kitchen table math, the sequel: Search results for Hawkins
Showing posts sorted by relevance for query Hawkins. Sort by date Show all posts
Showing posts sorted by relevance for query Hawkins. Sort by date Show all posts

Monday, August 13, 2012

Jeff Hawkins: memory is intelligence, intelligence is memory

I've just read the transcript of Hawkins' talk, which is posted along with the video.

Maybe this is overstating matters, but on a quick read-through of the transcript my impression is that pretty much everything Hawkins says it as odds with pretty much everything constructivist educators believe:

Intelligence is not behavior.

Intelligence is not computation.

Intelligence is memory.

Memory is memory of sequence.

The point of memory is to predict what comes next.
"[T]he neocortex is just memorizing."
"You cannot learn or recall anything outside of a sequence."
"[I]ntelligence is defined by prediction."
"[P]rediction of future inputs is the desired output."
The education establishment has for many years denigrated both memory and sequence in favor of critical thinking, problem solving, spiraling and history taught as themes instead of narratives.

Meanwhile actual experts persist in knowing stuff and in organizing the stuff they know in coherent sequences.

On the other hand, constructivists have picked up on the idea of 'pattern' and 'prediction' in a way instructivists arguably have not....but the injunction that students must 'look for a pattern' (math) and 'make predictions' (reading) seems often to be a means of avoiding sequence (math) and the kind of ordinary nouns-come-after-prepositions-type prediction Hawkins is talking about.

I don't know whether Hawkins is right, of course. Reading the transcript, I wanted to hear him talk about cognitive illusions and the invisible gorilla. (I don't remember whether he discusses illusions in his book with Sandra Blakeslee.)

From the transcript :
So what is the intuitive, but incorrect assumption, that's kept us from understanding brains? Now I'm going to tell it to you, and it's going to seem obvious that that is correct, and that's the point, right? Then I'm going to have to make an argument why you're incorrect about the other assumption. The intuitive but obvious thing is that somehow intelligence is defined by behavior, that we are intelligent because of the way that we do things and the way we behave intelligently, and I'm going to tell you that's wrong. What it is is intelligence is defined by prediction.

[snip]

The AI people said, well, the thing in the box is a programmable computer because that's equivalent to a brain, and we'll feed it some inputs and we'll get it to do something, have some behavior. And Alan Turing defined the Turing test, which is essentially saying, we'll know if something's intelligent if it behaves identical to a human. A behavioral metric of what intelligence is, and this has stuck in our minds for a long period of time.

Reality though, I call it real intelligence. Real intelligence is built on something else. We experience the world through a sequence of patterns, and we store them, and we recall them. And when we recall them, we match them up against reality, and we're making predictions all the time. It's an eternal metric.

[snip]

You're all being intelligent right now, but you're not doing anything. Maybe you're scratching yourself, or picking your nose, I don't know, but you're not doing anything right now, but you're being intelligent; you're understanding what I'm saying. Because you're intelligent and you speak English, you know what word is at the end of this -- (Silence) sentence.

[snip]

You still have that alligator brain. You do. It's your emotional brain. It's all those things, and all those gut reactions you have. And on top of it, we have this memory system called the neocortex. And the memory system is sitting over the sensory part of the brain. And so as the sensory input comes in and feeds from the old brain, it also goes up into the neocortex. And the neocortex is just memorizing. It's sitting there saying, ah, I'm going to memorize all the things that are going on: where I've been, people I've seen, things I've heard, and so on. And in the future, when it sees something similar to that again, so in a similar environment, or the exact same environment, it'll play it back. It'll start playing it back. Oh, I've been here before. And when you've been here before, this happened next. It allows you to predict the future. It allows you to, literally it feeds back the signals into your brain; they'll let you see what's going to happen next, will let you hear the word "sentence" before I said it. And it's this feeding back into the old brain that'll allow you to make very more intelligent decisions.

[snip]

So what is the recipe for brain theory? First of all, we have to have the right framework. And the framework is a memory framework, not a computation or behavior framework. It's a memory framework. How do you store and recall these sequences or patterns?

Sunday, June 10, 2007

On Intelligence



100 pages into On Intelligence and.... wow

If this guy is right, and he sounds right to me, our constructivist friends should quit while they're ahead. They are sooooooo not 21st century.
The book is a manifesto for a theory based on an elegantly simple premise: that intelligence is rooted in the brain's ability to access memories rather than in its ability to process new data.
Access memories.

Not process new data.

