Research on the Reform of College Teachers’ Teaching Methods Driven by Large Language Models

Abstract

The rapid development of large language models is profoundly transforming the knowledge supply mode, classroom interaction structure, learning support mechanisms, and evaluation logic in higher education teaching. Against this backdrop, the reform of teaching methods among university faculty is no longer a mere issue of technological application, but a systematic transformation involving teaching objectives, teaching organization, classroom relationships, evaluation systems, and teachers’ professional development. Based on the teaching practice of higher education, this paper synthesizes the policy orientation of educational digitalization, research on the educational application of generative artificial intelligence, and the reality of classroom reform in universities. It analyzes the historical background, value logic, and practical dilemmas of the reform of university teachers’ teaching methods in the era of large language models, pointing out that the current reform mainly faces challenges such as the risk of tool dependence, insufficient teachers’ digital intelligence literacy, lagging classroom evaluation, growing data ethics risks, and inadequate institutional supply. On this basis, it proposes that the reform should be systematically promoted from the dimensions of reconstructing teaching objectives, reshaping teaching processes, rebuilding evaluation systems, repositioning teachers’ roles, and improving institutional guarantees. The study holds that large language models will not replace teachers, but will reshape the boundaries of teachers’ teaching functions and their professional competence structures. Only by transforming from knowledge imparters to learning designers, thinking guides, process diagnosticians, and value gatekeepers can university faculty achieve a deep transformation of teaching methods and a substantial improvement in educational quality in the era of large language models.