Computer or Human

1. 53

   TruthDebaterposted 13 years ago

   Can a computer ever be made to outperform a human? Can a computer ever be made with the amout of variables a human has that allows them to adapt?
   
   I think a computer can be designed to live longer or perform in conditions that we can't perform in. But with all of the variables in the world, would it ever be possible for us to design computers to understand the variables, when we ourselves don't understand or know all of them? How do we design something smarter or with more intelligence than us?

   replyreport
   1. 96

      thooghunposted 13 years agoin reply to this

      Computers already outperform us in many areas, which is why we use them. In order to become some kind of super-human (how else can  we measure intelligence, but through relativity?), which is what I think you're getting it, I believe it will need to self-conscious, with our role in laying an adaptable and open framework which exceeds our own.
      
      We have our own "script" based on genes, where action and thought in its simplest form is comprised of very few variables (essentially a neuron firing or not), so I don't see why computers cannot be engineered with a more complex framework.
      
      With regards to emotion, and whether or not a computer can be called intelligent without it, a trip through the animal kingdom shows us that emtion is in no way a singular  human trait.
      
      A machine driven by genes which in turn are driven by self-preservation should develop emotions as well. Emotion, as I understand it, is a necessary consequence of self-consciousness.

      replyreport
      1. 53

         TruthDebaterposted 13 years agoin reply to this

         Neuroplasticity argues with you that we have a "script" based on genes. Those neurons firing can result in infinite variables, much like life evolving into a human from a single cell. If you had no experience or observation of those variables, you would never guess them. Much like not being able to know a neuron fires without the science of observation. There are unknown variables, my main point is, how do you design something to predict the future, when we can't predict the future?

         replyreport
         1. 96

            thooghunposted 13 years agoin reply to this

            Point taken. In a sense I agree that it may be impossible. But perhaps we should allow self-conscious machines to evolve freely and see if their guesses are better then ours.

            replyreport
            1. 53

               TruthDebaterposted 13 years agoin reply to this

               I enjoy and appreciate your perspective on this. I like the chess example where the computer beat the chess champion. In chess all of the variables could be figured out, there are a predetermined number of blocks in which the pieces can move on each given turn. In the universe, there are many unknown variables that can possibly never be predicted. Physics in the uncertainty principle states that we can never know the exact position and momentum of a particle. This gives clues that the universe may also have some type of neuroplasticity. I think we may only have computers on the same level of intelligence when we can design conscious life from scratch.

               replyreport
   2. 60

      RogerADposted 13 years agoin reply to this

      If you haven't ever read about it the Turing Test, which was created by Alan Turing after WWII, touches on this very subject of creating a machine that can understand and comprehend in the same way a human would.
      
      The general premise is whether or not it is possible for a machine to mimic human behavior/responses enough to fool a blind judge.
      
      The captcha, which is that little hard to read random phrase that you have to type in to prove you are not a computer is like a reverese turing test.

      replyreport
2. 61

   Csjun89posted 13 years ago

   Like they have mentioned before, chess can be an example
   
   But it is unlkely that they will reach us anytime soon
   
   As for the far future, who knows really?

   replyreport
3. 61

   chinweikeposted 13 years ago

   Neurologists made us to understand that no human being has used up to 3% of the capacity of the cells in the brain.
   
   Imagine what happens when this will happen.

   replyreport
4. 61

   SpanStarposted 13 years ago

   To begin with I like computers-(at least most of the time). The versatility of this machine and the flexibility along with the speed and accuracy makes it a remarkable invention. The idea that a computer can replace the beings that created it seems far-fetched to me. Since mankind has been around so long people take it for granted "humanity." We are however an absolute miracle. We may not be able to outperform a machine but our ability to think, to see, to visualize, our ability to reason and all the other mental complexities that we do are simply impossible for a machine to duplicate. Take for example when one looks at a Van Gogh painting some can see the beauty in his artwork. Beauty has no concept for a machine. Take any animal in the wilds or domestic and try to explain to that animal the concept of God-no matter how well of an explanation you give the animal simply doesn't have the capacity to grasp the concept. Physically ( we can create) a machine that performs exactly like humans but there is much more to being human and being physical.

   replyreport
5. 80

   Aficionadaposted 13 years ago

   The problem with Deep Blue, the set of IBM computer mainframes that beat Garry Kasparov at chess, was its huge inefficiency.  It could evaluate more than 200,000,000 possible moves every second, but it was a fire hazard while doing so.  They had to provide special heat-dissipating equipment so that it didn't burst into flames.  Kasparov could only evaluate five moves per second, but his brain (like ours) was highly efficient, using only about as much energy as a light bulb.  He didn't need to evaluate every possible move - only the most useful strategies.
   
   The designer of Deep Blue learned that its major flaw was that it couldn't "learn," so he designed TD-Gammon (backgammon-playing computer) to be able to learn from its mistakes.  With TD-Gammon, he did not program in all the possible moves (as he had with Deep Blue).  Instead, he programmed it to make predictions about outcomes, evaluate results, and try to decrease the error signal - the way the human brain learns.  [Info from How We Decide, by Jonah Lehrer]
   
   As impressive as these two examples of AI are, what else can they do?  I honestly know very little about the subject, so I am intrigued to learn anything I can.

   replyreport