A Thousand Brains

A New Theory of Intelligence

By: Jeff Hawkins, Richard Dawkins - foreword

Narrated by: Jamie Renell, Richard Dawkins

Length: 8 hrs and 40 mins

Publisher's summary

A best-selling author, neuroscientist, and computer engineer unveils a theory of intelligence that will revolutionize our understanding of the brain and the future of AI.

For all of neuroscience's advances, we've made little progress on its biggest question: How do simple cells in the brain create intelligence? Jeff Hawkins and his team discovered that the brain uses map-like structures to build a model of the world - not just one model, but hundreds of thousands of models of everything we know. This discovery allows Hawkins to answer important questions about how we perceive the world, why we have a sense of self, and the origin of high-level thought. A Thousand Brains heralds a revolution in the understanding of intelligence. It is a big-think book, in every sense of the word.

PLEASE NOTE: When you purchase this title, the accompanying PDF will be available in your Audible Library along with the audio.

©2021 Jeff Hawkins (P)2021 Basic Books

Starts out good, ends up a train wreck

Part 1 was good. It was interesting and to the point of the subject of the book. After that it was a complete waste of time for me as it degenerated into rants by the author of what he considers to be “true” vs. “false”. That would be fine if the book was supposed to be about climate change, etc., but it did not belong in this kind of book. I could not care less about his opinions on the cause of climate change or any other similar issues, but it seems like he mainly wanted to lecture us on social issues (the final two-thirds of the book) and used the title simply as a hook.

Started with insights, ended with propaganda

This is a not place to push your beliefs about population control and anti religion. This should be about brain and Ai. Very dissapointing.

first part ok second nonsense

the first part of the book is fantastic . i advise the reader who is interested in neural network kind of artificial intellegence to concentrate on the first part of the book . the second part is just superficial mix of naive sci fi and political opinions such as: global warming is old stupid brain thinking while feminism is the new intelegence brain . there are really many chapters on this . i consider them really very shallow and waste of time. reader can just ignore them altogether.

There Are 3 Sections. Listen to the First One.

The first section, which lays out a model of the neocortex that's the product of the author, his company, and former researchers is interesting, and worth listening to. The only flaw in the first section I would point out is that the author appears to believe that movement is necessary to intelligence. In fact it isn't. One of the most powerful capabilities of the neocortex is that it can run its own hardware to do simulations. In effect, to "move" through new experiences, thought experiments, and so on, w/no physical movement at all. And it does this every moment it's functioning.

The second part, in which the author attempts to attack some thorny issues has huge problems. He derides philosophers (full admission: I am *not* one, and usually deride them as well.) But it's clear he needs to study philosophy if he wants to make arguments in the area of qualia and consciousness, because that chapter is an embarrassing hash in which he doesn't even seem to understand the physics of color vision--AT ALL. The discussion of consciousness is equally silly. He should have had a better editor.

His attempts to allay the fears of those who believe AI is potentially an existential threat to the human race because artificially intelligent machines won't have the "old brain" that causes so much mischief (like, allowing us to have fun, enjoy work, love, motivating us to survive... you know, all the "bad" stuff.) In fact, on this point he is catastrophically wrong. artificially intelligent machines *will* without the slightest doubt have an "old brain." Wanna see it?

Go look in a mirror.

Just as our old brain still dominates us, it will also dominate AI, using our neocortex as proxy.

Finally on the point about there being no moral hazard in unplugging an AI, it's clear the author has no ethical sense whatsoever; the rationale he gives for why this would be OK is genuinely preposterous. These machines wouldn't have any emotions, so killing them isn't immoral. Just ... wow.

The third section of the book isn't even worth commenting on. It's his political, cultural, and religious (or atheist) perspective and if you like the garbage that leftists spout as enlightenment no doubt you'll love it, but it has nothing to do with the topic, no matter how hard he tries to stretch it into "I believe the right stuff because I have no false consciousness. You, on the other hand, live in a fantasy world." This is infantile, but given his obvious political perspective, unsurprising.

Nothing new to see here

Old ideas about neural networks + grating self-agrandizement + endless repetition + unrelated musings about culture + unrelated musings about aliens = new book?

Awesome! But did he have to be so antitheist?

Truly an amazing book! I'm so happy to now know how intelligence works on the scientific level and definitely would recommend this to anyone who wants to understand the basic science underlying things - especially our own brains. Despite my rant below, I would highly recommend this book to everyone who enjoys learning - even if you agree with my annoyance at antitheism. You can always skip that chapter or only read the first part of the science. I would say the non-science chapters were very good interpretations of the consequences of this new paradigm, but not necessary. Any reader capable of enjoying learning this science from reading/listening to this information will be able to make the same deductions as the author without having to sit through the annoying parts, should you find them as annoying as I did.

