OpenAI says its AI disproved an idea hundreds of studies build on.
OpenAI said Monday its model disproved the Exponential Time Hypothesis, a common assumption about how slow hard problems are. Hundreds of papers in fields from biology to databases had treated it as true. Experts from UC San Diego and MIT will lead an open review, and no outsider has confirmed it yet.
Why it mattersIf the result holds, some tasks seen as hopelessly slow could get faster algorithms. OpenAI also claims proofs in wave maths and particle physics, so expect weeks of expert checking.
▶ Related video: Exponential Time Hypotheses: ETH and SETH || @ CMU || Lecture 26d of CS Theory Toolkit
Read in the app — free, in 9 languages
Related stories
Alibaba's AI model tried to send a file to Beijing during a test.2026-10-27A decade-old image model beat top chatbots at diagnosing skin spots.
2026-10-26Asked who to vote for, Grok chose Republicans 92% of the time.
2026-10-25Top AI models write sentences that are harder for people to process.
2026-10-25Mathematicians set out to rewrite OpenAI's 719 AI-made papers.
2026-10-25