{"version":1,"type":"story","url":"https://digestai.news/story/openai-and-anthropic-disprove-27-year-old-computer-science-hypothesis","json":"https://digestai.news/story/openai-and-anthropic-disprove-27-year-old-computer-science-hypothesis.json","markdown":"https://digestai.news/story/openai-and-anthropic-disprove-27-year-old-computer-science-hypothesis.md","slug":"openai-and-anthropic-disprove-27-year-old-computer-science-hypothesis","headline":"OpenAI and Anthropic disprove 27-year-old computer science hypothesis","summary":"OpenAI and Anthropic announced this week that AI-generated proofs have disproved the **exponential time hypothesis (ETH)**, a long-standing assumption in theoretical computer science. The claim, if verified, could force major revisions to decades of research in complexity theory and algorithm design, according to the companies’ published work on GitHub in both natural language and Lean format for machine verification.\n\nPeter Jonsson, a professor at Linköping University, called much of his own work ‘dead’ after the hypothesis’s collapse, though he noted the discovery still advances scientific understanding. UC Santa Barbara and MIT researchers involved in the original hypothesis and its prior refutations have been consulted. The field remains cautious: independent verification is pending, and experts like Daniel Lokshtanov and Thore Husfeldt warn that roughly 50 of their papers could become obsolete or require urgent updates. Jonsson’s own presentation at the upcoming IEEE Symposium on Foundations of Computer Science may face immediate backlash from peers.","keyPoints":["OpenAI and Anthropic published AI-generated proofs disproving the 27-year-old exponential time hypothesis (ETH)","ETH underpins decades of research in complexity theory and algorithm design, per the companies’ Lean-format proofs","Independent verification is pending, but experts warn 50+ papers may need revision if the claim holds"],"whyItMatters":"A verified disproof of ETH would force a rewrite of foundational assumptions in computer science, affecting algorithm design and theoretical research globally.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":["OpenAI","Anthropic"],"models":[],"people":["Peter Jonsson","Daniel Lokshtanov","Thore Husfeldt"]},"firstPublishedAt":"2026-10-29T06:53:00Z","updatedAt":"2026-10-29T06:53:00Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"el-balad.com","title":"OpenAI and Anthropic Challenge 27-Year Hypothesis","url":"https://el-balad.com/17062026","publishedAt":"2026-10-29T06:53:00Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":{"title":"AI Models Refute Exponential Time Hypothesis","url":"https://digestai.news/thread/openai-model-allegedly-disproves-exponential-time-hypothesis","storyCount":2},"cite":{"text":"Digest AI, \"OpenAI and Anthropic disprove 27-year-old computer science hypothesis\", 29 October 2026, https://digestai.news/story/openai-and-anthropic-disprove-27-year-old-computer-science-hypothesis","publisher":"Digest AI","title":"OpenAI and Anthropic disprove 27-year-old computer science hypothesis","datePublished":"2026-10-29T06:53:00Z","url":"https://digestai.news/story/openai-and-anthropic-disprove-27-year-old-computer-science-hypothesis"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}