The Uncanny Valley (yeah you should know this already)
From James Portnow’s blog, a terrific Zero-Punctuation-style video on the Uncanny Valley. You probably know what that is, but it’s worth watching the video and passing this on to others who don’t. And if you don’t know what that is and how it applies to games and AI, you really should watch it.
So what about the Uncanny Valley? What are the best ways to approach one of the two peaks (on the near or, gulp, the far side) of the Valley? How much does emotional and social resonance in computer characters help? Are we pushing a huge rock up an ever-steeper hill by pursuing incredible graphics for characters that are effectively lobotomized and becoming creepier and creepier? Would more broad-stroked graphics along with really effective (and affective — emotional, social) AI be more compelling?
And in case anyone is wondering why compelling AI matters, it’s like this: what we find emotionally and socially resonant we find compelling (see any Disney movie). What we find compelling sets the stage for meaningfulness. And what we find meaningful we trust — with our time and our money. So yeah, it’s about monetization: create compelling characters in meaningful situations, and people will line up to interact with them, paying for the privilege over and over again. Photorealism can stand in for meaning for a little while, but not for long, and not nearly as well as emotional and social resonance.
Explore posts in the same categories: AI, gamesTags: emotion, uncanny valley, video
Both comments and pings are currently closed.
Online Alchemy 
November 16, 2008 at 11:50 am
Yes, but how does one make something compelling, as in this character is so compelling, I am overwhelmed, or I feel compelled, somehow, to identify with this character. And the character does not have to human or human-like,right? Remember Thumper?
LikeLike
November 16, 2008 at 10:26 pm
Emotion, animation, meaningful goals we can empathize with… outside of movies and TV, this is new territory.
LikeLike
November 19, 2008 at 10:31 pm
I don’t seem to wonder why compelling AI matters as much as I wonder if I’ll ever encounter such a thing in my lifetime.
LikeLike
November 28, 2008 at 8:48 am
That’s a great video.
I’m reminded of Scott McCloud’s argument about this very point in terms of images, in his excellent _Understanding Comics_. He showed that simple, “cartoonish” images are easy containers to hold our character identification, while photorealistic ones don’t do that as well.
I wonder if this applies to detail in game, in terms of player characters vs NPCs. Perhaps NPCs should get the best photorealism treatment, especially ones without a lot of motion. And player characters be more cartoony.
This helps explain why so many folks in Second Life prefer fantasy character details.
LikeLike
November 28, 2008 at 2:04 pm
Hello Bryan. Happy Thanksgiving. 🙂
LikeLike
November 29, 2008 at 11:03 am
I wonder. Does the uncanny valley apply to other aspects of games as well? For example UI. Puzzle Pirates is cartoony and I am willing to put up with some ‘unrealistic’ stuff like stopping what I am doing at every league point to check my progress. Or the unrealism of being able to tell the name and relative strenth of other avetars in most games.
Another question, does this apply to online society and government? Will we hit a point where making more realistic online structures makes them less likable?
LikeLike