Study: If Robots Want Us to Like Them, They Should Pretend They’re a Little Stupid
Hey, wise-guy robots, news flash: Humans would like you a lot more if you didn’t act like you knew everything all the time.
A study conducted by Cornell and Carnegie Mellon scientists found that people respond to robots better if they act less authoritative and use meaningless filler words to soften their communication.
“People use these strategies even when they know exactly,” Susan Fussell, associate professor of communication at Cornell University, wrote in a statement. “It comes off more polite.”
The study, titled “How a Robot Should Give Advice,” started with a cupcake lesson. Researchers videotaped inexperienced bakers making cupcakes as an instructor offered advice. These instructors were basically actors working from two different scripts: one that used very straightforward instructional language, and another that included discourse markers like “uhm,” “you know” and “maybe.” Then they made duplicate videos that superimposed a robot over the human with the original soundtrack. Subjects of the study watched the videos and answered a questionnaire judging the helpers based on their consideration, controlling and likability.
Ultimately, researchers found that both the human and the robot instructors were better liked when they hedged a little.
This theory has been tested time and time again throughout pop culture. In Star Wars, for instance, everyone sort of hated C-3PO for being so condescending and proper. R2-D2, who just beeped and made exclamation points, was much easier to stomach.
Similarly, everyone on Futurama always had a soft spot for Bender because they never felt threatened by him.
So take note, robots! You could be the belle of the ball if you only learned how to correctly incorporate the word “like,” into conversation. Feel free to take notes from Her.
Yahoo Tech is a brand new tech site from David Pogue and an all-star team of writers. Follow us on Facebook for all the latest.