【摘要】 生成式人工智能(Generative Artificial Intelligence, GAI)技术的快速发展已引起教育界的广泛关注和探讨。然而,与曾经信息技术赋能教育教学变革的发展路径相似,当前GAI融入教育教学的研究呈现出“重体系构建、轻学科实践”的“上宽下窄”局面。从学科教学视角出发,应当正视GAI赋能教育教学在研究主体上缺少学科研究者参与,在内容方法上坚持技术功能主义取向,在视野范畴上忽视学科教育复杂性的问题。应当从学科本位出发,制定GAI赋能学科教学的基本原则,包括以深化教学模式为目标、以彰显学科本质为前提、以场景差异定制为牵引、以构建“学习流”为方式。在这些基本原则指导下,以科学项目式学习场景为例,构建了生成式教学模式,并给出了相应的教学设计案例。未来, GAI赋能学科教学的研究应坚持:下即是上,立足学科本位考虑GAI的介入;小即是大,深入学科教学细节思考教学流程的重组;简即是繁,缩小智能体功能单位深化学科“学习流”设计;微即是宏,关注学科学习细节探寻生成式学习底层规律。更多还原

【Abstract】 The rapid development of Generative Artificial Intelligence(GAI) technology has attracted widespread attention and discussion in the education sector. However, similar to the path of information technology empowering educational teaching reform, current research on integrating GAI into education and teaching presents a situation of “broad at the top and narrow at the bottom,” focusing more on system construction and less on disciplinary practice. From the perspective of subject teaching, it is essential to address the issues of the lack of participation of discipline researchers in GAI-enabled education and teaching research, the adherence to a technical functionalist approach in content and methods, and the neglect of the complexity of discipline education. Starting from the discipline-based approach, the basic principles for GAI to empower subject teaching should be established, including aiming to deepen the teaching model, taking the demonstration of discipline essence as a prerequisite, customizing for different scenarios as a guide, and constructing the “learning flow” as a method. Guided by these fundamental principles, this study takes the scientific project-based learning scenario as an example, constructs a generative teaching model, and provides corresponding teaching design cases. In the future, research on GAI empowering subject teaching should adhere to the following principles: “what is below is above,” considering the involvement of GAI based on the discipline-based approach; “small is big,” delving into the details of discipline teaching to rethink the reorganization of teaching processes; “ simple is complex, ” reducing the functional units of intelligent agents to deepen the design of the “learning flow” in disciplines; and “micro is macro, ” focusing on the details of discipline learning to explore the underlying patterns of generative learning.更多还原