Announcement

Collapse
No announcement yet.

Team Creates Bot That Tricks Users Into Thinking Its Human

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Team Creates Bot That Tricks Users Into Thinking Its Human

    By Tyler Lee



    Can you tell when you are chatting with a computer or a human? Do you think you are able to perceive the differences in responses? Well, in a recent Turing Test, it seems that a computer managed to fool a about a third of the judges, which apparently makes it the first time a computer has successfully passed the Turing Test.

    For those unfamiliar with the Turing Test, it is a test named after Alan Turing, in which a computer/machine has to try and fool at least 30% of humans that the computer is a real person, and not just a machine giving out scripted responses.

    In this case, a Russian-based team claims to have created a program that has managed to successfully deceive 33% of the judges, making the test a successful one.

    33% of those who judged the program ended up believing that the machine was a real-life 13 year old boy from Odessa, Ukraine.

    Now we should note that in the past, there are instances where machines/bots have reportedly passed the test successfully, so what makes this particular program the first ever?

    Well, according to Professor Ken Warwick, a visiting professor at the University of Reading (which also organized the event), “Some will claim that the Test has already been passed. The words Turing Test have been applied to similar competitions around the world. However this event involved the most simultaneous comparison tests than ever before, was independently verified and, crucially, the conversations were unrestricted.”

    He adds, “A true Turing Test does not set the questions or topics prior to the conversations. We are therefore proud to declare that Alan Turing’s Test was passed for the first time on Saturday.”

    According to Vladimir Veslov, the man behind the program, the reason their program succeeded this time was due to the dialog controller which made it more human-like.

    It’s an interesting concept and if you’d like to check out the bot for yourself and see if it could possibly fool you into thinking it was a real boy, head on over to its website and take it for a spin.
    Last edited by dlevere; 06-09-2014, 05:55:37 PM.
    The Hackmaster

  • #2
    I just get an error that it can't find the bot.
    July 7, 2019

    https://www.4shared.com/s/fLf6qQ66Zee
    https://www.sendspace.com/file/jvsdbd

    Comment


    • #3
      http://web.archive.org/web/201401121...om/bot/bot.jsp
      The Hackmaster

      Comment


      • #4
        The archive.org link doesn't work right. Use this link:
        http://default-environment-sdqm3mrmp...beanstalk.com/
        Video Game Chat

        Comment


        • #5
          How did that thing fool anybody? I know I went to it knowing it was a bot, but that thing repeats the same things over and over word for word. You can't get 10 sentences without duplicates out of that bot. Seriously, people were fooled? Try anything and see if you don't get asked your specialty and where you are twice within a few sentences. It's a circle that repeats without change. It also keeps responding as if offended. The only way that could fool a person is if they didn't take the time to say a whole 10 sentences, and sentences fool it hard. Just ask it the same question twice. "Who is your grandma?" and you'll get a few different answers word for word.

          I remember talking to the iGod bot years ago. Now that one can be interesting, especially when it memorizes things people have said to it, like their name, address, and anything else about themselves. You can still easily fool any bot. Ask it the same exact question a few times, you'll probably get the same few answers over and over.
          Last edited by bungholio; 06-09-2014, 09:57:19 PM.
          July 7, 2019

          https://www.4shared.com/s/fLf6qQ66Zee
          https://www.sendspace.com/file/jvsdbd

          Comment


          • #6
            Goes to show how much of a life I have, talked to the thing for like an hour before I got bored. Robot boy might not be able to remember your name but he'll keep it real on Foreign Policies. Lol
            You Sheeople need to wake up. Baaaaa!

            Comment


            • #7
              Factbot: a bot that spouts viralish, truth-sounding lies

              By Cory Doctorow

              Shardcore, who gave us the programatically generated Hipsterbait tees, had advanced the art of autonomous, self-perpetuating Internet memes, with @factbot1, a bot that creates true-sounding, viral-ish lies ("Indonesians always turn left when exiting a cave", "In just one drop of Sesame seeds, 50 million bacteria can be present", "Morels were used as a Sesame seeds substitute during the Norwegian Civil War").

              Here's an essay that explains the project:

              The Internet is rife with misinformation, some propagated by ignorance, but increasingly as carefully crafted campaigns of deceit. The problem is that we think we're smart enough to spot the lies, and we're not.

              Factbot was built to play in this space.

              The facts themselves are based on 'real facts' I've harvested from the web. A fact is picked at random, the entities, numbers and dates are algorithmically manipulated, and lo, a new fact is born.

              While most of the facts are patently absurd, I find it most interesting when it skirts the fringes of authenticity. The code which runs Factbot is dumb, it has no knowledge of the world, nor any notion of when it stumbles into the realm of plausibility - it is *our* reaction to the fact which defines its possible veracity, not the actual relationship to 'real' truth. In some ways, the more it's tweeted, the truer it gets.

              Shardcore: Factbot, creator of the unbelievable truth(via Metafilter)
              The Hackmaster

              Comment

              Working...
              X