Books that have been the most influential on my thinking.

The reference book, on deep learning, this book was particularly useful to me since I had Aaron as a teacher who I could ask for insight on topics I wanted to explore further. And all topics are still as relevant today as when the book was published.

The reference book about reinforcement learning, I particularly liked the third part of the book “Looking deeper” about the psychology and neuroscience of reinforcement learning. Really insightful.

This book made me definitively understand the free energy principle and convinced me that it was the best current theory of cognition that we have. Probably the most important book of all those on my list, it really puts all the puzzle pieces together, they manage to explain a complex framework like active inference in a very clear way.

Dynamical systems are widely used in computational neuroscience. Therefore, in order to read the literature, one must have a good understanding of the mathematics behind nonlinear dynamical systems. This book is particularly effective in this regard!

As a computer science and artificial intelligence student wanting to work on brain-computer interfaces, I had to study neuroscience on my own. This book, which serves as a textbook for Stanford's biology/neuroscience program, is a fantastic resource that traces the history and biological concepts of modern neuroscience in a very educational way.

A short and easy-to-read essay by Jeff Hawkins on his theory of how the brain works, especially the neocortex. Even for someone who doesn't fully subscribe to his theory, it is a very interesting and well-written book.

Probabilistic graphical models are essential to understand the literature about inference, generative models, and a bunch of other topics, even in neuroscience active inference massively uses probabilistic graphical models and this is a very good resource to learn them.

Half autobiographical book on the state of the art and history of artificial intelligence by Yann Lecun, unfortunately only available in French at the time of writing, this book is very insightful on the field, research and the future of AI and Yann Lecun is one of my most significant influence overall.

Very good introduction to bayesian modeling of cognition, you can feel that the authors have put a lot of work into making it as pedagogical as possible. The illustrations, the exercises, and the way the book is written are perfect. I just wish the book had talked a little more about how Bayesian computation might be implemented in neural circuits.

A book about the psychology and neuroscience of what we understand today about the concept of "intelligence". It is a book that forces a paradigm shift and an awareness of the importance of areas such as AI or BCI in the final chapters.