to access it
Each year, the IMO invites elite pre-university level mathematicians to solve six extremely difficult problems in algebra, combinatorics (counting, sorting and sorting a large number of objects), geometry and number theory. The competition, which extends beyond humans, has also become a way to test and measure machine learning and artificial intelligence systems in advanced mathematical reasoning.
With the problems translated into a formal language that Google’s AI understands, AlphaProof solved two algebra problems and one number theory problem, not only finding the answer but also proving that the answer was correct. Google cited the number theory challenge as the most difficult of the competition, solved by only five of the human contestants. AlphaGeometry 2 solved the geometry problem, but neither model was able to solve the two combinatorial problems.
AlphaProof is an AI-based system that can train itself to prove mathematical statements using formal language. LeanCombining a pre-trained language model with the AlphaZero reinforcement learning algorithmAlphaProof previously taught itself to play and win at chess, shogi, and Go.
Also: Google’s new math app solves almost any problem with AI: Here’s how to use it
AlphaGeometry 2 is an improved version of Alpha geometryBased on Google’s Gemini artificial intelligence, this model can handle highly challenging geometry problems, including those covering object motions and equations of angles, proportions and distances.
In addition to testing the mathematical abilities of AlphaProof and AlphaGeometry 2, Google took advantage of the IMO to test a natural language reasoning system built on top of Gemini with advanced problem-solving capabilities. Unlike the other two models, this one does not require problems to be translated into a formal language.
While the achievement of these models may seem abstract, Google sees it as another step toward the future of AI.
“We’re excited for a future where mathematicians work with AI tools to explore hypotheses, test bold new approaches to solving long-standing problems, and quickly complete time-consuming elements of proofs — and where AI systems like Gemini become more capable at math and broader reasoning,” the company said in its post.