When computers mimic us, we love what we hear

September 29, 2005

Researchers have long known that mimicry from one person to another indicates positive intentions and emotions. A new study published in the current issue of Psychological Science finds that when artificial intelligence mimics us, we find it just as persuasive and likable.

Participants in the study listened to an argument given by an artificial agent that either mimicked the listeners' head movements at a four second delay or repeated the movements of another participant. Those listeners who were mimicked viewed their agents as more persuasive and likable than those who listened to agents that did not mimic them.

"In addition, participants interacting with mimicking agents on average did not turn their heads such that the agents was outside of their view," researchers Jeremy N. Bailenson and Nick Yee state. At times, those not being mimicked did turn their heads away. The researchers also found that although participants knew they were being spoken to be a nonhuman agent, most did not notice the mimicry.

The artificial or embodied agents consisted of a head and shoulders and came in both male and female forms/voices. They mimicked three dimensions of the participants head (pitch, yaw, and roll) and blinked randomly (as deemed by an algorithm based on human blinking) and exhibited lip movements driven by the amplitude of the recorded message. Along with this newfound knowledge that mimicry by a computer is persuasive like mimicry from person to person, the researchers leave us with a glimpse of what else technology has in store.

"Anyone who releases a digital representation of themselves to the outside word-- by posting a digital photograph, by leaving a cell phone recording of their voice… is leaving a footprint of their identity that can be subtly absorbed by people with both good and bad intentions."

Explore further: Cellular safety shelters allow TB agent to survive in infected individuals

Related Stories

Helping robots handle uncertainty

June 3, 2015

Decentralized partially observable Markov decision processes are a way to model autonomous robots' behavior in circumstances where neither their communication with each other nor their judgments about the outside world are ...

Graphene-coated heart valves could sidestep harmful drugs

February 18, 2014

Every year thousands of people are fitted with artificial heart valves to replace their own malfunctioning valve. Many of these patients, however, have to remain on drugs that stop blood clotting on these artificial valves. ...

Recommended for you

Roboticists learn to teach robots from babies

December 1, 2015

Babies learn about the world by exploring how their bodies move in space, grabbing toys, pushing things off tables and by watching and imitating what adults are doing.

Getting into the flow on the International Space Station

December 1, 2015

Think about underground water and gas as they filter through porous materials like soil and rock beds. On Earth, gravity forces water and gas to separate as they flow through the ground, cleaning the water and storing it ...


Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.