It’s getting harder and harder to tell the difference between computer generated images and the living, breathing human beings they are supposed to resemble — and this breathtaking image might be the closest we’ve come to replicating life on screen.
One theory says there’s an “Uncanny Valley” into which illustrations, robots, and other not-humans fall when they start to look more like real people. In other words, we have no problem seeing and identifying objects that we can clearly see are fake, or inanimate. But when man-made images or objects approach a certain level of realism, our brains start to get weirded out by the close-but-imperfect features and feel uneasy. The point between these two is known as the “Uncanny Valley.” For example, a corpse falls directly into the pit of the Uncanny Valley — it looks like something we recognize as alive, but is clearly, unsettlingly, not right.
The valley differs slightly for moving versus still images or objects.
Now two Japanese artists have drawn a schoolgirl that manages to make it through the Uncanny Valley on its way toward looking like a real person.
There are some nitpicks to make about the image — the collar of the girl’s shirt is far too white, for example — but you have to search for any issues to notice them. Otherwise, this image could easily pass as a photograph of an actual schoolgirl.
That image isn’t weird because its subject resides in the Uncanny Valley — it’s weird because it’s so believable. And, unlike CGI used to touch-up an image and make someone appear younger, this image creates its subject out of nothing. Here’s what lies beneath its digital skin:
That’s much closer to the Uncanny Valley. Because it depicts a person, it gets less leeway than the computer-generated animals from this year’s The Jungle Book or the orcs from the Warcraft movie, which are less realistic than this image.
If this can happen with computer-generated images, perhaps it could also happen with robots, which also tend to creep people out. It took a while for this to happen with still images and will probably take longer with robots, which also have to move and sound like we do before they can leave the Uncanny Valley, but it could still be a matter of “when” this will happen, not “if.”
