Creepy or what?

Yesterday evening I was prompted by the Kurzweil newsletter to look at reference 1, where I have been reading about Google’s work to make a computer much more like a human being, at least within the restricted domain of making appointments and reservations over the telephone. They look to be leveraging their investment in DeepMind, in the sort of deep learning which has cracked all the games like chess and go.

Now many of us have become used to talking to the computers which man the front ends of call centres. HSBC, for example, has one that checks who you are and finds out something about what you want before routing you through to a person. Speaking for myself, I find this sort of computer irritating rather than disturbing.

The Google answer is to work even harder to make the computer indistinguishable from a human and they have built something called Duplex. The idea seems to be that it sits between your mobile phone and the businesses with which you want to make an appointment (say a hairdresser) or a reservation (say a restaurant). You tell your mobile phone what you want to do and it works with Duplex to actually phone up the people concerned, make your appointment or reservation and keep your mobile phone posted. Your mobile phone will no doubt make an entry in your calendar and remind you about it from time to time. You, a very busy and important person, but not so busy and important as to run to a real secretary, might think that this was a good wheeze. You were able to fix up your evening without having to take more than a few seconds out from your busy and important day.

But if I was the person taking the booking at the other end, I might find this a bit creepy. I would no longer be able to tell whether I was talking to a computer or to a person, knowledge which I believe I ought to have a right to. I believe that any such computer ought to introduce itself in such a way as to make the fact that it was a computer clear. Going on to pretend to be a human is fine and I don’t think I would even mind that I would, from time to time, forget that I was talking to a computer. Rather in the same way that sometimes, when talking or thinking about my favourite soap on television, I forget that I am talking or thinking about a show rather than real life. I might have sounded a bit silly but no real harm has been done.

Rather in the way also that I think I read somewhere that it is considered bad form in Japan to take peoples’ photographs with a mobile phone without asking their permission, and that to this end their mobile phones are supposed to make a very audible click when you press the button.
Companies like Google seem to be all for blurring the difference between humans and computers.

But while I can understand their pride in their scientific and technical achievements, I also think that they are pushing us off down a slippery slope and that we, collectively, would do well to think about it and set up some ground rules. The sort of thing that the European Commission might usefully set up a committee about – but on which we no longer get a seat.

Behind this strong, reflexive, reaction, I think there is the thought that a human being is one of us. On a good day I can reach another human and be reasonably confident that I know where he or she stands, what he or she it up to. That he or she will have some ground rules about behaviour, about what is reasonable and permitted and what is not. That he or she has not been subverted by a computer virus or an alien. I can enjoy a bit of human contact, safe in that knowledge. While if I am talking to a computer, I want to know. Partly because I would feel silly talking to a computer as if it were a human and then realising the truth of the matter some way down the track. Partly because computers have long memories and private agendas and I might want to be careful about what I gave away.

It is also a strong reaction that plenty of people do not share and plenty of people seem to be quite comfortable with this blurring; they are quite happy to enjoy a bit of fake human contact. I associate to the elaborate sex toys that are becoming available, which most of us are rude about but which some of us clearly buy.

So perhaps things are not as simple as I had at first thought. So I opt for the European Commission solution: someone ought to do something about it.

I shall try to make a point of taking a look at reference 2 from time to time, to see what else Google are up to. At least, to give them their due, they seem to be quite happy to tell us quite a lot about what they are up to.

PS: in the foregoing I had originally been going to say ‘a real secretary with legs’, but decided that this was probably sexist. The idea is not so bad though: a secretary with genuinely human attributes. The sort of thing that the very rich can treat themselves to at their places of work. The sort of people that can still enjoy the privacy and comfort of their own offices.

Reference 1: https://ai.googleblog.com/2018/05/duplex-ai-system-for-natural-conversation.html.

Reference 2: https://ai.googleblog.com/.

Reference 3: http://www.kurzweilai.net/.