OpenAI claims its AI produced three research results in math, computer science and physics
Article timeline
The articles in this story. Times are ET.
- AI EraNewsOpenAI says its AI independently produced three new scientific findings
AIOpenAI says its internal AI model independently produced three new scientific findings, which the company calls milestones. The source text is truncated after this announcement, so the specific results, fields and verification details are not available.
- The Verge · AINewsPickResearchers criticize OpenAI's selective release of three math, computer science, and physics results
AIOpenAI announced results in mathematics, computer science, and physics, releasing each with a different strategy: the mathematics output publicly, the computer science result with an external group, and the physics result deferred to outside experts. Researchers praised the findings as major, but many criticized OpenAI for handpicking experts before wider scrutiny, and an advisory group urged immediate public release of the physics outputs.
- The DecoderNewsPickOpenAI asks outside experts to check AI-produced math, computer science, and physics proofs
AIOpenAI says its AI produced proofs in math, computer science, and physics, and asked outside experts to act as shepherds to check them. The computer science result claims the Exponential Time Hypothesis is false, which would undermine hundreds of papers built on it. OpenAI also translated the math and computer science proofs into Lean, though the article notes Lean only checks statements as written, not whether they capture the intended problem.
roonXroon says fighting an AI intelligence explosion keeps losing to new capabilitiesAIroon, an OpenAI researcher, writes that control over a runaway intelligence explosion will repeatedly reveal surprising new capabilities from mere bits and atoms, and that humans will keep losing. The post quotes danintheory's announcement of three discoveries in mathematics, computer science, and physics, linked to an OpenAI post on building scientific understanding with academic communities.
Yuchen JinXYuchen Jin lists AI math milestones and jokes OpenAI proved ETH is falseAIYuchen Jin posts a timeline of AI math achievements, from GPT-4 nearly solving grade-school math in 2023 to an IMO gold medal in 2025. He says that today OpenAI proved the Exponential Time Hypothesis is false, and jokes that P = NP may come next.
Amjad MasadXAmjad Masad jokes that OpenAI got three different answers on releasing discoveriesAIAmjad Masad posts that OpenAI asked mathematicians, computer scientists, and physicists how to release its discoveries and got three different answers. Mathematicians said to just post the work, computer scientists said to open source it while paying for their conference, and physicists said to leave it to them.
- Gizmodo · AINewsOpenAI claims three more math and physics results and outlines a new publishing plan
AIOpenAI says its internal AI models produced three new results: a disproof of the Exponential Time Hypothesis, a proof of the local smoothing conjecture tied to the Kakeya conjecture, and a claimed proof for certain interacting quantum field theories in four dimensions.
- The Verge · AINewsOpenAI announces three science discoveries and a new way to work with academics
AIOpenAI says its models produced three major results in mathematics, computer science, and physics, less than a month after an earlier batch of manuscripts. The company says the computer science result shows the Exponential Time Hypothesis is false, and external groups led by Russell Impagliazzo, Ramamohan Paturi, and Ryan Williams will verify it. The physics work, led by David Gross, has not been released in full.
Rohan PaulXPickOpenAI says its AI proved three long-standing research results in math, computer science and physicsAIRohan Paul relays OpenAI's claim that its AI produced proofs in math, computer science and physics. The claims include a disproof of the Exponential Time Hypothesis, a proof of the local smoothing conjecture that settles the Kakeya conjecture, and a Yang-Mills proof covering half of a Millennium Prize problem. The post presents these as the AI's claims, not established results.
Lucas Beyer (bl16)XLucas Beyer says an OpenAI-linked math result is finally workingAILucas Beyer says "can't believe it's actually working" without giving details. Andrew Curran's quoted post says OpenAI announced three new results in mathematics, computer science, and physics, including a claim that the Exponential Time Hypothesis is false.
Joshua AchiamXOpenAI announces three AI-assisted scientific discoveries in math, computer science and physicsAIOpenAI, quoted through Greg Brockman, announces three major scientific discoveries, one each in mathematics (harmonic analysis), computer science (complexity theory), and physics (quantum field theory). Brockman says AI is starting to help improve understanding of the universe, linking to an OpenAI post on building scientific understanding with academic communities. The author, Joshua Achiam, adds only the caption "Things are getting incredibly weird."
Greg BrockmanXGreg Brockman announces three AI-assisted scientific discoveries in mathematics, computer science, and physicsAIGreg Brockman says OpenAI is announcing three major scientific discoveries, one each in harmonic analysis, complexity theory, and quantum field theory. He says AI is beginning to help improve understanding of the universe and links to an OpenAI page on building scientific understanding with academic communities.
Andrew CurranXPickOpenAI announces three AI-generated discoveries in math, computer science, and physicsAIOpenAI announces three new discoveries in harmonic analysis, complexity theory, and quantum field theory. It says the manuscripts are best understood as raw research data, so it asked external experts to organize groups that explain the results. The experts advised a different release strategy for each field: public AI output for math, an external collaborator group for computer science, and deferring to experts for physics.
