MIT Press makes its AI & ML book library 100% free — 9 classics to start with

thisguyknowsai · x · 2026-09-22

MIT Press has made its AI & machine learning book library 100% free online. The thread lists 9 recommended starting points:

Foundations of Machine Learning (2nd ed.) — PAC learning, VC dimension, kernels, boosting
Learning Theory from First Principles (Francis Bach) — every result derived from scratch
Deep Learning (Goodfellow/Bengio/Courville) — the canonical reference
Understanding Deep Learning (Simon Prince) — modern, covers transformers and diffusion
Algorithms for Optimization — optimizers with code and intuition
Reinforcement Learning: An Introduction (2nd ed.) (Sutton/Barto) — the RL bible
Distributional Reinforcement Learning — modeling full outcome distributions
Multi-Agent Reinforcement Learning (Albrecht) — RL meets game theory
Agents in the Long Game of AI — cognitive architectures for trustworthy hybrid AI

All are free to read online at MIT Press.

More from AGI Musings