What I think is really interesting here is that all these questions surrounding this are focused on the human-AI interaction and whether the human should know t

What I think is really interesting here is that all these questions surrounding this are focused on the human-AI interaction and whether the human should know that they're talking to an AI when, in reality, what Google is after is completely removing the "human" aspect of that relationship. Google is betting that, in the future, we won't need people at all to do these kinds of things. In fact, they would love if every booking for every appointment was done seamlessly for the user without ever interacting with people. This Duplex technology is just a stop-gap to fill in that interim period where we're still using people for tasks that don't require specialization, are very mundane, and end in a very obvious success condition. All you need to make an appointment somewhere is a time/date and a confirmation that the appointment was made. Instead of an API, we currently have a "human" API that handles that. I guarantee that Google is banking on Duplex eventually bridging that gap so that, eventually, I can make an appointment anywhere simply by asking my personal assistant to do so and the "human" is completely unnecessary. These are the questions we should be looking at, in my opinion. This situation is only temporary. We need to focus on the future (that is most certainly coming) ideas of what we're going to do with our societies when even these simple jobs that we thought required that human touch are replaced by automation and AI.