AI作为“认知对手”：高中物理概念转变的教学设计——以“牛顿第一定律”为例

AI as a Cognitive Opponent: Instructional Design for Conceptual Change in High School Physics—Taking “Newton’s First Law” as an Example

摘要: 针对高中物理概念教学中学生迷思概念顽固、传统讲授难以触及认知重构的困境，提出利用生成式AI构建“认知对手”型智能体的教学设计思路。基于皮亚杰平衡化机制、维果茨基最近发展区理论与苏格拉底式提问，构建了“认知脚手架搭建→认知冲突触发→概念顺应与巩固”三阶段认知路径模型，并明确了教师主导、AI认知对手、学生主体的人机协同边界。以“牛顿第一定律”教学为例，运用四问设计方法开发了“亚里士多德式AI对手”智能体，呈现了完整的教学设计、对话引导结构及课堂实施片段。通过对学生对话的SOLO层次分析，揭示了概念转变的认知跃迁路径。课堂观察与描述性数据初步表明，该设计有助于触发学生的生产性认知失衡，并促使学生在反驳与论证中主动建构科学概念，为AI赋能物理概念教学提供了可迁移的设计范式。

Abstract: Aiming at the predicament that students’ misconceptions are stubborn and traditional teaching is difficult to touch on cognitive reconstruction in high school physics conceptual teaching, this paper proposes an instructional design idea of using generative AI to build a “cognitive opponent”-type agent. Based on Piaget’s equilibration mechanism, Vygotsky’s zone of proximal development theory and Socratic questioning, a three-stage cognitive path model of “scaffolding construction → cognitive conflict triggering → conceptual accommodation and consolidation” is constructed, and the human-machine collaborative boundary of teacher-led, AI cognitive opponent and student-subject is clarified. Taking the teaching of “Newton’s First Law” as an example, the “Aristotelian AI opponent” agent is developed using the four-question design method, presenting the complete instructional design, dialogue guiding structure and classroom implementation fragments. Through the SOLO hierarchical analysis of student dialogues, the cognitive leap path of conceptual change is revealed. Classroom observations and descriptive data preliminarily indicate that this design may help trigger productive cognitive disequilibrium and encourage students to actively construct scientific concepts in refutation and argumentation, providing a transferable design paradigm for AI-empowered physics concept teaching.