On Intelligence: How a New Understanding of the Brain Will Lead to the Creation of Truly Intelligent Machines

Written by: Jeff Hawkins, Sandra Blakeslee

On Intelligence: How a New Understanding of the Brain Will Lead to the Creation of Truly Intelligent Machines Book Cover
From the inventor of the PalmPilot comes a new and compelling theory of intelligence, brain function, and the future of intelligent machines

Jeff Hawkins, the man who created the PalmPilot, Treo smart phone, and other handheld devices, has reshaped our relationship to computers. Now he stands ready to revolutionize both neuroscience and computing in one stroke, with a new understanding of intelligence itself.

Hawkins develops a powerful theory of how the human brain works, explaining why computers are not intelligent and how, based on this new theory, we can finally build intelligent machines.

The brain is not a computer, but a memory system that stores experiences in a way that reflects the true structure of the world, remembering sequences of events and their nested relationships and making predictions based on those memories. It is this memory-prediction system that forms the basis of intelligence, perception, creativity, and even consciousness.

In an engaging style that will captivate audiences from the merely curious to the professional scientist, Hawkins shows how a clear understanding of how the brain works will make it possible for us to build intelligent machines, in silicon, that will exceed our human ability in surprising ways.

Written with acclaimed science writer Sandra Blakeslee, On Intelligence promises to completely transfigure the possibilities of the technology age. It is a landmark book in its scope and clarity.
feedback image
Total feedbacks: 70
19
31
14
4
2
Looking for On Intelligence: How a New Understanding of the Brain Will Lead to the Creation of Truly Intelligent Machines in PDF? Check out Scribid.com
Audiobook
Check out Audiobooks.com

On Intelligence How a New Understanding of the Brain Will Lead to the Creation of Truly Intelligent Machines Reviews

Jordan Lewis
I have plenty of praise for this book, but most important is its accessibility. Apart from necessary jargon, Hawkins speaks in layman's terms and is willing to crack a joke at the right time. That said, the author takes his work seriously, and is determined to get his readers prepared to understand the results of his neuroscience research. For that reason, On Intelligence isn't light reading. That the book taught me a great deal about intelligence while rarely requiring me to look up information I have plenty of praise for this book, but most important is its accessibility. Apart from necessary jargon, Hawkins speaks in layman's terms and is willing to crack a joke at the right time. That said, the author takes his work seriously, and is determined to get his readers prepared to understand the results of his neuroscience research. For that reason, On Intelligence isn't light reading. That the book taught me a great deal about intelligence while rarely requiring me to look up information or reread any part of the book is a testament to Hawkins' organization and hard work.

Any clarification of physical processes taking place in the brain is balanced with examples and analogies. I believe that someone can pick up On Intelligence knowing nothing about the brain (I knew very little), and within 100 pages be able to explain how the brain identifies sensory patterns, creates spontaneous order, leverages memory, and informs better and better predictions about the world around us.

As the inventor of the PalmPilot, Hawkins has a consistent hypothesis that grounds his research: understanding the algorithms that produce intelligence in the brain will guide us to develop more intelligent machines. The author's bold claim (that most neuroscientists, AI researchers, and software engineers have been misguided, trying to build a better, faster brain rather than develop top-down theories on the root of intelligence) is well-argued, and just as accessible as the rest of his book.

I don't believe Hawkins' work is the definitive view of intelligence; no doubt there are experienced researchers with competing theories. The author spent limited time developing his own rigorous experiments, instead preferring to apply deductive reasoning using the limited intelligence research available. To me, Hawkins succeeds more as a explainer than an expert.

At the least, On Intelligence is a great book for thinkers and dabblers. At the most, it is a powerful springboard to greater knowledge. I can't wait to read more on the topic.
Manuel Alfonseca
This book offers what the author considers a revolutionary framework to understand the workings of the brain. Unfortunately, the author gets carried by his own enthusiasm for his pet theory, to the point that he speaks as though all his ideas, suggestions and suppositions were facts, and sometimes it's difficult for the reader to notice the difference. One must read the appendix with proposals of experimental confirmation / refutation, to see clearly that most of the ideas in the book are actual This book offers what the author considers a revolutionary framework to understand the workings of the brain. Unfortunately, the author gets carried by his own enthusiasm for his pet theory, to the point that he speaks as though all his ideas, suggestions and suppositions were facts, and sometimes it's difficult for the reader to notice the difference. One must read the appendix with proposals of experimental confirmation / refutation, to see clearly that most of the ideas in the book are actually unsupported suggestions.

The author correctly states that our knowledge about the brain is very scanty. However, this does not make him doubt his belief that strong artificial intelligence is possible and impending. Nothing in the book made me change my opinion that this is not as likely as Jeff Hawkins and Ray Kurzweil find it.

On top of this, in the beginning of the book and in his discussion on conscience, the author makes a confession of materialistic faith that taints a supposedly scientific book. I have written on this in more length in my blog:
http://divulciencia.blogspot.com/2015...
http://populscience.blogspot.com/2015...

Finally, in his discussion on conscience, Hawkins mistakes two different meanings of the word: I am conscious of myself and I am conscious of seeing you. His thought experiment to prove that conscience is just declarative memory is taking the word conscience in its second acceptation, while we were supposed to be speaking about the first. This is an example of the straw man fallacy, as his materialistic declaration is an example of the begging the question fallacy. I wonder why materialists fall so frequently in logical fallacies. Perhaps because their knowledge of logic is very scanty?
Wersly
Best of this book: the chapter where Hawkins lays down his theory of cortical function and expands on its implications. Really thought provoking stuff; he's got a way of extrapolating and coming to big, far reaching conclusions from the nitty-gritty of cortical wiring and structure. This is largely due to Hawkins attaching an overarching 'theory of intelligence' to his descriptions and illustrations of cortical organization. I find this perspective much needed given the current state of science Best of this book: the chapter where Hawkins lays down his theory of cortical function and expands on its implications. Really thought provoking stuff; he's got a way of extrapolating and coming to big, far reaching conclusions from the nitty-gritty of cortical wiring and structure. This is largely due to Hawkins attaching an overarching 'theory of intelligence' to his descriptions and illustrations of cortical organization. I find this perspective much needed given the current state of science education. An example: Hawkins reiterates everything regarding cortical structure I learned in my first college neuroscience class; however, where my course dove into the specifics and failed to provide any big 'aha!' explanations, Hawkins comes through in his theoretical perspective.

Some drawbacks of this book: the chapter on memory is awful. Hawkins fails to sketch any of the actual machinery that drives memory, and instead puts together a whole sloppy chapter of horribly dull Malcolm Gladwell-esque anecdotes about remembering things. Given how wonderfully detailed he gets in his cortex chapters, this was pretty disappointing. Additionally, Hawkins' tone in the earlier chapters is pretty funny: in his writing about how he was rejected from both MIT and IBM, he comes off sounding as a pompous late-teenage hackathon reject ("I'm too good for them anyway!"). That is, his writing is a bit egocentric, and he likes to draw examples from his own life and discuss his achievements. This doesn't really bother me - if anything, I see it as a keen insight into Silicon Valley start-up science - but it could be a turn-off for those that like their science writing 'objective.'

Really cool book.
The Mouse That Roared :: Professional Marketing & Selling Techniques for Digital Wedding Photographers :: The Knowledge of the Holy :: The Complete Essays :: Metaphysics as a Guide to Morals (Vintage Classics)
G
If to put it short, Palm Pilot-inventor Jeff Hawkins book explains his memory-prediction framework theory of the brain and describes some of its consequences.
Well, that's makes sense to me, as I've learned from professor Wang in my first lectures of neuroscience: "Brain is just a surviving organ...".
Sure, but there is always the Homunculus, that little bastard that is preventing us to perceive our realm directly, straightforwardly. Every time we turn to study ourselves we get into mess with that If to put it short, Palm Pilot-inventor Jeff Hawkins book explains his memory-prediction framework theory of the brain and describes some of its consequences.
Well, that's makes sense to me, as I've learned from professor Wang in my first lectures of neuroscience: "Brain is just a surviving organ...".
Sure, but there is always the Homunculus, that little bastard that is preventing us to perceive our realm directly, straightforwardly. Every time we turn to study ourselves we get into mess with that little guy, who interferes with us,...with us, with the real owner of our body and we found ourselves trapped with lost framework, with no philosophical background that would support our theories.
To comprehend Mr. Hawkins theory, the reader must accept what Mr. Thomas Metzinger believes strongly that it is possible to solve the philosophical puzzle of consciousness only if we come to understand that to the best of our current knowledge there is no thing, no indivisible entity that is us, neither in the brain nor in some metaphysical realm beyond this world.
So, to enjoy and understand this book, start with eliminating the "Homunculus", accept that the voice and dialog occurring in our heads is normal communication between brain modules that competes for the action of the body to keep us alive, and join Mr. Hawkins in trip into unknown of neurobiology and neuroscience generally.
To comprehend intelligence and AI, one must acquire proper framework, and learn about the brain and entire nervous system thoroughly. From point of view of AI however, one vital part of the brain is essential to fully comprehend- the Cortex! This is the trip that will lead you deep into the 6-layered magic cover that is making us exactly what we are. Have a nice trip!
Miguel Duarte
This is a fascinating book. If you are interested in the topic of human and artificial intelligence, please do yourself a favor and read it. This books aims at describing a unified theory of how the brain works and intelligence arises. Jeff Hawkins developed a a passion for the brain quite early in his career, and has obsessed over it ever since. He starts the book by explaining that although there are thousands of people studying the brain, there is no unified theory of how it works as a whole. This is a fascinating book. If you are interested in the topic of human and artificial intelligence, please do yourself a favor and read it. This books aims at describing a unified theory of how the brain works and intelligence arises. Jeff Hawkins developed a a passion for the brain quite early in his career, and has obsessed over it ever since. He starts the book by explaining that although there are thousands of people studying the brain, there is no unified theory of how it works as a whole. Many little details have been unraveled, but scientists remain clueless of how it all comes together. He looked at the problem the other way around by creating a model of how the brain works that not only fits all these small pieces, but it also explains other things that we didn't understand before.

