DIGC335

The Uncanny Valley Vortex [podcast]

 

UncannyValleyChart

 

The Uncanny Valley is a strong factor as to why we feel apprehensive to advancing robots. Somewhere between faceless machines and humanoids, lies the uncanny valley.
Japanese robotisict Masahiro Mori, proposed that as more robots became human-like, the more acceptable and appealing they would become than their mechanical undertones. Until they became
too close to human, that people developed a sense of unease and discomfort (Lay 2015). The similarities become too unsettling, and cause negative reactions. 

 

 

Humanoids or Synthetics are probably the most obvious example of Uncanny Valley, intentionally built to be indistinguishable to humans. The ‘company’ behind the synths in the show Humans, built and produced their humanoids with the purpose of being “Closer to humans than ever before”. Running through the veins of the entire show was the social implications of such an invention and the tension between synths and people.

People have an innate capacity to relate to technology as human and anthropomorphize them, empathising and projecting human qualities upon them to foster our acceptance of them. When we forget they aren’t human, the human qualities we gave them suddenly become frightening, because they are doing them the wrong way. This confusion is less likely with RD-D2 than it is with a humanoid, because we know just by looking at R2 he won’t always react the way we expect. Yet there it is, we gave R2 a gender, a human quality. He has personality, like a human.

Why we experience it?

Uncanny Valley often emerges from our brains instinctive need to understand and relate what we are seeing to ourselves. It’s obvious to see how this backfires, when we are presented with technologies that blur the lines. Robots become unsettling to people when we believe they have the ability to sense and experience, but clashing with our intuition it is not a thinking machine.  

The uncanny valley can occur because of mismatches between the robots appearance to its behaviour. Like synchronising speech to its mouth moving, and corresponding facial expressions. Unease occurs, when there are considerable contradictions.

The minute facial expressions we use to determine trust and familiarity, elicited the same reaction from the robots as when we react to psychopaths (Lay 2015). The synths in Humans tend to lack these tiny cracks in their demeanour. Emotions can be simulated, but there is a clear distinction between the humans and the synthetics, typically their hyper-green eyes. The main cast of synths however that ‘possess’ conciousness, are forced to hide their distinctions with contacts. They can simulate acting human slightly better than a normal synth, but their body movement still gives them away if one focused.

The evolution of robots has been building and reaching for the goal of creating incredibly life-like, human-like, appearances to make them ‘soothing’ to real people, but why? Why do robots need to look exactly like people? They can too easily fall into the vortex of becoming uncanny, so why risk treading near it?

Another theory suggested the uncanny valley is in response to robots triggering a ‘primal fear’ of being replaced by something so similar to us (Lay 2015). Robots forcing us to examine our own mortality (MacDorman 2005), because humanoids will never experience death and life the same way as us.

How to remove it?

Robots are still being built with the perception that they should be more and more like us, the growing realisation is that the more robots appear human, from appearance to voice, and even feeling; our sense of their familiarity increases until we reach the valley, where we feel horror.

There is still the point on Masahiro Mori’s chart of the valley, before the dip, of familiarity and acceptance, which is where robot developers must aim to achieve the desired goal. We still want to relate and attach human qualities to robots, and still have the awareness they are not real. Geminoids are crafted robots that still retain their mechanical subtleties to ensure they fall slightly short of human mimicking by a few points (Solon 2011).

In place of avoiding falling into the uncanny valley, is the uncanny wall, that we will always be able to tell the difference between artificial and human (Lay 2015). As robots become more realistic, we will also become more sensitive and aware of when something is not right, and looking for the signs of human or not human.

The Uncanny Valley is a real influence on our perception of robots and their social roles. Not only playing off our perceptions of humanity and how we determine trust and understanding, but its ability to modify this trust with robots. Are you more likely to trust R2D2, or Roy Batty? Something very clearly defined as a robot, or another that blurs the line?

References

Lay S, 2015 ‘Uncanny Valley: Why We Find Human-Like Robots and Dolls so Creepy’
http://www.theguardian.com/commentisfree/2015/nov/13/robots-human-uncanny-valley

 

Wegner D, 2012 ‘Feeling robots and human zombies: Mind perception and the uncanny valley’http://www.sciencedirect.com/science/article/pii/S0010027712001278

Macdorman K, 2005 ‘Mortality Salience and the Uncanny Valley’ http://www.macdorman.com/kfm/writings/pubs/MacDorman2005MortalityUncannyValleyHumanoids.pdf

 

Bowman N, 2014 ‘The Uncanny Valley is Uncanny’ http://www.uxbooth.com/articles/the-uncanny-valley-is-uncanny

 

‘Dissecting Humanoids Among Us…’
http://www.umich.edu/~uncanny/

Macdorman K, 2005 ‘Androids as an Experimental Apparatus: Why Is There an Uncanny Valley and Can We Exploit It?’ https://www.researchgate.net/profile/Karl_Macdorman/publication/245406914_Androids_as_an_Experimental_Apparatus_Why_Is_There_an_Uncanny_Valley_and_Can_We_Exploit_It/links/543bcc2b0cf24a6ddb97a803.pdf

 

Solon O, 2011 ‘Japanese robot twins fail to bridge the ‘uncanny valley’’
http://www.telegraph.co.uk/technology/8494633/Japanese-robot-twins-fail-to-bridge-the-uncanny-valley.html

 

Ueyama Y, 2015 ‘A Bayesian Model of the Uncanny Valley Effect for Explaining the Effects of Therapeutic Robots in Autism Spectrum Disorder.’

http://www.ncbi.nlm.nih.gov/pubmed/26389805

 

Mathur M, 2015 ‘Navigating a social world with robot partners: A quantitative cartography of the Uncanny Valley’ http://www.sciencedirect.com/science/article/pii/S0010027715300640

 

Lay S, 2009 ‘Introducing the Uncanny Valley’ http://uncanny-valley.open.ac.uk/UV/UV.nsf/Homepage?ReadForm

Advertisements