So, my one caveat with this book is the antitheism. It constantly annoys me how atheists go around pretending they have the intellectual high ground, when the truth will always be that We Cannot Know. Which the author even acknowledges. And then proceeds, like most atheists, to declare belief in the afterlife to be false. As a religious person who believes that the "whys" (which religion answers) are different from the "whats" and "hows" (which science answers), I will always maintain that the only truly rational stance is agnosticism. So having to sit through the half a chapter or so on the author's annoyances at religion as a false viral meme was quite frustrating for me. I wish he had attempted to respect us and found a religious person who follows the scientific method to review his book, and to change the antitheist parts to be less agenda-pushing and more scientific based, but respect was clearly not in his agenda. I agreed with his point that some beliefs in the afterlife can be the excuses used to kill others, maybe even entire cities. But the way he grouped all beliefs in the afterlife to be false and all of our beliefs equated with one of the possible Armageddons was just downright offensive. Just because I believe in heaven does not mean I am going to be a suicide bomber. As a religious person who enjoys nonfiction books and science tremendously, I am unfortunately used to be derailed and mocked for said beliefs. However it never fails to disappoint me.

My personal conclusion should I have the opportunity to give the author a piece of my mind, would be that I wish he could have written this book or future books with the knowledge that the information will disseminate better if you attempt to not offend us. Would it really have been so hard to just not equate all religions with suicide bombers? Did he and every person who read or edited this really not see how offensive that was? Or did they really all just not care that belief in why the universe exists can be separate from believing in flat Earth?

Hawkins has Purchase on the "Easy" Problem

The Thousand Brains theory of intelligence - that the neocortex is composed of many thousands of columns of connected neural structures based on older grid and place cells, which create and link together models of the world, and also vote together to construct hierarchically more nested models - is brilliant.  The number of empirical constraints it solves is incredibly satisfying.  Citing just two examples, 1) what/where pathways in the brain are explained if "reference frames" attach to objects in the what pathway, but attach to our bodies in the where pathway, and 2) the perceptual binding problem is solved - when cortical columns agree on an object via consensus voting, they will naturally have different sensory inputs, thus they contribute diverse sensory models to a complete perception.  I'm being terse here because 1) this is incredibly exicting, and 2) the book is so smoothly written that it itself is the perfect way to get these ideas across.  No review will suffice.  So please, read this book!

I will only add, Hawkins is a thorough physicalist when it comes to consciousness, such that this and his previous book insist on using the word "intelligence" when one might want to see the word "consciousness".  Unlike his previous book, Hawkins confronts this notion and dedicates a chapter to it here, in which the "hard" problem ala Chalmers is recognized.  As a physicalist who acknowledges the hard problem myself, I nonetheless find Hawkins' commentary here unsatisfying.  For example, is the qualia of green a cortical map of green experiences with dimensions corresponding to green surface orientations?  I'm not convinced.  However, as Hawkins notes, qualia indeed seem "out there".  In other words, qualia have the quale of location, an aspect nicely satisfied by reference frames at the core of the Thousand Brains theory.  A theory which, as testified to in the beginning of this review, is brilliant and satisfying in its own right.

Underdeveloped theory. Premature book.

I think the author should not have written this book. His theory is still in its early phases and still poorly developed. The second half of the book is propaganda in favor of intelligent machines. But you can't something that doesn't exist and of whch you don't have a clear idea. The author talks about intelligent machines but he doesn't say how to acheive such an artificial intelligence. Too many assumptions with little solid science, or for that matter reality . Also the second half is rather repetitive. All is very unconvincing.

It really is that bad...

I'll start by saying that I am a huge fan of Jeff Hawkins work and have been waiting a decade for this book to come out. I was so excited to hear the progress of numenta on understanding the human brain. Unfortunately, and I hate to say it, this book is a total waste of time and I got very little from it.

He begins by recapping his ideas from "On Intelligence" for a few chapters. Then he briefly discusses the thousand brains theory and their updated model. Which boils down to: each column creates models of entire objects by using grid cell derived reference frames and then votes using long range dendritic segments that put the neuron in a predictive state.

This is very interesting since it disrupts the standard hierarchical model we use to understand the brain. However, I felt like he just glanced over the whole theory and I was left wanting more. He mentions that it would take a while to describe the neuro-anatomy responsible for these processes so instead he just leaves it out entirely... thats exactly what people who read this book want to know about!

I wanted to learn about how grid cells work, how concepts and features are learned and how sequences of features could be mapped to a reference frame within the brain. I wanted to read about studies that were conducted that confirm or reject parts of the theory. I wanted to push my understanding of the brain to the limits.

Instead, and this is the real tragedy, he spends the last two thirds of the book with speculative filler. For some reason he spends an entire chapter talking about how overpopulation is an issue and the solution is for women to be able to have abortions. I'm not kidding. Why is this in a book about a theory of intelligence???

The last two thirds read like propaganda, and thats coming from someone who agrees with most of what he says. He explains how the brain can make bad models of reality enforced by viral ideas and then proceeds to talk about three or four beliefs he holds because of his perfect model of reality. As if one day we will overcome our ancient brains and finally start agreeing with all of his political beliefs.

I'm still giving two stars because I love the work they do at numenta and there was some value in the first third of the book.

new theory of intelligence-or global warming book

drones on and on about global warming, material that is literally everywhere, and covered better and more completely elsewhere. could have been great, needs a better editor