At first I was a little taken aback because I felt Jeff Hawkins' writing was quite pretentious: here's a guy saying that everyone else is wrong and that he has the answer to everything. I gave him a chance, and I'm glad I did. The theory is supported by state of the art research (or at least it was at the time of publishing), and there are a lot of "holy ****, it makes so much sense" moments. I can guarantee that this book will change the way you think about yourself and how your brain works.

As an AI researcher and sci-fi lover myself, I can't help but identify with Jeff's goals of designing truly intelligent machines. I give him credit for thinking about the problem differently, and coming up with a very compelling theory. Only time will tell if he was right, but I'm certainly rooting for him.
Matthew Pacitto
A fascinating new framework on a theory of the neocortex, otherwise labelled a theory of intelligence. Although the complexity of the human brain seems insurmountable even with current technology, Hawkins makes a very compelling hypothesis about how intelligence operates. The most significant revelation is that how we've been likening intelligence to machines, or more recently computers, is essentially dead wrong. We don't compute, we remember. We don't process, we predict.

Although the framework A fascinating new framework on a theory of the neocortex, otherwise labelled a theory of intelligence. Although the complexity of the human brain seems insurmountable even with current technology, Hawkins makes a very compelling hypothesis about how intelligence operates. The most significant revelation is that how we've been likening intelligence to machines, or more recently computers, is essentially dead wrong. We don't compute, we remember. We don't process, we predict.

Although the framework is just a set of hypotheses at this point, they are a fresh enough look on their own to scream for experimental verification. I found the framework compelling enough that I almost kind of believe it, until I hear something better.

The section on 'Building Intelligent Machines' is very short, be warned there. This is not a book about AI, but it's a great book to make you ponder about intelligence in general. Hawkins is right on one count: we should better understand biological intelligence (the only source we know of) before we go ahead and try building artificial intelligence.

This book stands in stark contrast to the opinions of some other futurists like Ray Kurzweil, who practically believes fast enough CPUs will be enough to open-and-shut the case for human and superhuman level AI. According to this new framework, no, all the CPU cores in the world will never make a computer intelligent.

Not unless machines are designed to truly remember patterns, to predict future patterns, to build their own model of the universe (whatever universe that may be), and to truly understand.
Ayman
This book gives a whole new meaning to "Know thyself"! It's a journey through the anatomy and functionality of the human brain exploring many things we take for granted like hearing, seeing and thinking.
Armies of scientists working on AI for decades have not come close to doing what an average human simpleton does effortlessly, but the problem could be in the lack of a novel theoretical framework which, once discovered, could unlock a series of discoveries similar to what Einstein unlocked with This book gives a whole new meaning to "Know thyself"! It's a journey through the anatomy and functionality of the human brain exploring many things we take for granted like hearing, seeing and thinking.
Armies of scientists working on AI for decades have not come close to doing what an average human simpleton does effortlessly, but the problem could be in the lack of a novel theoretical framework which, once discovered, could unlock a series of discoveries similar to what Einstein unlocked with his relativity theory.
As Einstein started from the simple assumption that speed of life is constant and brained his way through the rest of the story till he reached the famous E=MC2, we could achieve something similar in neuroscience by starting from some simple premises about our brains, namely:
(1) the brain is composed of similar but highly connected computational structures
(2) The brain is organized in hierarchies of low level soldiers that identifies and anticipates an auditory or visual sequence or executes a motion sequence
(3) The brain is a memory-prediction machines, i.e., a system whose sole purpose is to match the relentless flow of sensory data to predictions about the future using stored memories.

As my brain was busy thinking about its inner structure, throughout the book, I had the feeling that I want to flip my eyes inwards and look inside my frontal loop :)

Highly recommended for my computer science geeks out there
Mikal
Hawkins book highlights how difficult it is to make predictions in areas of deep experience. Fundamentally the cortex theory is clearly presented but belabored. Sadly, Hawkins makes no efforts to cite or reference sources, leaving the reader to trust Hawkins at his word or do due diligence on their own to review the neuroscience community's perspective of his work.

On Intelligence has two major goals: define intelligent machines and the roll of intelligence machines in future society and to defi Hawkins book highlights how difficult it is to make predictions in areas of deep experience. Fundamentally the cortex theory is clearly presented but belabored. Sadly, Hawkins makes no efforts to cite or reference sources, leaving the reader to trust Hawkins at his word or do due diligence on their own to review the neuroscience community's perspective of his work.

On Intelligence has two major goals: define intelligent machines and the roll of intelligence machines in future society and to define the cortex and how it works as a foundation for building intelligent machines that replicates them.

Hawkins does an above average job at both, which ultimately hinders his goal. He spends too much time describing the hierarchy of the cortical region (with too few diagrams) and too little time proving his basis of the feasibility of intelligent machines- and further supporting his blanket statements that intelligent machines are net positives and not a genuine threat to our society.

But the reality is the cortex learns false patterns and does not do a great job at many types of reasoning, Hawkins does little to explain how a machine built on the model of our biology will overcome our logical fallacies.

Here we are ten years after the book was written and the impact and echoes of the perspective appear to have gone cold.
Don Skotch Vail
The adrenaline ran through my veins as I read this book, because I loved it so much. I think he is onto something, although I suspect he got some of the details wrong. When I tried to map out what he was describing, somethings didn't look like they would pan out. E.g. "names" flowing up and down the cortical regions were very vague, and how they could still be static names while getting less specific at each region was confusing to me.

He doesn't do a great job of describing how he think the cort The adrenaline ran through my veins as I read this book, because I loved it so much. I think he is onto something, although I suspect he got some of the details wrong. When I tried to map out what he was describing, somethings didn't look like they would pan out. E.g. "names" flowing up and down the cortical regions were very vague, and how they could still be static names while getting less specific at each region was confusing to me.

He doesn't do a great job of describing how he think the cortex works, although he tries. Apparently, you can download the software for from his company's website so I guess only high level details matter for the book. He says it takes about a year to get good at it.

He is pro-Chinese Room experiment. I was annoyed that he takes Searle's side that computers will never be able to simulate intelligence / consicousness, and yet his company creates software using his proposal that is intended to simulate visual intelligence.

Basically he says here are the data structures involved in the cortex, and that is most of everything that is needed for intelligence, but you just won't be able to simulate or implement these in the computer. You'll need a new device that is not a computer. Then he goes off and simulates it on a computer. I wonder how the rationalizes this contradiction.
May
The book gets 5 stars for having been written in 2002 and only just now coming to the point at which the future has surpassed some minor aspects of what he's saying.

I really liked the way he addresses the topic of intelligence and what it really means to think about it in a non-linear way. If your going down this track of thought, On Intelligence sticks out.

P. 51 For example, a framework of mechanistic development of systems where the location is the difference vs. modularizing in a way that d The book gets 5 stars for having been written in 2002 and only just now coming to the point at which the future has surpassed some minor aspects of what he's saying.

I really liked the way he addresses the topic of intelligence and what it really means to think about it in a non-linear way. If your going down this track of thought, On Intelligence sticks out.

P. 51 For example, a framework of mechanistic development of systems where the location is the difference vs. modularizing in a way that does not really understand the flexibility created by nature via the brain. Brilliant if you're thinking at all of building systems.

P. 69 The idea of the forgotten time element and what that means to visual and auditory inputs via the brain vs. what we are programming. Very insightful.

P. 81 The way your brain differentiates important vs. unimportant in storage into long term memory (for example, memorizing a song by tonal modulation vs. perfect pitch) and the ability to recall even if the pitch differs. Really great observation.

Lots of great stuff on pattern recognition and how to think of the brain as a pattern recognition machine.

Some of his thoughts on speech recognition has come a long way as with a few other ways that people now think of AI.

Still, great book for a beginner like me.
Vikram Kalkura
Now I am not scared of Robots it machines taking over us in near future. They can never overtake what our brain functions. They can just be faster than what our brain thinks but can never beat it.
If you want to know more about how your brain functions or how complex your brain is, then it's a good read. The 6 layers in your cortex that completely runs your body and mind is fascinating.
Good explanation on why human is the only living being that can talk and had so many languages. And why anim Now I am not scared of Robots it machines taking over us in near future. They can never overtake what our brain functions. They can just be faster than what our brain thinks but can never beat it.
If you want to know more about how your brain functions or how complex your brain is, then it's a good read. The 6 layers in your cortex that completely runs your body and mind is fascinating.
Good explanation on why human is the only living being that can talk and had so many languages. And why animals can't though Dolphins and other mammals have bigger brain than us.
2 things - pattern recognition and prediction has been explained really good.
Being this an audible, you lose an hour of it since it has diagrams and charts which can't be seen and hence difficult to relate. 3 stars because there is a vast region in our brain but all he spoke was about neocortex. Overall a great insight after reading the brain that changes itself.
Filip
A new and clearly formulated theory of intelligence, provided by someone with an strong computational science background, "On Intelligence" has given me a great deal of thinking material and is one of those books that serves as a cornerstone of your thinking whenever studying anything new in this topic area.

