Enhancing Knowledge Exchange: The Impact of ChatGPT on Dedicated Users in Stack Overflow

In the past decades, information technology has been crucial in facilitating knowledge exchange activities by enabling connectivity and communication. This phenomenon is exemplified by the success of Wikipedia, open-source software development (e.g. Linux or GitHub), online learning (e.g., Coursea), and online communities (e.g., Reddit, Quora, and Stack Exchange). However, with the emergence of ChatGPT, an application of generative AI, the viability of such a collective intelligence model remains a matter of contention. Proponents argue that people can leverage generative AI to contribute higher-quality knowledge, while opponents question whether the demand for human-generated knowledge remains if generative AI can offer answers or solutions promptly. In this research, we collected data from Stack Overflow, the largest online Q&A community for programmers, and empirically examined the impact of generative AI on dedicated knowledge contributors, primarily focusing on the quality of their contribution. Our findings revealed that the quality of knowledge contributed from dedicated users improved after the introduction of ChatGPT, which was attributed to 1) the advancement of complementary explanations accompanying the technical solutions and 2) the provision of conclusive answers. Additionally, we found that “generalists”, dedicated users contributing diverse knowledge, benefited more from generative AI.