Assessment of ChatGPT-generated programming code based on exercises in an introductory programming course

Abstract

This study aims to assess the quality of the programming code generated by ChatGPT and compare it to the programming code used in an introductory programming course. ChatGPT is a software that produces programming code based on descriptions entered by users. Our study compares this generated code with a code we use in our programming course for the same programming exercises. ChatGPT is not a single software, and it is built from scratch. Instead, ChatGPT is built on a multitude of software that all contribute to the development of ChatGPT. This multitude of technologies falls under a general term called “OpenAI”. To better understand the generated code and the assessment we intend to provide in this study, knowing these technologies that preceded ChatGPT and contributed to its formation will be helpful. Thus, we provide a literature review of the technologies that started before ChatGPT and then compare the code generated by ChatGPT with the code that we use in our course for the same assignments or problems.