Tuesday, February 28, 2017

Another Zombograph

I warned you there would be another zombograph. Well here it is. The graph below is supposed to illustrate the concept of the uncanny valley, an easy concept to understand once you have gone through it but harder when you see it in a form like this. The simplest way to explain the concept, I think, is to give an example, and we all saw one recently in the digitized versions of Peter Cushing and Carrie Fisher in Rogue One: they look very much like human beings, but they're not, and your brain knows it. And is creeped out by it. Which is why human characters should not be digitized in movies and cast next to real people.

So in the graph below, a less human likeness is on the left and a more human likeness is on the right. That's the x-axis. The y-axis seems to me to be mislabeled, as it is the axis showing not "familiarity"--whatever that means--but positive or negative emotional response to objects with a supposedly more or less human appearance. The uncanny valley is that area where an object approaches human appearance but is obviously not human. That provokes a very negative response. The extreme dip in the curve (into negative territory) is the uncanny valley.

A graph of the uncanny valley, from Wikipedia.

Now, there are lots of things wrong with this graph and possibly something wrong altogether with the idea of the uncanny valley. I can tell you, if I had turned in something like this to my wildlife ecology professor, he would have had words for me. To begin with, there is no label and no author's name. Next, the graph is confusing. What does "familiarity" mean? What are the units by which it is measured? I assume the horizontal gray line to be the divider between positive and negative "familiarity." If so, why isn't it labeled "0" (zero)? Likewise, the x-axis, labeled "human likeness," is without units of measure, unless "50%" and "100%" are those units. If so, what does that mean? Does a humanoid robot supposedly have a 50% human likeness? Does a stuffed animal have a likeness that is more than 50% human? According to whom?

That leads to a more serious problem, not only with the graph but with the thinking behind it, for aren't all non-living objects actually 0% human? No matter what they look like or what they do, aren't they all equally non-human? And what does it say about a scientist who seemingly believes that something non-human can approach the human when there is actually an infinitely wide gap--an unbridgeable discontinuity--between human beings and all non-living objects, regardless of their appearance or how well they are animated? I know this graph is supposed to show a recovery of positive feelings after a passage through the uncanny valley, but do human beings really respond favorably to a non-living object that looks very nearly like us? Or is that just wishful thinking on the part of scientists who dream of the day when robot relationships will replace human relationships?

Finally, in regards to the graph, healthy persons, puppets, robots, stuffed animals, prosthetic hands, and corpses are all real things. Zombies are not. Why is there a non-existent thing on this graph? What kind of science is that? Going back to my wildlife ecology professor--if I had turned in a graph showing, for example, some kind of comparison of large, terrestrial North American wildlife species and had included Sasquatch on my graph, I would have received a talking to behind closed doors. But here we have zombies and nobody seems to object. Anyway, this is just an example of the uses of zombies in academia. It appears to be politically neutral, so we can be thankful for that at least. But it also appears to show a lack of intellectual or scholarly rigor as in so many of the papers in the zombibliography from the other day. And it shows that zombies have in fact been scientified. Or maybe there's something more behind the concept of the uncanny valley . . .

* * *

Don't get me wrong: I think that the uncanny valley is a useful concept and that it very likely describes something real. If it is real--if we shrink with revulsion from things that look human but are not--then it seems likely to me that our feelings for the uncanny preceded their description by science (or quasi-science) and that they are of use to us as we find our way in the universe. And I don't think we should be trying to bridge the uncanny valley, as some people seem intent on doing. On the contrary, we should strive to keep it deep and wide. If we don't we won't be able to recognize the monsters among us--or in us.

* * *

Very human-like robots are a big thing in Japan, a country that seems to have forgotten how to have human relationships. The originator of the concept of the uncanny valley is also Japanese. His name is Masahiro Mori (b. 1927). I know almost nothing about him, but it is ironic that his last name evokes the Latin word meaning "to die" and nearly echoes the Latin phrase "memento mori," meaning "remember you must die." Dr. Mori seems to believe in the potential for robots to achieve buddhahood or a state of enlightenment. With all of that in mind, I would like to quote from a quote of a quote from a paper by W.A. Borody on Dr. Mori and his concepts of the uncanny valley and the Buddha in the robot:
"What is this, Channa?" asked Siddhartha. "Why does that man lie there so still, allowing these people to burn him up? It's as if he does not know anything."
"He is dead," replied Channa.
"Dead! Channa, does everyone die?"
"Yes, my dear prince, all living things must die some day. No one can stop death from coming," replied Channa.
The prince was so shocked he did not say anything more.
--From The Fear and Terror Sutra (Bhaya-bherava Sutta)
translated from the Pali by Thanissaro Bhikkhu

It goes without saying that robots don't die. In fact, a robot made of advanced technology may very well carry on the personality and character of its maker, whether that maker be a person or a whole nation.

* * *

In addition to originating the concept of the uncanny valley, Masahiro Mori has studied the purported relationships between religion and robotics. He believes, like I said, that robots are capable of achieving buddhahood or a state of enlightenment. I'm not sure that we have an equivalent concept in the West, although there are some people who believe that robots will someday become self-aware, thus rendering Western religions problematic, if not obsolete. Eastern religions may not have that problem. An illustrative quote of a quote from Dr. Borody's paper:
Unlike Christian Occidentals, the Japanese don't make a distinction between man, the superior creature, and the world about him. Everything is fused together, and we accept robots easily along with the wide world about us, the insects, the rocks--it's all one. We have none of the doubting attitude toward robots, as pseudohumans, that you find in the West. So here you find no resistance, just simple quiet acceptance.
--Osamu Tezuka (1928–1989),
Japanese manga artist and cartoonist

If the continuation of a culture is a measure of the usefulness of its ideas, then I think we have to take things like this with a grain of salt. After all, the Japanese are very rapidly not reproducing themselves out of existence.

