从效率增强到主体重建：人机协同教学中教师角色的再定位

摘要

生成式人工智能正在深度进入高校教学过程，知识讲解、资料生成、作业批改、学习反馈和课程评价等环节均可被算法工具部分承担。由此带来的问题并不只是教师工作效率的提升，也不只是教学工具的更新，而是教师角色、教学关系和学生主体发展的重新界定。已有研究分别从机器教学、人机协同教学、人本教育、教师数字化转型和人工智能教育治理等角度展开讨论，但对教师如何从效率增强工具的使用者转向学生主体发展的支持者，仍缺少系统分析。文章认为，人机协同教学中教师不可替代的价值不在于与人工智能竞争知识供给速度，而在于诊断学生理解障碍、设计有意义的学习情境、维持教学中的价值判断和责任边界。为此，教师角色需要从知识传授者转向认知诊断者，从资源提供者转向学习设计师，从评价执行者转向价值引导者。高校应通过教师人工智能素养培训、人机协同课堂设计规范、人工智能使用反思日志、教师专业共同体和慢思考环节设计，推动人机协同教学从效率增强走向主体重建。

Abstract

Generative artificial intelligence is becoming deeply embedded in higher education teaching. Knowledge explanation, material generation, assignment grading, learning feedback, and course evaluation can now all be partially undertaken by algorithmic tools. The resulting issue is not merely an improvement in teachers’ work efficiency or an update of teaching tools, but a redefinition of teachers’ roles, teaching relationships, and students’ subject development. Existing studies have discussed this issue from the perspectives of machine teaching, human-AI collaborative teaching, humanistic education, teachers’ digital transformation, and AI governance in education. However, there remains a lack of systematic analysis of how teachers can move from being users of efficiency-enhancing tools to becoming supporters of students’ subject development. This paper argues that the irreplaceable value of teachers in human-AI collaborative teaching does not lie in competing with AI in the speed of knowledge delivery, but in diagnosing students’ learning obstacles, designing meaningful learning situations, and maintaining value judgment and responsibility boundaries in teaching. Accordingly, teachers’ roles need to shift from knowledge transmitters to cognitive diagnosticians, from resource providers to learning designers, and from evaluation executors to value guides. Universities should promote this transformation by strengthening teachers’ AI literacy training, establishing norms for human-AI collaborative classroom design, requiring reflective logs on AI use, building teachers’ professional communities, and designing slow-thinking activities, so as to move human-AI collaborative teaching from efficiency enhancement toward subject reconstruction.

参考文献