OpenAI has announced that its AI models have made several scientific discoveries in mathematics, computer science and physics. The San Francisco-based company said on Monday that the three new discoveries concerned harmonic analysis, complexity theory and quantum field theory.
The company, which is valued at $852bn, described the findings in its announcement. It said the results came from its models working alongside researchers, and it presented them as a step toward turning raw data into scientific understanding.
AI models prove the local smoothing conjecture
Among the discoveries, the AI models proved the local smoothing conjecture, one of the central propositions of harmonic analysis, in every spatial dimension for the first time, according to OpenAI. Harmonic analysis is a branch of mathematics that studies how complex signals and functions can be broken down into elementary waves.
Researchers had previously proved the theory only in three dimensions. The new result extends that work to all spatial dimensions, the company said.
Claims on complexity theory and quantum physics
OpenAI’s announcement also said its models disproved the Exponential Time Hypothesis, which relates to the time it takes to solve complex computational problems. It added that the models provided “mathematically rigorous [proof of the] existence” of quantum chromodynamics, a field that studies the interaction of quarks and gluons.
The company said it had reached out to scientists outside the firm to help turn these outputs into established knowledge. The findings are therefore still being examined by people beyond OpenAI.
Mathematician welcomes the breakthrough but expects upheaval
Alex Iosevich, a professor of mathematics at the University of Rochester who specialises in harmonic analysis, said OpenAI appeared to have “cracked” the conjecture in a “historic development”. He said the conjecture was formulated by his PhD adviser, Chris Sogge, which made the news personal for him. “The conjecture was formulated by my PhD adviser, Chris Sogge, so it is near and dear to my heart, and the news was a shock at first,” Iosevich said.
Iosevich said he now felt joy and optimism about the field. He argued that the scientific enterprise as a whole would make large advances because of the AI revolution. He also said he expected theoretical mathematics to “change drastically” because of AI.
“I feel a twinge of sadness as a result,” Iosevich said. He added that many theoretical mathematicians he knew shared the feeling, but he believed it would soon give way to a recognition of what could now be achieved. He also noted that many problems in pure mathematics had come out of concrete physical considerations, and that these links could now be explored quickly.
OpenAI details how the findings will be released
OpenAI said it chose several different methods for releasing the findings after consulting three expert groups, which it called “shepherds”. The groups working on the mathematics and computer science problems chose to make the AI-produced data public immediately.
The physics group decided to delay its release so it could write its own interpretation of the findings, OpenAI said. The company acknowledged that every option carried trade-offs. “All of these approaches involve tradeoffs,” it said.
OpenAI said the collaborations allowed experts to help unpack the results and place them within the wider academic community. However, it also acknowledged that choosing the initial shepherds shaped who became involved first and created an imbalance of information. The company said it welcomed feedback on how to broaden participation.
Anthropic’s enzyme claim draws mixed reactions from scientists
OpenAI’s announcement is the latest in a series of claimed scientific breakthroughs credited to AI. Last month, Anthropic, the creator of Claude, announced the discovery of a new enzyme that was said to resemble CRISPR, the gene editing mechanism used in treating conditions such as cancer and sickle cell disease.
Anthropic’s announcement drew mixed reactions among scientists. Some researchers hailed the finding as an exciting development, while others suggested the company had exaggerated its significance.
