In medieval times human-like figures run by hidden mechanisms were used in churches to make peasant worshippers believe in a higher power. These mechanisms created the illusion of self-motion (moving without assistance) which made people assume that the figures were somehow alive. As humans, we readily perceive emotional states in everything that is vaguely lifelike (anthropomorphism).
So, what happens if it actually looks real?
The first robot when walking into the Robot exhibition at the Science Museum here in London is a small baby hanging on the wall. It’s moving like a toddler and it looks like one, too. At first, you might have some feelings of love and tenderness but those are quickly replaced by rejection and even disgust. Once your brain registers the swindle it becomes almost painful to look at it.
Freud, the uncanny valley, and death
Freud attempted to explain this phenomenon in 1918 with an essay titled “The Uncanny” which he published in his own psychoanalytic journal, Imago.
As you can guess, according to Freud it all comes down to death. We have a primitive urge to avoid death and the way we try to achieve that is by creating copies of ourselves. The copies aren’t necessarily physical. The immortal soul, for example, was the first double of the body. Otto Rank, one of Sigmund Freud’s closest colleagues called it “an energetic denial of the power of death”. Some would argue that procreation is another way of escaping death.
Freud states that those almost-real objects only remind us of the fact that we are all going to die. The original purpose of their creation, namely to escape death by creating a copy becomes clear because of their imperfection. So when we make a robot that looks almost life-like but is imperfect, the robot becomes “an uncanny harbinger of death.” (Freud).
The uncanny valley
In 1970 Masahiri Mori, a robotics professor at the Tokyo Institute of Technology first coined the term uncanny valley in an article for the Japanese journal called Energy.
Mori noticed that at first, the more his robots appeared human the more people reacted positively towards them. They found the human characteristics charming. As Mori wrote himself humans would “naturally respond… with a heightened sense of affinity.” However, with technology evolving and robots looking more life-like, people started to feel uneasy and even repulsed.
Mori names this eerie sensation the uncanny valley, which basically says that if something doesn’t look human but has a few human qualities we find it endearing. If you give it too many human qualities and it starts looking like an imperfect simulation, people will find it creepy or even revolting. You can read an interview with the now almost 90-year-old professor from 2012 here.
We are fooled for a moment by appearance only to quickly understand that we are not looking at a human. This cognitive dissonance produces an uncanny sensation. The sensation most are feeling by looking at the toddler mentioned earlier.
With robots that look and behave like humans around the corner, the uncanny valley has sparked a debate whether scientists should overcome it or if we should just go for a more mechanical appearance. Mori himself doesn’t see any benefit in making robots look like humans. In his mind, they should look and behave differently than we do.
Is Mori right when he cautions us not to aim beyond the valley? Many people and scientists are of that opinion, but…
…why are we so scared of human-like robots? Is the real reason because they threaten our uniqueness, our special place in the universe?
Here are the 2 of the robots at the Science Museum exhibition I found most uncanny: