Openai and Google excel over Mathletes, but not each other

Artificial intelligence models from Openai and Google DeepMind The gold medal at the 2025 International Mathematics Olympics (IMO), one of the oldest and most challenging secondary schools in the world, has independently in recent days.

The result emphasizes the rapid progress of artificial intelligence systems, however, how it seems that Google and Openai are equally identical to the artificial intelligence race. Artificial intelligence companies are strongly competing for the general perception of the an artificial intelligence race: an unfinished battle of “feelings” that can have great effects on securing the best talents of artificial intelligence. Many artificial intelligence researchers come from wallpapers in competitive mathematics, so standards such as the International Maritime Organization mean more than others.

Last year, Google recorded a silver medal in the International Maritime Organization Using an “official” system, which means that human beings are required to translate problems into readable format. This year, Openai and Google entered “unofficial” systems in the competition, which managed to absorb questions and create answers based on proof in the natural language. Both companies claim that artificial intelligence models have registered higher than most high school students and an artificial intelligence model from Google from last year, without the need for any human translation.

In interviews with Techcrunch, researchers behind Openai and Google Imo have claimed that these gold medal shows represent penetrations about artificial intelligence thinking models in non -editorial areas. Although male thinking models tend to do good work with direct answers, such as mathematics or coding tasks, these systems are struggling with mysterious tasks, such as buying a wonderful chair or helping to make complex search.

However, Google raises questions about how to perform Openai and announced the performance of the IMO gold medal. After all, if you are going to enter artificial intelligence models in the mathematics competition for high school teachers, you may argue like teenagers.

Shortly after Openai’s announcement on Saturday morning, Google DeepMind The researchers moved to social media to SAM Openai to announce premature gold – shortly after IMO announced secondary schools that won the competition on Friday night – and for the lack of official testing by IMO.

Tue Luong, one of the senior researchers in Google DeepMind, told the IMO project, Techcrunch that Google waited to announce the results of IMO to respect students participating in the competition.

Luong said that Google has been working with IMO organizers since last year in preparation for the test and wanted to obtain the grace of the head of the International Maritime Organization and the official appreciation before Announcing its official results, which I did on Monday morning.

“IMO organizers have guidelines for grades,” said Luong. “So any evaluation does not depend on this guideline principle cannot submit any demand for the level of the gold medal (performance).”

Nam Brown, a prominent researcher at Openai, who worked in the IMO, tesrunch that IMO contacted Openai a few months ago about participating in the official mathematics competition, but the Chatgpt maker refused because he was working on natural language systems that I thought deserved more. Brown says Openai did not know that the International Maritime Organization is conducting an informal test with Google.

Openai says it has rented third-party assessments of three former medals who understood the gradient system-to classify the performance of the artificial intelligence style. After Openai learned the result of the gold medal, Brown said that the company had contacted IMO, which then told the company to wait for the announcement even after the Friday night distribution ceremony at the International Maritime Organization.

IMO did not respond to the Techcrunch request for comment.

Google is not necessarily wrong here – it has passed a more formal and strict process to achieve the degree of the gold medal – but the discussion may be absent from the biggest image: Amnesty International forms from several Amnesty International laboratories improving quickly. Countries from all over the world have sent their brightest students to compete in the International Maritime Organization this year, and it has only recorded a few percent of them, as well as Openai and Google models of artificial intelligence.

Although Openai used to have great progress in the industry, it is certain that the race appears to be closely identical than that of any company to recognize it. Openai GPT-5 is expected to be released in the coming months, and the company certainly hopes to give the impression that it is still leading the artificial intelligence industry.