OpenAI says GPT-6 Astra led the proof behind a new candidate lattice definition of the Standard Model
UC Santa Barbara physicist Nathaniel Craig says GPT-6 Astra turned his idea into a construction and led the detailed proof of a candidate lattice definition of the Standard Model. The preprint is on arXiv; Craig calls it a beginning that still needs expert scrutiny.
OpenAI says its new model, GPT-6 Astra, helped a physicist build a detailed mathematical proof toward a problem that has stumped physics for more than 50 years. On Nov. 15, OpenAI published the first guest post on its new science blog, written by Nathaniel Craig, who leads the particle theory group at the University of California, Santa Barbara, and is a visiting scientist at OpenAI.
Some background. The Standard Model is the theory physicists use to describe the fundamental particles and three of nature's four forces (electromagnetism, the strong force and the weak force). Its predictions match experiments across eighteen orders of magnitude, from the faint 21-centimeter radio glow of hydrogen to collisions at the Large Hadron Collider. But it has a long-standing gap: it can only be calculated by treating interactions as small corrections added piece by piece. No one has a rigorous, complete definition that doesn't lean on that approximation. One route is to chop spacetime into a grid, called a lattice, and define the whole theory there so computers can simulate it exactly. The catch is that one class of Standard Model particles, chiral fermions (roughly, particles whose left-handed and right-handed versions feel forces differently), has been extremely hard to put on a lattice.
What happened? Craig posted a preprint on arXiv on Oct. 6, "A gauge-invariant measure for lattice chiral fermions," running 60 pages plus 63 of supplementary material, with a mathematical companion and a verification report written in Lean, software that lets a computer check a proof step by step. It proposes a "candidate" lattice definition of the Standard Model. In the OpenAI post Craig says he had an idea, which was to carry the force fields onto a series of ever-finer grids and compare particle states on neighboring grids, but he couldn't prove it worked in every required case. He says Astra turned the idea into a working construction, showed the comparisons converge to a definite answer, and "took the lead in developing the detailed proof." Three lattice field theorists, Maarten Golterman, Andreas Kronfeld and Yigal Shamir, looked over the proof and asked probing questions.
Don't read this as "problem solved." Craig himself calls it a candidate definition, "much more a beginning than an end" and "an answer in principle rather than in practice." Much work remains before it can be used in realistic simulations, and he is now checking whether the construction satisfies reflection positivity at finite lattice spacing, a condition tied to probabilities adding up in a quantum theory. Whether other physicists accept it will depend on peer review and independent checks.
Zoom out. OpenAI's blog notes that in September an internal system produced a proof on the existence and smoothness of solutions to the Navier–Stokes equations (which describe fluid flow), and in October the company released a batch of mathematical results. Those claims largely come from OpenAI itself, and outside verification comes later.
What does it mean for regular people? A complete definition of the Standard Model won't change your life tomorrow, but it is groundwork for precisely calculating the basic properties of matter. The nearer change is in how research gets done: a proof that once took one person years to grind out may now have its details driven by an AI while human experts judge whether it holds. That could speed up discovery, and it makes "how do we check a proof written by an AI?" a new question.
Share this article
What do you think of this story?
Anonymous feedback, up to 5 reactions per article.
Reader comments
Comment without signing in, or sign in to display your member name.
Discuss the technology and the story. Hate, politics and war, threats, abuse, personal accusations, sexual content, ads, links and personal information are not allowed.
No comments yet. Be the first to share.