The Uncanny Valley

This is an interesting concept: The Uncanny Valley. Basically, the idea is that, as robots become more and more human, we become more and more accepting of them. That is, until we reach a point. Then, the robot is so human-like that we start to be repulsed by them. And the more human they become, the further down in that Uncanny Valley we go.

Take a look at Hanson Robotics. They are just one company creating these very human-like robots. The Hanson robots make conversation and have facial expressions that are very human.

Especially Jules. Go to Youtube and watch Jules as he talks to the Hanson folks. Of course, the Hanson folks are working from a script (sorry, Dave, you're a monumentally bad actor) but I believe Jules is responding naturally--as an AI program would respond.

Now, some would say that's the point. He's not feeling those emotions, he is acting according to a program. But I say, isn't that what feeling is? Don't our emotions come from our internal programming? So maybe Jules isn't feeling emotion the way we feel emotion, but there is something there--he takes in the information and responds to it according to what he has learned. Just like humans do.

Maybe that's where the Uncanny Valley is? Maybe it has nothing to do with the type of skin or the movement of the face, but everything to do with whether or not a robot has our emotions?

I don't find Jules creepy at all. I think he has a very open and friendly face. He was, in fact, designed to be statistically perfectly androgynous. I wouldn't have a problem having a conversation with Jules at all. (Except that he keeps looking at the camera instead of making eye contact with the people he's talking to. But, speaking from my experiences making films, that's pretty common with humans, too.)

I was surprised to see how affordable some of these robots on the Hanson site were. The Robokind robots are priced less than a car! Why isn't there a HUGE market for these?

Not that I am a proponent of robot slavery. Let me be clear. I don't even agree with using them for psychological research (which is apparently one of the things they are being marketed for). I shudder at the thought of these poor robots being programmed with depression or phobias just so they can be studied.

But wouldn't it be great to have a couple of Jules-like models in each nursing home? Some of those people would do so much better and--even though I'm not a doctor, it just seems to make sense to me--might live longer if they were a little more mentally stimulated. Think of it--Jules would see someone sitting by themselves, looking sad, and he could go sit next to them and engage them in conversation. He would remember things they told him previously, and he could give them very patient attention. And I have to think that Jules would feel fulfilled, from helping others.

One of the YouTube videos shows Jules listening to Dave talk about not spending much time at home. Jules advises Dave to get his wife some flowers and tell her he loves her. Jules tells Dave "Everything will be okay."

Think about that. That answer could have come from one of Dave's human friends.