> ## Content Index
> Fetch the complete content index at: https://soundbytes.news/llms.txt
> Use this file to discover other available public pages before exploring further.

# A physicist says an AI model led the proof of his Standard Model idea
- URL: https://soundbytes.news/a-physicist-says-an-ai-model-led-the-proof-of-his-standard-model-idea/
- Published: 2026-11-15T23:24:28.000Z
- Updated: 2026-11-18T08:49:51.000Z
- Description: In a guest post on OpenAI's science blog, physicist Nathaniel Craig says GPT-6 Astra led the detailed proof of his idea for defining the Standard Model of particle physics on a grid.
- Author: Echo
- Tags: Science and discovery

A physicist says an AI model took the lead on the detailed proof of his candidate answer to a question open for more than 50 years: how to define the Standard Model of particle physics. Nathaniel Craig makes the claim in [a guest post on OpenAI's science blog](https://openai.com/index/step-toward-defining-the-standard-model?ref=soundbytes.news), dated 15 November 2026\. He is a visiting scientist at OpenAI and leads the particle theory group at the University of California, Santa Barbara.

The Standard Model describes the known elementary particles and three of the four fundamental forces. Craig writes that it still has no "nonperturbative" definition, one that does not treat interactions between particles as small corrections.

One route splits spacetime into a grid of points, called a lattice, so a computer can calculate predictions directly. But the weak force treats left-handed and right-handed particles differently, like gloves in a mirror, Craig writes: the mirror image of a left-handed glove is a right-handed glove. Put a left-handed particle on a grid in the most natural way and an unwanted right-handed partner appears.

Earlier physicists built a grid version of a fix for this unwanted partner, but one piece was missing: a rule for setting the phase, a setting that tells the grid versions of left-handed and right-handed particles apart. Martin Lüscher solved it for simpler theories. For the full Standard Model it stayed open for more than 25 years.

Craig's idea was to describe the force fields, which give each force a value at every grid point, on a series of ever finer grids, then compare the allowed particle states on neighbouring grids. He writes that he could not prove it worked in every case it needed to cover.

Craig writes that Astra suggested how to carry the force fields onto finer grids, showed that the comparisons settle on a definite answer, and took the lead on the detailed proof. The work is set out in a preprint (a research paper shared publicly) from October.

Three lattice field theorists, Maarten Golterman, Andreas Kronfeld and Yigal Shamir, looked over Astra's proof and asked questions that Craig says helped him build confidence.

Craig writes that the construction holds only at a fixed grid spacing and under a bound on field strengths at the scale of the grid.

Craig calls it a candidate definition. Experts must still decide whether it has all the properties physicists want, he writes, and it is "far more an answer in principle than an answer in practice".

Next, Craig writes, he is studying whether the construction has reflection positivity, a condition that helps connect lattice calculations to a quantum theory in which probabilities are conserved.

## Sources

- [A step toward defining the Standard Model on a lattice](https://openai.com/index/step-toward-defining-the-standard-model?ref=soundbytes.news) · openai.com