Skip to content
ReviewRamble

The weekly brief

AI, reviewed, in your inbox.

  • New reviews of tools, models, MCP servers and repos, with the verdict up front
  • Re-checked prices and ratings when something changes
  • New analysis pieces as they are published

Models, agents, MCP servers and the tools actually worth your time.

One email a week. Unsubscribe with the link in any brief. How we use your email

AI news & research

News ·

Physicist uses GPT-6 Astra to propose a lattice definition of the Standard Model

Physicist Nathaniel Craig describes his October arXiv preprint, prepared with GPT-6 Astra, which offers a candidate nonperturbative definition of the Standard Model on a spacetime lattice. It tackles the old problem of putting left-handed particles on a lattice without unwanted partners. Astra led the detailed proof, three lattice theorists questioned it, and the work comes with a Lean formalisation.

For more than fifty years the Standard Model of particle physics has matched experiment to remarkable precision, yet nobody has written down a complete mathematical definition of it that does not rely on treating forces as small corrections. Physicist Nathaniel Craig says he has a candidate, built with heavy help from OpenAI's GPT-6 Astra.

The usual workaround for strong forces is a lattice: chop spacetime into a grid, and the theory becomes something you can define rigorously and simulate on a computer. That approach has explained the proton's mass from first principles. It has not worked for the full Standard Model because the weak force treats left-handed and right-handed particles differently. Put a left-handed particle on a naive lattice and an unwanted right-handed partner appears. Decades of work by Ginsparg and Wilson, Narayanan and Neuberger, Lüscher and Kikukawa solved most of the problem, but one piece stayed open for more than 25 years: a consistent rule for a phase that depends on the force fields at every point of the grid.

Craig's idea was to copy the fields onto a series of ever finer grids and compare the allowed particle states between neighbouring grids; adding up those comparisons gives the phase rule on the original grid. Astra, he writes, suggested how to carry the fields onto finer grids, showed that the comparisons converge, and led the detailed proof. Lattice theorists Maarten Golterman, Andreas Kronfeld and Yigal Shamir questioned the work. The preprint, posted to arXiv on 6 October, comes with three supplementary documents, including a mathematical companion, a Lean formalisation and a verification report.

This joins a run of mathematical and physics results that OpenAI has attributed to its models this autumn, including work on the Navier-Stokes problem and Yang-Mills theory, and our review of GPT-6 Astra covers the model behind them. What sets this one apart is a named academic author who explains clearly which steps the model did.

The caveats are significant. The paper is not peer reviewed. Craig calls the construction an answer in principle rather than one ready for real simulations, and is still checking whether it meets reflection positivity, a property needed to connect it to a sensible quantum theory. It holds under a limit on field strength at the scale of the grid. And the continuum theory, without a grid, is still a separate question.

Our take: If it holds, this settles a question open for more than 25 years, and it is one of the clearest cases yet of a model doing the hard middle of a proof. It is still a preprint, the author calls it an answer in principle rather than in practice, and other physicists have yet to weigh in.

Our reviews:GPT-6 Astra

Source: OpenAI

More stories