This paper discusses the application of generative artificial intelligence (GAI) in foreign language teaching, particularly its practical effects in French intensive reading courses, English-medium instruction classrooms, and English writing courses. Shan Zhibin's team demonstrated how to integrate GAI tools with teachers' professional judgment to optimize the design and evaluation of continuation tasks, thereby enhancing students' language proficiency. Fang Fan's team introduced GAI chatbots to construct a translanguaging practice space, fostering student autonomy and critical thinking. Sun Peijian's team found through experiments that GAI feedback significantly improved the quality of college students' English writing. Li Mang's team explored the role of educational digitization in driving educational transformation and emphasized the teacher's key role in motivating learner agency.

1 Introduction

The continuation task is a language teaching activity that combines reading and writing, promoting language learning through synergistic effects. While widely used in English teaching, its practice and research in non-English language teaching remain insufficient. This study focuses on a second-year French intensive reading course, exploring the design, implementation, and optimization of continuation tasks from a human-AI collaboration perspective. The research aims to address two major challenges in implementing continuation tasks in intensive reading courses: the appropriateness of task design and the effectiveness of post-writing feedback. By combining AI tools with teachers' professional expertise, the study designs continuation tasks aligned with the course objectives and investigates how human-AI collaboration can enhance the relevance and effectiveness of feedback.

2 Literature Review

The continuation task is grounded in the theory of synergistic effects, facilitating the application and transfer of language knowledge through learners' deep cognitive interaction with the source text. Empirical studies support its effectiveness in improving lexical diversity, syntactic complexity, form-meaning mapping ability, and reading proficiency. The core of continuation tasks lies in deeply processing the context of the source text to achieve appropriate transfer of language forms. Combining language comprehension and production, the task enhances language acquisition efficiency and assessment reliability while positively guiding instruction. In intensive reading courses, continuation tasks should focus on the contextualized reuse of taught knowledge, ensuring correct matching of language forms and contexts to consolidate learning outcomes. AI tools in foreign language education provide technical support through context generation, multimodal resource creation, and instant feedback systems, reducing teacher workload and enabling personalized adaptation of teaching resources. In writing feedback, AI tools offer multidimensional diagnostic analysis, guiding learners to understand the cognitive logic behind language rules. However, effective AI tool application requires teachers to possess human-AI interaction negotiation skills, precisely defining task boundaries in the teaching process to maintain instructional control and achieve a collaborative human-AI co-teaching model.

3 Instructional Case Design and Implementation

This section introduces the design and implementation of a continuation task in a second-year French intensive reading course. The task is based on Michel Tournier's Friday or, The Other Island, aiming to help students comprehend the text's meaning and Robinson's emotional shifts, review grammatical phenomena, and learn relevant vocabulary. The core objective of the task design is to promote the contextualized transfer of target vocabulary, with AI tools generating the source text to provide students with new authentic contexts for continuation. The task was conducted in small groups outside class hours, requiring students to read the text and continue the story with coherent and logical plots. Evaluation and feedback followed a dual mechanism of AI preliminary review and teacher final review, focusing on emotional expression, language coherence, and alignment with the original context. Data collection included teacher-AI interaction logs, student continuation texts, AI feedback, and teacher feedback to ensure traceability of the research process.

4 Findings and Discussion

The study found that students' performance in transferring target vocabulary in continuation tasks was below expectations, possibly due to early task implementation and variations in source text design. Students performed relatively well in meeting the basic requirements of the writing task, producing complete and logically coherent plots while excelling in creativity and language use. Students held complex attitudes toward AI feedback, acknowledging its fairness but finding it difficult to match the quality of AI-generated texts. AI feedback, which emphasized strengths, contrasted with traditional teaching methods but also caused discomfort. A limitation of AI feedback was its lack of understanding of students' developmental trajectories, preventing personalized feedback. Teacher feedback demonstrated professional value by providing more accurate diagnoses and improvement suggestions. The optimal feedback solution should be based on a complementary human-AI collaboration mechanism, combining AI's instant surface-level feedback with teachers' in-depth diagnostic evaluations.

5 Conclusion and Implications

This study, set in a French intensive reading course, explored the application of human-AI collaboration in embedded continuation tasks. AI tools can quickly generate source texts that meet teachers' requirements, providing students with contexts for continuation. Embedded continuation tasks in intensive reading are open-ended and reflect students' language learning progress. Teachers can use human-AI collaboration to generate source texts with clear focus and highlighted target language points, guiding students in imitation. While AI tools can rapidly generate post-writing feedback, their lack of specificity prevents them from replacing teachers' in-depth diagnostic evaluations. Teachers' long-term support and professional judgment remain crucial for student learning. Human-AI collaboration combines AI's efficiency with teachers' expertise to jointly promote student learning. The use of AI tools highlights the irreplaceability of teachers in learning support and developmental guidance. Future research could encourage students to build personal learning portfolios and engage in regular interactions with AI tools to achieve continuous language proficiency improvement.