Statistical mechanics for networks of real neurons

It is an old dream of the physics community to use the language of statistical physics to describe the emergent behavior of neural networks in the brain. This is at the origin of models for artificial neural networks and the resulting revolution in artificial intelligence. In the past twenty years there also has been dramatic progress in statistical physics approaches to networks of real neurons, and this has gone hand in hand with progress in experimental methods for exploring these systems. Encouraged by one of the RMP editors, and by responses to our talks at APS meetings, we decided to try and review these developments, and our paper appeared in RMP 97, 045002 (2025). This talk will give an overview of the physics, a few anecdotes about the process of writing, and some perspectives on where the field is going.