* * *

It seems to me that one reason we have a sense of the uncanny is so that we can recognize those things around us that are without souls and so that we can differentiate them from ensouled beings like ourselves. The scientist, atheist, and materialist has no use for an idea like that. Practitioners of Eastern religions or members of Eastern cultures may not have the same sense that we do. And maybe that's why the Japanese don't seem to mind having "relationships" with robots and why robots are replacing human beings in the roles of lover, friend, and caretaker in Japan. Robots are simply doing what human beings there are failing to do. Maybe, when there are no longer any Japanese people, Japan will survive as a nation populated (or robulated) by human-like Japanese robots.

* * *

Japan and Germany have different but related problems. Both tried and failed to destroy themselves during World War II. Their prospects for doing so now have improved, for both are in demographic decline, with Japan approaching demographic collapse. I can't diagnose the German problem, but it seems to me a combination of self-loathing, nihilism, secularism (or atheism), materialism, and hedonism. The Japanese problem isn't exactly clear to me either, although Shintoism, the predominant belief system in Japan, is essentially atheist in orientation. (No pun intended.) The Japanese people also have a reputation for being stoic. You might say that both Germany and Japan now exist in a pre-Christian state. In any case, both are on the path towards annihilating themselves in the original sense of the word, meaning reduced to nothing. People in both countries see the threat. The proposed solution for demographic collapse and depopulation in each seems to be different. I can imagine a time in the not very distant future when Japan will be a nation of a hundred people and a million robots. Germany, on the other hand, seems intent on passing itself and its two thousand years of history and culture on to Muslims, who will destroy it just as well as Germans have seemed intent on doing for so long now.

* * *

So there is an uncanny valley in visual terms. Is there also an uncanny valley in auditory terms? Do we recoil from the almost-human voice? I do, whether it's a computer voice from the national weather service or a robot on the phone. Towards the end of her life, my mom lost her ability to speak. She could have used a computer to speak for her but she didn't want it. I think I understand her reasons, although she never told us why she didn't want it: as a human being, she wanted to speak in the voice of a human being or not at all. So will there come a time when robot voices will serve some of the purposes of the human voice? Will we try to speak soul-to-soul to a thing without a soul as with the robot analyst in THX 1138? Will we have robot Facebook friends, for instance, who will mimic very closely the sympathies and sentiments of real people, moreover, who will serve the purposes of somehow affirming our value as human beings and building our self-esteem through a digital intermediary? I don't doubt that there are people working towards these goals. Our acceptance and use of robot analysts, friends, confidants, and lovers would only confirm to me that we have, as Albert Camus put it, a worm in our hearts. It would also confirm to me that any system of belief that does not recognize our humanity and the existence of the human soul is a literal dead end.

* * *

One last thing: I have speculated that zombies are the monster of the twenty-first century and that they are likely to remain so for a long time to come. I have considered the possibility that robots or androids will succeed zombies, and there are indications of that happening. I suppose it depends on which future we prefer, the apocalyptic or the dystopian. But there is a lot of death involved in all of this, and of things beyond death. Maybe the robot will prove to be simply a technological zombie--a thing without a soul that looks human, lives among us, and saps from us--little by little and without our realizing it--our humanity. Beginning as slaves (robot means slave) and without their own will, robots may one day be able to reproduce themselves, creating in the process robot-zombie hordes and precipitating a robot-zombie apocalypse, as in the Terminator movies. Maybe that's why the zombie is on the graph above, actually on the same curve as robots but well along in its supposed likeness to human beings. Remember the zombies in Israel in World War Z? There they are at the base of the walls around humanity, like creatures at the lowest point in the uncanny valley, clawing their way to get at us, keen in their desire to subsume us and destroy us. The difference here is that there appear to be human beings ready to give them a boost, to bridge the valley so that robot-zombies might exist among us, eventually to . . . ?

Text copyright 2017 Terence E. Hanley


  1. But what if robots or AI's became human enough to be, essentially, "ensouled?" Or humans enhanced themselves with cybernetics?

    1. Hi, Jordan179,

      I'm not sure how anything made by human beings, or made by things made by human beings, can become ensouled. After all, we--ensouled beings--were created by a Creator. Only that Creator can create an ensouled being. We are not capable of doing such a thing, however highly we might think of ourselves and our abilities.

      As for "enhanced" humans: if we are human beings, then we are ensouled, despite whatever accoutrements we might wear, including cybernetic parts. That brings up a question though. Maybe this is your question: how far does that go? At what point does a cybernetically "enhanced" human become no longer human? Maybe this analogy is the same: When does a nation or a people become no longer such a thing? If there is one cell and a billion switches, does that still make a person? If there is one human and a billion robots, does that still make a nation?

      All of this may prove moot: I have a feeling that we will never be able to "enhance" ourselves the way some people imagine. Reality has its way of setting us straight. Witness our current situation.

      Thanks for writing.