Photo: Courtesy of Princeton AI.
Over the weekend, a computer made history.
In the roughly 65 years since its invention, no computer has been able to pass the Turing Test — a test that is supposed to determine whether a computer can successfully trick a person into thinking that it's a fellow human being.
This iteration of the test involved a group of interrogators who spoke with the computer via text for five minutes. They then judged whether or not they were speaking to a computer.
On Saturday, a Russian-made supercomputer passed this test at an event organized by the University of Reading — but barely. Disguised as a 13-year-old Ukrainian boy named Eugene Goostman, the computer successfully duped about 33% of the 30-person panel of judges, just over the 30% mark required to pass.
As The Verge's Dante D'Orazio notes, it doesn't matter whether Eugene answered questions correctly. "This is no HAL from 2001: A Space Odyssey. For instance, the Turing Test doesn't hinge on whether the computer's responses are correct or not — it only involves the 'humanness' of its answers."
"Some will claim that the Test has already been passed," University of Reading professor Kevin Warwick said in a statement. "However this event involved the most simultaneous comparison tests than ever before, was independently verified, and, crucially, the conversations were unrestricted."
The implications of a human-acting computer, of course, are a little unsettling. "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 cybercrime," said Warwick. "The Turing Test is a vital tool for combatting that threat. It is important to understand more fully how online, real-time communication of this type can influence an individual human in such a way that they are fooled into believing something is true...when in fact it is not."