36 - Adam Shai and Paul Riechers on Computational Mechanics

Summary

In this episode, host interviews Adam Shai and Paul Riechers about applying computational mechanics—a physics subfield for predicting random processes—to understand and scale transformer models. They explain how computational mechanics differs from other approaches, describe the fractal geometry of belief‑state representations, and explore its implications for AI safety and world‑model claims. The discussion blends technical insight with practical research directions, highlighting their recent work and future avenues in the field.