Though AI has made impressive advances, it still falls well short of replicating human cognition. This book connects neuroscience and AI by exploring the billion-year evolution of the human brain. It identifies Five Breakthroughs marking leaps in human intelligence, which provide a map for navigating our technological future.

I LOVED this book. It does more to unravel the mysteries of the human mind than anything I've read before. The narrative is interesting and easy to read. If you have any interest at all in how intelligence and learning have evolved over the past billion years, this book is a must-read.

Thanks, NetGalley, for the ARC I received. This is my honest and voluntary review.

A Brief History of Intelligence is a "big history" narrative following the evolution of human intelligence. Penned by Max Bennett, an entrepreneur in artificial intelligence, the book has two parallel tracks. 1) Describe the five major breakthroughs in the evolution of human intelligences and 2) compare those breakthroughs with the current state of knowledge and history of artificial intelligence. I imagine part of the exigence of this book was for Bennett to collect and organize knowledge on human intelligence in order to leverage this knowledge to advance his work in the field of AI. Because of this, the earlier portion of the narrative is more robust than the later portion. And I imagine he has kept some of his more tantalizing ideas out of the book.

Although this work lacks the verve and panache of other sweeping book that it'd like to be compared with (Sapiens, Behave, etc), Bennett delivers an eminently digestible story on what in reality is an unwieldy subject. He does this without unwarranted oversimplification or wanton speculation (though it may have made for a more exciting book, if he had). Bennett's strength as a writer appears to be organization and distillation. He drills down to essential insights on the evolutionary neurobiology, draws out the AI takeaways and comparisons, recaps and then moves on. It's yeoman science communication. For those interested in this subject, Bennett has saved readers a great deal of time by assembling this book.

When it comes to the five evolutionary breakthroughs, Bennett's insights are not mind-bogglingly clever or original. Every hard-working undergraduate in evolutionary biology will have come to the same conclusions over their coursework and reading. However, I don't want to sell Bennett short. His framing is cleverly clear and in the context of developments in AI, very topical.

So what are these breakthroughs in neurobiology and how did intelligence result? First, the very ancient common ancestors of human, bilaterians, needed a way to navigate the environment. Thus, a nervous system capable of steering emerged. Second, these early organisms had to figure out where they could obtain resources and how they could avoid predation. So another adaption emerged, reinforcement learning. The nervous system encoded valence into stimuli from the environment (food = good, predatory = bad, etc). As the evolutionary arms war between predator and prey heated up over time, the brains of ancestral species developed a way of simulating events: If I turn this way, I might find food but if I turn the other way, I may find a mat. This sort of thing. At this level, primitive brains were beginning to model the world around in order to learn from it. Once modeling the world took off, brains started to model themselves. The striking thing is there is good evidence for this recursion occurring well before humans. And then the final breakthrough and probably the most obvious one is language. The ability to symbolically label and standardize inter-individual exchange with grammar enable the accumulation and distribution of knowledge (culture).

There are still a lot of questions that both Bennett and science have left unanswered but this broad framework is sense-making in productive ways. I think a lot of reader can benefit from the book. Even myself, as someone with advanced training in neuroscience, genetics and the like still found a lot of the work interesting and compelling. I do with the later portion of the book was more robust, and it would have behooved the work to be a bit more rigorous citations, but overall I'm happy to recommend this one to anyone interested in the brain or artificial intelligence.

A Brief History of Intelligence: Evolution, AI, and the Five Breakthroughs That Made Our Brains by Max Solomon Bennett is an in-depth yet easy to read and understand journey of how our brains have evolved throughout time as well as our understanding of five specific breakthroughs that really unlocked an even deeper understanding of how we think, act, and understand the way our brains function. Max starts off at the very beginning explaining basic brain function, takes us to the time in the world before brains existed, where he introduces the first breakthrough. This section is the most in depth explanation of how animals have evolved and what that looked like as well as brain functions for survival, and our understanding of how the brain worked during this time frame.
Each section then builds on the next breakthrough of understanding and the research and evolution of our understanding not only of the brain, but how that has helped change our way forward in our exploration all the way up to the beginning stages of creating artificial intelligence, and the pitfalls of our understanding AI, from creating AI from going to running simulations to adaptive artifical thinking and reasoning, from beating chess, to beating the next hardest game Go, from inventing TD-Gammon to DeepMind to Watson to AlphaZero to ChatGPT to whatever we discover next, all of this based off of the studying and breakthroughs of our understanding of how the brain works to unlock our progress to the next big thing. It's a fascinating deep dive and yet so easy to understand the way that Max Bennett writes it. I highly recommend!!!
*I received a copy of this book from NetGalley. This review is my own opinion*