Michael Goldthorpe: Are we losing the plot in the uncanny valley?

If you boil down the marketing buzzwords of 2018, I reckon you land on two: authenticity and automation. The first is an acknowledgement that people buy from people and a movement to make businesses seem human and socially responsible. The latter is our passion to drive efficiency by introducing robots. The third big theme of 2018 is constant contradiction. But let’s go back to the robots.

What’s a robot?
Ask three different people to define a robot and you’ll get three different answers. But the prevailing thinking is that robots are programmable and autonomous. They sense what’s going on and make relevant decisions according to their input sensors. Much like a traffic light. The first of these was trialled in 1868. It blew up and injured a policeman. But traffic lights as we know them have been a fixture of road safety and traffic efficiency for well over 70 years. Simple, autonomous machines that do a useful job. That’s the benchmark. 

Dumb robots. Smart robots and AI.
Dumb robots do what they’re told. I programmed one in primary school on a computer that had less than half the memory of the average email. It asked you your name, checked its database and established that you were “a noo noo head”. Unless, of course, you were called Michael – then you were obviously “a genius”. That’s a dumb robot. Smart robots can make decisions and actively ‘learn-on-the-job’. This artificial intelligence has infinite applications in diagnostic medicine, law, and yes, customer service and marketing.

What’s an uncanny valley?
In aesthetics, the uncanny valley is a theoretical relationship between the degree of an object’s resemblance to a human being and our emotional response to that object. Specifically, the more a robot looks like a human, the less we like them. It’s a concept identified in the 70’s by a Japanese guy called Masahiro Mori. Essentially the idea is that people respond better to people-like characteristics in robots – right up until the point that they’re nearly human when there’s a visceral WTF?

Why are we replacing humans with “humanish”?
As Vodafone is resolutely third to market with a fancy AI toy to look after its customers, someone needs to ask the obvious question. Why are we hell-bent on faking human when we have perfectly good humans who are good at it? And yes, I get the scalability of a machine over headcount. But even then, why spend so much time and money crafting a scary almost-human that 40 years of science tells us people are unlikely to like?

The tech is genius. The execution is freaky. The point is lost.
Don’t get me wrong, I love the technology. An artificially intelligent robot that can solve my problems on the spot and save me a wait in the queue is awesome. But we don’t need the tech to fake humanity. Because it can’t. And don’t get me started on the growing question as to why all AI Interfaces are designed as conventionally attractive subservient women (hat-tip Vaughn Davis) Surely if we’re inventing the future we can do it in a way that moves past 1920’s stereotypes. 

More important than any of that, if the ultimate goal is customer service and brand authenticity, surely there are better investments than ‘human-ish’ she-servants. These robots might be super clever and hyper efficient, but they also state, by their very existence, that the brand doesn’t care enough to pay a human to look after its customers.

The secret of automation is smart back-end process. The genius of automation is making automated stuff feel human. The promotion of the robots from the back-cupboard to front-of-house is a nosedive into the uncanny valley – and somewhat missing the point.

That’s what I reckon, what do you think?

  • Michael Goldthorpe is a managing partner at Hunch.

About Author

Avatar photo

Michael Goldthorpe is Managing Partner at Hunch.

Comments are closed.