The core premise of the book rests on findings in neuroscience about the nature of the human brain. The most important study mentioned, it seemed to me, was Mountcastle's 1978 study, "An Orga A new and clearly formulated theory of intelligence, provided by someone with an strong computational science background, "On Intelligence" has given me a great deal of thinking material and is one of those books that serves as a cornerstone of your thinking whenever studying anything new in this topic area.

The core premise of the book rests on findings in neuroscience about the nature of the human brain. The most important study mentioned, it seemed to me, was Mountcastle's 1978 study, "An Organizing Principle for Cerebral Function", which demonstrated the rather astounding fact that the human cortex is highly uniform and its segments are not specialized, but highly flexible in use. The rest of the book is highly influenced by this finding and goes on to explain Hawkins' own "memory-prediction" model of intelligence.
Todd Martin
In "On Intelligence" Jeff Hawkins provides an interesting theory of the mind and intelligence. His theory centers around the idea that brains perform two main functions: pattern recognition and prediction. These two features help to explain why brains, even though they are small and slow are still able to perform some amazing feats.

He also discusses computing, artificial intelligence and neural networks and why these techniques have yet to produce intelligent computers.

The chapters discussing H In "On Intelligence" Jeff Hawkins provides an interesting theory of the mind and intelligence. His theory centers around the idea that brains perform two main functions: pattern recognition and prediction. These two features help to explain why brains, even though they are small and slow are still able to perform some amazing feats.

He also discusses computing, artificial intelligence and neural networks and why these techniques have yet to produce intelligent computers.

The chapters discussing Hawkins theory of brain function will only be of interest to neuroscientists. It was odd that in presenting this material Hawkins presents precisely no experimental evidence to support his claims. It's possible that none exists, and he somewhat makes up for this by listing some experiments that could be done in the appendix to the book.
Zarathustra Goertzel
As another commenter noted, the full title is, "On Intelligence (and Condescension)."

A lot of the content isn't particularly new or insightful. However some chapters are written well and portray a nice imagery >_<.

Expect the first 2-3 chapters to be full of condescending descriptions of how "artificial intelligence" has utterly failed up until now.

His descriptions in chapters 4, 5 and more-so 6 can be good =]

And his responses on consciousness and creativity are ok, typical responses.
The As another commenter noted, the full title is, "On Intelligence (and Condescension)."

A lot of the content isn't particularly new or insightful. However some chapters are written well and portray a nice imagery >_<.

Expect the first 2-3 chapters to be full of condescending descriptions of how "artificial intelligence" has utterly failed up until now.

His descriptions in chapters 4, 5 and more-so 6 can be good =]

And his responses on consciousness and creativity are ok, typical responses.
The future of intelligence too, really.
If not for the pompous attitude about other attempts to understand intelligence, it could be a pleasant, light introduction to the ideas for someone who doesn't know them already :O
P Michael N
This is a great book with a compelling model for AI. Big data is all the rage at the moment and it has accomplished a lot of great things that we see today that were not all that great when the book was written, so things like vision and speech recognition etc. have gone beyond what a lot of AI researchers thought would be possible today. Still, a better model is required for AI that doesn’t rely on so much data because as great as big data is, it doesn’t scale very well, we don’t have tons of d This is a great book with a compelling model for AI. Big data is all the rage at the moment and it has accomplished a lot of great things that we see today that were not all that great when the book was written, so things like vision and speech recognition etc. have gone beyond what a lot of AI researchers thought would be possible today. Still, a better model is required for AI that doesn’t rely on so much data because as great as big data is, it doesn’t scale very well, we don’t have tons of data on everything and even when we do have data in some narrow areas, there are issues. Great read, and that it needs an update after such a short time is testament to the rapid development in this field.
Bohdan Trotsenko
Sadly, I read this book only in 2017, 10 years after it has been published.
I'm still astonished by the ideas, which are so simple and obvious, and that I failed to observe on my own while examining how I think.
I'm proud that I did find something very relevant and what this book doesn't cover but comes close :)
I my collections of books that helped me uncover the mystery of a human mind this one tops the list. Rica Cater's books go second...
Jonathan
A gem. An absolutely clear description of how Hawkins hypothesizes brain mechanics & thought.

The only issue I had was with the chapter describing consciousness. It had notes of shallow hubris.

Highly recommend.
Jason
WOW, that's some heavy stuff. Learning how your brain works...
Last chapter is the best for sci-fi people and parents. So many possibilities.
Goes great w/ the other book from Bill Gates suggested reading list: 13 things that Don't make sense.

Ramkumar Ramachandra
Pop science. He uses very little data and argues very vaguely- in some parts, he's almost coercing the data to fit his interpretation.

Very light read. Might fascinate and entertain a bit, but not teach.
Khalid
An amazing book with a nice framework about how our brains work. A must read for any Brain-Enthusiast!
Juanmi
Key for understanding the different abstraction layers of how likely our brain and intelligence works. Key for understanding new developments in artificial neural networks too
djcb
Jeff Hawkins, famous for creating the Palm Pilot and some other technology, is also deeply interested in neuro-science. Furthermore, he thinks that the most viable path to artificial intelligence is through understanding (parts of) the human brain, rather than through bottom-up programming (as in classical-AI) or various neural-network (NN) approaches (though inspired by early brain research, the NN approach took its own path with little regard for the advances in neurology).

So, what's one to do Jeff Hawkins, famous for creating the Palm Pilot and some other technology, is also deeply interested in neuro-science. Furthermore, he thinks that the most viable path to artificial intelligence is through understanding (parts of) the human brain, rather than through bottom-up programming (as in classical-AI) or various neural-network (NN) approaches (though inspired by early brain research, the NN approach took its own path with little regard for the advances in neurology).

So, what's one to do then? Jeff Hawkins came up with his own theory about how the brain works, of which the core is that your brain works in hierarchically ordered memory, where each layer n abstracts the information from layer n-1 and involves layer n+1 when there's something unexpected. Crucially, layer n gives feedback to layer n-1 so the 'unexpected' becomes less so -- "learning".

So from, say, the retina the next layers register movement, edges, and further up the hierarchy (hard to describe what each of these layer perceives in normal language) and depending on how "unexpected" the information is, it could bubble up to some concept ("dog").

Hawkins makes it sound plausible, but is true? Though Hawkins is a wealthy technologist outside academia but he is not a crank. He has seriously thought about this, and comes up with testable predictions; moreover, he tries to implement his model in a computer - through his company Numenta.

The book is from 2005, but I don't think we can see to what extent his model is correct -- however, the Numenta software based on his model competes with software based on "traditional" NN-models -- the "deep learning" software of today, which is much stronger and useful than their 2005 predecessors against which Hawkins argues.

Regardless, interesting read. Hawkins can come off as a tad... self-confident here and there... but on a scale of 1 to Wolfram/Taleb, it's not too bad.
Anirudh
I found this book fascinating with regards to the ideas it proposes regarding neuroscience, intelligence and intelligent machines. I have no background in this field and in that context, I found the theories contained within it to be worthy of consideration. The writing style was generally good and the author painstakingly ensures that this is accessible to the general public.

Jeff Hawkins proposes a theory that he says is radical and has the potential to boost both neuroscience and AI. I dont ha I found this book fascinating with regards to the ideas it proposes regarding neuroscience, intelligence and intelligent machines. I have no background in this field and in that context, I found the theories contained within it to be worthy of consideration. The writing style was generally good and the author painstakingly ensures that this is accessible to the general public.

Jeff Hawkins proposes a theory that he says is radical and has the potential to boost both neuroscience and AI. I dont have the qualifications or expertise to vet these claims, but considering that 10 years after the book's publication we still do not have 'intelligent' machines makes it at least a little suspect.

That said, this book is certainly well-researched and Jeff Hawkins has certainly spent a great deal of time and effort into ensuring that he knows what he is talking about.

Will be interested to see what a neuroscientist or researcher in the field has to say about the theory and its ramifications.

My only complaint with the book itself was the fact that the author proceeds far too slowly and methodically. He puts forward a point or argument and after 2 analogies, I have a pretty clear idea as to what he is talking about. There are 3 or 4 more analogies for the same argument after that, which made me want to throw the book at the wall. This beating of the dead horse was my only crib with the book, which is why I give it 4 stars.
Paulynn Yu
I've been interested in neurobiology lately as I delve into the field of machine learning and think about my past readings in psychology / emotional science. I was looking for a book that would give me some working model of how the cortex works and how it is comparable to how machine / deep learning is being practiced today.

This book did a 1st class job in giving a working model of the cortex; I appreciate how there are large sections of just the biology part without being yet muddled with phil I've been interested in neurobiology lately as I delve into the field of machine learning and think about my past readings in psychology / emotional science. I was looking for a book that would give me some working model of how the cortex works and how it is comparable to how machine / deep learning is being practiced today.

This book did a 1st class job in giving a working model of the cortex; I appreciate how there are large sections of just the biology part without being yet muddled with philosophical debates and applications. I am studying how the brain works generally, but it was great to have a focused book on the cortex, difficulties in understanding the brain, memory-prediction hypothesis of intelligence.

He holds the Chinese Room Argument in the book. And while I appreciate and agree with the analogy, I was slightly taken aback about his claims on current inferiority of machine learning technology in many areas one of which is image recognition which I find is untrue. Additionally, I think his claims on human consciousness and whether humans have a soul or we die with our brain crosses the line on what you can claim scientifically about the spiritual realm.

It's an educational book for sure, but do take some claims with a pinch of salt.
Dan Riaz
Great read for those interested in learning more about the brain, and theories of it's underlying structures. Note, this book is primarily focused on the neocortex, the thin layer resting atop the brain that scientists believe is responsible for so much of higher-level human thought (including consciousness and self-awareness).

