Med thumb 9591024162 589ea4c499 o

When babies smile, (most) people beam right back. So small and squishy and untainted by the world are they that it’s hard not to catch their innocent joy. But perhaps those kewpie-eyed chubsters aren’t as guileless as they seem. A frightening new study, published this month in the journal PLOS ONE, suggests there’s intention behind those infectious coos and grins. 

Why would a study on smiling infants inspire fright? Well, in an effort to investigate the purpose behind babies’ grins, computer scientists, roboticists and developmental psychologists from the Machine Perception Laboratory at the University of California, San Diego programmed 13 toddler robots to mimic the behavior of babes, and then studied interactions between the artificial children and college students.

robot babies smaller

The researchers wanted to test their hypothesis that babies smile with the purpose of making other people smile back. The robot babies confirmed the hunch, and then some. Not only do babies flash toothless grins for strategic gain, they do so with total efficiency. The robot babies appeared to time their smiles perfectly — just as comedians time punchlines — so that they hardly had to crack a smile, but still elicited prolonged, ear-to-ear grins in return. 

Psychologists have previously observed patterns of smiling during infant-and-mother interactions, but this study elevated understanding of the meaning underlying those patterns. Previously, as study author Daniel Messinger said in a release, they “couldn’t say what the mother or infant is trying to obtain in the interaction. Here we find that infants have their own goals in the interaction, even before four months of age.”

To build robots that mimicked the behavior of real, flesh-and-bone babies, researchers used “optimal control theory,” which lets scientists program robots to “perform a specific behavior based on specific goals.” So, in this case, researchers used the method to work backwards, so to speak, and determine babies’ goals based on observed behavior patterns. They used previously collected smiling-pattern data from a study that documented when and how often mother-infant pairs smiled during in-person interactions. The study is part of a National Science Foundation initiative using robots to understand human development. 

The findings offer new insight into the motive behind previously opaque infant behavior. But researchers said they don’t yet know whether babies are conscious of their mastermind ways. “We are not claiming that a particular cognitive mechanism, for instance conscious deliberation, is responsible for the observed behaviors.”

Still, we’re onto them. Next time a robot baby beams at you, don’t fall into the trap. Purse those lips and move on. 

Here's a video from UC San Diego on programming a robot to smile.