Lewis Tunstall says AI is unlikely to crack unified field theory

Hugging Face researcher Lewis Tunstall says candidate theories for unifying physics already exist, so the central problem is experimentally distinguishing among them.

He says an AI-driven breakthrough in fundamental physics seems extremely unlikely, though he would welcome being proven wrong. A first report on X names the topic as unified field theory.

A later X post by Tunstall says a thread explains why theoretical physics conjectures and theories are harder to resolve than mathematical ones. He links to arXiv:2610.16733, posted by @WKCosmo, without summarizing its argument.