The Turing test was proposed to evaluate if a human can not distinguish the (written) interaction between a human and a machine. You type something and you get your reply. And you go on trying to find out who is actually answering your messages? Another human or a computer?
The interaction through “written” text (typed, actually) was intended to get rid of the difficulty of emulating a human voice. Since what matters is the content of the interaction it made sense.
However, we are used to voice interaction and the voice tone, inflection, empathy is telling us a lot, it let us read between the lines… The progress made in voice syntheses in these last few years is making artificial voice almost undistinguishable from a real one. Add to that the possibility to craft a content that is undistinguishable from the one that a real person would create and you are opening up a can of worms.
We are already confronted with fake news. We now have technologies that can make these fake news credible, by having them voiced by a trusted person. Your friend calls you on the phone and push you to invest in a stock, as he did with great success. Unfortunately it is not your friend talking, although it is his voice (at least that’s your perception) and the reasoning, the way he answer to your questions all tells you its him.
The demonstration provided by Google at their event (see clip below) of a chatbox, Google Duplex, that can proactively assess your need and help you out, like fixing an appointment to the hairdresser, checking your agenda (that’s easy) for a free slot and making sure you get that stylish look before the party and conversing over the phone with the hair saloon, left everybody speechless, and quite a few concerned for the reasons I mentioned above.
We are rapidly entering into a new space, the one of interaction with intelligent autonomous systems that are becoming part of ourselves, we are growing into symbiotic autonomous systems and it is happening faster than we were expecting.