Skip to main content

Quest for the World's First "Thinking" Computer

But not today.

In fact, there’s clearly no thinking computer in accordance to rules laid out by British mathematician Alan Turing. In his 1950 seminal paper entitled Computing Machinery and Intelligence, Turing states that a thinking computer is impossible to distinguish -in text-based dialogue- when compared with human responses. He predicted that by the end of the 20th Century, there would be a 30% chance that computers would be mistaken for humans in a five-minute conversation.

To test this theory, Reading University conducted the Turing Test yesterday, an exercise by human judges to determine if a computer is actually thinking as Turing dictated. This report by Will Pavia of Times Online dictates his experience with the many entrants. As one of the judges, he sat down and conversed with each artificial contestant. The winner walks away with the coveted Loebner Prize and possibly deemed as the "first thinking computer."

While many impressed Pavia, one in particular seemed too human. Named Eugene Goostman, this AI had to convince all five judges -in five minutes worth of text messaging- that is was human. Eugene talked about its guinea pig called Bill, its parents and reading Kurt Vonnegut’s Slaughterhouse Five. "I was fooled. I mistook Eugene for a real human being. In fact, and perhaps this is worse, he was so convincing that I assumed that the human being with whom I was simultaneously conversing was a computer."

Pavia goes on to talk about other entrants in his article, one chatting playfully while another responded to his questions with a question. By the end of the day, no computer passed the Turing Test although an AI named Elbot won the Loebner Prize, convincing "a quarter of the judges."

While it seems as though thinking computers are still a work of fiction, its interesting to see scientists and computer enthusiasts alike strive for the same goal. Naturally, there’s always fear of the thinking computer becoming "aware" as dictated in movies such as WarGames, the Terminator and even the subliminally sinister computer Hal from 2001: A Space Odyssey. Still, who wouldn’t want a robot friend to converse with?

"There too, was Rollo Carpenter, 43, a computer scientist from Devon, whose program Jabberwacky has spent years developing a conversational style via millions of web chats," said Pavia, referring to one of the human contestants. "Some of its conversational partners confide in it every day; one conversation, with a teenaged girl, lasted 11 hours."

Perhaps in our lifetime, a computer might actually begin to think in accordance to Turing’s theory. They might even be able to chat longer than five minutes. But when that eventually happens, what then?

  • eklipz330
    there should be a link of the conversation

    this thing obviously had to be designed to answer those specific questions, i mean i still wouldn't consider it a thinking computer unless of course it can learn instead of being updated... i doubt this thing would have personality, it was probably based on a real human being...edit, it most def was

    but that would scare me. irobot/eagle eye will happen -_-
  • Check out this Web 2.0 approach to chatbots:

    Just as Deep Blue brute-forced it in chess with speed, the idea behind the Chatbot Game is to brute-force it with a huge number of user-submitted Google-like chat rules.
  • Grims
    I don't even understand the point of theses, they just run off answer files and scripts.
  • jhansonxi
    When computers start to think then software engineers won't be needed to program them. Instead they will be programmed by lawyers.
  • kittle
    @Grims - The point is "because we can"

    As to the usefulness - that will come all to quickly. Consider how many people spend time in "chat rooms" Just text based responses flying around. Thats one possibility.
  • This could be great instead of the very annoying automated systems for customer support