Learning from the Future, Episode 6 with Nova Spivack « Live on RapidStage!

Get Feed

Description

A 5 minute video with Nova Spivack. In the interview he talks about the semantic web (putting meta data into data), the web is the database (linked data) and that Facebook will become a search engine. In ten years he expects reasoning to be online. Not Artificial Intelligence, but assisting agents. He doesn’t believe in Artficial Intelligence.

Original URL

http://rapidstage.wordpress.com/2008/04/07/learning-from-the-future-episode-6-with-nova-spivack/

http://rapidstage.wordpress.com/2008/04/07/learning-from-the-future-episode-6-with-nova-spivack/

Comments

• Public Comments

  • David Scott "Lightman" Lewis

    23 months ago

    
    
    

    Isn't Ray Kurzweil the only person who still believes in AI? ;-) But I still love him, even follow his supplements regimen (not really, but I try).

    Isn't Ray Kurzweil the only person who still believes in AI? ;-) But I still love him, even follow his supplements regimen (not really, but I try).

    • Some HTML is allowed.
    • Reply
    • Cancel
    • Submit
    Commented in: Web 3.0 - Semantic Web

  • François Dongier

    23 months ago

    
    
    

    "Human consciousness is and will remain a mystery. Machines will excel at automation, but will never be conscious."
    Well well... I just hope you're wrong here. Concentrating on artificial stupidity (automation) may be necessary in the short-term, but hopefully we won't have to stop there.

    "Human consciousness is and will remain a mystery. Machines will excel at automation, but will never be conscious."<br />Well well... I just hope you're wrong here. Concentrating on artificial stupidity (automation) may be necessary in the short-term, but hopefully we won't have to stop there.

    • Some HTML is allowed.
    • Reply
    • Cancel
    • Submit
    Commented in: Artificial Intelligence

    • Nova Spivack

      23 months ago

      
      
      

      It would be great to be able to make conscious machines, but I just can't see how that will ever happen. The more you explore consciousness the more of a mystery it becomes. It certainly does not seem to be a physical material substance or aggregation, and it in fact is unfindable even subjectively, yet at the same time it is undeniable. The only other things like this are space, time and energy -- they are fundamental axioms of our universe. You can't really detect them directly. Consciousness seems to be like this too. I doubt we can create consciousness any more easily than we can create space, time or energy.

      It would be great to be able to make conscious machines, but I just can't see how that will ever happen. The more you explore consciousness the more of a mystery it becomes. It certainly does not seem to be a physical material substance or aggregation, and it in fact is unfindable even subjectively, yet at the same time it is undeniable. The only other things like this are space, time and energy -- they are fundamental axioms of our universe. You can't really detect them directly. Consciousness seems to be like this too. I doubt we can create consciousness any more easily than we can create space, time or energy.

      • Some HTML is allowed.
      • Reply
      • Cancel
      • Submit
      Commented in: Artificial Intelligence

      • Scott White

        23 months ago

        
        
        

        I personally think it's way too early in the study of the mind and brain to make any conclusive judgements about consciousness. We just don't know what we don't know.

        I personally think it's way too early in the study of the mind and brain to make any conclusive judgements about consciousness. We just don't know what we don't know.

        • Some HTML is allowed.
        • Reply
        • Cancel
        • Submit
        Commented in: Artificial Intelligence

        • Nova Spivack

          23 months ago

          
          
          

          Well my perspective on this comes from several pools of experience: (1) studying computer science, cognitive science and AI, (3), working on a number of AI systems professionally, (2) Buddhist scholarly and contemplative traditions which have been investigating this particular topic ("what is consciousness?") for 2500 years. It would take hours to really explain my reasoning on this point, and I've written about it elsewhere in any event. But suffice to say, while I agree the study of the mind and brain are in their infancy (in the West), I still am confident that consciousness is not going to ever be "explained" (apologies to Daniel Dennett) let alone be synthesized.
          
          Let me offer an analogous case: suppose I said that space-time can be replicated on a computer someday. Is it possible, given what we know about computers and space-time today, to conclusively state that this will never happen?
          
          There has been a lot of speculation in the emerging field of "digital physics" (particularly from people working on cellular automata, and more recently on loop quantum gravity, as well as folks from the virtual reality camp) that our entire universe may in fact be a giant computer simulation running on some cosmic computer. That is the principle behind the film, The Matrix, for example.
          
          In order to evaluate this question we might ask, is there anything about space-time (in other words, all space and time) that is NOT merely reducible to binary data and Turing Machine equivalent computation?
          
