The Imparts and Pedagogical Challenges of AI-Generated Codes on Computer Programming Education at Introductory Level: Curriculum Design Recommendations

Abstract

The variety of generative artificial intelligence (AI) applications has led to a sharp rise in interest in their application in Learning programming and the process of writing programs. From the reviewed literatures, researchers are yet to find best ways to integrate AI-generated programs in the pedagogy of teaching and learning programming especially in the introductory level where understanding the fundamentals is crucial. The researchers and teachers need to think about ways to maintain academic integrity and guarantee a rigorous and equitable assessment of students' programming skills in this era of Large Language Models (LLM) rather than just viewing AI as a danger to academic integrity. How are computing educators planning to modify their courses in response to the increasing proliferation of AI code generation and explanation tools, considering the current state of affairs and looking into the future when AI capabilities are likely to improve? This paper harnessed the experiences of using LLM (AI code generating tools) for teaching and learning computer programming at the introductory level based on the perspective of the students, the teachers and the researchers. This paper also gathered various perspectives on the above question based on the existing works, position papers and workshops by the computing community. Recommendations on the possible best ways to integrate LLM or AI code-generating tools were made. By providing recommendations for teachers and curriculum developers on how to successfully integrate these tools into teaching and learning programming, this paper will advance knowledge of the possible advantages and consequences of using them in educational environments.