In what may be the first seeds of the eventual robot uprising, researchers at the University of Wisconsin have released a study that suggests that people are less likely to forgive an AI that gives bad advice than they are to forgive real humans who give bad advice.

The researchers conducted 14 different trials where participants were tasked with forecasting scheduling for hospital operating rooms, something none of the participants had previously done. During the trials, the participants were given advice that they were told came from either an “advanced computer system” or “a person experienced in operating room management.”

After giving correct advice multiple times, the researchers then gave the participants intentionally bad advice, both as the AI and as the human advisor. As it turns out, people don’t put up with mistakes from computers.

Participants who were given bad advice from the AI quickly ignored many of its instructions and attempted to complete the assignment without its help. Meanwhile, participants who were given bad advice by the human expert were more forgiving, and many continued heeding the experts advice despite the mistake.

“This has very important implications because time and time again we are seeing humans being replaced by computers in the workplace,” said Andrew Prahl, one of the authors of the study. “This research suggests that any potential efficiency gains by moving towards automation might be offset because all the automation has to do is err once, and people will rapidly lose trust and stop using it – this is one of the few studies out there that really show the potential downsides of automation in the workplace.”

There are likely a number of factors at play between the different ways the participants treated the AI and human advisors, and a lack of sympathy for a computer program versus a real person is only one possible explanation.

Photo by miss_rogue