You might think the best way to make a robot seem more “human” is to reproduce human features as precisely as possible, like in this YouTube video:
But most people are creeped out by robots this “real.” We’re actually more comfortable interacting with less realistic robots that exhibit some human traits, like this adorable robot named Leo:
So why is this less realistic robot so much more endearing? A fascinating article in this week’s New York Times Magazine may offer some answers:
If a robot had features that made it seem, say, 50 percent human, 50 percent machine … we would be willing to fill in the blanks and presume a certain kind of nearly human status. That is why robots like Domo and Mertz are interpreted by our brains as creaturelike. But if a robot has features that make it appear 99 percent human, the uncanny-valley theory holds that our brains get stuck on that missing 1 percent: the eyes that gaze but have no spark, the arms that move with just a little too much stiffness. This response might be akin to an adaptive revulsion at the sight of corpses. A too-human robot looks distressingly like a corpse that moves.
Domo and Mertz are among several robots discussed in the Times article, which covers a lot of ground in considering whether robots can ever be made into real substitutes for humans.
But despite the issue of the too-realistic robot, one grad student actually claims she would prefer a robot to a real boyfriend, if it could be made to simulate caring about her, since this behavior would be much more reliable than the real thing. But another student countered with this:
“Anyone who tells you that in human-robot interactions the robot is doing anything — well, he is just kidding himself…. Whatever there is in human-robot interaction is there because the human puts it there.”
The Times reporter had been especially impressed with a video of Leo apparently showing that he had mastered the “false belief” task, where a graduate student Jesse Berlin looked for an object in the wrong box, and Leo seemed to know both what he was looking for and how to help him get it. But all was not as it seemed:
Leo did not learn about false beliefs in the same way a child did. Robot learning, I realized, can be defined as making new versions of a robot’s original instructions, collecting and sorting data in a creative way. So the learning taking place here was not Leo’s ability to keep track of which student believed what, since that skill had been programmed into the robot. The learning taking place was Leo’s ability to make inferences about Gray’s and Berlin’s actions and intentions. Seeing that Berlin’s hand was near the lock on Box 1, Leo had to search through its internal set of task models, which had been written into its computer program, and figure out what it meant for a hand to be moving near a lock and not near, say, a glass of water. Then it had to go back to that set of task models to decide why Berlin might have been trying to open the box — that is, what his ultimate goal was. Finally, it had to convert its drive to be helpful, another bit of information written into its computer program, into behavior. Leo had to learn that by pressing a particular lever, it could give Berlin the chips he was looking for. Leo’s robot learning consisted of integrating the group of simultaneous computer programs with which it had begun.
In other words, Leo had to be given a massive head start by its programmers in order to solve the problem any five-year-old can easily manage. What’s more, the robot can only be programmed with one instruction set at any given time — so it can’t do the false-belief task, for example, on the same day it does the button-pushing task.
Clearly, however, once the concept is proven, it’s a much simpler matter to integrate the entire set of possible behaviors into a single robot. Watching these technologies develop is riveting stuff, and I’d highly recommend reading the entire Times article.
As a bonus, here’s one more video (not included in the Times story), featuring COG, a robot discussed extensively in the article: