Categories: SecurityWorkspace

Turing Test ‘Beaten’ By Program Pretending To Be Teenage Boy

Updated: It’s been claimed the Turing Test, designed to see whether artificial intelligence could fool people into believing computer programs are human, has been passed for the first time ever by a computer program. On balance, these claims are dubious, but what follows is what we published on Monday. Skip to the comments for discussion.

The program, named Eugene Goostman, posed as a 13-year-old boy during an event at the Royal Society in London on Saturday, commemorating the 60th anniversary of the death of Alan Turing, the codebreaker considered one of the founders of modern computing.

Eugene had conversations with the judges, 33 percent of whom were convinced the computer was a real boy after interactions with the machine, the University of Reading announced yesterday.

The judges simply sat at a terminal with a split screen, sending and receiving messages from a human and a computer program. Four other programs based on supercomputers tried to beat the Turing Test and failed.

The “first true pass” of the Turing Test?

It’s been claimed the Turing Test has been passed before but organisers said this was the only one that followed all the rules correctly.

“This event involved the most simultaneous comparison tests than ever before, was independently verified and, crucially, the conversations were unrestricted,” said Professor Kevin Warwick, a Visiting Professor at the University of Reading and Deputy Vice-Chancellor for Research at Coventry University.

“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.”

Professor Warwick also noted that the success of the competition also had negative connotations. Some the world’s most prevalent protections against bots used for criminal purposes, such as logging into people’s Internet accounts, use takes on the Turing Test, such as CAPTCHA.

“Of course the Test has implications for society today. Having a computer that can trick a human into thinking that someone, or even something, is a person we trust is a wake-up call to cyber crime.”

Eugene was born in 2001, the creation of Russian Vladimir Veselov and Ukrainian Eugene Demchenko, who focused on producing a believable character.

“Our main idea was that he can claim that he knows anything, but his age also makes it perfectly reasonable that he doesn’t know everything,” said Veselov.

“We spent a lot of time developing a character with a believable personality. This year we improved the ‘dialog controller’ which makes the conversation far more human-like when compared to programs that just answer questions. Going forward we plan to make Eugene smarter and continue working on improving what we refer to as ‘conversation logic’.”

Turing was recently pardoned by the UK government, overturning his 1952 conviction for homosexuality.

UPDATE: Various claims have emerged to suggest the program did not pass the Turing Test, as the character of Eugene had been decided beforehand and the program should have been able to score as well as humans.

Fancy a quiz? Try our Turing Test!

Thomas Brewster

Tom Brewster is TechWeek Europe's Security Correspondent. He has also been named BT Information Security Journalist of the Year in 2012 and 2013.

View Comments

  • "The event was organized by Kevin Warwick at Reading University. If you've spent any time at all in the tech world, you should automatically have red flags raised around that name. Warwick is somewhat infamous for his ridiculous claims to the press, which gullible reporters repeat without question. He's been doing it for decades."

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

    • Yes, dammit. We didn't spot the presence of Captain Cyborg till it was too late.
      Hopefully we avoided the worst of the hype but we are sorry we ran the story.

      As has been detailed elsewhere, this was a chatbot, not a supercomputer, and previous chatbots have done just as well or better at versions of the Turing Test.

      The people running the test gamed the system by pretending that the bot was a child whose first language was not English. Even so, they didn't do very well.

      Peter Judge

Recent Posts

X’s Community Notes Fails To Stem US Election Misinformation – Report

Hate speech non-profit that defeated Elon Musk's lawsuit, warns X's Community Notes is failing to…

1 day ago

Google Fined More Than World’s GDP By Russia

Good luck. Russia demands Google pay a fine worth more than the world's total GDP,…

1 day ago

Spotify, Paramount Sign Up To Use Google Cloud ARM Chips

Google Cloud signs up Spotify, Paramount Global as early customers of its first ARM-based cloud…

2 days ago

Meta Warns Of Accelerating AI Infrastructure Costs

Facebook parent Meta warns of 'significant acceleration' in expenditures on AI infrastructure as revenue, profits…

2 days ago

AI Helps Boost Microsoft Cloud Revenues By 33 Percent

Microsoft says Azure cloud revenues up 33 percent for September quarter as capital expenditures surge…

2 days ago