A Thousand Brains: A New Theory of Intelligence

by Jeff Hawkins

Richard Dawkins (Foreword)

For all we hear of neuroscience's great advances, the field has generated more questions than answers. We know that the brain combines sensory input from all over your body into a single perception, but not how. We think brains "compute" in some sense, but we can't say what those computations are. We believe that the brain is organized as a hierarchy, with different pieces all working collaboratively to make a single model of the world. But we can explain neither how those pieces are differentiated, nor how they collaborate.

Neuroscientist and computer engineer Jeff Hawkins argues that it's so hard to answer questions about the brain because our basic picture of how the brain works is wrong. In A Thousand Brains, Hawkins takes a radically new approach to the brain, with stunning implications.

Hawkins' proposal, called the Thousand Brains Theory of Intelligence, is that your brain is organized into thousands upon thousands of individually computing units, called cortical columns. These columns all process information from the outside world in the same way, and each builds a complete model of the world. But because every column has different connections to the rest of the body, each has a unique frame of reference. Your brain sorts out all those models by conducting a vote. The fundamental job of the brain, therefore, is not to build a single thought, but to manage the thousands of individual thoughts it has every moment.

With this powerful new framework, Hawkins is able to reassess some of neuroscience's most stubborn problems, like why pain needs to be painful to be useful, how we can understand that our perspective of a thing changes as we move around it, and why we might be conscious but individual pieces of our body aren't.

And once you understand how the brain works, it is a lot easier to make one yourself. Hawkins is, above all, an engineer, and A Thousand Brains outlines how a new understanding of intelligence could lead to truly intelligent AI. Hawkins explores how we might create machines that can learn on their own, why we need not fear superintelligent systems, and how human and machine intelligence may someday merge.

Combining cutting-edge theoretical neuroscience with an ambitious program for tomorrow's digital minds, A Thousand Brains heralds a revolution in the study of intelligence. It is a big-think book, in every sense of the word.

• ISBN10: 1541675819
• ISBN13: 9781541675810
• Publish Date: 18 March 2021
• Publish Status: Active
• Publish Country: US
• Imprint: Basic Books
• Format: Hardcover
• Pages: 288
• Language: English

Reviews

[Jeff Sexton]

Amazing Discussion Marred By Myopia In Its Final Act. This book, by the guy that created the Palm Pilot (who has since turned to study neuroscience, which he had wanted to do from the beginning apparently), describes the intriguing new theory of how the brain works that he and his team have crafted very well. Hawkins does a truly excellent job of making the advanced theoretical neuroscience he works with approachable by all, from those who have barely ever heard of the word "neuroscience" to his colleagues and competitors in the field. In discussing the neuroscience leading up to the "thousand brain" concept and in discussing how the "thousand brain" idea directly impacts computing and artificial intelligence, Hawkins is truly amazing. The perils come in the third act, when Hawkins begins to apply the theory and what he believes it could mean directly to humans. Here, he begins to sound both Transhumanist and Randian in his claims of absolute certitude that certain beliefs are false - even while actively ignoring that by the very things he is claiming, there is so much that we simply cannot know - and therefore, logically, there can be no true certitude on these claims. While it was tempting to drop the overall work another star specifically for how bad this particular section is, ultimately the sections of the book leading up to that point are so strong that I simply can't go quite that far. So read this book through Parts I and II, just be aware up front that Part III is the weakest section of the book and could easily be skipped entirely. Recommended.