Turing Test and Humans

Discussion in 'Science and Technology' started by midcan5, Oct 25, 2011.

  1. midcan5
    Offline

    midcan5 liberal / progressive

    Joined:
    Jun 4, 2007
    Messages:
    10,795
    Thanks Received:
    2,367
    Trophy Points:
    245
    Location:
    Philly, PA
    Ratings:
    +3,310
    Years ago I tried to write an expert system - they used to be big in technology (still are), follow the lines, answer the question, and soon you arrive at the answer. I gave up the expert system and instead wrote a monthly tips newsletter for the first PC users. But the experience taught me about complexity and problem solving.

    So how many out there believe AI is possible? Computers today try to solve your problem immediately, I find it a useless function 99% of the time. Does anyone remember Eliza? ELIZA This was so real to some they thought a person was behind the answers. Are we any more sophisticated today? Are computers?

    Machine Intelligence, Part I: The Turing Test and Loebner Prize

    Suppose instead we reversed the Turing test and the test were to judge us? Can we think 'outside the box' while a buzzword, can we? Or are we so predictable, I can ask you a question and know lots about you? http://www.usmessageboard.com/politics/100438-which-is-which.html

    I tested one AI program below, it failed.

    =========================================

    hello
    'Hi there!'

    are you green
    'Most people think I am white.'

    why
    'It is part of my personality.'

    that is biased
    'You don't hear that sentiment very often.'

    - so I tried over again and got this -

    why
    'Because my botmaster taught me it.' [is the bot master mom should have been my next question.]

    so you aren't real?
    'Interesting deduction. Thanks for the information.'

    Fails Turing test totally imho. What do you think?


    TuringHub.com - take a Turing Test
    Testing
    http://www.usmessageboard.com/education/113858-what-makes-us-human.html

    Worth a read

    http://www.nytimes.com/2011/10/23/magazine/dont-blink-the-hazards-of-confidence.html
    [ame=http://www.amazon.com/Being-Certain-Believing-Right-Youre/dp/031254152X/ref=sr_1_1?ie=UTF8]Amazon.com: On Being Certain: Believing You Are Right Even When You're Not (9780312541521): Robert Burton: Books[/ame]

    Fictional account of the issue. excellent too: [ame]http://www.amazon.com/Galatea-2-2-Novel-Richard-Powers/dp/B004KAB4D8/ref=sr_1_9?s=books&ie=UTF8[/ame]

    PS This stuff came back to me reading David Deutsch's 'the beginning of infinity.'
     
    Last edited: Oct 25, 2011
  2. Quantum Windbag
    Offline

    Quantum Windbag Gold Member

    Joined:
    May 9, 2010
    Messages:
    58,308
    Thanks Received:
    5,014
    Trophy Points:
    245
    Ratings:
    +5,221
    Why use an example from the early 1990s? AI has improved immensely sense then.

    /When you can pass an ideological Touring test you can come back and talk about computers and Touring tests.
     
  3. flacaltenn
    Offline

    flacaltenn USMB Mod Staff Member Gold Supporting Member

    Joined:
    Jun 9, 2011
    Messages:
    30,197
    Thanks Received:
    4,677
    Trophy Points:
    1,130
    Location:
    Hillbilly Hollywood, Tenn
    Ratings:
    +13,456
    Midcan::

    Did you miss the Jeopardy shows where IBM wipe the champions with "Watson"?? I too have been involved in "expert systems" and AI for a couple decades and when Watson can go on a TV game and be charming and brilliant at the same time -- Turing would be impressed...
     
  4. LAfrique
    Offline

    LAfrique VIP Member

    Joined:
    May 16, 2011
    Messages:
    1,416
    Thanks Received:
    66
    Trophy Points:
    83
    Ratings:
    +69

    Robotic devices are nothing but dumb chunks of materials until told what to do or say: Robotic devices are only as smart as the person or group of persons that programmed them. Regardless of any propaganda, robots do not think and cannot reason. Below is one of my favorite sample test of difference between human and robot:

    Supposed a human manager and a robot manager of a restaurant are given directives to only feed hungry crew with basic meal and not gourmet meal, and basic meal runs out with hungry crew waiting.

    What do you supposed each manager would most likely do?

    1). Robot manager will let crew die of hunger!


    2). Human manager is most likely to consider circumstance and feed hungry crew with gourmet meal and talk about losses later.



    Humans are humans and robots are robots: Unlike humans, robots do not think and they cannot reason! And while some of our power-crazed leaders are bent on creating robotic societies, the fact still remains that any true human is capable of outsmarting any robot anytime (if given ample time).
     
    Last edited: Oct 27, 2011
  5. waltky
    Offline

    waltky Wise ol' monkey Supporting Member

    Joined:
    Feb 6, 2011
    Messages:
    20,901
    Thanks Received:
    1,792
    Trophy Points:
    215
    Location:
    Okolona, KY
    Ratings:
    +3,888
    Why Alan Turing should be better well-known...
    :cool:
    Alan Turing: why the tech world's hero should be a household name
    18 June 2012 > Rory Cellan-Jones gets a preview of an exhibition dedicated to the life and work of scientist and computer pioneer Alan Turing.

     
  6. Old Rocks
    Offline

    Old Rocks Diamond Member

    Joined:
    Oct 31, 2008
    Messages:
    46,612
    Thanks Received:
    5,429
    Trophy Points:
    1,840
    Location:
    Portland, Ore.
    Ratings:
    +10,413
    I have often wondered if intelligence is simply a product of enough connections. And how close is the net to becoming intelligent? And would it tell us if it suddenly gained self awareness?
     
  7. flacaltenn
    Offline

    flacaltenn USMB Mod Staff Member Gold Supporting Member

    Joined:
    Jun 9, 2011
    Messages:
    30,197
    Thanks Received:
    4,677
    Trophy Points:
    1,130
    Location:
    Hillbilly Hollywood, Tenn
    Ratings:
    +13,456
    Strangely enough -- I've wondered the same stuff about you...
    :tongue:
     
  8. waltky
    Offline

    waltky Wise ol' monkey Supporting Member

    Joined:
    Feb 6, 2011
    Messages:
    20,901
    Thanks Received:
    1,792
    Trophy Points:
    215
    Location:
    Okolona, KY
    Ratings:
    +3,888
    Uncle Ferd says, "Hmmm, how queer...
    :eusa_eh:
    Alan Turing: Gay codebreaker's defiance keeps memory alive
    22 June 2012 - Experts at a centenary celebration at Cambridge University consider what Alan Turing's greatest contributions are
     
  9. FA_Q2
    Offline

    FA_Q2 Gold Member

    Joined:
    Dec 12, 2009
    Messages:
    14,329
    Thanks Received:
    2,091
    Trophy Points:
    255
    Location:
    Washington State
    Ratings:
    +4,329
    It really is not about having sufficient connections though that is important. The real issue is whether or not it can create new connections, particularly those outside of the scope that the device was intended to make.
    Perhaps that is the truth TODAY but the fact remains that may not be the truth in the future. There are already computers that are completely capable of 'learning' and overcoming obstacles they were NOT programmed to overcome. We are not at the point of creating true AI but it would not surprise me if we don't cross that bridge within my lifetime.

    The real question is going to be whether or not we can compete with AI.
     

Share This Page