A Computer Program Has Passed the Turing Test For the First Time

Jun 9, 2014

By Pranav Dixit

 

This is big. A computer program has successfully managed to fool a bunch of researchers into thinking that it was a 13-year-old boy named Eugene Goostman. In doing so, it has become the first in the world to have successfully passed the Turing Test.

The test is named after computer pioneer Alan Turing. To pass it, a computer program needs to dupe 30 percent of human judges in five-minute, text-based chats, a feat that until now had never been accomplished.

“Eugene” was created by a team based in Russia, and passed the test organized by the University of Reading just barely, by duping one in three judges. It should also be noted that a chatbot successfully pretending to be a 13-year-old boy for whom English is a second language ain’t exactly Hal 9000. There’s no artificial intelligence at work here; it’s more clever gamesmanship by Eugene’s creators.

8 comments on “A Computer Program Has Passed the Turing Test For the First Time

  • 4
    Nunbeliever says:

    Am I the only one who thinks that the judges should have been other teenagers. I mean, adults generally have a very hard time understanding the language, irony and humor teenager use. Yes, it’s of course hard for a teenager to pretend to be an adult. But, I would say it’s just as hard for an adult to pretend to be a teenager. I read and article that argued the creators of this bot were sneaky, because teenage boys barely can’t form a coherent sentence in real life. This is of course ridiculous. Yes, they might not the vocabulary and eloquence of an adult. But, they have their own subtle codes and style that is very hard to mimic.



    Report abuse

  • This website should not be construed as endorsing that which it reposts; to think that misunderstands the purpose. When http://www.richarddawkins.net reposts a story on another web page, that is not them “falling for it”; it’s providing a single place for our members to comment on stories from across the Internet. We no more “fell for” the claim that this program passed the Turing test than we “fell for” the countless religious criticisms of Richard Dawkins that have been reposted here over the years. This comparison is apposite not just because the latter would be a clear counterexample to the reposted-therefore-endorsed inference, but because in both cases the comments nigh on universally disagree with the original. Seriously, look at the posts on this thread, or any of the threads I just compared it to. Do any of the former, or many of the latter, “fall for” the nonsense we’ve reposted?



    Report abuse

  • Not the only one NunbeIiever, I first noticed something amiss when Professor Dawkins Tweeted the Turing medium had been breached and qualified; that he would be more impressed if the computer would have engaged in deep philosophical discussion!

    I wanted it to be true too. It was not so.

    Yet, as you say, language is arbitrary, often complicates and detracts from elementary expression; so may it be said, that the article could have been classified as pseudo-science – not as the science so construed.



    Report abuse

  • The Turing test has one specific flaw : it depends on human interpretation in order to test it.
    In other words, if you have a very gullible human being, the test will pass, but if it’s a skeptical person who can ask though questions, it will be a lot harder for the machine to pass the test.

    So basically this test just proves that humans can make mistakes.



    Report abuse

Leave a Reply

View our comment policy.