While the author discusses widely known concepts such as plasticity, the primary focus is on the anatomy of the neocortex, and a theory of the neural architecture that at Great read for those interested in learning more about the brain, and theories of it's underlying structures. Note, this book is primarily focused on the neocortex, the thin layer resting atop the brain that scientists believe is responsible for so much of higher-level human thought (including consciousness and self-awareness).

While the author discusses widely known concepts such as plasticity, the primary focus is on the anatomy of the neocortex, and a theory of the neural architecture that attempts to explain it's remarkable ability to create intelligence.

While this may sound like a dull affair, the author is incredibly provocative.

The general simplicity of his ideas elicits an intuitive understanding of his core theory. Without going into great depths, his key thesis is that intelligence is nothing more than memory, and more specifically our ability to leverage our memories to make predictions about the world around us.

Definitely offers a new perspective on the value of experience.

Not much else to say here other than go and buy it if you enjoy scientific theory, and are curious about the inner-workings of the human brain. Won't be disappointed.

Jina
I actually found this book incredibly fascinating. While I admittedly have a very basic understand of the brain, it’s interesting to think about the neocortex working the way Jeff Hawkins suggests. He goes into great detail on how the neocortex most likely behaves under certain circumstances ; explains things such as imagination, thought, creativity and stereotyping; and acknowledges that these are manifested in every one differently not only because the neocortex evolves according to our life e I actually found this book incredibly fascinating. While I admittedly have a very basic understand of the brain, it’s interesting to think about the neocortex working the way Jeff Hawkins suggests. He goes into great detail on how the neocortex most likely behaves under certain circumstances ; explains things such as imagination, thought, creativity and stereotyping; and acknowledges that these are manifested in every one differently not only because the neocortex evolves according to our life experiences, but also because everyone has a slightly different brain for genetic reasons. We’re visually drawn to people with deformities because our neocortex has detected a prediction error. Once you start to try to break down your human experience and see the world the way he has described you are, it’s a bit mind boggling. While other scientists might find Jeff Hawkins overly obsessed (to the point of delusion) with the functions of the neocortex, I can’t say that I finished the book with the same impression.
Kyle
This book explores a new model of intelligence. The book is fascinating, and presents a compelling solution to the intelligence problem. The book is easy to follow, and presents the author's theory in a concise manner. The theory explored makes a lot of sense and overall I would absolutely recommend this book to anyone who is interested in artificial intelligence and cognition in general.

The reason I did not give this book 5 stars is the author's attitude. Throughout the book there is an air of This book explores a new model of intelligence. The book is fascinating, and presents a compelling solution to the intelligence problem. The book is easy to follow, and presents the author's theory in a concise manner. The theory explored makes a lot of sense and overall I would absolutely recommend this book to anyone who is interested in artificial intelligence and cognition in general.

The reason I did not give this book 5 stars is the author's attitude. Throughout the book there is an air of condescension, the author views his intelligence model as the last word in the field, and seems to view other theoretical neuroscientists and the field as a whole as being ignorant for largely not recognizing his theory as "fact".

Early on the author compares his theory to Einstein's theory of relativity, and while he may be correct, the self association comes off as self absorbed and egotistic.

The author does not hold a doctorate in any neuroscience field.
Pradeep Kumar
The book kept me interested throughout. Jeff Hawkins presents his theories around four attributes of ne0cortex handling inputs from external world through sequences, hierarchies, auto-association and invariant forms. He gives day to day examples and makes it simple for anyone to understand the theory he is proposing. He ends the book summarizing on how far we are from building the AI system that is more close to real intelligence.
Good read for someone who is interested in the area of AI. The i The book kept me interested throughout. Jeff Hawkins presents his theories around four attributes of ne0cortex handling inputs from external world through sequences, hierarchies, auto-association and invariant forms. He gives day to day examples and makes it simple for anyone to understand the theory he is proposing. He ends the book summarizing on how far we are from building the AI system that is more close to real intelligence.
Good read for someone who is interested in the area of AI. The ideas might help build a better AI system.
Probably adding example's of current AI system, then comparing their working with the human brain and highlighting in detail the missing parts would have kept me more interested in some of the pages in the middle.
Aditya Asopa
While trying to look very foresightful and novel, the book manages to put forth a bold hypothesis for a general explanation for working of cortical networks, it barely hints to a mechanistic physiological explanation to the entire basic process. Read this book for some conciliatory ideas and out of box thinking but don't expect details. Hawkins tries to compensate this lack of physiological details by suggesting some behavioural experiments based on the predictions of his model described in the While trying to look very foresightful and novel, the book manages to put forth a bold hypothesis for a general explanation for working of cortical networks, it barely hints to a mechanistic physiological explanation to the entire basic process. Read this book for some conciliatory ideas and out of box thinking but don't expect details. Hawkins tries to compensate this lack of physiological details by suggesting some behavioural experiments based on the predictions of his model described in the book.

I suppose a deeper insight can be gained from reading Jeff Hawkins' published research papers. the writer has also cared to provide some references in the end, many of which are big books.

Read it for ideas not descriptions of them.
Terralynn Forsyth
Offers a comprehensive and thorough overview of the parallels between neuroscience and computer intelligence without the extensive length. Offers a technical section in one of the chapters, which I greatly appreciated, as someone who isn't has well-versed in the science of the brain. This book served as a good compliment to "How to Create a Mind", although the two possess slightly different arguments. Listened on audio while driving, but will be purchasing the physical book shortly, as I didn't Offers a comprehensive and thorough overview of the parallels between neuroscience and computer intelligence without the extensive length. Offers a technical section in one of the chapters, which I greatly appreciated, as someone who isn't has well-versed in the science of the brain. This book served as a good compliment to "How to Create a Mind", although the two possess slightly different arguments. Listened on audio while driving, but will be purchasing the physical book shortly, as I didn't feel like I absorbed as much of the information as I could have.
Masa Nishimura
Relating prediction not the behavior to intelligence was interesting. As a software engineer, I know the current way of thinking is behavior: design the system that responds to the input in the perfectly expected way. AI is a different way of looking at engineering. Aside from that, I was not convinced his problem statement that we need to build a system that perfectly simulates the brain neocortex in the first place. He should have dedicated more chapters on that topic.
Colin Das
In the pursuit of diving deeper into psychology and the mechanics of our mind, I found this absolutely fascinating and enjoyable. Jeff Hawkins does a great job of approaching "how the mind works" from various perspectives (biological, psychological, computational) and finds the key components: connections and hierarchies in the neocortex.

This seemed to nicely complement the famous Thinking, Fast and Slow.
Victor Negut
This is a great book. It has a very useful perspectives. My only issue is with how Hawkins describes the average neuroscientist or the average AI programmer. He seems to be creating a very naive image of the state of the AI field. Having said this, the book is 14 years old and the field has certainly evolved quite a lot in that time.
Jackie
I really like the narrator, Stefen Rudnicki. He uses just the right emphasis. However, the book refers to charts and pictures from the paperback version, and describes them. I am in need to see these charts, so will buy the book. I am fascinated by the explanation of an original view of how the brain works, and how it compares with brain function of animals.
Shawn Ridenour
Interesting book, I enjoyed the premise, it makes sense.

