Heterogeneous Effects of Generative artificial intelligence (GenAI) on Knowledge Seeking in Online Communities

• 成果类型: Article
• 署名作者: Quinn, Martin; Gutt, Dominik
• 署名单位: Erasmus University Rotterdam; Erasmus University Rotterdam - Excl Erasmus MC
• 刊物名称: JOURNAL OF MANAGEMENT INFORMATION SYSTEMS
• ISSN/ISSBN: 0742-1222; 1557-928X
• DOI: 10.1080/07421222.2025.2487313
• 发表日期: 2025-04-03
• 页码: 370-399
• 关键词: generative AI; GAI; GenAI; knowledge seeking; commitment-based theory; information foraging theory; knowledge-sharing communities; CORE-PERIPHERY TENSION; organizational commitment; incentives; INNOVATION; turnover; IMPACT

摘要

Generative AI (GenAI) may fundamentally reshape how users seek knowledge in online knowledge sharing communities. Although prior work found an overall decrease in knowledge seeking in online communities upon the availability of GenAI, the underlying dynamics across user groups have remained unexplored. This study addresses that gap. Drawing on commitment-based theory, we hypothesize that casual users-motivated by cost-benefit considerations-are more likely to reduce their question-posting activity than highly committed members. Using a difference-in-differences analysis, we find that ChatGPT's arrival leads to a substantial drop in questions on StackExchange, primarily driven by casual users (about 18.2%). Motivated by information foraging theory, we reveal heterogeneous downstream effects of GenAI on question characteristics. In particular, we find that the questions by casual users become more complex and novel, while those by intensive and top users do not. These results highlight the importance of heterogeneous user motivations in shaping platform dynamics, underscoring that while GenAI may diminish overall participation, it may also increase the value of the remaining content. Our study offers insights for knowledge sharing communities, managers, and stakeholders reliant on user-generated data, providing a nuanced view of GenAI's disruptive influence.