There is another PhD graduate in our sHAI group! On November 5th, 2024, Marieke Wieringa successfully defended her dissertation at Radboud University, exploring a fascinating question: Can robots feel pain?
When Robots "Suffer"
Marieke and her team conducted experiments where participants interacted with robots in different ways. Some watched videos of robots being mistreated, while others were asked to shake a robot themselves. When the robot showed no reaction, participants had little hesitation. But when the robot displayed distress - whimpering sounds, pained expressions - people hesitated.
In one of her experiments, participants had a choice: perform a tedious task or shake the robot. When the robot (aka Cozmo) remained silent, most people had no issue shaking it. But when it indicated to be in discomfort they were more likely to chose the boring task instead.
The Risk of Emotional Manipulation
This emotional connection has real-world consequences. Marieke warns that businesses may exploit this vulnerability, much like how Tamagotchis once triggered emotional attachments. Imagine a future where a virtual pet begs for food - and you have to pay to feed it.
Her solution? Regulations on when and how robots and AI should display emotions. While banning emotional AI entirely isn't the answer - after all, it can be useful in therapy and social settings - we must stay vigilant. We like to think we are rational beings, but in the end, we are guided by emotions. And that’s a good thing - otherwise, we’d be just like robots.
Marieke's work also received great attention in the Dutch press, see for instance here: Hoe voelt het om een robot te mishandelen? ‘Mensen geloven graag dat ze ook pijn voelen’ - NRC.
🚀 Congratulations, Marieke, on your defense and fascinating research!
Commenti