That would mean.... knowing stuff is the essence of intelligence.
Hawkins complains that thinking machines do not exist because scientists have been sidetracked. Too many subscribe to the widespread view that the brain is essentially a powerful computer, constantly processing and integrating incoming data. [ed.: processing and integrating incoming data - kind of like Piaget! except that's not what makes you smart! knowing stuff makes you smart!] Many computer experts believe that thinking machines will arrive once there are processors as powerful and as integrated as human neurons. There is a problem with this brain-as-computer analogy, however: Computers are already faster than brains. A neuron can manage 200 operations per second; a modern computer can race through 1 billion per second.

fast and stupid
Processing speed doesn't matter in the brain, says Hawkins, because the basis of thought is not data manipulation but memory retention and prediction. The brain, he says, accesses previous experiences, compares them with existing circumstances, and predicts what is most likely to happen next. When a ball is thrown, for example, we know from experience where it is most likely to land and move our hands to that spot. It's a simple action, but it has proved nearly impossible to build a robot smart enough to perform it. "The brain doesn't compute the answers to problems," he says. "It retrieves the answer from memory."
Very cool.
Intelligence, posits Hawkins, is essentially the capacity to remember and then predict patterns in the world.

Source:
Redefining Smart
Business Week
November 8, 2004
I joined the mailing list.

Jeff Hawkins' Bold Brainstorm
21st century skills
On IntelligenceWayne Gretsky explains the world to you

Sunday, August 12, 2012

Jeff Hawkins TED talk

Haven't watched it yet, but it's probably worth the time.

Here's the blurb:
Treo creator Jeff Hawkins urges us to take a new look at the brain -- to see it not as a fast processor, but as a memory system that stores and plays back experiences to help us predict, intelligently, what will happen next.

Jeff Hawkins pioneered the development of PDAs such as the Palm and Treo. Now he's trying to understand how the human brain really works, and adapt its method -- which he describes as a deep system for storing memory -- to create new kinds of computers and tools.

Wednesday, July 25, 2007

cortical homunculus



image posted at Mind Hacks

This model shows what a man's body would look like if each part grew in proportion to the area of the cortex of the brain concerned with its sensory perception.
Natural History Museum


I'm reading Rodolfo Llinas' I of the Vortex.

The book is pretty spectacular, possibly life-altering.

First time I've ever felt queasy reading about neuroscience, though.





update from Mr. Person:

omg

I'm feeling faint.


developing intelligence summary/review
Chris Chatham home page
interview with Rodolfo Llinas
recent publications
I of the Vortex (MIT Press)

The Future of the Brain by Steven Rose
On Intelligence by Jeff Hawkins
Jeff Hawkins posts

Tuesday, February 26, 2008

should I install the wavy letter thingie?

I see Jason is back.

Should I go ahead and install the wavy-letter-thingie that prevents spam?

I hate to do it because of the FWOT aspect, but I think we're at that point.

Speaking of the wavy-letter-thingie, Jeff Hawkins' book explains why it works.

Tuesday, July 3, 2007

yet another surprise from the animal world

Probably at least half the animal studies I read open with some variant of we thought only people did this stuff.

As a matter of fact, the entire history of research on animal behavior could be summed up as a protracted exercise in goalpost moving.

(I'm joking.)

btw, I now believe animals have language.

I can't prove it. I can't even argue the case competently.

But I believe it. I guess I believe it thanks to my Bayesian brain (and a glimmer of an insight that Jeff Hawkins' argument will lead to the conclusion that animals have language....) I've read SOOOO many articles that start out with "We scientists thought only humans do this" that if someone forced me to bet money on whether animals do or do not have language, I would have to bet yes. I just don't see enough, "We scientists always thought animals couldn't do X and, sure enough, animals can't do X" articles to make me bet "no."

In other words, I'm seeing a trend.

(or a pattern!)

Thinking that animals have language is heresy, in case you didn't know.* Ed, as it turns out, did not. I dropped my animals-have-language bomb the other morning at breakfast and he said, "Didn't I read an article saying animals have language?"

answer: no

You did not "read an article" saying animals have language. You may have read a book coauthored by your wife saying that animals (might) have language, or something close to it.

It's funny how one person's heresy is another person's misconception.

Anthropologists have long recognised altruism as a vital element in allowing complex social groups to form. This raised the question of when it first evolved. True altruism has always been seen as a uniquely human trait in that only people were thought capable of deliberately helping others, knowing there would be a cost to themselves.

Warneken and his colleague Brian Hare carried out the study on 36 chimpanzees at the Ngamba Island Chimpanzee Sanctuary, in Uganda.

The animals were not allowed to interact with any humans they knew or had received food from.

In the first experiment, the chimpanzee saw a person unsuccessfully reach through the cage bars for a stick on the other side, too far away for the person but within reach of the ape.

The chimpanzees spontaneously helped the reaching person regardless of whether this yielded a reward or not. When the chimpanzees could see the person making no effort to reach the object, they did not help.

A second experiment was designed to make it much harder to offer help, with the chimps forced to climb 6ft to get the stick. There were no rewards but, again, the animals still helped.

A third piece of research looked at the apes’ willingness to help each other. One chimp was made to watch as a second animal tried to get into a closed room containing food.

The only way it could get in was if the watching animal removed a chain on the door. In each case they did so.

Elsewhere in the animal world there are many examples of apparent altruism. Dolphins, for example, will support sick or injured animals, swimming under them for hours at a time and pushing them to the surface so they can breathe.

source:
Chimps beat us to that human touch


yet another surprise from the animal world
are meerkats constructivists?


* It may not be heresy in the UK. Not sure.