생성형 AI 활용 대학 수업에서 학습자가 경험하는 모순과 대응 양상: 활동이론의 관점에서

초록

As the use of generative AI by university students continues to increase rapidly, tensions are emerging between traditional systems of higher education and the integration of new learning tools. Learners often adopt generative AI without sufficient reflection on its potential benefits and drawbacks, and institutional norms or guidelines for appropriate use have yet to be established. This study aimed to analyze the contradictions experienced by students in university courses that incorporate generative AI tools, as well as the strategies they employ to resolve these contradictions, through the lens of activity theory. In-depth interviews were conducted with nine undergraduate students from a university in Seoul, and the data were analyzed thematically. The findings identified six key contradictions arising in the context of generative AI-supported learning: (a) mixed interpretations of the role of generative AI (contradiction within the tool), (b) tension between learning efficiency and learner agency (contradiction within the subject), (c) superficiality and inaccuracy of AI-generated responses (contradiction between tool and object), (d) lack of AI literacy and usage skills (contradiction between subject and tool), (e) absence of clear guidelines and evaluation standards (contradiction between rules and object), and (f) instructor-centered classroom dynamics (contradiction between community and object). To navigate these contradictions, learners employed a range of strategies, which were categorized into three overarching types: (a) establishing an autonomous learning stance and reorganizing learning strategies, (b) recognizing the limitations of AI tools and adapting question formats accordingly, and (c) engaging in collective verification of AI-generated content. Based on these results, the study provides practical implications for the design and implementation of effective generative AI-integrated university instruction.

키워드