UNIT.City — місце, де люди працюють... КРАЩЕ! Обирай свій простір просто зараз 👉
Наталя ХандусенкоAI Eng
22 July 2025, 13:07
2025-07-22
Google and OpenAI's AI models beat the world's math prodigies. But it wasn't without controversy and nuance.
Google DeepMind and OpenAI’s AI models won gold medals at the 2025 International Mathematical Olympiad (IMO), a competition that drew 630 high school students from around the world. While both models won, there was some controversy between the companies, as one of them wanted to boast about its victory before the official results were announced. There’s also the issue of which AI models were used to compete in the competition.
Google DeepMind and OpenAI’s AI models won gold medals at the 2025 International Mathematical Olympiad (IMO), a competition that drew 630 high school students from around the world. While both models won, there was some controversy between the companies, as one of them wanted to boast about its victory before the official results were announced. There’s also the issue of which AI models were used to compete in the competition.
International Mathematical Olympiad
IMO is considered one of the most difficult competitions for high school students worldwide.
Participants take two exams of four and a half hours each over two days. The task involves solving 6 questions.
To earn the gold title, students must score a total of 35 points out of a possible 42.
Models from Google DeepMind and OpenAI solved 5 out of 6 questions, passing the threshold for the final prize. 67 students also received gold medals, Interesting Engineering writes .
But there is one caveat. The AI models of both companies that won are not publicly available. And those that are available to users performed below average. The researchers tested the questions using Gemini 2.5 Pro, Grok-4, and OpenAI 04 — none of them scored more than 13 points in the test.
In addition, a controversial situation has arisen between Google DeepMind and OpenAI.
Google was officially invited to participate by the IMO, but OpenAI was not. However, the latter tested its model with questions and announced its result over the weekend, before the IMO released the official results.
Google, on the other hand, waited for the official results of the contest before announcing the feat on its blog on Monday.
Complicating matters is that the IMO has asked AI companies to refrain from announcing any results until the competition closes on Monday.
Google reacted angrily to OpenAI's actions.
“We respected the IMO Council’s initial request that all AI labs share their results only after the official results have been verified by independent experts and the students have been given a fair and deserved standing ovation,” Google DeepMind CEO Demis Hassabis said at X on Monday.
“I’m completely exhausted… barely alive.” Polish programmer becomes first in the world to beat OpenAI’s ChatGPT at algorithmic programming competition in Tokyo: how it felt
Як нейромережі бачать вільну та незалежну Україну? Тест dev.ua
Нейронні мережі для генерації зображень бачать світ по-своєму, їхню логіку зрозуміти часом зовсім неможливо. Але таки хочеться. На честь Дня Незалежності України редакція dev.ua вирішила провести невеликий експеримент.
Ми задали чотирьом різним нейронним мережам п’ять однакових запитів: «прапор України», «День Незалежності України», «український Крим», «перемога України» та «українці». Отриманими результатами ми ділимося з вами нижче.
