Eight years ago, Karl McDorman stayed late at Osaka University and received a fax from a colleague about 1 am with an essay in Japanese written in the late 1970s. Since McDorman was involved in the creation of hyper-realistic androids, the reading was extremely interesting.
The more a robot or cartoon character looks like a human, the more we like him, but only up to a certain point. A still from the groundbreaking cartoon The Polar Express, which failed at the box office.
The author argued that people are afraid of artificial creatures that are too much like humans. This phenomenon is known as the “uncanny valley”.
McDorman and his comrades hastily translated the text into English, believing that it would not go beyond the circle of specialists in robotics. But the term went to the people. For example, with his help, journalists began to explain the unpopularity of the movie blockbuster "Polar Express" and humanoid robots.
If an explanation could be found for this effect, Hollywood and robotics could make millions of dollars. But when researchers began to study the phenomenon, citing the work of McDorman himself, nothing came of it. The psychological mechanism of the "ominous valley" remains unrevealed until now.
The essay was written by the Japanese robotics engineer Masahiro Mori and was called Bukimi no tani - Valley of Terror. Before McDorman, few people knew about this theory outside of Japan.
The first work of McDorman himself on this topic was devoted to an idea proposed by Morey: we feel uncomfortable because robots, similar to humans, appear dead and thus remind us of our own mortality. To test this hypothesis, McDorman used the so-called fear management theory, which argues that reminders of death are at the core of our behavior: for example, it makes us cling more strongly to our beliefs, including religious ones. McDorman asked volunteers to fill out a worldview questionnaire after showing them photographs of humanoid robots. Those participants who saw robots defended their views of the world with greater zeal, that is, androids really reminded people of death.
But this explanation is obviously not enough. The gravestone also reminds us that we are mortal, but it does not cause supernatural fear. Therefore, new theories soon emerged. Some researchers are trying to get to the evolutionary roots of this feeling: they say, our ancestors tried not to mate with unattractive partners. Others argue that through disgust we defend ourselves against pathogens. Christian Keissers from the University of Groningen (Netherlands) suggests that a humanoid creature seems to us to be sick, and since it is also very similar to us, then there is a high probability of picking up something bad from it.
Promotional video:
Of course, neither hypothesis holds up to scrutiny. There are many disgusting and unsympathetic things around, but they do not evoke in us a particular inexplicable feeling, this very "ominous valley." For example, we know perfectly well that a person who sneezes on the subway can infect us, but we do not experience supernatural fear going down the escalator.
It wasn't until 2007 that Thierry Chaminade of the Institute for Advanced Telecommunications Research (Japan) and his colleagues looked into the brains of people watching images of computer-generated humanoid characters. The more the object resembled a person, the stronger was the activity in that area of the brain that is responsible for the ability to understand the mental state of another person, which plays an important role in empathy.
In 2011, Aishe Saigin of the University of California, San Diego (USA) and her colleagues conducted a similar experiment. The volunteers who were in the tomograph were shown videos in which mechanical robots, humans and humanoid robots (it was known in advance that they caused the same fear) performed the same movements. The sight of a realistic android significantly increased the activity in the visual and motor centers of the cortex. Probably, the brain had to strain additionally to associate the movements of the robot with the appearance.
The evolution of the Cylon from the toaster to Caprica illustrates the average man in the street about the development of robotics.
It is assumed that in the motor areas of the cortex there are mirror neurons, which are sharpened for specific tasks and are able to activate when we see someone else performing a similar task. And there is evidence that these neurons are involved in empathy (this hypothesis is disputed). Perhaps the eerie feeling is triggered by the very system that is associated with the ability to feel what the other feels. The appearance of a humanoid robot or a character drawn on a computer at the first minute suggests that this is a person, but the next moment his movements are given out in him a fake. That's when fear arises.
It should be noted that in his article, Mori used the neologism “shinwakan” as the opposite of “uncanny”. McDorman translated this with the word "familiarity," which reflects the fact that the object is familiar to us; later there was a variant of "likeability" (ability to please). Now Mr. McDorman believes that "shinwakan" is a kind of empathy. Last June, he published a new translation that he hopes will correct a misunderstanding among Anglophone researchers of the "Sinister Valley" due to the 2005 inaccurate translation.
In cognitive neuroscience, empathy is often divided into three categories: cognitive, motor, and emotional. Cognitive (cognitive) is, in fact, the ability to understand another point of view, to understand why another person acts in one way or another ("social chess", in the words of McDorman). Motor empathy is the ability to imitate movements (facial expressions, postures), and emotional empathy is simply what we call empathy, the ability to feel what others are feeling. And Mr. McDorman boils down to the question of what kind of empathy is suppressed in the "sinister valley."
Now at Indiana University (USA), Mr. McDorman shows volunteers videos of robots, computer characters and humans in situations ranging from harmless to dangerous. Viewers are then asked to rate the happiness and unhappiness of the commercials. The most difficult thing is to determine the emotional state of the characters who find themselves in the "ominous gorge". This seems to mean that empathy is suppressed in this case. That is, at the cognitive and motor level, everything is fine, but we cannot show sympathy for such characters.
A curious and very similar result came from psychologists Kurt Gray from the University of North Carolina and Daniel Wegner from Harvard (USA), who, through a survey, found out that of all the potential functions of computers and robots of the future, the greatest fear in humans is their ability to feel our emotions. Probably, the researchers conclude, in humanoid robots we see the shadow of the human mind, into which we will never penetrate. In other words, it's not just our inability to empathize with creepy robots and computer characters, but also that we can't, and they can!
Empathy suggests that the person we empathize with has a self of his own. Therefore, as long as we realize that we are facing a robot or a virtual character, and not a person, we will not get out of the “sinister valley”, even if someday robots appear that are outwardly absolutely identical to humans. Think of Caprica and the other humanoid Cylon from the TV series Battlestar Galaktika.
Perhaps Mori understood all this perfectly. In one interview, he was asked if he believes that one day humanity will learn how to create robots on the other side of the "evil valley". "What for?" was the answer.