席卷而来的生成式人工智能（AIGC）引发了全球人才能力结构需求的变化，未来的AIGC必将在社会分工中承担重要角色，人类与AIGC协同工作已成趋势。教育研究者已从多维度对这一新兴技术展开了探讨，一线教师对该技术应用于教学也有强烈的现实需求。如何培养适应AIGC时代的个体，如何利用AIGC赋能教学，AIGC是否会增加学生的依赖性降低其主体性等问题随之而来。因此，本研究聚焦于探讨AIGC对教学要素的影响，从教学实际出发，为一线教师提供可操作的基于AIGC的人智协同教学模式，并探讨AIGC参与教学对学生的影响。本研究围绕以下三个问题展开：基于AIGC的人智协同教学要素及其结构是什么？基于AIGC的人智协同教学模式及实施策略是怎样的？人智协同教学模式下学生交互特征及能力变化如何？针对第一个研究问题，研究者基于混合教学七要素框架，探讨了将AIGC作为主体要素之一加入教学系统后对其他各要素及要素间关系的影响，构建了人智协同教学八要素框架。针对第二个研究问题，本研究通过对72位有丰富混合教学经验的一线教师的访谈记录进行三级编码后，分析出当前教学实践对AIGC的真实需求，结合AIGC的能力特征及创新扩散条件，构建了基于AIGC的人智协同教学模式。该教学模式将首要教学原理与自我调节学习理论为依据设计了七个教学环节，研究对其开展了两轮的迭代应用。第一轮初步应用中，研究者对合作教师进行AIGC技术和人智协同教学模式的培训后，让教师自主依据该模式进行教学设计与实施，研究者在此过程中观察、答疑，但不干预教学。第二轮改进应用中，研究者始终与合作教师紧密联系，严格基于该模式进行教学设计与实施。两轮迭代后，提炼出了人智协同教学模式的关键：8种人智协同类型，面向教师的“协同共析、协同设计、协同共教、协同共评”，面向学生的“协同共知、协同共解、协同共创、协同共评”，并依据八要素框架总结了12条模式实施条件与有效策略。针对第三个研究问题，学生在该教学模式下与AIGC协同时，表现出在理解中创造，在创造中深度理解的认知趋势；人智协同话语序列复杂度与任务内容的难易、复杂程度相关；学生交互体现出了较强的主体意识与批判性思维；以小组的形式在线下与AIGC交互时，话语序列更为复杂和动态，后期又趋于简单；学生的AIGC协同能力和自我调节学习能力均有所提升。
The emergence of Artificial Intelligence Generated Content (AIGC) based on the Large Language Model technologies has triggered a shift in the talent capability structure across various industries globally. In the future, AIGC will undoubtedly play a significant role in the division of labor within society, making collaboration between humans and AIGC inevitable. In light of this groundbreaking technology, educational researchers have embarked on multi-dimensional discussions, and there is a strong practical need for its application in teaching among frontline educators. Questions on how to utilize AIGC to better cultivate students, how to develop individuals who are adaptable to the AIGC era, and whether AIGC might increase students‘ dependency while diminishing their agency have arisen, with empirical research yet to provide answers. Therefore, this study focuses on exploring the impact of AIGC on teaching elements, offering frontline teachers an operational Human-AIGC collaborative teaching model, and examining the effects of AIGC-based Human-AI collaborative teaching on students.This research addresses the following three questions: What are the elements and structures of AIGC-based Human-AI collaborative teaching? What constitutes the AIGC-based Human-AI teaching model, and what’s the process like, what’s the implementation strategies of the model? How do student interaction characteristics and capabilities change under this teaching model?For the first question, this research, based on a framework of seven elements of blended teaching, discussed the impact of incorporating AIGC as an agent element into the teaching system on other elements and their interrelationships, thus constructing an eight-element framework for AIGC-based Human-AI collaborative teaching. For the second question, by conducting three-level coding of interview records from 72 vocational education teachers with extensive experience in blended teaching, the study identified current teaching practices‘ needs for AIGC. Integrating AIGC‘s capability features and conditions for innovation diffusion, a AIGC-based Human-AI collaborative teaching model was constructed. This model, integrates first principles of instruction with self-regulated learning theory into seven teaching segments, which was iteratively applied in two rounds. In the initial application, researcher trained teachers in AIGC technology and the AIGC-based Human-AI collaborative teaching model, allowing them to design and implement teaching independently, with researcher observing and answering questions but not intervening. In the improved application round, researcher maintained close contact with the teacher, strictly implementing the teaching design according to the model. After two iterations, key aspects of the AIGC-based Human-AI collaborative teaching model were distilled: eight types of AIGC-based Human-AI collaboration, four aimed at teachers - collaborative analysis, design, teaching, and assessment，and four at students - collaborative knowing, understanding, creation, and assessment, summarized under the eight-elements teaching system along with 12 conditions and effective strategies for model implementation. For the third research question, students showed cognitive trends of creating in understanding and deeply understanding in creating while collaborating with AIGC. The complexity of Human-AIGC discourse sequences correlated with the difficulty and complexity of tasks; student interactions exhibited strong agency and critical thinking. When interacting with AIGC in group discussion offline, the discourse sequences were more complex and dynamic, later more simplified; students‘ collaborative capabilities with AIGC and their self-regulated learning abilities were both enhanced.
