▲ AI OpenAI ChatGPT
The mathematical community has been thrown into shock as artificial intelligence (AI) has solved hundreds of problems in one go, including difficult challenges that mathematicians have wrestled with for a long time.
The U.S. daily The Wall Street Journal (WSJ) reported on the 7th (local time) that OpenAI has released new solutions to about 300 math problems spanning various fields such as algebra and mathematical logic.
Among them are advanced contents for three of the five "Millennium Prize Problems," which are representative mathematical challenges, including the proof of the quasi-Riemann hypothesis.
Alex Kontorovich, a distinguished professor at Rutgers University, commented on the proof of the quasi-Riemann hypothesis, saying, "If a human had accomplished this, they would have unquestionably and immediately received the Fields Medal."
The mathematical community is shaken by the AI's challenge.
Earlier this week, when rumors circulated that OpenAI would flood the market with mathematical proofs, mathematicians rushed to publish their own research results ahead of time to beat the release.
Academia has previously expressed a negative stance toward AI's attempts at mathematical challenges.
Last month, when OpenAI revealed that it had solved the Navier-Stokes equations problem—which carried a 1 million dollar prize—25 Fields Medalists issued an open letter warning that AI's math problem-solving risks undermining the true purpose of the discipline of mathematics.
The Institute for Advanced Study in Princeton pointed out, "Human understanding of mathematics is of the utmost importance," adding, "Now, AI is producing mathematical arguments in situations where even the human who entered the prompt cannot understand or verify them."
Other cited problems include the unauthorized learning of others' existing research and the use of proprietary models not made public to the broader mathematical community to tackle challenges.
In response, OpenAI has established an independent advisory committee including prominent mathematicians and stated that it will find ways to support mathematical understanding through AI.
However, it rejected the mathematical community's request to stop testing advanced math problems with its own AI models, explaining that it is "crucial for accelerating the development of tools to advance the field."
(Photo: Yonhap News)
※
Copying, redistribution, and unauthorized use in AI training are strictly prohibited.