Insulting robot squashed human performance


Researchers at Carnegie Mellon University, USA, have created a ‘trash talking’ robot that reduces human performance while playing the game ‘Guards of Treasures’.

Engineers ensured the robot, named Pepper, used mild insults such as; “I have to say you are a terrible player,” and “over the course of the game your playing has become confused.” Even though the robot used such mild language, humans who played a game with the robot performed worse while being insulted and better when being encourage.

Aaron Roth, lead author, said that some of the 40 study participants were technically sophisticated and fully understood that the machine was the source of their discomfort.

“One participant said, ‘I don’t like what the robot is saying, but that’s the way it was programmed so I can’t blame it,'” said Roth, who conducted the study while he was a master’s student in the CMU Robotics Institute.

The research team discovered that human performance ebbed regardless of the individual’s sophistication. “This is one of the first studies of human-robot interaction in an environment where they are not cooperating,” said co-author Fei Fang, an assistant professor in the Institute for Software Research.

Fang continued: “We can expect home assistants to be cooperative…but in situations such as online shopping, they may not have the same goals as we do.”

The researchers of this study wanted to explore the uses of game theory and bounded rationality in the context of robots. They designed a study in which humans would compete against a robot in a game called ‘Guards of Treasures’. This is a typical game used to study defender-attacker interaction in research on security games.

Each participant of the study played the game 35 times with the robot. Although the human players’ rationality improved as the number of games played increased, those who were criticised by the robot didn’t score as well as those who were praised.

Laboratory Supplies Directory - Now Live


  1. Only creativity will remain for humans. We should be rushing to CRISPR babies to all be Mozarts, writing symphonies at age 6, out of nowhere.

  2. It seems like this would be common sense… Studies of the environment and factors that affect human behavior are just as wacky as the weird world of physics- but it should be common knowledge that an abusive environment shuts down the brain and affects the belief/action process. This why you have to be so careful about who you let into your life…;-))

  3. As a former competitive pool player, this is called “Sharking” and can be very effective in throwing off your opponent. I consider it bad sportsmanship.

    • Naw. That isn’t sharking. That is just trash talking. Sharking is when you play below your ability levels at first so your opponent is given a false sense of superiority. Then you up the bet, keep the games close, and start winning.

      I was not a pool hustler in that sense. However I did find a great joy in getting people to bet on pool things that seemed easy but turned out to be impossible or looked impossible but were easy.

  4. It’s clear that intense negative feedback will discourage anyone and constrain production and improvement, but seriously, are these people this feeble so that mild critique, and from a robot at that, is so discouraging they are unable to make improvements in their strategy and improve the outcome?

  5. Sounds like the experiment didn’t have control group(s) for comparison. For example: 1) play 35 times against robot insulting you, 2) robot not insulting you, 3) human insulting you, 4) human not insulting you. If the performance was degraded by insulting, regardless of human or robot, it says more about the human sensitivity to being insulted, not about AI or robots.


Please enter your comment!
Please enter your name here