I kinda hated the self aggrandizement off it, more focus on the theory and consequences would have made it a better read. But, Jeff did be sure to give plenty of due it's due, so it was bearable.
Tudor Ștefănescu
well, I found it harder to read than I expected it - but I think I got the main idea: most of our thinking can be modeled using a hierarchical memory-prediction system. Which is humbling and at the same time liberating...
Savanna
Good overview of hierarchical model of intelligence useful for understanding some ML techniques used today. Little standoffish but generally intriguing read with subject presented in intuitive manner.
Harmeet Singh
Read this book after reading Ray Kurzweil’s “How to create a mind” so could not help comparing the two. Both books give detailed explanations on working of the “new brain” or neocortex touching upon complex topics like consciousness. Authors do a very good job of keeping the topics crisp.
Brandon Warman
Jeff puts the mind into a perspective that I had only really seen described in a similar way by today's AI engineers. This makes sense since this book was written over 10 years ago and lays the groundwork for understanding the brain in a way that will allow us to build "intelligent" machines.
ferhat
Simple theory of how brain does its magic.
Joshua Wright
If you're interested in Cutting Edge treatments of the concept of consciousness, look no further. Jeff Hawkins has a very refreshing, and unique perspective on the issue.
Jake Stevens
Interesting theories. Would love to see an updated version, since so much as happened in the field since the book was written
Alex
This book takes complex and cutting-edge scientific ideas and explains them with great clarity. Completely changes the way I viewed the operation of my mind.
David Knip
Great read. Would love to have another book with some updated information!
Devashish Arora
How to create a mind by Ray kurzweil is much better
Charlie Gorichanaz
will need to reread a few chapters and let it digest
Kit
I don’t particularly enjoy the “what everyone gets wrong” Narrative style. I couldn’t get through it. I think the main idea is interesting, but could have easily been an article in wired.
David
A little dated, but a well thought out description of how the mind may work and how machines may someday mimic it.
Si
A book with many fascinating insights into how the mind works, sadly flawed with too many brash assumptions and glossing over important issues.
Let’s get the negatives out of the way first:
On p41 he makes the big, big assumption that only neocortex houses intelligence. This is pretty brash, given how we’re still learning about how the mind works and so much other stuff I’ve read shows how all the parts of the mind influence the others. That said, he hadn’t defined intelligence at that point.
And t A book with many fascinating insights into how the mind works, sadly flawed with too many brash assumptions and glossing over important issues.
Let’s get the negatives out of the way first:
On p41 he makes the big, big assumption that only neocortex houses intelligence. This is pretty brash, given how we’re still learning about how the mind works and so much other stuff I’ve read shows how all the parts of the mind influence the others. That said, he hadn’t defined intelligence at that point.
And to say on p43 that the mind is produced only by the brain period is also bold given all the research about body-mind, and the nervous system around the body which many think has a lot more to play in the makeup of the mind than we intuitively assume. It has been argued that the mind wouldn't function without the body, c.f. the feedback stuff he mentions. And this seems to be an unnecessary assumption.
Again on p51-52, though this may be splitting hairs a bit: the underlying algorithm doesn't contradict the idea of different ways of processing sight, sound, etc. to continue the computer analogy, it's all logic gates underneath, all number calculations, but built on tip of this are many different computer languages.
On the good side I laud his aims – he made his money building personal gadgets and is now spending his money on his intelligence institute.
So his aim is to correctly understand how the mind works – most other people have got it wrong, and no one has a grand theory of how the mind works like he does. Ahem, yes, bit grandiose and martyr-like. The book is that, i.e. a grand theory of how the mind works. That’s the first part of his work. The second is to figure out how to truly put this algorithm to work inside a machine. He says that while the AI industry has come up with some great applications, they missed the point as they started implementing AI before they fully understood the intelligent brain they were aiming to base it on.
He posits that the mind can take any input – we have sight, hearing, etc. and learn to process it. It works the same for each: it takes in data over a period of time. Sight is not a snapshot – the eye has three “saccades” every second; it takes in a little part of the field of vision each time and builds up a picture over time. Similarly and more intuitively with hearing – we process a series of sounds over time. It wouldn’t make any sense if we simply had a snapshot of time. And so with touch – if you wake up touching something you can’t figure out what it is until you’ve moved along it, i.e. a sequence of touch over time.
Then the cortex is made of 6 layers of neurons, each holding data that are an abstraction of the data in the lower layer. So for example when you hear music the lowest layer will hear notes, the next layer will put those notes into riffs, end so on. Or when you’re reading you’ll get letters at the lowest level, morphemes, words, phrases, then more abstract understanding at the top.
When we’re learning something, say reading, the simple part, i.e. the letters will go right to the top layer and we’ll be aware of that. As we learn, the letter bit goes down a layer and we can think more of the words, and as we get a bit more adept the words are all we need to consider at the top layer, and so on. So as we practise something, it gets so that we need to consider less at the top level
Unless of course there’s an “error”. Only errors filters up the chain, say if you are walking into your house, the way you always do, but you suddenly notice that a floorboard is loose – that will shoot up the layers until the higher layers become aware of it. Otherwise all the actions are pretty autonomous.
So the abstraction process is what the brain excels at, and something I’ve always intuitively thought the brain did too, so good to see someone else confirming the theory.
Now given his system for how the mind works, the final chapter on applying this algorithm to all walks of life is most inspiring. The idea of this algorithm’s ability to learn given any input is powerful indeed. So just plug it into a camera and a car, for example, and it would use that same algorithm to figure out driving. And once we spend the time training one system we can simply copy it and refine it. It would truly revolutionise our world.
Latest reviews here: http://unfebuckinglievable.wordpress....
Alessandro Perilli
This book contains a truly radical idea and incredible premonitions considering it was written in 2005. I highly encourage everybody approaching/working in Artificial Intelligence to read or re-read it.
Calvin Holst
On Intelligence is a theoretical work regarding the function and physiology of the human brain by Jeff Hawkins. Hawkins is the inventor of the Palm Pro, one of the creators of the microprocessor, and is considered one of the most prominent figures of electronics and neuroscience research in the past 20 or so years. Hawkins begins the book by explaining the shortcomings of AI (artificial intelligence) in the search for intelligent machines. In this section he argues that the defining feature of i On Intelligence is a theoretical work regarding the function and physiology of the human brain by Jeff Hawkins. Hawkins is the inventor of the Palm Pro, one of the creators of the microprocessor, and is considered one of the most prominent figures of electronics and neuroscience research in the past 20 or so years. Hawkins begins the book by explaining the shortcomings of AI (artificial intelligence) in the search for intelligent machines. In this section he argues that the defining feature of intelligence is the ability to recognize, predict, and apply patterns in the stimuli experienced. Using one of his favorite methods of demonstration, Hawkins describes a hypothetical scenario in which a person in a closed environment is given information in a language they do not know, along with instructions on how to change that information to a different language they do not know. The argument is that the system is inefficient, and despite being completely accurate in the translations, the person performing them has no actual understanding of what they are doing. This is the basis of Hawkins’ argument against AI, and it simultaneously shows how truly intelligent machines will not only perform complex tasks but will also understand them. Following this Hawkins begins an in depth description of the function of the human brain.
This is the first point where I found myself getting lost in the more abstract neurological theories. The chapters about neural networks and the cortex were the most difficult to follow, as they were more about the biological structure of the brain and how its structure relates to function. While the function of the structures is well explained, the structures themselves were difficult to understand based off of Hawkins’ descriptions. I found it strange that the more easily understood analogies were used to describe the functions of these complex systems but not to describe their biological structures. Despite my difficulty with these sections I found that the overall structure of the chapters made it simple to form an understanding of Hawkins’ theory.
While Hawkins spends a large portion of the book describing the functions of the brain, the real importance is in how these concepts will be applied to technology to create machines which have a true capacity for understanding. In his discussion of the moral and social implications of the creation and application of intelligent machines Hawkins addresses concerns regarding the classic “robot uprising” scenario which seems to surround the topic of engineered intelligence. Hawkins is very reassuring in his insistence that creating intelligence does not mean creating artificial humans, a very important distinction to be made for readers in a society where the most common depictions of artificial intelligence are humanoid androids and supercomputers. His belief is that intelligent machines will be designed for singular tasks, and by the time of their realization it will be completely within our control how human-like or not they are.
Whether you follow neurological science or science fiction, On Intelligence is a great read which offers complex yet accessible insight on the challenge of creating intelligent machines. The diction and tone are like that of a TED talk, walking the line between more casual layman-friendly analogies and the academic vocabulary of neuroscience and electronic engineering, making it possible for almost all audiences to understand and enjoy this informative, in-depth look at the reverse engineering of human intelligence. Hawkins’ finale is short and sweet, challenging readers (and the human race as a whole) to complete the daunting task of designing truly intelligent machines.
Nicolas Johnston
Jeff Hawkins, inventor of several handheld devices such as the Palm Pilot and the Treo, breaks new ground in his nonfiction book, On Intelligence as he tackles a new theory about how the human brain works. In his book Hawkins explores controversial ideas and concepts like his Memory-prediction framework, which suggests the brain combines old and new information to make predictions about what will happen in the future. Hawkins expertly incorporates theories like this one to explain how and why at Jeff Hawkins, inventor of several handheld devices such as the Palm Pilot and the Treo, breaks new ground in his nonfiction book, On Intelligence as he tackles a new theory about how the human brain works. In his book Hawkins explores controversial ideas and concepts like his Memory-prediction framework, which suggests the brain combines old and new information to make predictions about what will happen in the future. Hawkins expertly incorporates theories like this one to explain how and why attempts to build intelligent machines have failed as well as highlight how mankind must “extract intelligence from the brain” in order to succeed at building a truly intelligent machine sometime in the near future.

Hawkins’ book is far from conventional. He acknowledges that some of his views are merely speculation and cannot be proven in absolute certainty and that many aspects of his theory about the human brain may contain missing parts or be eventually proven incorrect by science. Hawkins also bases much of what he talks about from the work of others. He thoughtfully credits the work of brain researchers in other fields because some parts of his theory were based off of the research that they did. “New ideas are often old ideas repackaged and reinterpreted.” he says. What he means by this is that some of the ideas he talks about aren’t new at all. Instead, Hawkins brings multiple theories together for the purpose of creating one large theory. Additionally, in contrast to many other works of literature, Hawkins fully expects some negative reaction and controversy in the scientific community and among his readers concerning what he writes about in his book, in fact he welcomes it, which is something that many writers try to avoid. One example of this is given near the end of the book when he asks, “Should we build intelligent machines?” Many people fear building technology that may one day slip out of our control. Hawkins disagrees, saying “brainlike memory systems are going to be among the most useful tools we have yet developed.” Still, Hawkins presents his readers with something to think about, which is good for overall reader engagement in the book.

Possibly of the greatest aspect of Hawkins’ book is that it’s easily comprehensible. Hawkins does a phenomenal job of explaining the inner workings and structures of the very complex and mysterious human brain while using in an informal, simplified form of language for people who may not have an extensive knowledge of how the brain works. In addition to using diction that is equipped for a diverse audience, his frequent use of analogies and personal anecdotes also play a part in keeping the audience informed on concepts that might otherwise be considered too complex for non-scientists to understand.

Regardless of interest level in computing or neuroscience, this is definitely a book worth reading. It is engaging throughout, discussing various mysteries of the brain as well as giving the reader a fresh perspective on how the brain works. In a world full of new smartphones and tablets, it has the potential to change the way we make these technologies down the road. More importantly though, it can change the way we view our own brains, to quite literally change the way people think about thinking.
Carol Peters
Really interesting stuff I am glad I learned about inside a book three times longer than it needed to be.
Evan
On Intelligence is non-fiction book written by Jeff Hawkins, who is an inventor and a neuroscience researcher. The main purpose of his book is to relate brain theory to artificial intelligence and to demonstrate that if we apply what we know about the brain, then we can make truly intelligent machines.

