Humans Struggle to Turn Off Pleading Robots

There is a growing market of interactive robots that provide humans with companionship and help. They even take up typically human roles such as security guards and receptionists. As robots become more and more humanised, where do we draw the line in our relationships with the robots? Is compassion for robots overshadowing human interaction and true sentience? Studies have established that we are easily drawn in by the simulation of social cues and human traits exhibited by machines. A recent experiment carried out by German researchers from the University of Duisburg-Essen shows that the simulated cries of a robot can emotionally manipulate people to such an extent that they will refuse to switch them off.

The study consisted of 89 volunteers, all of whom were required to complete two tasks with the help of a small anthropomorphic robot called NAO. These tasks included asking NAO questions and organising a weekly timetable. The volunteers were told that the tasks were designed to improve NAOs learning algorithms, but in actuality, the real test was how capable the participants would be at switching the robot off as it begged them not to. In approximately 50% of experiments, the robot begged for its life, saying to the volunteers that it was afraid of the dark. When NAO desperately asked for mercy, the human volunteers were likely to decide they could not switch off the robot. Of the 43 participants who heard NAO’s pleas for its life, about 30% of them could not bring themselves to switch NAO off and refused to do so. The remaining 30 people participating in the study took twice as long to agree to turn NAO off compared to those who did not witness NAO’s begging.

Those who refused to switch NAO off had an array of reasons when asked why. A few people were shocked by the begging, which made their decision difficult, and other participants felt morally responsible for the robot, not wanting to do the wrong thing. For the most part, people chose not to switch NAO off simply because the robot asked them not to. In the paper documenting the study, the researchers conclude that the volunteers were “triggered by the objection, people tend to treat the robot rather as a real person than just a machine by following or at least considering to follow its request to stay switched on”. Now that robots of this calibre exist, a new social dynamic is introduced as humans have to ability to converse in their language with robots, and inevitably people forge some attachment to other social entities whether human or not.

Research resembling that of the NAO experiment included a study carried out in 2007. The experiment revolved around a cat-like robot who also begged for its life. Unlike the case with NAO, all participants did turn off the robot after quite a moral struggle. This is perhaps because the participants were in the company of the scientists and were more focused on obeying them as an authority figure than the robot, whereas NAO was left alone with the participants, making for more of a bonding session between man and machine.

Human empathy for machines has been identified in other studies too. For example, the first test of humans and robots interacting on an emotional level was carried out in 1966. Professor Joseph Weizenbaum from the Massachusetts Institute of Technology created a therapeutic computer program called “Eliza”. Eliza acted as a psychologist, asking participants questions about their wellbeing and feelings. As a result of this emotional interaction, people started treating Eliza as if she were human. It has been found that if a human can relate to a simulation of human life, there is little difference in the way they treat them in comparison to other human beings.

The German researchers from the experiment with NAO found that people predominantly prefer interacting with a robot that they feel is similar to themselves regarding personality. Robots can have such an impact on people if they accurately simulate human behaviour. They have such an effect that people can even end up following the commands of robots if they are acting as authority figures.

Robots have always been associated with the future and dystopian visions of the universe where robots threaten to overrule humans. This literary and cinematic trope is no longer all fantasy. Given their emotional capacity and their growing ability to interact with humans as if they were human themselves, the robots featured in films like Ex Machina, whereby incredibly convincing humanoids are objects of human affection, are not so fantastical. Relationships between man and machine, whether in the workplace or more controversially, romantically, are no longer solely present in storylines belonging to fiction novels or sci-fi films. A world shared by human and humanoid is ever increasingly becoming our future, as robots like NAO are already pulling on the heartstrings of the people it interacts with.

+ posts

CIF Presents TWF – Ems Lord


Related articles

The Future of Marketing: Automation vs Innovation

Does AI Understand Your Brand Voice? AI is dropping jaws...

AI Act – New Rules, Same Task

The first law for AI was approved this month...

Time to Ditch Traditional Tools for Cloud Security

Reliance on cloud technologies has significantly expanded the attack...

AI Show – Episode 3 – Guy Murphy

In this third episode of The AI Show! Host...

6 Ways Businesses Can Boost Their Cloud Security Resilience

The rise in cloud-based cyberattacks continues to climb as...

Subscribe to our Newsletter