Human-Machine Collaboration: Boundaries and Practical Paths of Generative AI in Spoken English Teaching

Abstract

Grounded in the philosophy of technology and humanistic pedagogy, this study focuses on the application boundaries of generative AI in oral English teaching, aiming to clarify the division-of-labor logic in human-machine collaboration. Based on Stiegler’s “technological prosthetics” and “pharmacology” models, as well as humanistic teacher-centered theory, the study reveals that the auxiliary value of generative AI lies in mechanical, standardized, and low-culture-loaded tasks. In contrast, the uniqueness of human teachers is rooted in domains beyond AI’s reach, such as transmitting emotional resonance, deeply interpreting cultural contexts, and guiding dynamic thinking. A three-dimensional boundary criterion and a four-quadrant model are proposed to define the practical scope of collaboration, where technical proficiency determines the feasibility of AI intervention, while the irreplaceability of teachers defines the necessity of their roles. Additionally, a dual-path approach is proposed: developing teachers’ critical AI and cultural pedagogies, and designing tools with cultural diversity to mitigate bias. The study concludes that the the essential boundary is ethical, safeguarding “human subjectivity and cultural nature,” thus offering a framework for responsible integration.