This book focuses on answering several questions:
What is intelligence?
How does the brain do it?
How can we replicate it in our technology?

At the start, Hawkins explains what the brain can do, and w On Intelligence is non-fiction book written by Jeff Hawkins, who is an inventor and a neuroscience researcher. The main purpose of his book is to relate brain theory to artificial intelligence and to demonstrate that if we apply what we know about the brain, then we can make truly intelligent machines.

This book focuses on answering several questions:
What is intelligence?
How does the brain do it?
How can we replicate it in our technology?

At the start, Hawkins explains what the brain can do, and why AI researchers have had a hard time duplicating the brains functions. He also provides several important definitions, such as what intelligence really is. He then moves on to how the brains work, offering a complex and in depth analysis of the brains systems and how they connect and handle information. The last part contains predictions and a description of the future for intelligent machines.

The first few chapters of this book are the most user-friendly. Hawkins is able to explain relatively complex theories in a easy to understand way. He uses examples and little anecdotes to help the reader to understand the concepts he is putting forward. For example, when Hawkins explains how time is an important part of memory, he uses the example of the Alphabet. He states “You know the alphabet. Try saying it backward. You can’t because you dont usually experience it backward.” Little examples like these help keep the book flowing steadily.

Another aspect of this book that makes it accessible to a larger audience is the author's diction. That is to say that the author uses words and phrases that are easy to understand and read. Any term that is not easily understood is thoroughly explained. His definition of intelligence is that “Intelligence is measured by the capacity to remember and predict patterns in the world”. Simple definitions like this can make a book as complex like this appeal to many different people.

Chapter six is where this book gets very technical and complicated. This section gives a very detailed analysis of how the neocortex works. While this portion of the book is crucial to brain theory, it can, at times, seem like a college textbook. This section is harder to comprehend, and appeals to a very narrow audience, which is made up of neuroscientists and AI researchers.

While this book may at times seem daunting, it is a must read if one is interested in the brain or how it works. After reading this book, the reader will come away with a new outlook on the brain, an appreciation for its complexity, and that he future of intelligent machines is very, very real.
Danielle
Firstly, this guy talks about how much he loves brains and studies brains and looks at brains that I couldn't help but imagine him as a zombie. In neuroscience, we mostly discuss "the brain" as an indeterminant entity (to use his own terminology). Secondly, I get that you're a smart guy and you invented the Palm Pilot. Nicely done. I'm sure you've reaped the benefits of that (with which you started your own neuroscience institute?). A BRIEF intro about the author would have been perfectly approp Firstly, this guy talks about how much he loves brains and studies brains and looks at brains that I couldn't help but imagine him as a zombie. In neuroscience, we mostly discuss "the brain" as an indeterminant entity (to use his own terminology). Secondly, I get that you're a smart guy and you invented the Palm Pilot. Nicely done. I'm sure you've reaped the benefits of that (with which you started your own neuroscience institute?). A BRIEF intro about the author would have been perfectly appropriate. Let me rephrase. A brief, 3RD PARTY intro would have been best. You don't have anything to prove. People know what a Palm Pilot is and will be sufficiently awed. Get over yourself.

Now, to business. This book felt like two separate entities to me. There was the first 5 chapters that had to do with general theories of mind and the current problems with past and current AI/neural net systems research, then there was chapter 6 and the appendix that gets really specific about which cortical layers are doing what. The general theories are really interesting. Since I mostly study disease pathology, I actually have little knowledge of what intelligence research entails. The points he brings up regarding problems with that research all make sense to me from what I do know. I think that if you just read through chapter 5, this book has everything you need to understand the brain in philosophical sense. This part gets a 4

So . . . Chapter 6, or as other reviewers have called it, the confusing chapter. At some point during my education, I did have to learn about the layers of cortex and their projections and things, all of which I promptly forgot. There's a lot going on there. The best way to look at it is in pictures and diagrams (which he has to an extent). My main complaint is that the ideas between these diagrams are so mashed together that I had trouble staying focused. It seriously took me longer to get through chapter 6 than all of the other chapters combined. This guy has been thinking about these things for 20+ years and it shows in his blatant disregard for the normal reader's attention span. I read scientific papers all the time, and I still had trouble getting through. The progression of thoughts in chapter 6 was too rushed, and the diagrams were too simplified during the later stages. It would have been helpful to see a short series of diagrams next to each other for each stage of memory that he described. This part of the book gets 3 stars, but since it's the minority of the book, 3.5-4 is my best guess for a rating.
R
This book was released in 2005. Still there are only close to 250 ratings on this book. The ratings, if not excellent, are good but before deciding to read the book I was wondering why this book may not have gained enough momentum in ten years? It may be that people didn't read it at all, or those who did, didn't review it because they couldn't really express what this book was about.

Readers may not have been able to understand the book because almost half of it talks about theory of brain syste This book was released in 2005. Still there are only close to 250 ratings on this book. The ratings, if not excellent, are good but before deciding to read the book I was wondering why this book may not have gained enough momentum in ten years? It may be that people didn't read it at all, or those who did, didn't review it because they couldn't really express what this book was about.

Readers may not have been able to understand the book because almost half of it talks about theory of brain system. As you go through that part, it feels like you are sitting in an algorithms class where the instructor tried to simplify the concepts but in the effort to do so, made it completely mundane and tiresome. The author hints on the fact that this part can be skipped but if you skip it, you will lose the continuity of what the book is trying to talk about, which bring us to the question - what is the book all about?

Well, the author starts with a conjecture that the book will be about why current state of Artificial Intelligence may not be able to design intelligent machines. For some time, the author talks about himself and then tries to explain the basic details of Neural Networks. If you already know about Neural Networks and Machine Learning, the part of the book would make sense. Then the author talks about how the brain works (possibly to explain how it is different from Neural Networks). It is at this point that the author goes into gory details of hierarchical structure, which although makes sense seems irrelevant to the line of argument.
Nevertheless, by the time this argument ends, you are left with only one more chapter where author tries to condense the gist of the entire argument by talking about what the intelligent machines should look like, how it should be designed and what would it achieve. In between, there is some explanation about cautiousness and awareness which although the author explains clearly has no relevance to the topic in discussion.

In the end, you will be utterly confused about the book. What was it about - Neural Networks, Psychology, Biology, Author? Oh, the book was about how in its current state of Artificial Intelligence may not be able to do what a human brain does - it could probably be explained in 50 pages. Ironically only about 50 pages are devoted to that. Rest of the pages could have been blank as well!
Elizabeth Humphries
I'm a bit conflicted by this book. On one hand, Hawkins' exploration of intelligent computers (and how to get there) is fascinating. He makes several interesting points about the speed of transistors versus the speed of neurons, as well as the processing speed of computers and humans in terms of recognizing "mistakes" in patterns or memories. It's hard to argue with him that the current view on the framework of intelligent machines (and how they should be set up/programmed) is wrong, especially I'm a bit conflicted by this book. On one hand, Hawkins' exploration of intelligent computers (and how to get there) is fascinating. He makes several interesting points about the speed of transistors versus the speed of neurons, as well as the processing speed of computers and humans in terms of recognizing "mistakes" in patterns or memories. It's hard to argue with him that the current view on the framework of intelligent machines (and how they should be set up/programmed) is wrong, especially for a reader with only passing familiarity in the field.

However (and it's a big however), it's hard to trust Hawkins' scholarship when you realize he made at least two errors in his statements about general biology. They're small, but they are disturbing. He ignores a rather extensive field of study looking at spatial memory and problem solving in corvids when discussing the role of the neocortex in intelligence. Yes, mammals are intelligent by his definition, and they have a neocortex. Corvids (birds such as crows, ravens, and jays) are also intelligent by his definition, and they do NOT have a neocortex. In fact, they have a demonstrated spatial memory that is better than any measured in humans. This does not invalidate his proposed theory, but it does raise questions he conveniently ignores. (I can't decide if he skipped it deliberately or because he is unaware of the research. Either is a bit concerning.) Additionally, he uses a thought experiment in which he invokes a single-celled animal. Animals are, by definition, never single-celled. Maybe it's a slip of the pen, or a detail he doesn't consider important. As one of the definitions of an animal involves cephalization (movement of sensory organs towards the head region in development) which is most definitely a fact relevant to neuroscience, his misuse raises a red flag for me. Can I really trust the scholarship of a man who makes errors about such basic and fundamental biological facts?

I don't think outsiders are incapable of formulating paradigm-changing theories in science. Science history is filled with such examples (like Dalton, who revolutionized chemistry despite being a geologist).

This book was written in 2004. I do wish I could heard Hawkins' thought on Siri!
Ira Therebel
I am no expert in neither brain activity nor building intelligent machines. This is why it is hard for me to say anything about the idea itself, nor would it really matter.

I did find it very interesting. I am pretty sure I learned about hierarchical brain functioning in my Cognition class a few years ago. I don't want to make a mistake to think that it is 100% right or that this kind of thinking is indeed superior to what is done by AI people. As Jeff Hawkins said, it is impossible for computers I am no expert in neither brain activity nor building intelligent machines. This is why it is hard for me to say anything about the idea itself, nor would it really matter.

I did find it very interesting. I am pretty sure I learned about hierarchical brain functioning in my Cognition class a few years ago. I don't want to make a mistake to think that it is 100% right or that this kind of thinking is indeed superior to what is done by AI people. As Jeff Hawkins said, it is impossible for computers being exactly like humans, so maybe having their intelligence work in a different way than ours would make more progress. Humans are not really perfect to try to make everything work like us.

