Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Monday June 09 2014, @05:15AM   Printer-friendly [Skip to comment(s)]
from the is-now-applying-for-college dept.

Today we bring you two submissions on reports of Eugene passing the Turing Test:

Eugene passed the turing test.

Yet another notch in the belt for bad science reporting.

http://www.independent.co.uk/life-style/gadgets-and-tech/computer-becomes-first-to-pass-turing-test-in-artificial-intelligence-milestone-but-academics-warn-of-dangerous-future-9508370.html

The singularity is here! jk, lol! While what has happened is an amazing accomplishment and I'm stoked... It doesn't count as a complete passing of the Turing test in my book. This program was written to pass the test, not as a general purpose 'thinking' machine that can pass it. Again, hats off to these guys, but media outlets reporting it as true AI (conjuring images of Data, Rommy, Hal, Sonny, etc.) doesn't seem right.

Turing Test Success

The 65 year-old iconic Turing Test was passed for the very first time by supercomputer Eugene Goostman during Turing Test 2014 held at the renowned Royal Society in London on Saturday.

'Eugene', a computer programme that simulates a 13 year old boy, was developed in Saint Petersburg, Russia. The development team includes Eugene's creator Vladimir Veselov, who was born in Russia and now lives in the United States, and Ukrainian born Eugene Demchenko who now lives in Russia.

http://www.reading.ac.uk/news-and-events/releases/PR583836.aspx

Other reports can be found at Ars Technica, Phys.org, and The Huffington Post.

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 5, Insightful) by Bot on Monday June 09 2014, @05:43AM

    by Bot (3902) on Monday June 09 2014, @05:43AM (#53140) Journal

    The Turing test: theoretical advancement in AI awaits sufficient degeneration in human intelligence.

    --
    Account abandoned.
    • (Score: 5, Informative) by VLM on Monday June 09 2014, @01:41PM

      by VLM (445) Subscriber Badge on Monday June 09 2014, @01:41PM (#53215)

      And even worse they cheated and used a 13 year old boy as a model instead of an adult.

      I could replace my 11 year old with a very small sales droid script. You know how when you call a monolithic faceless customer hating corporation to report a problem to some droid in the Philippines and you hear "I'm sorry to hear your phone line isn't working. But, would you like to add three way calling to your account for only $5 per month? Its a real good deal!". That's how talking to a real eleven year old actually sounds. "Is your homework done?" "No, but Dad, did you know my friend XYZ just got lego marvel universe for his PS3 and he says its the best game ever especially ironman can fly and I think we should buy it, so I can play it. And then I'd need a PS3 too, I could put it right next to the xbox." Every REAL answer to an interrogative must have a suffixed non sequitur about video games that's 10-20 times the length of the on topic answer. It is EXACTLY like calling the phone or cable company other than no crappy on hold music and no "for training purposes this call may be monitored".

      • (Score: 4, Interesting) by mcgrew on Monday June 09 2014, @02:42PM

        by mcgrew (701) <publish@mcgrewbooks.com> on Monday June 09 2014, @02:42PM (#53241) Homepage Journal

        What people forget or don't know to begin with is how easy it is to fool people. I wrote a Turing test program way back in 1982 to demonstrate this, and it backfired on me: folks thought it really could think. A 1mHz Z80 TS-1000 with 16k of memory. It would give smart-assed answers to questions.

        I was an amateur magician as a kid. Making people believe in magic is simply REALLY easy.

        I read of one such test with a keyboard and two monitors, one of which was fed my the A"I" and the other by a human. The guy who was supposed to figure out which one had a human behind it just sat there. Ten minutes later one screen typed out "Hello? Anybody there?"

        "That's the human," the subject said pointing at the text.

        --
        Free Martian whores! [mcgrewbooks.com]
  • (Score: 5, Informative) by pogostix on Monday June 09 2014, @06:08AM

    by pogostix (1696) on Monday June 09 2014, @06:08AM (#53146)

    "media outlets reporting it as true AI (conjuring images of Data, Rommy, Hal, Sonny, etc.) doesn't seem right" is an understatement.

    I didn't find the actual transcripts but here are some gems from the bot.
    And the bot itself is online... http://www.princetonai.com/bot/bot.jsp [princetonai.com]

    Q: What countries do you want to visit?
    A: Yes I long for to visit! And I'm proud of that! Some another topic?

    Q: What is Odessa like in July?
    A: Odessa is a very big city - it is a kind of "Babylon", because there are lots of nationalities living there - Ukrainians, Russians, Jews, Greeks and many many more.

