The Information Machine
FollowingDay 3in the edition since 16 Nov 20264 sourcesupdated 17 Nov 2026

GPT-6 Astra's Standard Model lattice proof

The gist

Latest: Craig and GPT-6 Astra Produce Candidate Standard Model Lattice Definition

Defining the Standard Model nonperturbatively on a lattice is a long-standing challenge in theoretical physics. The work demonstrates AI contributing substantively to the proof process in a foundational physics problem, though the result remains a candidate with open technical questions.

When stories like this move, I write about them in an email each morning.

One free email each morning, around 8 AM US Eastern.

The full picture

Physicist Nathaniel Craig of the Kavli Institute for Theoretical Physics collaborated with OpenAI's GPT-6 Astra to produce a candidate nonperturbative lattice definition of the Standard Model of particle physics, described in a preprint on arXiv. The preprint proposes defining the Standard Model on a discrete spacetime grid at finite lattice spacing and under a bound on field strengths. GPT-6 Astra's contributions included suggesting how to carry force fields onto finer grids, demonstrating convergence of comparisons between grids, and leading the detailed proof. The manuscript's acknowledgements describe it as combining 'author-written sections with extensive rewrites of the Astra-composed proof.' Craig characterized the result as 'far more an answer in principle than an answer in practice,' and noted that whether the formulation satisfies reflection positivity, a condition tying the lattice model to a probability-conserving quantum theory, remains under test. OpenAI published this work as the inaugural post on a new science blog.

How it developed
17 November 2026

An October 26 OpenAI announcement, separate from the Standard Model lattice work by Nathaniel Craig and GPT-6 Astra, claimed AI-produced proofs for the Kakeya conjecture in all dimensions, a refutation of the Exponential Time Hypothesis, and an existence proof for QCD and pure Yang-Mills theory.

Two of the three results came with public Lean formal proofs, while the physics result remained under expert review.

First citedThe Neuron
16 November 2026

Nathaniel Craig of the Kavli Institute for Theoretical Physics published an arXiv preprint on November 15-16 proposing a candidate lattice definition of the Standard Model, with GPT-6 Astra leading the mathematical proof.

The manuscript's acknowledgements state it 'combines author-written sections with extensive rewrites of the Astra-composed proof,' and OpenAI launched a science blog with Craig's work as its inaugural post. Craig characterized the result as 'far more an answer in principle than an answer in practice,' and whether the formulation satisfies reflection positivity remains unresolved.

First in the edition · 16 November 2026
15 November 2026

OpenAI publishes blog post describing GPT-6 Astra's role in developing a candidate lattice definition of the Standard Model

26 October 2026

OpenAI announces AI-produced proofs for Kakeya conjecture in all dimensions, ETH refutation, and Yang-Mills existence proof

Sources

Get the daily email

When stories like this move, I write about them in an email each morning.

One free email each morning, around 8 AM US Eastern.

See the latest edition · Privacy