Physicist Nathaniel Craig of the University of California, Santa Barbara, a visiting scientist at OpenAI, has proposed a candidate way to define the Standard Model of particle physics on a lattice, a grid of points standing in for continuous spacetime. The work, described in an OpenAI blog post and detailed in a paper prepared with the aid of OpenAI's GPT-6 Astra model, may answer a question open for more than fifty years.
The Standard Model describes the known elementary particles and three of the four fundamental forces. Yet physicists still lack a nonperturbative definition of the full theory, meaning one that does not treat its interactions as small corrections to a simpler problem. Lattice field theory, which slices spacetime into a finite grid so the theory can be defined rigorously and simulated on computers, has worked well for the strong force. The weak force, however, is chiral, meaning it treats left-handed and right-handed particles differently. Putting such theories on a lattice tends to produce unwanted extra particles, called doublers.
Getting the phase right
Earlier work introduced overlap fermions, a lattice construction that keeps only fermions of the desired handedness and removes the doublers. But one piece was missing for more than 25 years: how to fix the phase, a wave-like quantity attached to each fermion, consistently as the force fields vary across the grid. Craig's idea was to represent the same force fields on a succession of finer grids and compare the allowed fermion states on each pair of grids. The differences supply small phase corrections that add up to a rule on the original grid. Astra suggested how to carry fields onto finer grids, showed that the comparisons converge to a definite answer, and took the lead in developing the detailed proof. Outside lattice theorists reviewed the proof and asked questions.
The result is a candidate nonperturbative lattice definition of the Standard Model, and Craig stresses it is a beginning rather than an end. Experts must still judge whether the prescription has all the required properties, and much work remains before it can support realistic simulations. The paper, titled A gauge-invariant measure for lattice chiral fermions, was posted on arXiv in October and is a striking example of an AI model contributing directly to frontier theoretical physics.