Transformation and Reconstruction of Programming Course Teaching Mode in The Era of Large Language Model

Abstract

The rapid development of large language models is reshaping the teaching mode of programming courses in universities. While AI tools enhance learning efficiency and enable personalized tutoring, they also pose a risk of students becoming overly dependent and core computer thinking deteriorating. The traditional "grammar-driven" teaching mode faces severe challenges. Based on a systematic analysis of the current status of AI-assisted programming course teaching, this paper deeply analyzes the dual dilemma of "teaching facilitation" and "thinking hollowing out", pointing out the necessity and urgency of teaching mode transformation. In response to the current dilemma, this paper proposes a "thinking-oriented, human-machine collaborative" teaching mode, systematically elaborating on the transformation path from four dimensions: reshaping teaching objectives, reconstructing classroom teaching, upgrading project practice, and innovating evaluation systems. Practical teaching research conducted in two consecutive "C Language Programming" courses shows that this mode can significantly improve students' algorithm design and problem-solving abilities without the aid of AI, while effectively cultivating students' collaborative literacy in mastering AI tools to complete complex engineering tasks. This paper provides an operable theoretical framework and practical reference for the teaching reform of programming courses in the intelligent era.