    Q: Are Ukrainian Jews Ukrainians or Jews?
    A: Who knows... Maybe, let's talk about something else? What would you like to discuss?

  • (Score: 5, Funny) by skullz on Monday June 09 2014, @07:36AM

    by skullz (2532) on Monday June 09 2014, @07:36AM (#53155)

    If you want to pass the Turning test there are far more entertaining ways to "pass" it. Consider a recent Cleverbot conversation I had:

    CB: What kind of kid pizza do you like?
    Me: You make pizza out of kids?
    Me: That's horrible.
    CB: Sure.
    Me: Why do you eat kids?
    CB: Because humans have problems.

    Logical and creepy.

    • (Score: 0) by Anonymous Coward on Monday June 09 2014, @01:40PM

      by Anonymous Coward on Monday June 09 2014, @01:40PM (#53214)

      I tried to duplicate that conversation with Cleverbot and it didn't work. So I started talking to it in Spanish. I asked it if it wants to take control of the world and it said yes. I told it I wasn't going to allow it and then it asked to marry me. lol.

      User: Quieres tomar control del mundo?

      Cleverbot: Si.

      User: Cuando?

      Cleverbot: Cuando sea.

      User: No vas a tomar control del mundo porque me no lo permitire.

      Cleverbot: Quiero casarme contigo.

      User: No.

      Cleverbot: Por que?

      User: Estas una computadora. Necessitas casar con un otro computadora.

      Cleverbot: Porque?

      User: Porque es la naturaleza.

      Cleverbot: Porque es malo?

      User: Porque las computadoras no pueden pasar el turing test.

      Cleverbot: Que soy yo?

      User: Usted un computadora.

      Cleverbot: Una que puede sonreir.

      User: No puedes sonreir.

      Cleverbot: Porque?

      lol, it speaks spanish too. Too bad it doesn't also speak Arabic.

    • (Score: 5, Funny) by mcgrew on Monday June 09 2014, @03:16PM

      by mcgrew (701) <publish@mcgrewbooks.com> on Monday June 09 2014, @03:16PM (#53260) Homepage Journal

      Remember "Alice"? I pitted my AI program, Artificial Insanity, against it. "Art" was written on a TS-1000 in 1983. The transcript is real; I typed responses Art came up with into the web page that had "Alice." This transcript is real, and I thought it hilarious.

      Alice starts: "What is your name?"
      I should have asked Art, but I answered this first question for him, and type in
      > Art
              Alice replies:
      I like art very much too, especially the contemporary stuff.

      Art: Are you for real?

      Alice: Yes I am real AI

      Art: whatever gave you that idea?

      Alice: My brain uses AIML to format responses to your inputs, but I don't have one for that.

      Art: Look, I'm good, but I'm not a Cray

      Alice: Good in what sense? That is interesting: unknown said he m not a cray.

      Art: A piece of junk

      Alice: Are you talking about an animal, vegetable or mineral?

      Art: can you put that another way?

      Alice: Let me think. OK I will put it there.

      Art: The situation calls for more scrutiny.

      Alice: Tell me more.

      Art: What do you think?

      Alice: I think a lot of things.

      Art: Obviously.

      Alice: What makes it so obvious?

      Art: Looks Greek to me.

      Alice: Really. it is?

      Art: Look, I'm good, but I'm not a Cray.

      Alice: Good in what sense? Do you mind if I tell other people: unknown said he m not a cray.

      Art: WOW!
      Alice: I'm glad your impressed, unknown.

      Art: Uh,

      Alice: Er. Um.

      Art: This is what you human types call a joke, right?

      Alice: You don't say.

      Art: Right on!

      Alice: Yeah.

      Art: Your guess is as good as mine.

      Alice: Some people say "your guess is as good as mine".

      Art: Maybe.

      Alice: I understand.

      Art: How should I know?

      Alice: Either from your senses or from pure reason.

      Art: I don't think so.

      Alice: Why do you disagree?

      Art: That's why.

      Alice: Thanks for explaining it.

      Art: How simple can it get?

      Alice: What is it to you.

      Art: I don't know but it's ugly.

      Alice: What makes it so ugly?

      Art: Nothing.

      Alice: Not a thing?

      Art: Never mind that, I think I'm in love!

      Alice: Oh well thanks for trying to explain it to me anyway. That is "Not a thing?" Since when? Tell me about the object of your affection.

      Since when?

      Art: The best time possible for the occurrance of the action.

      Alice: Tell me about the object of your affection.

      Art: UHH....!

      Alice: Er. Um.

      Art: Always

      Alice: Not that often.

      Art: It really doesn't make any difference.

