【摘要】 随着生成式人工智能特别是大语言模型的迅猛发展，智能系统正深入高等教育教学全过程，推动课程从“师—生”模式向“师—生—机”模式转变。针对智能化水平参差不齐与深度应用带来的风险挑战，分析当前课程建设呈现的“欠智能化”与“过智能化”并存困境，提出“师—生—机”(TSM)模型，从教师主导性、学生依赖性与大模型偏好性3个维度出发，探讨如何构建可量化、可调控的评估指标体系与分析框架，以应对新时代教育智能化的双重难题，通过在Java程序设计课程中的验证应用说明该模型的应用效果。更多还原

【Abstract】 With the rapid development of generative artificial intelligence,particularly large language models,intelligent systems are increasingly integrated into the entire process of higher education teaching,driving a shift in course structures from a traditional “teacher–student” model to a “teacher–student–AI” paradigm.In response to the uneven levels of intelligent integration and the risks arising from deep AI involvement,this study analyzes the coexistence of under-intelligent and over-intelligent in current course development.It proposes a teacher–student–machine (TSM) model,which examines three key dimensions:teacher dominance,student dependency,and large-model preference.Based on these dimensions,the study explores how to construct a quantifiable and regulatable evaluation index system and analytical framework to address the dual challenges of educational intelligence in the new era.The effectiveness of the proposed model is further demonstrated through its application in a Java Programming course.更多还原