Physicist credits GPT-6 Astra with leading proof work on Standard Model proposal
Nathaniel Craig’s new account explains the AI’s role in an October preprint tackling a longstanding physics problem. The result remains a candidate construction.
Loading page…
Nathaniel Craig’s new account explains the AI’s role in an October preprint tackling a longstanding physics problem. The result remains a candidate construction.
Listen to this story
Craig's October 6 arXiv preprint proposes a mathematical construction for defining the Standard Model on a four-dimensional spacetime lattice, and he credits GPT-6 Astra with developing the detailed proof from a strategy he had not been able to establish across all required cases. The proposal tackles a longstanding obstacle in representing chiral particles without unwanted mirror partners, but it remains a candidate definition—not an accepted resolution or an experimental result. The work is a notable example of AI contributing to proof-heavy theoretical research; the construction's validity remains to be tested by scrutiny.
Straightforward lattice rules introduce unwanted right-handed partners; overlap fermions address that doubling but leave a phase-setting problem.
Craig says the corresponding problem for the full Standard Model remained open for more than 25 years after Lüscher's prescription for simpler theories.
The construction is scoped to four-dimensional Euclidean lattices and gauge fields meeting a condition the paper calls admissible.
Physicist Nathaniel Craig says GPT-6 Astra helped turn his strategy for a longstanding particle-physics problem into a concrete construction—and led the detailed proof. In a November 15 guest post on OpenAI’s science blog, he describes their work toward defining the Standard Model on a spacetime grid. The October preprint presents a candidate, not an established resolution.
Craig leads the particle theory group at the University of California, Santa Barbara, and is a visiting scientist at OpenAI. His account describes a strategy he already had, but had not been able to prove worked across every required case.
The research itself predates the new explanation. Craig submitted A gauge-invariant measure for lattice chiral fermions to arXiv on October 6. It proposes a mathematical construction for a class of theories that includes the Standard Model, rather than a new experimental result.
Astra helped turn that strategy into a construction and took the lead in developing the detailed proof.
Nathaniel Craig, writing on OpenAI’s science blog
The Standard Model describes known elementary particles and three fundamental forces: the strong force, the weak force and electromagnetism. Its successful predictions often rely on perturbation theory. Physicists begin with a simpler problem without interactions, then restore those interactions through a series of small corrections.
That approach cannot answer every question. A proton’s mass comes mainly from the motion and interactions of its quarks and gluons, the particles carrying the strong force. Turn that force off and the proton falls apart. Treating it as a small correction is therefore not a useful starting point.
A lattice replaces continuous spacetime with a grid whose spacing sets the shortest distance the theory can resolve. On a finite grid, physicists can define interactions without treating them as small corrections—the meaning of nonperturbative—and calculate predictions on computers. Lattice methods have already helped explain the proton’s mass from first principles.
Extending that success to the whole Standard Model runs into a problem of handedness. The weak force treats left- and right-handed particles differently. The theory is therefore chiral: like a single glove, it does not look the same under a mirror reflection.
Straightforward lattice rules introduce an unwanted right-handed partner when physicists put a left-handed particle on the grid. This doubling produces the wrong particle content for the weak force. Earlier work developed overlap fermions, a way to represent the particles while removing those unwanted partners.
But removing the extra particles leaves another choice: setting a mathematical phase that depends on the force fields across the grid. Martin Lüscher established a prescription for simpler theories. According to Craig, the corresponding problem for the full Standard Model had remained open for more than 25 years.
The preprint compares mathematical descriptions of left-handed particles at successive lattice spacings. Its abstract says this produces the properties needed to construct a smooth, gauge-invariant fermion measure—the mathematical ingredient at the center of the proposal. The stated scope is four-dimensional Euclidean lattices, with a condition the paper calls admissible gauge fields.
Craig’s arXiv submission also lists a mathematical companion, a Lean formalization and supplementary calculation records. Those materials accompany a result the paper explicitly calls a candidate nonperturbative lattice definition of the Standard Model.
Loading discussion...
Join the conversation
Explain how you would weigh the original idea against developing the proof.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.