USING PERSUASIVE ROBOTS FOR POSITIVE CHANGE

austin lee Photo by Krista Rayford
coconut Photo by Krista Rayford

This article was originally published in ‘Trends in Technology,’ a special advertising section in the 3/24/17 edition of the Cincinnati Business Courier.

By Dr. Austin Lee
Assistant Professor, Department of Communication

Intelligent and sociable robots are not just on sci-fi movies but are very much the reality now. They have already begun to change our lives.

In a Silicon Valley hotel, a robotic butler named Botler greets guests and makes room deliveries. A library in Connecticut uses a robot to help patrons locate materials. Such robots—which are capable of communicating with humans by following the rules of social interactions—are called social robots.

As social robots permeate wide segments of our daily lives, they may become capable of influencing our attitudes, beliefs, and behaviors. In other words, robots may be able to persuade us, and this is already happening. When I visited Japan last summer for a conference, I ran into a humanoid robot in the hotel lobby trying to sell me a cruise package. I was almost sold.

I believe that if we can teach robots the rules of social interactions, they can effectively persuade humans, and this can be applied in many contexts. For example, we may develop a robot that promotes environmentally friendly products. We may have robots at grocery stores to recommend healthy recipes. Or we may develop a robotic health coach that keeps track of our diet and exercise. In other words, we may utilize robots for positive change.

To test this idea, my research team in the College of Informatics at Northern Kentucky University examined the persuasive potential of social robots. Specifically, we examined if persuasive strategies in human communication can be used by robots. Among many persuasive strategies, we chose to use the norm of reciprocity. Reciprocity is a social norm in which someone does something for you and you then feel obligated to return the favor. If you scratch my back, I will scratch yours. The norm of reciprocity is deeply ingrained in our social interactions.

As Cicero said, “There is no duty more indispensable than that of returning a kindness.” So we believed that this norm can be applied even when humans interact with robots.

We conducted an experiment in our lab. We invited 60 participants to play a trivia quiz with a robot, similar to the Jeopardy! show in which IBM’s artificial intelligence Watson defeated human contestants. In our experiment, instead of competing against humans, the robot helped them by searching keywords and suggesting the best answer. However, this was just a front. The true purpose of the experiment was to see if participants reciprocate the robot’s favor if they are helped. Half of participants were randomly assigned to a helpful robot condition, suggesting mostly correct answers, and the rest were assigned to a not-so-helpful robot condition. After the five-minute trivia quiz was complete and when participants were about to leave, the robot asked participants if they could help the robot for 15 minutes.

The result was unequivocal.

When participants did not receive help from the robot, only 33 percent agreed to help. By contrast, when participants received help from the robot, 60 percent agreed to help. Note that they helped the robots for 15 minutes, because they were helped by the robot only for five minutes. The result clearly shows that robots can effectively persuade humans to help them by utilizing the norm of reciprocity. We also applied different types of persuasive strategies, including foot-in-the-door in which a small request is made first and is then followed up with a larger, actual request. When participants complied with the small request (helping the robot for 15 minutes), 70 percent of them also agreed with the larger, subsequent request (30 minutes). However, when the robot only asked for the larger request, only 10 percent complied.

This finding is exciting because it can be applied in many social contexts to offer a number of potential benefits. First, it enables robots to complete a task more efficiently and effectively. Robots will soon work in teams with humans. Persuasion skills are necessary for such collaboration and team building. Second, it helps robots build credibility and gain trust from humans. For example, a search-and-rescue robot may need to build credibility quickly in order to convince disaster victims to follow important instructions.

Finally, it benefits our society by encouraging positive change. For example, a robot may be used to increase awareness for energy conservation. It may also persuade people to stay away from substance abuse and help us deal with the heroin epidemic plaguing our community.

My research team is currently designing campaigns to promote healthy behavior on campus. I look forward to using robots to creative positive change in our society.