• Google and OpenAI's AI models won gold medals at the International Mathematical Olympiad (IMO).
• This is the first time AI has crossed the gold-medal scoring threshold in the competition's history.
• Both models solved five out of six problems, showcasing significant improvements in reasoning capabilities.
• About 11% of human participants also achieved gold medals, with five perfect scores awarded.
• The achievements herald potential collaborations between AI and human mathematicians in the future.

Google and OpenAI announced that their artificial intelligence models won gold medals at the International Mathematical Olympiad (IMO), a significant development that underscores advancements in AI's mathematical capabilities. This achievement represents the first instance where AI systems have surpassed the gold-medal scoring threshold typically reserved for human students at the prestigious competition, which draws participants under 20 years of age from around the world, according to Reuters and South China Morning Post.

Both AI models successfully solved five out of six problems at the IMO, utilizing general-purpose "reasoning" frameworks that process mathematical concepts through natural language understanding. This contrasts sharply with earlier AI methodologies that relied heavily on formalized computation, marking a distinctive shift in the capabilities of AI in mathematical problem-solving, as mentioned by Reuters and Bangkok Post.

Junehyuk Jung, a professor at Brown University and visiting researcher in Google's DeepMind unit, expressed optimism about the future of AI in mathematics, stating, "I think the moment we can solve hard reasoning problems in natural language will enable the potential for collaboration between AI and mathematicians," highlighting that these advancements could lead to the resolution of unsolved problems in the field within a year, according to Reuters and Bangkok Post.

OpenAI's success came from an experimental model designed to enhance "test-time compute," allowing its AI to perform extensive reasoning tasks. Noam Brown, a researcher at OpenAI, noted the high computational costs involved in achieving these results, which they believe represent a significant leap forward in AI's problem-solving skills. This has led to a broader expectation that such models may soon exceed their current capacities across diverse fields, including physics, as reported by Bangkok Post and CBS News.

In total, 67 out of 630 participants in the competition received gold medals, showcasing both human and AI achievements. Notably, five students achieved perfect scores, further underscoring the competitive nature of the ongoing mathematical prowess trialed through the IMO. Google DeepMind CEO Demis Hassabis called attention to the quick progress in AI capabilities, which may soon involve collaboration with human mathematicians in more complex problem-solving scenarios, as indicated by Reuters, South China Morning Post, and Bangkok Post.