One supercomputer achieved a milestone in artificial intelligence — tricking judges into believing it was human and allegedly becoming the first computer to pass the Turing Test.
“Scientists at the Royal Society carried out a test devised by the man known as the father of computing, Alan Turing. A computer program mimicked a conversation and fooled 30% of the humans listening to it.” (Via BBC)
Researchers at the University of Reading developed the program. Their winning machine mimics a 13 year-old Ukrainian boy named "Eugene Goostman."
According to The Daily Caller, he comes complete with the immature personality of a teen to match. The humans talking to the computer were judges at the Turing Test 2014 competition. CBC explains the history behind the test. (Via The Daily Caller)
“The Turing was devised by second world war code breaker and computer pioneer Alan Turing. He argued that a machine that could fool people into believing that it was human was thinking.” (Via CBC)
In 1950, Turing speculated artificial intelligence could be achieved at that benchmark of duping 30% of humans. The team from Reading says the achievement signals a new era for combating cyber crime and delivering better customers services to those online. (Via Wikimedia Commons / Elliott & Fry, University of Reading)
Others are not so impressed.
“A machine has passed a test by tricking more than 30% of judges that the machine is not a machine but a human, a teenage human, 13, but still a human nonetheless — if 13-year-olds are human.” (Via Fox News)
However, Shepard Smith might be on to something. Critics wonder if Eugene’s traits and backstory helped the computer too much. The bratty teen portrayed by the computer claims to know everything, but had the excuse of not knowing everything because he’s only 13. Plus, English is his second language, meaning any computer errors could be written off as something lost in translation.
Wired also gives a lengthy counter argument to the “thinking” machine and points a finger at the dated Turing Test. Writer Adam Mann wonders if 30% is really an impressive goal? By today’s school standards -- that’s an F Minus.
“For most modern-day artificial intelligence experts, the Turing Test has long since been usurped in importance. Today, we have programs that show quite interesting intelligent-like behavior, such as Netflix’s suggestion algorithm, Google’s self-driving car, or Apple’s Siri personal assistant.” (Via Wired)
The Reading researchers say the test was independently verified as “passed.” They also noted they were never given any preset topics or questions and that the computer’s answers were its own.