Physicist Nathaniel Craig proposed a candidate nonperturbative definition of the Standard Model on a lattice, and writes that GPT-6 Astra “took the lead in developing the detailed proof” (OpenAI). His account, published Nov 15, 2026, is the first guest post on OpenAI’s new science blog (OpenAI), which OpenAI started as a place for scientists to share what they discover with AI (OpenAI).
- years the Standard Model has gone without a full nonperturbative definition
- 50+
- years the phase rule for the full Standard Model stayed open after Lüscher’s work
- 25+
- outside lattice field theorists looked over the proof
- 3
What happened
Physicists usually calculate with the Standard Model by switching its forces off and adding them back as small corrections. Some things, such as the mass of the proton, can’t be reached that way. Putting spacetime on a grid, a lattice, gives a rigorous definition at a given grid spacing and works for the strong force. The weak force treats left- and right-handed particles differently, and on a lattice that produces unwanted partner particles and leaves a phase that has to be set correctly (OpenAI).
Craig, a visiting scientist at OpenAI who leads the particle theory group at UC Santa Barbara, had a strategy he could not prove in every case: carry the force fields onto finer grids, compare the allowed fermion states on neighboring grids, and add up the differences for the phase rule. He writes that Astra suggested how to carry the fields onto finer grids, showed the comparisons converge, and led the detailed proof. Lattice field theorists Maarten Golterman, Andreas Kronfeld and Yigal Shamir looked over the proof and asked questions (OpenAI). The preprint, “A gauge-invariant measure for lattice chiral fermions”, was posted on arXiv on Oct 6, 2026, with no journal reference listed.
Why it matters
Craig calls the result a candidate that is “far more an answer in principle than an answer in practice” and says experts still need to decide whether it has the properties physicists want (OpenAI). He writes that he is now thinking about whether it satisfies reflection positivity at finite lattice spacing, and the construction leaves open how to define the theory in continuous spacetime. The post doesn’t say whether the paper has been submitted for peer review.




