stub Robot ‘Thinks Out Loud’ During Thought Process - Unite.AI
Connect with us


Robot ‘Thinks Out Loud’ During Thought Process



Image: Pepper the robot

Italian researchers have designed a robot that can ‘think out loud,’ which enables users to hear its thought process and gain deeper insight into the technology’s decision-making processes. 

Antonio Chella is co-author of the research, which was led by first author Arianna Pipitone at the University of Palermo. 

“If you were able to hear what the robots are thinking, the robot might be more trustworthy,” Chella says. “The robots will be easier to understand for lay people, and you don’t need to be a technician or engineer. In a sense, we can communicate and collaborate with the robot better.”

Inner Speech in Robots

Humans rely on inner speech when looking for clarity and moral guidance, and it helps in the decision-making process. The researchers set out to see how this could impact robots, so they developed one called Pepper, which can speak to itself. The researchers then had people set a dinner table with Pepper while following etiquette rules, and they studied how Pepper’s self-dialogue skills affected human-robot interactions. 

After studying the robot, the researchers found that Pepper is better at solving dilemmas when relying on inner speech. One experiment saw a user ask Pepper to break the etiquette rules by placing a napkin in the wrong spot, which led to the robot asking itself a series of self-directed questions. It then concluded that the user might be confused, so it confirmed the request before continuing to use inner speech. 

Here is the dialogue Pepper had with itself:

“Ehm, this situation upsets me. I would never break the rules, but I can’t upset him, so I’m doing what he wants,” Pepper said while placing the napkin in the wrong spot.

The user can understand Pepper’s thoughts as it uses its inner voice to solve a dilemma, which the researchers say could lead to human-robot trust. 

Framework for robot’s inner speech

Self-Dialogue and Higher-Task Completion Rate

When Pepper’s performance was compared with and without inner speech, it demonstrated a higher-task completion rate when utilizing self-dialogue. Inner speech enabled Pepper to outperform the international standard functional and moral requirements for collaborative robots, which are a set of guidelines that machines follow. 

“People were very surprised by the robot’s ability,” Pipitone says. “The approach makes the robot different from typical machines because it has the ability to reason, to think. Inner speech enables alternative solutions for the robots and humans to collaborate and get out of stalemate situations.” 

According to Pipitone and Chella, the new framework can lead to a better understanding of how self-dialogue can be used by robots to focus, plan, and learn. 

“In some sense, we are creating a generational robot that likes to chat,” Chella says. “Inner speech could be useful in all the cases where we trust the computer or a robot for the evaluation of a situation.” 


Alex McFarland is a Brazil-based writer who covers the latest developments in artificial intelligence & blockchain. He has worked with top AI companies and publications across the globe.