Playing Games with Robots Makes Them Seem More Humanlike, Study Finds

New research from the University of East Anglia, published in the Journal of Experimental Psychology: Human Perception and Performance, reveals that playing games with robots can make people perceive them as more humanlike. This finding holds significant implications for the future of robotics, especially as robots increasingly take on roles in care-giving, customer service, and various other settings. The study, led by Dr. Natalie Wyer from UEA's School of Psychology, aimed to understand how people interact with robots and whether these interactions can bridge the gap between machines and humans. With robots becoming more common in homes, hospitals, and workplaces, the researchers sought to determine if social engagements could make bots more relatable and acceptable to humans. In the experiment, over 100 participants interacted with a small, box-shaped robot named Cozmo. Half of the participants were encouraged to play a social game with Cozmo, while the other half interacted with it purely mechanically. Cozmo, known for its emotive capabilities, can display a range of expressions, including happiness, sadness, frustration, and curiosity, and is equipped with tank-like treads and a small lift arm. The results showed a marked difference in how participants perceived Cozmo. Those who played a social game with the robot viewed it as more humanlike compared to those who only engaged in mechanical tasks. Participants who engaged socially with Cozmo began to interpret its actions in a similar way to human actions, making systematic errors in their judgment of timing that typically occur when humans interact with each other. Dr. Wyer explained, "We found that when people interact socially with a robot, they start to perceive its actions more like they would with another human. This suggests that context is crucial. Even a non-humanoid robot like Cozmo can be seen as more humanlike and less like a machine if it is treated as a social actor initially." The study implies that social interaction is not just a superficial engagement but can deeply influence how we perceive robots. By fostering a sense of social connection, designers and developers can enhance the acceptance and integration of robots in various roles, making them more effective collaborators and companions. Dr. Wyer's findings highlight the importance of human-robot social dynamics in the broader context of AI and robotic technology. If robots are to become integral parts of our lives, they must be designed to engage in meaningful, social interactions that can trigger emotional and cognitive responses similar to those we have with other humans. This research adds valuable insights to the ongoing discourse on human-robot interaction and the ethical considerations surrounding the design and deployment of AI systems. It suggests that the initial "breaking the ice" phase through social interaction could be crucial in shaping human perceptions and comfort levels with robots. Industry insiders commend the study for its practical approach and findings. They believe that these results will influence future robot designs, emphasizing the need for social features that go beyond mere functionality. For companies focusing on robotics and AI, such as Boston Dynamics and Pepper Robotics, this research reinforces the idea that social engagement is a critical component of building successful AI systems. The study, titled "Observed Nonhumanoid Robot Actions Induce Vicarious Agency When Perceived as Social Actors, Not as Objects," further supports the notion that robots can evolve from tools to social entities, enhancing their utility and acceptance in human environments. As the field of robotics continues to advance, the emphasis on social and emotional intelligence will likely become a standard feature in AI development, ensuring a smoother transition and more harmonious coexistence with these technological marvels.