I like to write books that I would have wanted to read myself as a student. I really wish I had been able to read this book. It would probably have changed my life and my career, as I would have insisted on doing my PhD on this topic. Alas, when I was a student in the 1990s, the topic of this book was not something a reasonable PhD student would have worked on. There was hardly any literature on the topic; it would have been considered uncharted territory, if not suspicious. I hope the world has changed, and that this book may contribute to that change. With the huge increase in research on AI and computational neuroscience on the one hand, and affective neuroscience and mindfulness meditation on the other, I think the time is ripe to attempt a synthesis, which is the motivation for this book.
What I should emphasize is that this book is about a scientific theory, or rather, several scientific theories. It is not a book that teaches meditation; it has little to do with self-help and certainly constitutes no clinical guidance. Nor is it really a philosophical book in the sense that the word would be used in academic circles: while there is some philosophical speculation, the main paradigm is that of the natural sciences. It may be surprising that I seem to include artificial intelligence in the natural sciences, but here it is largely used as a computational model of the brain, even if sometimes on a very abstract level. The strong neuroscience component of this book further connects it to empirical science.
I have tried to write the book so that it is suitable for as wide an audience as possible. I believe anybody trained in computer science or neuroscience should be able to understand it. Scientific training in any discipline might be enough to understand the main ideas, and I hope that even members of the general public might find something interesting in it. Although not primarily intended as such, the book can also be used as a university-level textbook for advanced undergraduates or graduate students in computer science or cognitive science; it should also be suitable for computationally minded students in neuroscience or psychology.
This book was written while working in different institutions. Most of the work was done while a faculty member at the University of Helsinki (Department of Computer Science). Part of the writing was accomplished while a faculty member at University College London (Gatsby Computational Neuroscience Unit) as well as a research scientist at Université Paris-Saclay (DataIA Institute and Inria–Saclay-Ile-de-France, supported by grant ANR-17-CONV-0003). The work was further supported by a Fellowship from CIFAR (Learning in Machines & Brains Program).
Finally, I’m very grateful to Moritz Grosse-Wentrup, Riitta Hari, Marianne Maertens, John Millar, Tiina Parviainen, Jonne Viljanen, and, especially, Michael Gutmann, for most helpful comments on the manuscript.