You Almost Had Us Fooled, Eugene; Almost…

If there was one news item last week that got the better of the 2014 FIFA World Cup hype, it was that of a computer finally passing the Turing Test by making a third of the judges believe that it was a young boy from Ukraine. For those not in the loop about matters related to artificial intelligence or AI, the Turing Test is one of the most time-tested challenges to determine whether a machine can behave like a human or not. It was designed by mathematician Alan Turing in 1950 when he was confronted with the question – can machines think?

The ‘computer’ in question is actually what techies have termed as a chatbot, a chatting programme designed to simulate intelligent conversation with one or more humans. And this one, designed by a team of coders from Russia, pretended to be a 13-year-old boy named Eugene Goostman.

Now here’s the interesting thing – even though, maybe for a day, the Internet buzzed with the news with some calling it a milestone in the evolution of artificial intelligence, experts soon started punching holes into the claim.

Google’s engineering director and acclaimed futurist, Ray Kurzweil, believes that the milestone moment has not arrived yet. His biggest criticism of this event is the very thing that is the essence of the Turing Test – it is arbitrary in nature. This implies if a particular computer or programme manages to clear the Turing Test once, retaking the test will not necessarily yield the same result. This also means any claim about a computer passing this test will only be a temporary ‘win’ for AI. Being a futurist, Kurzweil believes that for machines to behave like humans, the process will have to be over a long period of time and hence, will not be based on one random test.

Kurzweil, too, interacted with Eugene. So what is his verdict? According to him, “he was not impressed.” The interaction reinforced his notion that Eugene is, in fact, not the AI we have been waiting for. Based on his chat with ‘Eugene,’ Kurzweil commented that this programme often fails to keep track of the conversation and often repeats phrases word for word. So, it is most likely that a qualified judge, trained in detecting fraud, will see through the façade.

Finally, the biggest red flag to the claim is the very person who organised the event – a certain Kevin Warwick from the University of Reading. For those who are clueless about the claims of Warwick, here are a few gems. In 2000, he claimed to be the world’s first and only ‘cyborg’ by having a chip implanted in his arm and then, a decade later, his lab claimed to have found the first case of a ‘human being affected by a computer virus.’

There was even a Kevin Warwick Watch site that kept tabs on all of Warwick’s bogus claims. His current claim also falls like a deck of cards where he states that Eugene beat the test by claiming to be a 13-year-old boy for whom English is a second language. An ideal Turing Test judges a machine without any such logical restriction or limit.

The bottom line is – Eugene Goostman beating the Turing Test has created some hype that may very well propel many coders, programmers and designers to work more on AI. At least, they know now that this test can be beaten, even if that is not always the case. And that could definitely pave the way for more dedicated and genuine efforts by machines trying to pass Alan Turing’s challenge. Till then, we can rest assured knowing that Skynet (of Terminator fame) is still far away.