Jeff Hawkins says himself in this book that he is sure a lot of what he says may end up being wrong. This happens in science pretty often with new hypotheses and theories. But this doesn't make this idea any less interesting. It could be a progress towards something great, even if only 40% of what he says is correct.

One thing I know that even though fish are not mammals and have a small brain, there have been studies done that have shown that fish can indeed learn, unlike it is said in the book.

It is the year 2012 now and the book was written in 2004. I am sure there is a lot of progress done by now and some of the details improved. I just had trouble finding it in the past hour. All I can find are the reviews of the book.

The book was written for people like me. Neuroscientists will most like find it too trivial. But us, general public who work in other fields, can appreciate how this book is written. I liked his way of explaining. He is great at providing examples and analogies to make one see what he is talking about. He is also good at simplifying the brain activity to make it easier to understand. My main problem was that he kept on repeating things. It seemed to me that he tried to go from a completely simplified view on his idea to the complete version of it by repeating it many times and always adding something new to it. I ended up being bored and wish he would break it down less.
Tclizzy
On Intelligence is a book that discusses Jeff Hawkins’ theory as to how the brain works and as a vessel to flout his superiority complex. If you are a person casually interested in the functions of the brain and the future of artificial intelligence who is willing to slog through what I felt was sub-par writing, this is a interesting read.
Hawkins proposes that the brain is essentially a device that uses past experiences to predict the future. He makes this proposal to support his goal of making On Intelligence is a book that discusses Jeff Hawkins’ theory as to how the brain works and as a vessel to flout his superiority complex. If you are a person casually interested in the functions of the brain and the future of artificial intelligence who is willing to slog through what I felt was sub-par writing, this is a interesting read.
Hawkins proposes that the brain is essentially a device that uses past experiences to predict the future. He makes this proposal to support his goal of making intelligent machines, stating that a deeper understanding of the brain is needed to make such machines. He writes how, “On intelligence develops a powerful theory of how the brain works, explaining why computers are not intelligent”. This is an interesting theory that is presented in an easy way that provides food for thought on a boring day. However, there, it all gets worse.
On Intelligence falls short of being a good book in its tone and style. It is not a difficult read, however it was mind numbingly dull in its diction.It was hard to stay focused on the book, which is disappointing considering how interesting the base idea is. A personal favorite book of mine, Bill Bryson’s A Short History of Nearly Everything has a similar genre of popular science that is actually entertaining to read, unlike this book.
Jeff Hawkins is a businessman who created “the first commercially successful handheld computing device” and more recently has worked in neuroscience towards the goal of producing intelligent computers.This book’s purpose is to support his theory on how to achieve this goal.
Hawkins, however, is the worst part of this book. It is expected that he would promote his theory and attempt to disprove others. However, his downright condescending tone toward other theories and the people behind them was off putting.
To bring it all back together, this book makes a great point, but it fails is execution. This book came as a major disappointment; so unless you are interested in the future of neuroscience and Artificial Intelligence, stay away from this book.
Mirek Kukla
Jeff Hawkins presents his 'memory-prediction framework' of intelligence, which roughly states that "prediction, not behavior, is proof of intelligence". He rejects the common (implicit) assumption that intelligence is defined by intelligent behavior (think: the Turing test of AI). Instead, Hawkings implores us to better examine the brain.

The brain, he argues, doesn't compute the answers to problems - it retrieves them from memory, where memories are stored in a memory hierarchy. Our brain stores Jeff Hawkins presents his 'memory-prediction framework' of intelligence, which roughly states that "prediction, not behavior, is proof of intelligence". He rejects the common (implicit) assumption that intelligence is defined by intelligent behavior (think: the Turing test of AI). Instead, Hawkings implores us to better examine the brain.

The brain, he argues, doesn't compute the answers to problems - it retrieves them from memory, where memories are stored in a memory hierarchy. Our brain stores invariant representations of the world, which means that relative attributes of the world, not absolute ones, are what matter. As an example, consider that the star spangled banner is invariant with respect to key - we recognize the song whether its in A sharp or D flat.

As an example of how the memory hierarchy works, consider the processing of sensory input. Sensory input is received by the lower regions of the hierarchy, and gets filtered up to the higher regions of the brain. Here processed data is matched to an 'invariant representations' of the world. This higher region of the brain, then, uses this representation of the world to make a prediction regarding future sensory input. A correct prediction constitutes understanding.

Hawkins' attempt to gleam understanding by examining the brain is valiant, but when it comes down to some of the gritty details, it gets a little confusing. I was left with the impression that Hawkins should either have skipped the technicalities, or gone to greater length to make the develop them.

Fortunately, the above complaint only pertains to a a brief twenty pages. The rest of the book is terrific. It's still too early to say whether or not Hawkins theory is correct, but I'm convinced that he's on to something. He does a great job of explaining his central thesis and fostering an intuition for it. It's all very concrete, and the implications are practical. A refreshing read, and you're left with a fuller understanding of how you tick.
Christopher Litsinger
This has been on my to-read list for a while, and after an enthusiastic review from a friend I finally got around to it.
The book is interesting, and contains lots of theories, but I had a difficult time deciding how trustworthy it is. Hawkins couldn't get his PhD proposal accepted by Berkley, so he used his Pam cash to start his own neuroscience research company. It's an interesting set of credentials, but it smells a bit too much like the hubris which technologists are often guilty of: "I'm goo This has been on my to-read list for a while, and after an enthusiastic review from a friend I finally got around to it.
The book is interesting, and contains lots of theories, but I had a difficult time deciding how trustworthy it is. Hawkins couldn't get his PhD proposal accepted by Berkley, so he used his Pam cash to start his own neuroscience research company. It's an interesting set of credentials, but it smells a bit too much like the hubris which technologists are often guilty of: "I'm good at tech so therefore I can learn everything about x and change the world."
Here's a typically off the cuff statement (after a section defining V1, V2 and V4 without talking about who or how they were discovered): "For these and other reasons I have come to believe that V1, V2, and V4 should not be viewed as single cortical regions." other reasons? hrm. Maybe talk about the research that backs this up?
The book also sometimes assumed I knew more than I did. Example: " Since many people have heard the term pattern classification used in AI and machine vision research..." uh, many?
And sometimes, language like "This is all it would take to stop the layer 3b cell from firing when the column correctly predicts its activity..." would throw me off -- is this speculation? Established theory?
Hawkins does present some interesting ideas, and I particularly like the idea of expert learning as pushing patterns lower in the cortex, but ultimately I just don't know whether this book represents true genius breakthroughs, well established theory, or just strange theories by an outsider that may have no real application. Perhaps it had some mix of all of those.
Jeff Gabriel
This is a really smart and engaging evaluation of how understanding our biological brains can help direct how we write smart software - at a fairly high level of abstraction. Which coincidentally is how the author identifies intelligence. He argues in a convincing way that physical layers of the cortex are used for increasingly broad symbolic abstractions which are then used for prediction. He states that this prediction capability is the primary marker of intelligence. This book is a bit more t This is a really smart and engaging evaluation of how understanding our biological brains can help direct how we write smart software - at a fairly high level of abstraction. Which coincidentally is how the author identifies intelligence. He argues in a convincing way that physical layers of the cortex are used for increasingly broad symbolic abstractions which are then used for prediction. He states that this prediction capability is the primary marker of intelligence. This book is a bit more than 10 years old and the industry certainly seems to have invested heavily in the science of prediction whether this was due directly to his influence or not. Having dabbled a little with data science technology I appreciated the framework for the clarity it can bring to the learn, model, test, predict, repeat steps in much of the predictive analytics space. Learning a bit of brain science was also enjoyable.

Finally one subject the author didn't spend a lot of time on it in this book, though apparently has elsewhere - his prediction that neural networks will not be the AI of the future. It really is his primary premise that while we should learn from the biological model we shouldn't think that computers will mimic the hardware used to implement that model. In fact a human brain has so many neurons connected in so many ways that a general purpose AI would be impossible to model with an artificial neural network. Further reading on how this argument may have changed in the last 10-12 years will be interesting follow-up.
J. Edward
A while back, I saw an episode of Wired Science on PBS, featuring Jeff Hawkins (he founded Palm Computing) talking about the area of study that's pulled him in repeatedly: neuroscience. His description of the neocortex, including its similarity in size and thickness to a cloth dinner napkin and that thin layer of cells' pretty much *being* the thing that makes us human intrigued me. So, I bought his book.

On Intelligence is the book on this list that took me the longest to actually get through. I A while back, I saw an episode of Wired Science on PBS, featuring Jeff Hawkins (he founded Palm Computing) talking about the area of study that's pulled him in repeatedly: neuroscience. His description of the neocortex, including its similarity in size and thickness to a cloth dinner napkin and that thin layer of cells' pretty much *being* the thing that makes us human intrigued me. So, I bought his book.

On Intelligence is the book on this list that took me the longest to actually get through. It's not particularly long or even hard to read. However, every chapter led me to ponder quite a bit. As a result, I tended to read this one in fits and starts over a few months.

The central premise is his theory and the science to back it up focuses on the general algorithm for the neocortex. Oversimplified, every portion of the neocortex just watches for and stores patterns, combining them and replaying them. That goes for sensory input, our own motor control, etc.

Ever since reading this book, I've been seeing more and more in day to day life that fits with this theory. Should his model for how the brain works turn out to be completely right, it will be huge, particularly in the area of computer-based artificial intelligence.

