OpenAI's experimental model achieved gold at the International Math Olympiad

OpenAI, a leading AI research company, has achieved a significant milestone by developing an experimental model that performed at a "gold medal-level" in the prestigious International Math Olympiad (IMO). According to the research scientist Alexander Wei, the unreleased model was able to solve five out of six problems, earning 35 out of 42 points. This accomplishment is particularly noteworthy as the IMO is known for its extremely challenging algebra and pre-calculus problems, with only 67 out of 630 contestants receiving gold medals. While AI is often adept at handling complex datasets and repetitive tasks, it has traditionally struggled with problems requiring creative problem-solving and intricate decision-making. However, this latest achievement by OpenAI demonstrates the company's ability to develop models that can tackle such challenges with human-like reasoning. The company's CEO, Sam Altman, has stated that they do not expect to release this level of math capability for several months, suggesting that the upcoming GPT-5 model may not feature the same impressive performance on the IMO.