Is Watson the most advanced super computer in existence?

Discussion in 'Science and Technology' started by DarthTom, Nov 18, 2013.

  1. rhubarbodendron

    rhubarbodendron Rear Admiral Rear Admiral

    Joined:
    May 1, 2011
    Location:
    milky way, outer spiral arm, Sol 3
    I deliberately didn't comment on the second half of your post because there are a few expressions I have to look up first and didn't get around to yet (the usual last week before holidays chaos at work, Christmas party at office, last minute cookie baking and suitcase packing - I'll drive to my parents on Sunday noon. Plus my dictionary weighs 8 lbs...).
    I'll look them all up tonight and then try to say something halfways intelligent. Promise!

    ETA: sorry that I gave you an overkill with botanical and zoological names in that other post (re: Si use in terrestrial life). Being a biologist, I have a pretty unfair advantage of you in that field.
     
  2. DarthTom

    DarthTom Fleet Admiral Admiral

    Joined:
    Apr 19, 2005
    Location:
    Atlanta, Georgia
  3. rhubarbodendron

    rhubarbodendron Rear Admiral Rear Admiral

    Joined:
    May 1, 2011
    Location:
    milky way, outer spiral arm, Sol 3
    As a general rule I indeed prefer a printed dictionary because translation software has the unfortunate tendency to always pick the wrong one if a word has several meanings. A brain is more reliable in that instance.
    As for the links you provided (thank you! :)) : to find such a dictionary one has to know the correct keywords to enter into the search engine. Else the situation is a little bit like explaining to a blind what yellow looks like. (Which reminds me - but this is totally off-topic - we used to have a poster here named "blind groping". Haven't seen him in years. Anyone of the senior members here know what happened to him?)

    I am not familiar with Turing machines (had to look them up) but from what I read I understand that a Turing machine is a mathematical model that simulates a computer's "way of thinking". While I agree that they both work in the same way (after all, what use would a simulation be if it was inaccurate?) I wouldn't go so far as to reverse the order and say computers are Turing machines. I may be overly careful in this case but I keep remembering that example a Greek philosopher made: if a human is defined as a naked, 2-legged creature, a plucked chicken would under that definition count as a human. Therefore I am always extremely careful about reversing conclusions.

    I agree about computers being self-consistent. They can not do anything that goes contrary to their programmes.

    With humans the matter is complicated and has been discussed by philosophers and biologists for the last 3 milennia. Admittedly, people occasionally do things that seem out of character. But only because we can't imagine a person might do something, doesn't necessarily mean it's not somewhere in their character range after all. Maybe it was there all the time, only dormant because it was not yet needed? It is a very difficult and controversial matter that basically rests on the question: how do we think? As long as the mechanical and biochemical mechanisms are not completely known, we can only speculate.
    Therefore it is currently rather a matter of faith whether you believe that humans can deviate from their programming or not.

    At the moment we are surely better than computers at combining informations and finding connections between seemingly unconnected things. This helps us finding solutions for problems we encounter for the first time. We have encountered different problems in the past and use our experience to analyze the new problem, divide it up into handy subproblems and solve them one by one.
    (off-topic: that mechanism was what I meant in the suicide thread when I said that parents should not clear their kids' path from all obstacles but rather teach them how one attacks and solves problems. If they are familiar with overcoming obstacles, it gives them a better chance to deal with problems later without panicking.)

    In order to teach computers how to be creative and find unlikely or unexpected solutions, we must first find out how we ourselves do it. This might take a while since for ethical reasons your possibilities of experimenting on living human brains are rather limited. There are quite promising attempts with nervous tissue cell cultures, though, but they have only the basic (mechanical) abilities and - for all we know - no intelligence and creativity. Unfortunately, physiological research goes very much slower than IT development. So, atm we biologists are like an iron ball chained to the computer specialist's ankles.