I fully expect to continue mulling this one over for months and years to come.
Darren
Interesting book by Jeff Hawkins on intelligence, and how our understanding of the way the brain learns can help us recreate intelligence artificially. If I read it correctly, I think it goes something like this:

Using the 6-layered cortex as a hierarchical information system, lower levels send collections of input to higher cortex layers as patterns (which Hawkins calls "names") , and the higher levels use the same cortical algorithms to decode the information (and return signal when necessary, Interesting book by Jeff Hawkins on intelligence, and how our understanding of the way the brain learns can help us recreate intelligence artificially. If I read it correctly, I think it goes something like this:

Using the 6-layered cortex as a hierarchical information system, lower levels send collections of input to higher cortex layers as patterns (which Hawkins calls "names") , and the higher levels use the same cortical algorithms to decode the information (and return signal when necessary, eg: from somatosensory cortex to the legs muscles to react to predatory stimulus and run away). This hierarchical grouping of information patterns into names allows for expected information to be ignored and only "new" or unexpected information to be passed to the next level. The design of different neurons and their duplex synapse pathways at each layer determine how effectively this information passes in each direction. New information that was NOT predicted gets passed up the next level until it matches something that "is" predicted, making sense of the input. If it makes it all the way up, then it is truly something NEW, and gets stored as a memory by working with the Hippocampus. The more these pathways fire together, the more likely this NEW input will become a pattern (name) that will be recognized the next time, and therefore "learned". Cool stuff.
Nathan
Quite an amazing book. It has definitely influenced many of my views on the mind and the brain. I recommend this book to everyone, even though there are some parts a bit harder to get through if you are less interested in the machine side of things.

Most of the fundamental ideas in this book about how the brain thinks and learns are easy to grasp and simple in concept, yet far reaching in implication. It can even change how you think about yourself and why you do the things you do. It's really gr Quite an amazing book. It has definitely influenced many of my views on the mind and the brain. I recommend this book to everyone, even though there are some parts a bit harder to get through if you are less interested in the machine side of things.

Most of the fundamental ideas in this book about how the brain thinks and learns are easy to grasp and simple in concept, yet far reaching in implication. It can even change how you think about yourself and why you do the things you do. It's really great to have the ideas laid out and described in such detail.

My only real criticism as a lay person is that Hawkins can come across as too into himself at both the beginning and end portions of the book. The actual content of the book is great though and I do think I agree with the basic principles. It just makes sense and is consistent with everything I know about how we behave. This is what it means to learn and what true intelligence actually is.

I would argue that when it comes to building intelligent machines he might devalue the importance of some human characteristics (emotion, purpose, desire) and the role they might play in creating intelligence. Yes machines don't require these things to be intelligent in a technical sense but I personally think they might be invaluable in creating the environment for intelligence to arise.
John Martindale
I found a review from audible by a fellow named Joseph that captured exactly my reaction to this book. Joseph wrote

"No doubt Jeff Hawkins is a brilliant cortex, given that each one of us and the world we live in is nothing more than the experience of an active cortex. But he is not a wise human, which, in my mind, is the greatest achievement of homo sapiens, not the ability to recreate intelligence in a machine.

It is telling that he admits to never studying the nature of consciousness, but in o I found a review from audible by a fellow named Joseph that captured exactly my reaction to this book. Joseph wrote

"No doubt Jeff Hawkins is a brilliant cortex, given that each one of us and the world we live in is nothing more than the experience of an active cortex. But he is not a wise human, which, in my mind, is the greatest achievement of homo sapiens, not the ability to recreate intelligence in a machine.

It is telling that he admits to never studying the nature of consciousness, but in one pithy statement of "fact", reduces it to one thing - the neo-cortex experience. When the cortex ceases to function, so does consciousness. It's not that he holds this "belief" that is troublesome, but rather the arrogance with which he holds his cortex-created model of the world to be "the truth."

This book is an excellent example of the scientist telling us how the trunk of an elephant works and the value of putting that information to work for us humans, but the consistent conclusion that the elephant IS the trunk is tiresome, offensive, and indicative of an immature soul.

Not being schooled in AI or neuroscience, I have no judgment on his theory - other than it is too reductionistic in general - and as one who has a passion for understanding as much as I can about being human, the discussion of how the brain works and what a model of this might be was enjoyable."
Vincent Russo
Jeff Hawkins was the inventor of the Palm Pilot, and he has since turned his creative and research endeavors into the realm of building intelligent machines, which led him primarily to the study of brains and neuroscience. This book is a overarching look at his approach to brain science, and how one would and should approach the task of creating intelligent machines.

The book criticizes some of the tried and true methods of artificial intelligence, as an admirable, but inevitably prone to failur Jeff Hawkins was the inventor of the Palm Pilot, and he has since turned his creative and research endeavors into the realm of building intelligent machines, which led him primarily to the study of brains and neuroscience. This book is a overarching look at his approach to brain science, and how one would and should approach the task of creating intelligent machines.

The book criticizes some of the tried and true methods of artificial intelligence, as an admirable, but inevitably prone to failure approach. Hawkins also takes on certain ideas not conventionally held in the domain of neuroscience, for instance, the fact that sight, sound, taste, etc. are all in fact manifestations of the same thing. In other words, the input of outer stimuli from the outside world get fed into the brain in terms of some information / signal, and it is the brain which processes this signal into the different aspects of sight, touch, etc.

Not being a specialist in neuroscience, it's difficult to tell precisely when Hawkins is making a worthwhile claim that doesn't go against hard evidence. It seems that there are many good ideas presented in the book, and I would be interested to hear someone's opinion who is more immersed in this field of study.
Russell
This books was not what I expected as I was thinking it was going to be about artificial intelligence in the software world. But, I was pleasantly surprised to find out that it is instead a book about Hawkins' personal theories as to how the brain works and how the popular AI approaches are (in his opinion) seriously missing the mark.

The reading flows pretty well, there are good examples and diagrams, but the most difficult aspect for me was how near it felt to a textbook. He writes very clearl This books was not what I expected as I was thinking it was going to be about artificial intelligence in the software world. But, I was pleasantly surprised to find out that it is instead a book about Hawkins' personal theories as to how the brain works and how the popular AI approaches are (in his opinion) seriously missing the mark.

The reading flows pretty well, there are good examples and diagrams, but the most difficult aspect for me was how near it felt to a textbook. He writes very clearly and candidly which helped keep my attention. In addition to my curiosity about neocortical learning patterns, his assertions and postulating on the workings of the mind kept me engaged.

If I had to guess, I would bet that Hawkins is an atheist. While the text is quite scientific in its analysis, he clearly draws on evolution. I respect his thought processes and summaries, but to me the most interesting aspect of reading this book was to combine my own ponderings of how the soul and mind work in harmony as I read his ideas. That was quite fascinating and I think it'd be great to have conversations about this book with others that also consider the integration spiritual elements in the human mind.
Bruce
Presents a new framework for understanding the function of the neocortex. The main idea - that the cortex functions as a hierarchical feedforward/feedback system - is quite compelling. Learning this was worth 4 stars. Flaws: Hawkins took too long to get into the meat of the book, was too soft on the science (for my taste), and finished with an airy, useless discussion of quasi-philosphical issues and applications of "intelligent machines." The new framework presents a host of scientific issues ( Presents a new framework for understanding the function of the neocortex. The main idea - that the cortex functions as a hierarchical feedforward/feedback system - is quite compelling. Learning this was worth 4 stars. Flaws: Hawkins took too long to get into the meat of the book, was too soft on the science (for my taste), and finished with an airy, useless discussion of quasi-philosphical issues and applications of "intelligent machines." The new framework presents a host of scientific issues (details!!!) that Hawkins ignores. One major issue, which Hawkins breezes past and almost certainly gets wrong, is the fitting of inherited behaviors into the new framework. He makes some pretty strong statements along the lines that the mind is tabula rasa upon birth. Wrong (see, e.g. new born pups go directly for food), and an assumption not remotely demanded by the new framework. From a geeky engineering standpoint, I'd have liked to see a mention of Kalman filters, a pervasive engineering technique (avionics to speech enhancement) that bears similarities to the author's theory. Could go on and on with science issues. Nonetheless, as mentioned, a more than worthwhile read because the theory is so powerful.
Preston Lee
Fascinating read on the nature of human intelligence, written by a scientist and engineer not originally trained with basic neuroscience theories past the high school level.

On Intelligence presents a model of human cognitive processing based on the hierarchical, tree-like structure of neurons throughout the neocortex. These "columns" of information, over time, develop feedback loops that yield a massive, temporal, parallelized pattern-recognition system. (e.g. The sound of a leaky faucet is iden Fascinating read on the nature of human intelligence, written by a scientist and engineer not originally trained with basic neuroscience theories past the high school level.

On Intelligence presents a model of human cognitive processing based on the hierarchical, tree-like structure of neurons throughout the neocortex. These "columns" of information, over time, develop feedback loops that yield a massive, temporal, parallelized pattern-recognition system. (e.g. The sound of a leaky faucet is identifiable by both sight AND sound. The concept is "invariant", and can be recognized as the same thing regardless of the human "sensor" that makes the observations.) These theories are presented with a slew of examples, and seed the idea that, in time, we may be able to change our focus on predicate logic and artificial neural network-based machine learning approaches to something more resemblant of our innate biology.

I highly recommend Hawkins' On Intelligence to anyone with an interest in the human brain, biological or artificial intelligence, or machine learning. It is not a "technical" book--making it great for computer geeks--but neither is it a "quick read". Expect frequent trips to Wikipedia. :)

Overall: 4.5
Leave Feeback for On Intelligence: How a New Understanding of the Brain Will Lead to the Creation of Truly Intelligent Machines
Useful Links