AI reality check?

Discussion in 'Intelligence & Machines' started by kmguru, Apr 16, 2008.

Thread Status:
Not open for further replies.
  1. kmguru

    kmguru Staff Member

    Messages:
    11,757
    Opinion: Time for another AI reality check?


    Bill Schweber
    (04/02/2008 2:58 PM EDT)
    URL: http://www.eetimes.com/showArticle.jhtml?articleID=207001349

    I love to read stories about how software and hardware that will emulate the brain is just around the corner. Of course, this "corner" always seems to be 2, 5, 10, or even 20 years away. A recent example was presented at the International Solid-State Circuits Conference (ISSCC), "Keynoter says chips will someday mimic human brain," where computer expert Jeff Hawkins said great progress has been made, and gave some specifics.

    I love these stories because they actually prove that we really haven't got a clue about how the brain works. Look at research papers from brain and artificial intelligence experts: Despite the progress, there are embarrassing gaps in understanding.

    Consider the Defense Advanced Research Projects Agency competition, which brings together very innovative and proficient teams to develop a self-driving vehicle using megaflops of processor power, kilowatts of dissipation, millions of lines of code and an array of sensors to do what just about any person can do, under a wider set of circumstances and environments: drive a car.

    All we need is a three-pound, 75-watt processor supplemented by a few peripheral actuators with position and motion feedback. And unlike Darpa vehicles, our input sensor is just a visible-light array without sonar, thermal, laser, radar or other supplemental imaging support.

    We don't know if the brain actually translates and stores information, and, if it does, in what format. Does it digitize all the images, sounds and experiences? If so, at what resolution? (Highly unlikely: There's no MSPS, high-resolution A/D and D/A subsystem.)

    How does the brain organize, sort, search, pattern-match and retrieve memories and information? How does it do this so quickly? How does it implement background searches so the name of that movie you are trying to recall finally "pops" into your memory an hour later?

    How does the brain figure out how to do what has to be done? How does it learn to learn, and dynamically reconfigure itself for different tasks and priorities? How is it able to recall and regenerate sounds, images, movies or songs with so little apparent effort?

    Why can we drive, run and catch a ball with relative ease? There is no megaflop processor and no organized code listing. Even if we postulate, for simplicity, that the brain is a parallel processor with near-infinite memory, how does it do what it does?

    Clearly, the brain is configured and works in ways we don't grasp. All those live-action brain scans, which we now see everywhere, obscure that reality. Perhaps someday, artificial intelligence plus radically new computer architectures will allow us to build brain-like systems.

    But until we acknowledge our gaps in understanding—and they are huge—more powerful processors, slightly different architectures and better software won't do it.
    --------------------------------------------

    Very interesting...
     
    kmguru, Apr 16, 2008
    #1

  2. Google AdSense Guest Advertisement



    to hide all adverts.
  3. Xelios

    Xelios We're setting you adrift idiot Registered Senior Member

    Messages:
    2,447
    I've been reading The Holographic Universe by Michael Talbot lately, and though most of the book just tries to convince the reader of the theory without actually explaining much he does make some interesting points.

    When you make a hologram by splitting a laser beam you end up with a piece of holographic film that's just unintelligable interference patterns. Shine a laser through it again and you get a 3D image of the original object. Cut the film into smaller pieces and each piece will show the same intact object, though it gets fuzzier as you cut the pieces smaller. You can even make a more complex hologram that will deliver a different image depending on the angle at which the laser hits the film.

    He contends the brain (well, the whole universe really) works the same way, through interference patterns. The electrical impulses being fired around by billions of neurons act as waves which interfere with each other, and consciousness is like this reference laser, reconstructing memories and other things from the interference.

    Though he never explains how this all works (not finished the book yet) the general idea could explain a lot. Memories don't seem to be stored in any one place in the brain, they're everywhere all at once. Hearing works on the same principle, the ear seems to have its own reference beam that allows the brain to decifer the position of sounds in 3D space (google Holophonic sound). And there's tons more examples in the book.

    Point is exactly what the article says, we still have no idea how the brain really works, and without understanding that we have no hope of duplicating it on a completely different base (like binary logic). We have a good idea how the hardware works, the chemicals and neurons and so on, but we know nothing about the software running it all.

    Besides, there's no reason to think AI running on 1's and 0's would be anything like us at all, we may not even recognize it as intelligent if we ever succeed in creating one.
     
