How Deep is Uncanny Valley?

When the first Toy Story movie came out in 1995, the masses were in awe. At the time, Toy Story was the first and only movie made entirely in CGI.

CGI, or Computer Generated Images, essentially means that all the drawing in an animated film is done digitally. Before the invention of CGI, big film companies like Disney painstakingly hand-drew every frame, and compiled all of them on a film reel. Sound (before the computer) was recorded separately, and then added onto the film reel, sometimes even played off a different device.

When adapted for television or broadcasting services, the sound would be put into the broadcast along with the images.

A quick overview of how it works: the information is broadcast from the initial source, the radio/television station or towers. Then, it’s taken by satellite and picked up by one of those dishes on top of the building. From there, the information is sent to a receiver in the television unit.

Okay, so that’s a really brief overview of how television broadcasting works. At any rate, once computers started to develop further into commercial use, it didn’t take long for companies to start and capitalize on it.

Television Takes A Turn

So now we’ve got Disney off at the races with the very first fully CGI movie ever. Here’s a quick frame just to get the gist:

There’s an obviously plastic-y look to the whole style, but CGI permitted the animators to go deeper in textures and make the whole scene way more realistic than they could with hand-drawn animations.

Toy Story worked out pretty well in terms of animation success, not really falling into the uncanny valley. So now here’s a better question: what’s the uncanny valley?

Understanding Uncanny Valley

To start off, uncanny valley is the group of CGIs that look creepy, or more simply, wrong. This happens because the goal of CGIs is to make something look more realistic, either more life-like or maybe more human.

However, sometimes this doesn’t work out, and the balance between cartoon and human is blurred. Us humans, we like to see an image that we can distinguish as either ‘human’ or ‘cartoon’, but in CGIs, the images aren’t always like this.

BUT WAIT! An image that falls into uncanny valley doesn’t have to be CGI; can refer to a robot or android, other human (yes, really, people can fall into uncanny valley!) or even an AI.

At any rate, this is what we call the uncanny valley.

Welcome to the uncanny valley.

This is one of the several characters from a movie called ‘The Polar Express’. The movie in particular is well renowned for its unintentionally uncanny images.

You can tell in this particular animation that the image clearly is cartooned, but still has a semi-realistic look about him. However, the facial dimensions are off, and the hair has a fake, artificial look to it. These are aspects of a potential uncanny valley image.

Qualities of an Uncanny Valley image/item:

1. It’s not totally real. Either it is or it isn’t, but if the draw between cartoon and photograph or robot and human isn’t clear, it can be pretty terrifying for the viewer.
2. The facial features aren’t in places they should be. Human faces have a particular layout that’s pretty standard, and although it varies between individuals, uncanny valley shifts this beyond where it should be.
3. There’s a glossy look to the face or skin. Humans aren’t overly shiny on a regular basis. To make a CGI look “normal” (if the goal is to make it human), don’t shine up the surfaces of characters.
4. A particularly blocky face (or body). Even if we’ve got high cheekbones, they don’t make a right angle that protrudes from our face. Chins aren’t pointed polygons. We have natural curves to our face, not lego bricks!

That’s the same issue here, the face is too blocky for what we’d consider normal and human. Looking at the ears especially, there’s an uneven transition from ear to face.

1. Movements are jerky and unnatural. Particularly found in first person shooter games, the characters move from place to place with a strange jerk to the limbs. However, this doesn’t apply to every CGI, as many aren’t necessarily moving or are only freeze frames.

So that’s essentially what uncanny valley is and how you can find it in CGI media. But have you ever wondered why we consider particular images (and not others) to fall into this uncanny valley?

Mind Games With CGI

I like this image, since it displays way more than your traditional uncanny valley chart would. It adds in the ideas of what goes beyond the healthy person, which we haven’t taken too much though into just yet. That aside, the uncanny valley charts usually don’t display CGI per se, but rather actual things we can interact with in real life (like a stuffed animal or prosthetic hand).

For right now, let’s just analyze from 0% to 100%:

This sums it up pretty darn well.

So that’s what the valley is, but what makes us recognize things as “uncanny?” As is, there isn’t too much we understand about what makes us register this oddness, but there are some circulating theories:

• Our brains search for patterns. Many studies have demonstrated that there’s a crazy amount of correlation between what we see and understand. That’s just to say that we try to recognize images and things and classify them if you will. So when we come across an object, person, or thing that we can’t understand and mentally categorize (like the uncanny toaster — is it a face? A toaster? We’ll never know!), it freaks us out.
• It’s back to what we understand as humans. In a separate study conducted, a group of people were asked to identify which of robots (ranging from mechanical/factory working ones to human-like androids) they would prefer to choose a personal gift that a human would enjoy. People tended to choose those that appeared more human-like — except for the ones that fell into the uncanny valley.

Researchers have tracked the sense of uncanny valley to be created by the parts of the brain used to understand and evaluate things like facial expressions, including the visual cortex and medial prefrontal cortex (the part of the brain that looks for human likeness), as well as several other sections of the brain.

To the far left is located the prefrontal cortex and to the far right is the visual cortex.

On top of that, our brains search images automatically for human faces. Just take a peak at the image below and see if you can spot something that looks human, like a face.

In this image, if you look at the rock formation, you can see rocks that resemble a face. Although not initially intended to look like anything aside from a pile of rocks, our human brains will search for human faces in every given image. We’re built to make sense of our surroundings. That’s why when we see something vaguely resembling a human face, it freaks us out.

Our Future With Uncanny Valley

Okay, going back to that insane chart with the double valleys on either side of the “healthy human”.

It’s time to evaluate our future with this understanding of uncanny valley. As humans, we’re traveling into the future at essentially light speed, and technology is developing faster than ever. So let’s understand what goes beyond a healthy human on the chart.

1. Transhuman: the more “evolved” human that has advanced mentally and technologically. Here, BCIs (or Brain Computer Interfaces) would be included in the column. BCIs are technologies that connect to the brain and can give humans abilities that they otherwise wouldn’t have.
2. Near posthuman: the state beyond human. Okay, so this doesn’t really exist yet, but once we’ve gained further advancements as a race, we’ll be considered (philosophically) a near posthuman.
3. Radical posthuman: the state even (further????) beyond human. This one’s not just thinking outside the box, it’s throwing the box out the window and detonating it with firecrackers. Essentially, this one’s the biggest unknown on the list, leaving it’s definition very much up to either the future or philosophy.

Something to consider:

If we ever do achieve past “human” or “transhuman”, how would their uncanny valley chart look? Or would they even have one?

Takeaways:

1. CGIs, or Computer Generated Images are digitally rendered graphics
2. Certain CGIs or robots trigger parts of the brain which cause us to be creeped out by the image
3. Despite this, there’s still a lot for us to understand about uncanny valley, and a lot the theory leaves up to philosophy

Thank you for reading my article! I hoped you enjoyed it! Please follow me on LinkedIn here: https://www.linkedin.com/in/amelia-settembre-4850aa194/, and subscribe to my monthly newsletter here: (coming soon ;D).