This site may earn affiliate commissions from the links on this page. Terms of use.

Sound Wave iStock

Yesterday, at I/O 2018, Google showed off a new digital assistant capability that's meant to better your life by making unproblematic deadening phone calls on your behalf. The new Google Duplex characteristic is designed to pretend to be human, with enough man-similar functionality to schedule appointments or make similarly inane phone calls. According to Google CEO Sundar Pichai, the phone calls the company played were entirely real. You lot can make an argument, based on these sound clips, that Google really passed the Turing Test.

If you haven't heard the audio of the two calls, you should give the prune a mind. Nosotros've embedded the relevant part of Pichai's presentation below.

I suspect the calls were edited to remove the place of business concern, but apart from that, they sound like real phone calls. If you listen to both segments, the male person voice booking the restaurant sounds a bit more like a person than the female does, but the gap isn't large and the female vocalisation is still noticeably better than a typical AI. The female person speaker has a rather robotic "At 12PM" at one point that pulls the overall presentation downwardly, simply past that, Google has vastly improved AI spoken communication. I suspect the same technologies at work in Google Duplex are the ones we covered virtually six weeks ago.

So what's the Turing Test and why is passing it a milestone? The British computer scientist, mathematician, and philosopher Alan Turing devised the Turing test as a ways of measuring whether a computer was capable of demonstrating intelligent behavior equivalent to or indistinguishable from that of a human. This broad formulation allows for the contemplation of many such tests, though the general test instance presented in word is a conversation between a researcher and a computer in which the computer responds to questions. A tertiary person, the evaluator, is tasked with determining which individual in the conversation is homo and which is a machine. If the evaluator cannot tell, the auto has passed the Turing test.

The Turing test is non intended to be the final word on whether an AI is intelligent and, given that Turing conceived it in 1950, plainly doesn't accept into consideration after advances or breakthroughs in the field. There accept been robust debates for decades over whether passing the Turing examination would stand for a meaningful breakthrough. Just what sets Google Duplex autonomously is its excellent mimicry of human oral communication. The original Turing test supposed that any discussion betwixt reckoner and researcher would take place in text. Managing to create a voice facsimile close plenty to standard human to avert suspicion and rejection from the company in question is a significant feat.

As of right now, Duplex is intended to handle rote responses, like asking to speak to a representative, or elementary, formulaic social interactions. Even so, the program'due south demonstrated capability to deal with defoliation (as on the 2nd call), is still a significant step forwards for these kinds of voice interactions. As bogus intelligence continues to improve, phonation quality will amend and the AI will become meliorate at answering more than and more types of questions. We're obviously still a long way from creating a conscious AI, but we're getting amend at the tasks our systems tin handle — and faster than many would've thought possible.

Update 5/xviii/eighteen 3:40pm:This entire demo has since been thrown into question.