The "trolley problem" in relation to robots pointed to the danger of endowing machines with excessive anthropomorphism.
A well-known thought experiment from the field of ethics suggests choosing which way to direct a heavy trolley - in the direction where it will inevitably kill one person, or in another, where another will be the victim. Different versions of Trolley Problems deal with different combinations of potential victims: a group of criminals or one innocent person, a healthy person or a group of hopelessly ill patients, a great surgeon or a child, and the like. Professor Rick van Baaren at the University of Nijmegen in the Netherlands and his team presented volunteers with a similar dilemma for robots.
Indeed, robots are increasingly integrated into our lives, and many of them are becoming increasingly anthropomorphic. It is important for psychologists, sociologists and many others to understand how our attitude towards them is formed and what it depends on. European scientists talk about this in an article published in the journal Social Cognition. In the first experiment, the participants solved the Trolley Problem with reference to humans and robots, anthropomorphic and not. During the second stage, psychologists examined what factors influence the anthropomorphic perception of the robot.
“The more 'human' the robot is perceived to be - or, more precisely, the more emotionality is attributed to the machine, the less often our subjects agreed to sacrifice it,” says one of the authors of the work, Markus Paulus. Perhaps these are quite expected results, but scientists make rather bold conclusions, pointing out the need for reasonable restrictions when creating anthropomorphic machines. “Attempts to 'humanize' robots shouldn't go too far,” says Marcus Paulus. "Otherwise, they will create a conflict with their original goal of helping people."
Sergey Vasiliev