Cool Cool / Items

Brain, Mind, Consciousness and Learning: Jeff Hawkins: Brain science is about to fundamentally change computing

Get Feed
Brain, Mind, Consciousness and Learning: Jeff Hawkins: Brain science is about to fundamentally change computing
Description
To date, there hasn't been an overarching theory of how the human brain really works, Jeff Hawkins argues in this compelling talk. That's because we still haven't defined intelligence accurately. But one thing's for sure, he says: The brain isn't like a powerful computer processor. It's more like a memory system that records everything we experience and helps us predict, intelligently, what will happen next. Bringing this new brain science to computer devices will enable powerful new applications -- and it will happen sooner than you think.
Original URL
http://brainandlearning.blogspot.com/2008/03/jeff-hawkins-brain-science-is-about-to.html

Comments

  • Public Comments

    • Aldo Bucchi Aldo Bucchi
      16 months ago


      The HTM theory is one of the most intriguing, disruptive developments I have seen around.
      I am a steady follower of Numenta and have been using vitamin D since the first release.
      The thing really works...

      HIGHLY RECOMMENDED.
      Cool

    • François Dongier
      16 months ago


      Quotes: "Brains can't understand brains... Very Zen like."

      "What is the intuitive but incorrect assumption that's kept us from understanding brains?"

      Koch & Tononi (http://www.twine.com/item/11vg7qm8h-1p5/ieee-spectrum-can-machines-be-conscious) proposed a nice new "Turing test" based on image understanding which I think is fully consistent with Hawkins' theory of intelligence as prediction: computers able to make good predictions would be able to understand the content of an image.
      Tononi's example is a picture of a scene in which we see all sorts of clues that make it clear that we're in a store; we see two persons talking to each other; we also see the gun in the guy's hand, we see the fear on the other guy's face, his hands up, and then we naturally come to "interpret" the whole thing as an ordinary scene of burglary. Prediction is all over the place in this understanding. We see that it is a store because we predict that someone could enter it, pick things up from the shelves, pay for them, etc. We predict the bad guy pulling the trigger, with salesman blood flooding the place. In other words we match the scene with similar patterns stored in our brains, and this is how we come to give it a meaning.
      Artificial Consciousness: Edelman vs Buddha

    • Javed Alam Javed Alam
      16 months ago


      Here is a famous quote from Sir Arthur C. Clarke:

      "The Information Age offers much to mankind, and I would like to think that we will rise to the challenges it presents. But it is vital to remember that information— in the sense of raw data— is not knowledge, that knowledge is not wisdom, and that wisdom is not foresight. But information is the first essential step to all of these."

      - Arthur C. Clark
      Artificial Consciousness: Edelman vs Buddha

      • François Dongier
        16 months ago


        Interesting. Not sure I see your point here. Are you saying that we need to collect more information? Hawkins is telling us the opposite, i.e. that we now have enough data and that what we need is a theory, a paradigm, a set of concepts (an ontology if you like). In the theory of intelligence he is proposing, things like 'prediction', 'memory', 'pattern matching', 'hierarchical temporal memory', etc are central concepts. How does this relate to the data/knowledge/wisdom/foresight quote?
        Actually this quote was already somewhere in my head, but just with the data/knowledge/wisdom triplet. Foresight is a new element. And clearly foresight is about ability to make predictions. Is this what you had in mind? If so, an interesting thing is that, in this quote, foresight is presented as the "highest" level of intelligence (beyond perception, knowledge and even wisdom, whatever that is). I think that in Hawkins' framework, prediction is used as an explanation of our "basic" mental abilities: perception itself (ability to not just sense our environment but give meaning to it) is explained in terms of the predictions we make.
        Artificial Consciousness: Edelman vs Buddha

    • Javed Alam Javed Alam
      16 months ago


      Aldo

      I get your reference to Numenta http://www.readwriteweb.com/archives/numenta_artificial_intelligence.php

      How is it connected to "vitamin D"?
      Cool

    • Javed Alam Javed Alam
      16 months ago


      Francois

      Good questions. I really do not have an answer to your question but that does not stop me on pontificating on these issues. So here it goes:

      What Hawkins is saying that Neuroscience is piling up too much data about how different stimuli create brain activation patterns and there is very little effort to actually have models or laws like we have in physics. Actually that is not true totally. There is now lot of information about "Fear" and "Brain Reward System" and how it effects behavior. Same goes about "Attention".

      Brain being the ultimate complex adaptive system with billion neurons and trillion interconnection has so far defied any laws that are based upon reductionist science. We simply can not have nice clean laws like laws of Newton or Einstein to model the brain. That is the conclusion so far.

      It has however done one thing interesting. That is reopening the question "what is mind" and "what is consciousness" because Neuroscientist can see the brain activation pattern but they can not locate in the brain where these things are located. Same goes for "free will". With Benjamin Libets experiments it is now shown convincingly that actual action starts way before our conscious brain knows about our intent. The well known concepts from psycho analysis such as "ego" and concept of "I" are also being questioned. The best one can say that there are multiple I's in the brain that do the job when they are needed. The closest one come to a unified "I" is Prefrontal cortex that has functions similar to what we would consider as "I". Most of the brain processing is outside our conscious awareness that is well established by cognitive neuroscientist. At least they piled up enough data to show this part convincingly.

      I think that the traditional way of doing science will be supplemented by "data analysis" and "data mining" and better algorithms for pattern recognition (Artificial Neural Nets??). What we have now is too many algorithms chasing not enough data. Data provides context to a problem. Computer problem solving is context free. It is more like a directed liner or non-linear mapping.

      However, brain being a system of interconnected neurons can be approximated by scale free network theory. There definitely are clusters within brain where connectivity is very high. For example hypothalamus that sits in the middle and directs traffic either to the new brain at the top or the older reptilian brain at the bottom. I am sure as time goes by we will have more accurate picture. That is to establish how brain stores information in neuronal activation system. How imagination/hallucination is created out of simple firing of neurons and how psycho active drugs can change this pattern. That would be the ultimate never ending recursive loop where brain is working on itself to never find itself

      Considering all this AI in the old traditional sense has already evolved and computational neuroscience is creating some very interesting models. I did some work in that area long time ago when ANN become popular in early 90's. I think we will be abandoning linear causality theory and will finally accept non-linear causality that will allow for abrupt phase changes of systems.

      The parallels between brain organization and the emerging world wide web are very striking. The network structure studies by Watts and Barbarossa and many others have lot of potential of providing insight into the behavior of these networks but not the certainty we are looking for.

      For complex systems certainty is the first thing that gets sacrificed. If they were predictable why would we call them complex system that would be a mis- characterization. Wouldn't it??
      Artificial Consciousness: Edelman vs Buddha

    • Gabriel Kent Gabriel Kent
      16 months ago


      great vid.
      Brain, Mind and Consciousness
    Add a Comment
Report This

Twine is about discovering, collecting and sharing the content that interests you. Learn More

Join Twine

Stats

First Posted By

Javed Alam

First Comment By

Aldo Bucchi

Tags

Community Tags

Who's Interested In This?

Forgot your password?