【摘要】 本研究旨在报告一项教学试验，以法语专业二年级精读课为场景，探索人机协同视角下读后续写任务的设计、实施与优化路径。研究发现，在任务设计中，基于大语言模型的人工智能（artificial intelligence，简称AI）工具能高效生成语境文本，为教师减负。续写任务本身具有开放性，难以定向激活特定语言结构的复用，但能反映学生的阶段性语言发展。在写后反馈中，AI工具表现出即时性优势，但缺乏对学生学习历程的把握，因此难以提供个性化指导。教师的诊断性评价仍具有不可替代性。本研究建议融合AI效率与教师教学知识，通过人机协同优化读后续写任务设计与反馈机制，助力学生语言能力持续发展。更多还原

【Abstract】 This pedagogical experiment, conducted in an intensive reading course for second-year French majors, investigates the design, implementation, and optimisation of continuation tasks through a human-AI collaborative lens. Findings demonstrate that LLM-based AI tools can efficiently generate contextualised input texts, thereby alleviating teachers’ workload in task design. Although continuation tasks, due to their open-ended nature, cannot reliably provide opportunities to use specific linguistic structures, they effectively reveal learners’ developmental progress over time. Regarding post-writing feedback, AI tools provide timely and comprehensive comments, but fail to offer personalised guidance due to their inability to track individual learning trajectories. In contrast, teachers ’ diagnostic assessments, grounded in longitudinal classroom observation, remain indispensable. This study proposes a synergistic approach that combines AI’s operational efficiency with teachers’ pedagogical expertise, aiming to optimise the design and feedback of continuation tasks and to foster the sustained development of learners’ language proficiency through human-AI collaboration.更多还原