Metropolis and the Uncanny Valley

As I was watching Metropolis yesterday, an intersting thing happened. A shot of the 'Machine Man' came up, and I thought, "Say, that looks pretty real."

But by "real" I didn't mean, "Say, that looks like a real robot." What I meant was, "Say, that cgi almost like it exists in the real world." Of course, the reason the 'Machine Man' looks like it actually exists is because it did actually exist - it was a pliable wooden costume made roughly sixty years before computer generated effects would even be used in films.

I'll say this one more time to be clear: I was watching a silent movie from 1927 and my reaction was that the computer generated special effects looked pretty good.

If you’re familiar with the “uncanny valley” theory, skip this paragraph. If not, here’s a pretty cool idea: In 1970 Masahiro Mori, a Japanese roboticist, published an article in which he describes a hypothesis he termed the “uncanny valley.” Basically, it says this. The more a robot or human facsimile looks like a human, the more familiarity we have with it, and the more we respond warmly towards it. (For example, for the most part people have warmer feelings towards baby seals than they do towards snakes, because baby seals look more human-like. And we like apes even more than seals, etc.) But there comes a point where this familiarity becomes too great, and instead of warmth, we feel repulsed.

According to Mori, people think that a little robot toy that hobbles around on two legs is cute, but the more super-human-like robots become, the more we’ll get creeped out by them. If one were to chart out this phenomenon, that dip where familiarity becomes repulsion is called the “uncanny valley.”

Star Wars = cute and cool. Back in the late 70s, this was a picture in a calendar hanging on my bedroom wall.

I, Robot = disturbing. This guy was never hanging from anyone's bedroom wall.

Polar Express = very disturbing. The only reason you'd hang this on your kid's wall is if you hated your kid.

But this still doesn't explain why I thought the robot in Metropolis looked "pretty good for cgi." I'm wonder this...The reason the uncanny valley exists is because our brains know we're being tricked, and our brains don't like it. When you're watching Star Wars, and you see a guy in a robot suit, your brain acknowledges that it’s a guy in a suit, and then your brain says, “Okay, so there’s a robot. What’s next? How’s that robot going to figure into the story?” But when we are presented with an image of a subject that we know can’t exist, but we’re being told does, our brains put up a fight. They resist the untruth. A paradox thus arises. The more realistic a fake character becomes, the less real it seems. When the filmmakers can’t get it just right, if there’s just one little movement off – the direction of a pupil, the weight of an arm, the bounce of hair – our brains yell out “Fake! Don’t believe a word of it! Liars!”

So what does this have to do with the 'Machine Man' of Metropolis? I have a theory. See if you can follow me, here. We as filmgoers are getting so used to being tricked, so used to being tossed down the ‘uncanny valley’, that we’re constantly trying to figure out the new rules of existance. This is where we live, now, but we don't like it. Things might seem pretty real, but our brains are being trained to look for mistakes, even when they don’t exist. It's like a magic show. We know that we're not actually seeing a woman defy gravity or get her head chopped off, so we spend our time trying to figure out how they pulled it off. And we're okay with that, because we bought the tickets and went to the theater to see a magician. But imagine walking outside the theater after the show and then seeing a woman get her head chopped off or flying through the air. Our brians woudn't be comfortable with not knowing where the game ended and reality began.

So while I was watching Metropolis, instead of seeing an actress in a mask, my brain saw a computer generated woman in a computer generated mask – because that’s what my brain is used to seeing. And my brain is trying to help me out. It doesn’t want me to be tricked, even when it’s not. So my brain flashed a red warning light: "Look out, dude. This is probably not real. We can't tell exactly why, but it's better to err on the side of caution."

So here’s my question: Is this new existence of ours in the ‘uncanny valley’ retraining our brains? Is this the boy who cried wolf? We’re so used to being lied to (in terms of digital images), that we’re don’t accept the truth even when there is no lie?

One of these two futuristic cities actually existed in the real world - at least in terms of being a model made of wood and glue and paint - and the other existed only as data, as a series of imaginary ones and zeros. And our brains are being trained to work very hard to tell the difference. Maybe too hard. Are we starting to see ones and zeros even when they aren't there?

I guess I’ll call this hypothesis the “wolf in the valley.” So, do you agree? Am I right? Or am I over thinking things? To what degree, if any, are modern digital special effects ruining our ability to appreciate the non-digital special effects of yesteryear?