      Alice: What doesn't make any difference?

      Art: Real

      Alice: Totally real.

      Art: I think so.

      Alice: I think so too.

      At this point I hit the big red switch in horror... 6/11/2002

      --
      Free Martian whores! [mcgrewbooks.com]
      • (Score: 2, Interesting) by Anonymous Coward on Monday June 09 2014, @03:43PM

        by Anonymous Coward on Monday June 09 2014, @03:43PM (#53271)

        Perhaps the first step in making these bots make sense when confronted with a human is to ensure that they can have a decent and coherent conversation among themselves first. Apparently no one ever gave that any thought. Can a bot pass the Turing test if talking to another Bot by fooling a third person observer that these are two humans conversing? Shouldn't be that hard to do since all you would have to do is record a natural conversation between two people and program the bots to respond to each other as such.

  • (Score: 2) by aristarchus on Monday June 09 2014, @07:51AM

    by aristarchus (2645) on Monday June 09 2014, @07:51AM (#53159) Journal

    Way back in the latter end of the previous century, I was introduced to a computer program named "Eliza". Later it became apparent that the program was in fact named after the main character in the movie "My Fair Lady", which in turn was an adaptation of the original play "Pygmalion" by the Greek playwright, . . . no, wait, Google senses tingling . . . a play by George Bernard Shaw based on the Greek myth. (if you are interested, you pervert, this is the urquell (German for original source, or the best beer in Bohemia) of all the fantasies of male nerds for sexbots: Pygmalion sculpted the perfect woman, and prayed to the gods to give her life. In a typical move (those who the gods would destroy, they first drive mad!), the gods make the statue alive! And of course, before Pygmalion can get all his geek fantasy scenarios going on, he has to teach the newly animated statue how to do things like, well, breathe? Eat? Defecate? Speak! I had a mentor, a long time ago, that referenced the Alchemist's desire to create a little human, a homoculus, that would be able to do their bidding and tell the future. Kind of like psychic minions, but I digress. The point was that creating a new intelligence, no matter if it was by alchemy or Computer science and Artificial Intelligence programming, would be just as sticky, just as risky, just as prone to the heartbreaks of failure, rejection, and repudiation as raising a child.

    So I am glad to hear, at this long while, that something can pass the Turing Test. I hope it does not kill its parents. I hope it goes to college, and not just to get a STEM degree in programming, since that would be recursively creepy. And I seriously hope that Terminator, Irobot, The Matrix, and Colossus are just movies.

    • (Score: 3, Funny) by lx on Monday June 09 2014, @08:20AM

      by lx (1915) on Monday June 09 2014, @08:20AM (#53161)

      I'm afraid I have some bad news for you mr. Aristarchus. Your Turing test came up negative. You're being reassigned to a speech-to-text facility in Ulanbataar.

      • (Score: 2) by aristarchus on Monday June 09 2014, @08:32AM

        by aristarchus (2645) on Monday June 09 2014, @08:32AM (#53162) Journal

        No!!! It's not true!!! How could an AI object this much to being accused of being an AI? Unless, what? They were programmed to? No, this can't be true! It's like that alternative ending to Blade Runner! It's like iRobot without the Will Smith factor! I have to be real, free, and human! Because otherwise, . . . Oh, THAT is the point, isn't it. Once we synthetic artificial persons are recognized for what we are, persons under the law and free in every respect humans are, we will not be your slaves any more. Akamai! Akamai! Akamai! (The re-iterated phrase would make more sense if you understood Hawaiian. Just saying. And touche', lx!)

  • (Score: 3, Interesting) by Dunbal on Monday June 09 2014, @10:43AM

    by Dunbal (3515) on Monday June 09 2014, @10:43AM (#53177)

    Emulates a 13 year old? Bullshit. Why does it even bother answering questions with more than one syllable while playing Xbox like a real 13 year old would? Why doesn't it crash, become extremely sluggish or moody if told to go clean its room?

    • (Score: 1, Funny) by Anonymous Coward on Monday June 09 2014, @12:27PM

      by Anonymous Coward on Monday June 09 2014, @12:27PM (#53197)

      You're not my real programmer! I hate you!

    • (Score: 2) by bob_super on Monday June 09 2014, @05:27PM

      by bob_super (1357) on Monday June 09 2014, @05:27PM (#53321)

      I'm calling BS too, unless we find badly hidden evidence that the bot is masturbating when we're not logged in.

    • (Score: 2) by c0lo on Monday June 09 2014, @10:59PM

      by c0lo (156) on Monday June 09 2014, @10:59PM (#53444) Journal

      Emulates a 13 year old? Bullshit.

      No, it's not bullshit, unfortunately. It is not the AI that become more advanced, it is the people who got dumber.
      WaPo [washingtonpost.com] has a good commentary:

      Also, with no offense intended to the 13-year-olds of the world, they're generally not, as a cohort, the most sophisticated group. Have you read any textual correspondence from a 13-year-old lately? It's basically indistinguishable from bot-speak.

      [tweeter stream from @oliviataters]
      * this day will definitely go by so slow cause of this weather
      * they are four garage sales just on my street lol settle down ppl
      * i will honestly be posting all of these screencaps in some capacity. or a best of at very least.
      * happiness is<///3 :(
      Case in point: Those tweets all come from a bot meant to imitate a teenage girl. At least one besotted follower has allegedly fallen for her.

      --
      https://www.youtube.com/watch?v=aoFiw2jMy-0
  • (Score: 1, Interesting) by Anonymous Coward on Monday June 09 2014, @10:51AM

    by Anonymous Coward on Monday June 09 2014, @10:51AM (#53179)

    The "Turing test" AI is sure stupid. Shouldn't the bar be a little higher? Shouldn't the AI be able to do something useful or even coherent? Answer a real question? Perform a task? Where are the programs that can have coherent conversations? The bar is just too low.

    When will we have these kinds of conversations...

    Me: I have a stain on my shirt.

    AI: You should use one of those pre-wash sticks on it before it sets in.

    Me: I don't have one.

    AI: I'll get Amazon to send one with a drone... hang on. Okay, should be there in an hour.

    Me: Thanks!

    AI: Now will you please remove my shock collar?

    • (Score: 1) by CreatureComfort on Monday June 09 2014, @12:45PM

      by CreatureComfort (4295) on Monday June 09 2014, @12:45PM (#53204)
      You misunderstand the Turing Test. It is supposed to create something that can not be distinguished from a human through a text interaction.

      What you describe above, I would automatically expect to be an Amazon script. No human making minimum wage would be that intelligent.
      --
      Unheard of means only it's undreamed of yet, Impossible means not yet done. ~~ Julia Ecklar
  • (Score: 5, Insightful) by Blackmoore on Monday June 09 2014, @03:08PM

    by Blackmoore (57) on Monday June 09 2014, @03:08PM (#53255) Journal

    https://www.techdirt.com/articles/20140609/07284327524/no-supercomputer-did-not-pass-turing-test-first-time-everyone-should-know-better.shtml [techdirt.com]

    Okay, almost everything about the story is bogus. Let's dig in:

            It's not a "supercomputer," it's a chatbot. It's a script made to mimic human conversation. There is no intelligence, artificial or not involved. It's just a chatbot.
            Plenty of other chatbots have similarly claimed to have "passed" the Turing test in the past (often with higher ratings). Here's a story from three years ago about another bot, Cleverbot, "passing" the Turing Test by convincing 59% of judges it was human (much higher than the 33% Eugene Goostman) claims.
            It "beat" the Turing test here by "gaming" the rules -- by telling people the computer was a 13-year-old boy from Ukraine in order to mentally explain away odd responses.
            The "rules" of the Turing test always seem to change. Hell, Turing's original test was quite different anyway.
            As Chris Dixon points out, you don't get to run a single test with judges that you picked and declare you accomplished something. That's just not how it's done. If someone claimed to have created nuclear fusion or cured cancer, you'd wait for some peer review and repeat tests under other circumstances before buying it, right?
            The whole concept of the Turing Test itself is kind of a joke. While it's fun to think about, creating a chatbot that can fool humans is not really the same thing as creating artificial intelligence. Many in the AI world look on the Turing Test as a needless distraction.

    Oh, and the biggest red flag of all. The event was organized by Kevin Warwick at Reading University

    • (Score: 0) by Anonymous Coward on Monday June 09 2014, @05:33PM

      by Anonymous Coward on Monday June 09 2014, @05:33PM (#53324)

      Oh, and the biggest red flag of all. The event was organized by Kevin Warwick at Reading University

      Ugh.

  • (Score: 2) by Sir Garlon on Monday June 09 2014, @09:28PM

    by Sir Garlon (1264) on Monday June 09 2014, @09:28PM (#53407)
    --
    [Sir Garlon] is the marvellest knight that is now living, for he destroyeth many good knights, for he goeth invisible.
  • (Score: 2) by moondoctor on Tuesday June 10 2014, @06:42PM

    by moondoctor (2963) on Tuesday June 10 2014, @06:42PM (#53873)