What the faces of our robots tell us about ourselves

London CNN  — 

My first encounter at “Robots,” a new exhibition at the Science Museum in London, was with an animatronic baby. It is unquestionably life-like and perfectly baby-sized, safety-pinned into a crisp white nappy and hanging upright on a wall.

Every detail – the latex skin, the mat of baby black hair – seemed as realistic as a Ron Mueck sculpture. The left arm rose slowly, the mouth half-open, the eyelids flickering into a squint. I studied it intensely, half expecting a response – a cry, a gurgle – but without any desire to rescue the baby from the wall and cradle it. Side on, it was unmistakably a machine. An umbilical cord of shiny metal tubing fed into its spine.

“Robots” is as much about culture as it is about science. It answers a deceptively simple question that has been pondered for the last 500 years: How do we design robots we can happily interact with?

The question has become increasingly topical as humanoid robots multiply in the lab, with some likely to end up in our homes, schools, universities and clinics, as well as theme parks and museums.

robots science museum 4
Why do we design robots to look like humans?
01:10 - Source: CNN

The ‘uncanny valley’

Curator Ben Russell spent five years assembling over 100 humanoid robots for this show. He’s tracked down historic robots and automata and along the way, and managed to salvage a few of them. (One was made out of central heating components, another out of scrap metal and found rusting outside.)

“We like to anthropomorphize. We are the only species who do. We like to invent objects like us,” he says of the humanoids on display.

In 1970, a Japanese robotics researcher named Masahiro Mori posited a complex phenomenon known as the uncanny valley. His basic theory was that we respond positively to a robot as it becomes more human in look but only up to a certain point. And then suddenly, we are strongly repelled by it.

“Robots can reach a point where they become too much like us, are too corpse-like and creepy,” Russell says.

The robot appears nearly human, but not quite right. It induces the discomfort of being close to something that is ill, and reminds us of our own mortality.

Contemporary robot designers seem to have responded to this challenge in different ways.

Harry, a robot designed by Toyota in 2005, can play the trumpet.

The trumpet playing robot, Harry (2005), made by the Toyota car company, is plainly a white silicon humanoid robot but without any real facial features. He exists to entertain just like one of the old toy automata, and can play tunes like “What a Wonderful World.”

One of Russell’s favorite exhibits, Eccerobot (2009), was more realistic, with a design based on the 19th century medical textbook “Gray’s Anatomy.”

It’s human in shape, but without any kind of skin or proper face. All the innards are exposed and mimic the inner mechanics of the human body. Motors, cord, kite line and polymorph are substituted for muscles, tendons, joints and bones.

Eccerobot reminded me of one of those plastinated human bodies flayed and exhibited by German anatomist Gunther von Hagens, and as such, offered a mechanical anatomy lesson.

(I did find it humanly sympathetic in one respect: Eccerobot regularly seizes up with backache and has to be rested overnight.)

Robots across cultures

Russell introduced a Japanese communication robot called Kodomoroid as “one of the freakiest robots in the show.”

I didn’t disagree. With a helmet of black hair (almost a Mary Quant bob) and dressed immaculately in white smock and ballet flats, she seemed unnervingly real and yet also like a shop mannequin come to life. As with the animatronic baby, I examined her intensely. She too wasn’t quite right.

Her job, back in Japan, is to read the daily news at the National Science Museum.

According to Russell, the Japanese have embraced robots culturally more than any other country. In fact, about a third of the robots in the exhibition are from Japan.

Russell draws a connection with Japan’s dominant Shinto faith, in which there is no large between humans and inanimate objects. The sun, the moon, mountains and tree all have their own spirits or souls.

Kodomoroid, designed by Japan's Hiroshi Ishiguro Laboratories, reads the news.

Telenoid (2013), developed at Osaka University, is a communication robot, glistening white and bald with tapering limbs devoid of hands and feet. A child, operating it remotely by computer, can use it to communicate with someone in another country.

The claim is that Telenoid reproduces in a physical form the child’s movements and personality, as well as the voice. In trials, people have apparently been happy to talk to and cuddle the robot. They speak of the warmth of feeling in Telenoid’s eyes.

Conversely, robots are often seen as a threat in the West, and we’re still trying to overcome our suspicions.

Even the origin of the term “robot” was a bit sinister: It first entered the lexicon in 1921 via a dystopian play, “R.U.R.,” by Czech writer Karel Capek. (R.U.R stands for Rossum’s Universal Robots.)

The drama was set in a factory manufacturing humanoid robots from synthetic organic material. The robots rebel and wipe out the human race.

However, American robotics designer David Hanson has chosen not to worry about unnerving us and is already designing robots of uncanny realism with artificial intelligence and empathy, facial expression and the ability to chat. I was disappointed not to meet one; Hanson’s robots aren’t on display at the Science Museum.

“In the not-too-distant future, Genius Machines will walk among us. They will be smart, kind, and wise,” it reads on his website. “Together, man and machine will create a better future for the world.”

We shall see.

“Robots” is on at the Science Museum in London until Sept. 3, 2017.