生成型AI工具對做中學的影響

The Impact of Generative AI Tools on Learning by Doing

本文從Kolb的經驗學習理論出發，結合Dewey的經驗教育觀與von Glasersfeld的建構主義視角，探討生成型AI工具對「做中學」教育實踐的影響。研究指出，生成型AI工具不再僅是學習的輔助器具，而是形成具有「準主體性」的認知參與者，因此學習經驗的本質與轉化機制亦發生質變。亦即，生成型AI工具的介入促使學習歷程呈現分散認知與遞迴反思特徵，直接挑戰了學習理論中關於經驗、反思與概念建構的假設。本文進一步分析生成型AI工具在經驗學習四階段中的作用重構，提出AI參與下的知識建構已進入人機協同的協作時代，王體性、知識與認知主導權面臨重新定義的危機。最後，本文從AI素養教育的制度框架出發，提出既有政策與國際倡議大多仍聚焦於AI相關知識技能傳授，忽略反思與協作機制的設計，進一步呼籲教育體系應聚焦於學習者王體性的維護、生成型AI的適性轉化與知識建構的深度與廣度。

Grounded in David Kolb’s experiential learning theory and enriched by Dewey’s educational pragmatism and von Glasersfeld’s radical constructivism, this study critically examines the epistemological and pedagogical transformations brought about by generative AI tools in learning-by-doing practices. The research identifies that generative AI has evolved from a mere supportive tool to a quasi-subject in the learning process, altering the ontological structure of experience and knowledge formation. The intervention of AI reconfigures Kolb’s learning cycle􀈹especially reflective observation and abstract conceptualization􀈹into a recursive and distributed cognitive system, thus challenging traditional conceptions of learning subjectivity and knowledge ownership. This paper further analyzes the implications of AI-augmented co-construction in knowledge practices, emphasizing the emergence of dynamic subjectivity and cognitive hybridity. In response to current AI literacy frameworks, the study critiques dominant policies for prioritizing technical skills over reflective inquiry and co-learning practices. It concludes by proposing three core pillars for educational design in the AI era: preserving learner subjectivity, optimizing AI adaptability for personalized learning, and ensuring effective, transferable knowledge construction through recursive experiential cycles.