    Xelios, Apr 17, 2008
    #2

  4. Google AdSense Guest Advertisement



    to hide all adverts.
  5. redarmy11

    redarmy11 Registered Senior Member

    Messages:
    7,658
    Here's where we're at currently: partial simulation of rat brains at a costly one microprocessor per simulated brain cell:

    http://www.spiegel.de/international/spiegel/0,1518,466789,00.html
    http://bluebrain.epfl.ch/Jahia/site/bluebrain/op/edit/pid/18699

    Looks like that Eureka! moment is some time off yet. And yet:
    Fascinating stuff. Watch this space.
     
    Last edited: Apr 17, 2008
    redarmy11, Apr 17, 2008
    #3

  6. Google AdSense Guest Advertisement



    to hide all adverts.
  7. kmguru

    kmguru Staff Member

    Messages:
    11,757
    I do not think the brain works on the holographic theory. I think, it works on fractal math. Our chemical equilibrium such as sugar, oxygen saturation, heart rate, blood volume and pressure etc works on a mathematical basis that follows advanced control theory i.e. feedback, feedforward loops with right type of gain and error integration etc.

    So, my thinking is that the memories are data that can be reduced to a fractal math to store the mathematical representation of that time series data. Therefore it can be retrived faily well. The fuzziness comes from when the data is further compressed like those large movies that are compressed using Divx math. Except that humans can uncompress the data intelligently from the data stored separately. For example, there is no reason to store the picture of a pencil in a snapshot, a hundred times. Just the mathematical representation of the pencil or just the object description.

    Just some thoughts to noodle through....
     
    kmguru, Apr 17, 2008
    #4
  8. kmguru

    kmguru Staff Member

    Messages:
    11,757
    Too many cooks?

    It is like some one got a quad processor in 1950s and trying to reverse engineer it with vaccum tubes. Good luck with that....

    Please Register or Log in to view the hidden image!

     
    kmguru, Apr 17, 2008
    #5
  9. krokah

    krokah Registered Senior Member

    Messages:
    198
    What about the new teraflop processing chip that I read about not to long ago. Developed by HP and Microsoft I think.They were postulating that with this new processor it will radically change the concept of AI. It is so fast that it burns through current programs. It will be like have a Cray computer system on a wristwatch. The article stated that all hardware and software will have to replaced (don't we do this no a regular basis when we upgrade?). Hope it's not a lot of BS, article was posted in the Sacramento Bee in Sacramento California in 2007, not sure of the month.
     
    krokah, Apr 17, 2008
    #6
  10. kmguru

    kmguru Staff Member

    Messages:
    11,757
    I do not want to be the naysayer, but just raw horsepower would not cut it. We need software to go with it. That is where we do not know, how to write one. The Japanese have been trying for a long time, so is NASA. That does not mean, it will never happen....it means that one of these days....

    If you find some interesting software coming out of university labs...let us know. We need a black swan....

    Please Register or Log in to view the hidden image!

     
    kmguru, Apr 19, 2008
    #7
  11. Xelios

    Xelios We're setting you adrift idiot Registered Senior Member

    Messages:
    2,447
    I'll be studying in Saarbrucken, Germany this winter semester. On campus is the "Deutsches Forschungszentrum für Künstliche Intelligenz" (German Research Centre for Artificial Intelligence). I originally planned to get into security evaluation for my degree, but maybe I'll do AI instead. As long as I'm not responsible for Skynet later on

    Please Register or Log in to view the hidden image!

     
    Xelios, Apr 19, 2008
    #8
  12. Fabio4all

    Fabio4all Registered Member

    Messages:
    62
    exactly. Every science magazine I read seems to say science's biggest mystery is the human mind. How can we reverse-engineer and eventually recreate the human mind when we can barely speculate about it?
     
    Fabio4all, Apr 20, 2008
    #9
  13. kmguru

    kmguru Staff Member

    Messages:
    11,757
    Now that the cat is out of the bag...we will blame you for the skynet....

    Please Register or Log in to view the hidden image!



    Seriously, that is a good idea. PM me when you start working on a project. By the way, AI or close to it can help in the security matters specifically making firewalls and routers intelligent.
     
    kmguru, Apr 20, 2008
    #10
Thread Status:
Not open for further replies.

Share This Page