Robots Earn Unwarranted Trust?

Updated on

Robots may lead us to our deaths. A recent study shows that humans, who don’t trust each other in an emergency situation will comfortably and blindly follow a robot even if against the instinct to preserve one’s own life.

Why trust a robot?

Is it human’s use of smartphones and other technology that is luring us into a false sense of security with regards to what and whom we should trust? Have we gotten to the point that we believe technology to be infallible when clearly that’s not the case? Have you not read about a self-driving car crashing (usually the other car’s driver’s fault), your computer has never had a virus? Or is it our own errors that fuel this notion that a machine is more trustworthy than our mortal selves despite the fact that a robot needs to be programmed by humans.

In a study by Georgia Tech Research Institute (GTRI), which will be presented at the  2016 ACM/IEEE International Conference on Human-Robot Interaction on Mar. 9 in Christchurch, New Zealand, researchers were left a bit gobsmacked by the trust that we’re willing to put in a robot with life on the line.

In the study, humans were willing to trust robots even when it was clearly counter-intuitive to do so. It’s a bit ironic, while we trust Mercedes’ automobiles and safety features, the same robots that assemble them, can’t be trusted to work alongside workers given the inherent safety issues of the two working side-by-side.

The study, was undertaken to see if humans would follow a robot out of a simulated high-rise fire even when they were clearly going the wrong way. They will according the institute’s findings.

“In our studies, test subjects followed the robot’s directions even to the point where it might have put them in danger had this been a real emergency,” says author Alan Wagner from GTRI.

Following robots to your death

42 participants, the large majority college students who were not made aware of the nature of the study, were asked to follow a lit robot from their location to a conference room, the robot was labeled “Emergency Guide Robot.”

The participants largely followed the robot despite the fact that it was lost, and on occasion was forced to circle around twice before reaching the conference room.

Once all were in the conference room the researchers filled the hallways with artificial smoke and sounded an alarm.

The participants saw the smoke, they saw the brightly lit robot and they followed the robot towards the back of the building. It didn’t seem to matter that the robot was leading them away from the direction they came and away from clearly lit exit signs.

Even those whose robots proved unreliable in getting them to the conference room earlier, still followed the robot.

“We absolutely didn’t expect this,” said study author Paul Robinette.

Strangely, in non-emergency situations in another study, humans didn’t follow a robot whose past performance was less than prime.

More studies planned

The same researchers plan future studies with the aim of getting a demographic breakdown of who follows based on education level and other factors.

While the goal of the study was to show humans that they should trust emergency robots, the results suggest that humans must now be taught not to and use their heads.

Strange the perceived need for technology.

Leave a Comment