p>Researchers from the Singapore University of Technology and Design studied how young children aged three to five interpret information delivered by adults and by AI-controlled robots. The findings were published in the scholarly journal Child Development.
In the experiment, preschoolers from several early learning centers in Singapore interacted with both a human teacher and a humanoid social robot produced by SoftBank Robotics. Each agent provided labels for common objects, offering correct or incorrect descriptions of items such as a ball, a book, and other everyday objects.
After the labeling tasks, the participants were asked whether they trusted the human teacher or the electronic device more. The results indicated that the children more often favored the person over the device, even when the person had given mistaken descriptions.
One of the study’s co-authors, Li Xiaoqian, noted that children do more than memorize names. They assess the reliability of the information source based on past trustworthiness, which helps shape how they decide whom to learn from in the future. This selective social learning reflects a developing understanding of credible information sources among young learners.
The researchers emphasized potential takeaways for pedagogy as classrooms increasingly incorporate robots and digital learning tools. As AI-powered assistants become more common, educators can design experiences that help children distinguish reliable sources while maintaining positive interactions with technology.
Initial observations suggest that very young learners may not automatically view robots as trustworthy if they have limited exposure to intelligent machines. Over time, as interactions with smart devices grow, children could come to see robots as capable and dependable sources of knowledge. These shifts in perception may influence how future generations approach information and learning in both formal and informal settings.
In a related historical note, there were discussions at Moscow State University about when neural networks might begin to influence the homework routines of Russian schoolchildren. This historical context underscores the longstanding interest in how automated systems intersect with education and assessment. (Source: Child Development; related discussions from Moscow State University on educational technology and policy)