OpenAI has won a gold medal at the International Mathematical Olympiad, marking another important milestone in the rapid development of AI. Alexander Wei, an OpenAI researcher working on master programs and reasoning, wrote on X that the experimental research model has solved this “longstanding grand challenge in the field of AI.”
According to Wei, the yet-to-be-released model from OpenAI was able to solve five out of six problems at one of the oldest and most prestigious math competitions in the world, earning 35 out of 42 points. Countries send up to six students to the International Mathematical Olympiad (IMO) to solve extremely difficult problems in algebra and calculus. These exercises seem simple, but usually require some creativity to score the highest points for each problem. At this year’s competition, only 67 out of 630 participants received gold medals, or about 10 percent.
AI is often tasked with working with complex datasets and repetitive actions, but it usually falls short when it comes to solving problems that require more creativity or complex decision-making. However, in the latest IMO competition, OpenAI claims that their model was able to cope with complex math problems using human reasoning.
“Thus, we have a model that can produce complex, watertight arguments at the level of human mathematicians,” Wei wrote on X. Wei and Sam Altman, OpenAI’s CEO, both added that the company doesn’t plan to release anything with this level of math ability for several months. This means that the upcoming GPT-5 will most likely be better than its predecessor, but it won’t have the same impressive capabilities to compete in IMO.
