“Fields Award for being human”… Mathematics community tenses as Open AI solves over 300 difficult problems

(Los Angeles = Yonhap News) Correspondent Kim Kyung-yoon = The mathematics world was shocked as artificial intelligence (AI) solved hundreds of problems at once, including difficult problems that mathematicians had been struggling with for a long time.

The American daily Wall Street Journal (WSJ) reported on the 7th (local time) that OpenAI has released new solutions to over 300 mathematical problems across various fields, including algebra and mathematical logic, to the public.

Among these, progress is made on three out of five ‘Millennium Prize Problems’, which are representative difficult problems in the mathematics world, including proof of the quasi-Riemann hypothesis.

Alex Kontorovich, a professor at Rutgers University, commented on the proof of the Jun-Rimann hypothesis, saying, "If a human had accomplished this, he would have immediately received the Fields Medal, without a doubt."

The mathematics world is shaken by the challenge of AI. As rumors circulated earlier this week that OpenAI would produce a large number of mathematical proofs, mathematicians rushed to publish their research results faster.

So far, academia has taken a negative stance on AI's difficult challenges.

Last month, when OpenAI revealed that it had solved the 'Navier-Stokes problem' for which a prize of $1 million was at stake, 25 Fields Medal winners wrote an open letter, warning that AI's solution to mathematics risks damaging the original purpose of the discipline of mathematics.

The Princeton Institute for Advanced Study pointed out that “human understanding of mathematics is the most important,” and that “AI is now presenting mathematical arguments in situations where even the humans who entered the prompt cannot understand or verify it.”

Another problem was considered to be learning from someone else's existing research without permission or using one's own model that has not been released to the general mathematics community to challenge difficult problems.

Accordingly, OpenAI established an independent advisory committee that included prominent mathematicians and said it would find ways to support mathematical understanding with AI.

However, it rejected the mathematics community's request to stop testing high-level math problems with its own AI models, explaining that it was "critical to accelerating the development of tools that will advance the field."