The Generative AI Divide: A Descriptive Analysis of Heterogeneous Adaptation Among Knowledge Contributors

How are engagement patterns within online knowledge communities changing in the context of generative AI? While prior research has documented an overall decline in platform activity following the release of ChatGPT, less is known about how different types of contributors have responded to this disruption. In this study, we conduct a descriptive, correlational analysis intended to document patterns and examine shifts in contributor engagement within online knowledge communities, using Stack Overflow as our focal case. We cluster 394,295 users into canonical contributor profiles based on pre-ChatGPT behavioral data. Using a two-step method combining graph-based community detection and guided Latent Dirichlet Allocation, we identify nine user roles and track their post-ChatGPT engagement. A matched panel comparison with a difference-in-differences-style specification documents bounded heterogeneity between clusters within a context of platform-wide decline. All contributor types exhibited reduced engagement, though the magnitude varied. Even the most resilient users, such as Steadfast Contributors, showed measurable declines, albeit comparatively milder ones. Other groups, including Post Retractors, Solution Seekers, and Thread Revivers, experienced steeper declines in both questions and answers. These findings document that while heterogeneity exists in how contributors responded, it is constrained: no group escaped the downward trajectory. We discuss potential interpretations for the evolving dynamics of knowledge sharing as generative AI reshapes participation across online communities.