Recently, various discussions have been conducted on teaching and learning methods considering the stage of cognitive development of elementary school students. Currently, most curriculums consist of conveying the concept of AI and experiencing intere...
Recently, various discussions have been conducted on teaching and learning methods considering the stage of cognitive development of elementary school students. Currently, most curriculums consist of conveying the concept of AI and experiencing interesting services, but discussions on essential educational goals are ambiguous. From this point of view, Minsky and Peppert, who initially discussed computing thinking and AI education for children, emphasize thinking about thinking as an educational goal and argue that programming, algorithms, and computers are efficient tools for expanding such cognitive capabilities. In this study, the perspective of “thinking about thinking” is designed and applied to elementary AI education, and whether it has a positive effect on children's ability to observe, express, and improve intelligent thinking. Among the ‘classification·group’ models of AI, unplugged activities for ‘thinking about thinking’ were designed and class experiments were conducted under the theme of image and sound recognition. As a result of analyzing the activity sheets and individual interview contents, the learner analyzed his or her intelligent behavior and judgment to solve the problem n-dimensionally and discovered various features for accurate algorithms. In addition, he recognized that ‘AI is an imitation of complex human intelligence'. In this study, a new perspective on educational goals and teaching and learning methods in elementary AI education and their effects were presented using the ‘thinking about thinking’ approach.