          In answer to this: We don't yet know whether the universe is in fact computable. Quantum mechanics is not yet fully understood, and the elusive unification of quantum mechanics and general relativity has not happened yet. It may be the case that universe is not computable at all, that it does not obey the laws of logic at a fundamental level, and/or that universal computations are formally undecidable from the perspective of our finite logic and mathematics. Space-time may be vastly more complex than can imagine, and indeed present discoveries indicate that it may even be infinite in every dimension. One thing we DO know however is that a computer, as we define it today, is not capable of handling anything that is infinite in a finite amount of time. Yet the possibly infinite realm of space-time is "computed" in every seemingly finite moment of time. How does this happen? One answer might be that the universe is a quantum computation running on a cosmic quantum computer because a quantum computer can do an infinite amount of computation in a finite amount of time. But this is not a sufficient answer to our problem because it is a circular argument that is equivalent to saying "you need the universe to compute the universe." A quantum computer is based on the underlying physical substrate that we call "space-time" for its actual computation. Thus if we state that space-time is a quantum-computation running on a cosmic quantum computer then we are saying that underlying computer is itself running on space-time. In other words it is saying that space-time is computed by space-time. So this line of attack does not answer the question at all; it just pushes it down a level. As the saying goes, "it's turtles all the way down."
          
          A similar argument can be leveled against the possibility of ever computing consciousness. Consciousness is not necessarily a finite phenomena. In fact, direct contemplative observation of consciousness indicates that it is infinite in scope: every individual's consciousness can be cultivated to encompass an infinite expanse of space and time. From a scientific perspective consciousness cannot even be directly measured, let alone detected. Stranger still is the emerging realization that there is a direct quantum interaction between observation and experiment, such that it is really impossible to say that consciousness is separate from the underlying fabric of space-time. If consciousness is somehow connected directly to space-time then it too is just as infinite and potentially non-computable.
          
          Some might argue that consciousness is even more fundamental than space-time, but it is enough to simply state that it is equally fundamental to establish the non-computability of consciousness. If consciousness is at least equally fundamental to space-time then any computer which can compute it must actually invoke it to do the computation, and that is a circular argument.
          
          I rest my case.

          Well my perspective on this comes from several pools of experience: (1) studying computer science, cognitive science and AI, (3), working on a number of AI systems professionally, (2) Buddhist scholarly and contemplative traditions which have been investigating this particular topic ("what is consciousness?") for 2500 years. It would take hours to really explain my reasoning on this point, and I've written about it elsewhere in any event. But suffice to say, while I agree the study of the mind and brain are in their infancy (in the West), I still am confident that consciousness is not going to ever be "explained" (apologies to Daniel Dennett) let alone be synthesized. <br /><br />Let me offer an analogous case: suppose I said that space-time can be replicated on a computer someday. Is it possible, given what we know about computers and space-time today, to conclusively state that this will never happen? <br /><br />There has been a lot of speculation in the emerging field of "digital physics" (particularly from people working on cellular automata, and more recently on loop quantum gravity, as well as folks from the virtual reality camp) that our entire universe may in fact be a giant computer simulation running on some cosmic computer. That is the principle behind the film, The Matrix, for example. <br /><br />In order to evaluate this question we might ask, is there anything about space-time (in other words, all space and time) that is NOT merely reducible to binary data and Turing Machine equivalent computation? <br /><br />In answer to this: We don't yet know whether the universe is in fact computable. Quantum mechanics is not yet fully understood, and the elusive unification of quantum mechanics and general relativity has not happened yet. It may be the case that universe is not computable at all, that it does not obey the laws of logic at a fundamental level, and/or that universal computations are formally undecidable from the perspective of our finite logic and mathematics. Space-time may be vastly more complex than can imagine, and indeed present discoveries indicate that it may even be infinite in every dimension. One thing we DO know however is that a computer, as we define it today, is not capable of handling anything that is infinite in a finite amount of time. Yet the possibly infinite realm of space-time is "computed" in every seemingly finite moment of time. How does this happen? One answer might be that the universe is a quantum computation running on a cosmic quantum computer because a quantum computer can do an infinite amount of computation in a finite amount of time. But this is not a sufficient answer to our problem because it is a circular argument that is equivalent to saying "you need the universe to compute the universe." A quantum computer is based on the underlying physical substrate that we call "space-time" for its actual computation. Thus if we state that space-time is a quantum-computation running on a cosmic quantum computer then we are saying that underlying computer is itself running on space-time. In other words it is saying that space-time is computed by space-time. So this line of attack does not answer the question at all; it just pushes it down a level. As the saying goes, "it's turtles all the way down."<br /> <br />A similar argument can be leveled against the possibility of ever computing consciousness. Consciousness is not necessarily a finite phenomena. In fact, direct contemplative observation of consciousness indicates that it is infinite in scope: every individual's consciousness can be cultivated to encompass an infinite expanse of space and time. From a scientific perspective consciousness cannot even be directly measured, let alone detected. Stranger still is the emerging realization that there is a direct quantum interaction between observation and experiment, such that it is really impossible to say that consciousness is separate from the underlying fabric of space-time. If consciousness is somehow connected directly to space-time then it too is just as infinite and potentially non-computable. <br /><br />Some might argue that consciousness is even more fundamental than space-time, but it is enough to simply state that it is equally fundamental to establish the non-computability of consciousness. If consciousness is at least equally fundamental to space-time then any computer which can compute it must actually invoke it to do the computation, and that is a circular argument. <br /><br />I rest my case.

          • Some HTML is allowed.
          • Reply
          • Cancel
          • Submit
          Commented in: Artificial Intelligence
  Add a Comment

Here's a list of the HTML tags we allow: