Jeff Hawkins' "On Intelligence": A Fascinating Read

[Ratzinger]

Has anybody read Jeff Hawkins „On intelligence“?

I found it a fascinating read and his proposed ideas very plausible. Unfortunately there is not much more further reading since literature and research in artificial intelligence is mostly crap (since it is not or no more aiming at understanding and constructing human-like intelligence, only trying to simulate it), and neuroscience so far only dealt with where, what, when questions but not with how the brain works. That at least is what Hawkins claims.

But I would like to know and read more about Hawkins approach. So can anybody here give some comments?

 

[Ratzinger]

So wait a minute, no one of you has read this book? Not even heard of it?

You are kidding me?!

So if you don't believe that this is going to be an important book, check out the reviews. http://www.onintelligence.org/reviews.php

That's what nobel prize winner Eric kandel says

"Jeff Hawkins has written an original, thought-provoking and, with the help of Sandra Blakeslee, remarkably readable book that presents a new theory of the functions of the cerebral cortex in perception, cognition, action and intelligence. What is distinctive about his theory is the original way existing ideas about the cerebral cortex and its architecture have been combined and elaborated based on an extensive knowledge of how the brain works - what Hawkins calls Real Intelligence in contrast to computer-based Artificial Intelligence. As a result, this book is a must-read for everyone who is curious about the brain and wonders how it works. Many sections of this book, especially those on intelligence, creativity and minds of silicon, are so thoughtful and original that they are likely to be required reading for college undergraduates for years to come."

 

[selfAdjoint]

In spite of the rave review, proposed hypotheses about the stucture of the cortex are a dime a dozen. It baffles me why anybody would engage in large scale hypothesis building just now, when we are at last getting a flood of new experimental information about all those functions and are expecting to get much more in the near future. Can't they just wait to see what the data says before starting their hares?

 

[Ratzinger]

I see your point, and you are right to be sceptical about grand explain-it-all theories when it comes to human mind. But please read the book. The book is cheap, and it is a short and an easy read. I would just love to know what you think about it after reading.

 

[neurocomp2003]

i've read the book. But i was immersed in the literature earlier than that book was written...so it was rather naive to me...another book you might considering reading is steve grands old book called "Creation"(not a book on religious views but a book on programming AI) or the Child development book by lisa eliot entitled "wahts going on up there". You should look up the Berkeley facility which Hawkins was once or still is the director of.
Also the blue gene project by IBM and a group in switzerland.

 

[Hercule Poirot]

I've read on intelligence and it was fascinating for me given that I'm new to the subject. The author is negative about AI or what have become of it, but regarding his theory of prediction I can't really judge since my knoweldge of brains is little (I'm planning to take a course in the upcoming semester to remedy this)

En passant by Steve Grand, i just went through his resume in his website. It's interesting that he hasn't any academic degree; nevertheless his book seems to be a good read, thanks for the tip neurocamp2003

 

[Ratzinger]

I like to continue advertising Jeff Hawkins approach here by pointing out that there is now an excellent wiki entry giving you Hawkins ideas in a nutshell.

http://en.wikipedia.org/wiki/Memory-prediction_framework

 

[zoobyshoe]

It baffles me why anybody would engage in large scale hypothesis building just now, when we are at last getting a flood of new experimental information about all those functions and are expecting to get much more in the near future. Can't they just wait to see what the data says before starting their hares?

Amen.

...

 

[neurocomp2003]

becuase they require the code to build it.

 

[Ratzinger]

Let Jeff Hawkins http://mitworld.mit.edu/video/316" to you!

 

[selfAdjoint]

Let Jeff Hawkins http://mitworld.mit.edu/video/316" to you!

Why? Because hearing his voice will persuade me of things his written prose didn't? That's pure magical thinking. I repeat my former objection: https://www.physicsforums.com/showpost.php?p=1021416&postcount=3.

 

[Ratzinger]

Why? Because hearing his voice will persuade me of things his written prose didn't?

Perhaps.

Then who are your favorites in neurosciences? What or whose less magical approaches do you favour? Who should be read and heard instead?

 

[selfAdjoint]

Perhaps.

Then who are your favorites in neurosciences? What or whose less magical approaches do you favour? Who should be read and heard instead?

Well, I am on record here as liking http://psyche.cs.monash.edu.au/#metzinger".

I also like Susan Blackmore, whose opinions you can find at various sites on the web. She used to be a "magical" consciousness believer, but strictly schooled herself out of that philiosophy which had come to seem narrow and unsatisfactory to her. I can relate to that.

 

[Ratzinger]

I own Susan Blackmore's conciousness book which I find terrific. Great account of all approaches to the conciousness problematic eventhough she seems to sympathize a little bit to much with Dennett.

But I still fail to understand what you hold against Hawkins.

 

[eehiram]

If you're still there

I read the first four chapters of Jeff Hawkins' On Intelligence and would like to discuss it too.

I've thought about the definition of intelligence as prediction + behavior many times. At first I felt that intelligence encompasses more than prediction: imaginition, problem-solving, linguistics, artistic creation, etc. Simply opening a door or looking around a room is only one vein through which the mind's intelligence flows through.

Nevertheless, I kept reading, and I summed it up for my junior college librarian in an email. She said she felt behavior was more important than Jeff Hawkins had accounted for.

I disagreed. I felt simply lying in a dark room and thinking about something was still intelligence (perhaps even a more pure form of it), and following thought with action was more of an Aristotelian argument about knowing and doing the good. Albeit Aristotle did a lot of thinking himself as a philosopher, I'm sure he followed it with action much of the time and felt it made the thoughts more relevant to do so.

Hence computers can be intelligent in processing information without any moving parts to interact with objects. My computer lacks a printer most of the time, so it really does nothing but display text and numbers for me. Is it intelligent? Certainly, otherwise it would not be able to calculate e or 1/pi according to the programs I wrote. Might thought processing eventually become an end to itself, without needing to perform any action to make the conclusion worth all the effort? I think maybe so.

And yet, and yet...behavior is still important -- for now -- in demostrating what belief we carry inside or what conclusion we draw.

Well, I hope you find this relevant to the book. This is just my take on the sections I referred to.

o| Hiram

 

[neurocomp2003]

Just a few things you might want to think about when reading that book...
[1] Adaptive Learning
[2] Supervised Learning/Child Development
[3] How does intelligence arise from child dev.?

 

[eehiram]

Can someone elaborate a little more than that? Just briefly will suffice (just not as briefly, however).