Turing Test Passed - What Does It Mean For The Future Of A.I.?

Ever since the Turing test was first proposed in Alan Turing's 1950 paper "Computing Machinery and Intelligence," it has remained a rather illustrious target for researchers. The test is meant to serve as a benchmark not necessarily for intelligence, but the expression of human behavior indistinguishable from that of a human. The test pits human judges against both human and machine subjects, and the judges are to guess at which of the subjects are human, and which aren't. The machine is said to have passed the test if it can fool the judges at least 30% of the time.



That milestone has finally been passed (although 14 years later than Turing had guessed) when the chat bot named Eugene Goostman successfully fooled 33% of judges into thinking they were talking to a human being. Many were quick to point the flaws of this particular test though. For one, Eugene Goostman is portrayed as a 13 year old Ukranian Boy. As I'm sure you can imagine, it's easier to convince people that you're a 13 year old that speaks English as a second language than to convince people that you're an adult with a much better vocabulary.



Nonetheless, this is still the closest anyone has come to passing the test. As for what this means for the field of artificial intelligence? Not as much as you might hope. The Turing test doesn't serve as a comprehensive measure of intelligence, or any intelligence for that matter. The test only measures how closely a machine can behave like a human, not whether or not it can actually think and reason - a very critical distinction in A.I., and one that will likely take us a lot longer to be able to tackle in any meaningful way.