The Uncanny valley—picking up on subtle differences that make human-like robots feel wrong—is well known. When I taught bioengineering I decided to get the students to explore the other side of that valley. How should we treat robots once they become indistinguishable, in both looks and actions, from real humans?
The goal was simple: make the students think about where bioengineering might be not in five or 10 years, but in 50 or 100. After all, they will be leading us there.
“Here is a robot,” I would say, holding up a yellow toy, “who is the great grandfather of a robot so like a human that you cannot tell otherwise”. Then I asked for a gut response: “Is it okay to turn that future robot off permanently?”
“Yes” went to one side of the room, “No” to the other, in a roughly even split. The two groups then argued their cases. What they said was much less interesting than what happened next: I showed the students an image of the robot. It was not, as most had assumed, an adult human but a baby.
Without fail around 20% of those willing to turn off the robot would suddenly flee across the room. They could not countenance turning off a baby robot. One particularly vocal student, who’d argued Yes then later fled to save the baby, was so thoroughly discombobulated by the whole debate that she couldn’t even manage to turn off the little yellow toy.
To beep, or not to beep? turned out to be a deeper question than I anticipated.
The exhibition Robots, exploring man’s 500-year quest to recreate himself in mechanical form, is at the Museum of Science and Industry from 19 October 2017 to 15 April 2018