Applying AI to Quantum Field Theory

In 2019, Giuseppe Carleo and his team pioneered the application of machine learning to quantum physics. This book is an introduction to how AI revolutionizes quantum field theory (QFT), from scalar fields to complex gauge theories describing quarks and gluons.

The narrative unfolds in three acts. First, readers discover the mathematical kinship between neural networks and quantum fields—the renormalization group maps onto information flow through neural layers, while gauge symmetry provides blueprints for AI architectures. Through Python code, readers build networks that discover phase transitions without being taught physics, demonstrating AI's ability to rediscover fundamental principles from data alone.

The second act examines how AI addresses each type of quantum field. For scalar fields, neural networks identify exotic phases that traditional methods miss. For fermions, architectures like FermiNet achieve chemical accuracy while sidestepping computational barriers. For gauge fields, flow-based models conquer critical slowing down that has limited simulations for decades. Key breakthroughs include MIT's gauge-equivariant flows, which reduce autocorrelation times by a factor of 100, DeepMind's solution to 30-electron molecules, and the discovery by transformers that million-term scattering amplitudes can be expressed as a single equation.

The final act envisions AI not just calculating but creating physics systems like MELVIN, designing quantum experiments that no human has imagined. Language models solve bootstrap equations. Neural networks propose routes to grand unification. The book culminates in a convergence of quantum computers and classical AI—a partnership that could crack QFT's deepest mysteries.

By teaching AI nature's symmetries, we're creating systems that reveal patterns invisible to human analysis—AI intelligence is offering a different way of interrogating reality.

Written as an introduction for physicists curious about AI and ML, as well as for AI and ML experts interested in fundamental physics, the book strikes a balance between rigor and practical implementation, offering both conceptual frameworks and tools for the quantum field theory revolution.