A Theory of (the Technological) Mind: Developing Understanding of Robot Minds

child development;social cognition;theory of mind;uncanny valley;robots;trust in testimony;Psychology;Social Sciences;Psychology

Brink, KimberlyGelman, Susan A ;

University of Michigan

关键词: child development;    social cognition;    theory of mind;    uncanny valley;    robots;    trust in testimony;    Psychology;    Social Sciences;    Psychology;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/146010/kabrink_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

The purpose of this dissertation is to explore how children attribute minds to social robots and the impacts that these attributions have on children’s interactions with robots, specifically their feelings toward and willingness to trust them. These are important areas of study as robots become increasingly present in children’s lives. The research was designed to address a variety of questions regarding children’s willingness to attribute mental abilities to robots: (1) To what extent do children perceive that social robots share similarities with people and to what extent do they believe they have human-like minds? (2) Do attributions of human-like qualities to robots affect children’s ability to understand and interact with them? (3) Does this understanding influence children’s willingness to accept information from robots? And, of crucial importance, (4) how do answers to these questions vary with age? Across a series of five studies, I investigated children’s beliefs about the minds of robots, and for comparison adults’ beliefs, using survey methods and video stimuli. Children watched videos of real-life robots and in response to targeted questions reported on their beliefs about the minds of those robots, their feelings about those robots, and their willingness to trust information received from those robots. Using a variety of statistical methods (e.g., factor analysis, regression modeling, clustering methods, and linear mixed-effects modeling), I uncovered how attributions of a human-like mind impact feelings toward robots, and trust in information received from robots. Furthermore, I explored how the design of the robot and features of the child relate to attributions of mind to robots.First and foremost, I found that children are willing to attribute human-like mental abilities to robots, but these attributions decline with age. Moreover, attributions of mind are linked to feelings toward robots: Young children prefer robots that appear to have human-like minds, but this reverses with age because older children and adults do not (Chapter II). Young children are also willing to trust a previously accurate robot informant and mistrust a previously inaccurate one, much like they would with accurate and inaccurate human informants, when they believe that the robot has mental abilities related to psychological agency (Chapter III). Finally, while qualities of the robot, like behavior and appearance, are linked to attributions of mind to the robot, individual differences across children and adults are likely the primary mechanisms that explain how and when children and adults attribute mental abilities to robots (Chapter IV). That is, individuals are likely to attribute similar mental abilities to a wide variety of robots that have differing appearances and engage in a variety of different actions. These studies provide a variety of heretofore unknown findings linking the developmental attributions of minds to robots with judgments of robots’ actions, feelings about robots, and learning from robots. It remains to be seen, however, the exact nature of the mechanisms and the child-specific features that increase children’s willingness to attribute mental abilities to robots.