When we think of the robots of tomorrow, we think of something more human-like. Over half a century ago, Professor Masahiro Mori introduced the term uncanny valley to describe “people’s negative reaction to certain lifelike robots.” In his graph, our affinity to a robot does not grow linearly from least to most human likeness. At a certain point where a robot is almost human but not perfect, there is a pitfall in comfort. Could this mean we have a collective subconscious fear of technology that’s becoming too real?
"The Uncanny Valley" published by Masahiro Mori in 1970 explores how our affinity for robots changes as they become more human-like.
This paper has had a big impact on human-computer interaction and is especially relevant today's AI landscape.
Read on: https://t.co/oH52uz7F8U pic.twitter.com/28EBUcdlJ8
— Fermat's Library (@fermatslibrary) April 23, 2024
There are contrasting theories on why we experience the uncanny valley, but its existence has been corroborated, in part or fully, by many studies since Mori’s original essay. Most theories say this stems not from technology but from self-preservation instincts like pathology avoidance, mortality fear, and the blurring of the human concept. If you saw a corpse, the details won’t add up to your concept of what a human is and avoid it. Other studies even show monkeys can experience the same unease when presented with increasingly monkey-like images, and they, too, stop having it once the model is perfect enough.
How to avoid it
Technology today has worked out ways to keep our products uncanny valley-free. The phenomenon does not only affect robotics. Any visual cue can invoke this feeling, especially if movement is involved—art, animation, virtual reality, video games, etc. Understanding how to avoid the uncanny valley is an important part of all these fields. While some argue it is bound to decrease as new generations are more and more used to it, today’s customers will naturally avoid products that involve it.
Deadpool director Tim Miller suggests that CGI is on the cusp of crossing the uncanny valley – meaning digital characters may soon be indistinguishable from human ones.https://t.co/qRwQBdqOnf pic.twitter.com/C2AClq21Yj
— Film Stories: magazines, podcast, live shows, web (@filmstories) October 2, 2024
As a tech enthusiast, it is a relief that the uncanny valley is not a subconscious fear of technology. It is an evolutionary response and a challenge for developers and artists. One must consider customer expectations when designing a robot, 3D model, or animation. No one expects something that looks human but speaks in a synthetic voice, his eyes staring at one point, or moves in a non-natural way, nor will someone expect a high-definition human face texture on a low polygon model. If budget or other issues don’t allow for full realism, where it’s indistinguishable, settle for a stylized or cartoonish look.
Photo Credit: The feature image is symbolic and has been taken by Maximalfocus.
Sources: IEEE Spectrum / Pnas / eNeuro / Katrin Wolf
