Connect with us

Education

Humans Defeat AI at 2025 International Math Olympiad

Published

on

Despite years of AI breakthroughs, human competitors outperformed top AI models from Google and OpenAI at the 2025 International Mathematical Olympiad (IMO), held in Queensland, Australia.

For the first time, artificial intelligence models reached gold-medal level in the world’s top math competition for students under 20. Yet, AI still fell short of perfection.

Five human contestants achieved flawless scores of 42 points. In contrast, neither Google’s Gemini AI nor OpenAI’s advanced reasoning system could match that feat.

Google announced that its upgraded Gemini chatbot solved five out of six IMO problems, earning 35 points enough for a gold-medal status.

“We can confirm that Google DeepMind achieved a gold medal score of 35 out of 42,” said IMO President Gregor Dolinar, as quoted by Google.

“The solutions were clear, precise, and often easy to follow.”

OpenAI also shared similar success. Its experimental reasoning AI matched Gemini’s score with 35 points.

“We tested our model under the same rules as human participants,” said OpenAI researcher Alexander Wei. “Three former IMO medalists graded each proof independently.”

Still, only five out of 641 human contestants from 112 countries earned perfect scores something no AI has yet done.

The milestone marked major progress for Google. Last year at the 2024 IMO in Bath, UK, its model only achieved a silver medal score and required days of computation.

This year, the Gemini model completed its solutions within the 4.5-hour exam limit.

The IMO revealed that private tests were conducted by tech firms using the real 2025 problems.

However, the organizers could not verify how much computing power was used or whether humans assisted the AI submissions.

See also  Editorial: The never-ending fallacy Port Harcourt Refinery's rehabilitation

“It’s exciting to witness how far AI has come in mathematics,” Dolinar stated. “But there’s still a gap between top AI and the best human minds.”

While AI continues to improve in logic and reasoning, this year’s Olympiad proved that human brilliance in mathematics remains unmatched for now.


For Diaspora Digital Media Updates click on Whatsapp, or Telegram. For eyewitness accounts/ reports/ articles, write to: citizenreports@diasporadigitalmedia.com. Follow us on X (Fomerly Twitter) or Facebook

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Latest from DDM TV

Latest Updates

Nigeria outshines Africa, keeps WHO drug regulation crown strong

ADC vows to rescue nigeria as APC, PDP face 2027 backlash

why “retirement” may be an early ticket to decline, experts warn

Huge FAAC allocations to LGAs spark nationwide calls for accountability

Tambuwal’s arrest exposes contradictions in Nigeria’s anti-corruption war — Obi

Why I’m backing Tinubu for a second term — Soludo

Anambra guber: Don’t use my pictures in campaign posters — Ngige warns APC

Fidelity Bank Launches System To Enhance Cross-Border Payments Across Africa

Panic in Eastern Ukraine as Trump Considers Land Deal With Russia

Ekiti guber: Imole political movement inaugurates state working committee, unveils vision

Subscribe to DDM Newsletter for Latest News

Get Notifications from DDM News Yes please No thanks