AI in Business and Society
Loading...
Paper Number
1998
Paper Type
Completed
Description
Large Language Models (LLMs), e.g., ChatGPT, is expected to reshape a broad spectrum of domains. This study examines the impact of ChatGPT on question aksing in Q&A communitits via the natural experiment. Safe-guided by supporting evidence of parallel trends, a difference-in-difference (DID) analysis suggests the launching trigger an average 2.6% reduction of question-asking on Stack Overflow, confirming a lower-search-cost-enabled substitution. Our further analysis suggests that, this substitution effect has resulted in more longer, less readable and less cognitive and hence more sophisticated questions on average. Finally, the insignificant change in the score given by viewers per question suggests no improvement in the question quality and decreased platform-wide engagement. Our moderation analysis further ascertain the types of individuals who are more susceptible to ChatGPT. Taken together, our paper suggests LLMs may threaten the survival of user-generated knowledge-sharing communities, which may further threaten the sustainable learning and long-run improvement of LLMs.
Recommended Citation
Xue, Junzhi; Wang, Lizheng; Zheng, Jinyang; li, yongjun; and Tan, Yong, "ChatGPT Is A User-Generated Knowledge-Sharing Killer" (2023). ICIS 2023 Proceedings. 8.
https://aisel.aisnet.org/icis2023/aiinbus/aiinbus/8
ChatGPT Is A User-Generated Knowledge-Sharing Killer
Large Language Models (LLMs), e.g., ChatGPT, is expected to reshape a broad spectrum of domains. This study examines the impact of ChatGPT on question aksing in Q&A communitits via the natural experiment. Safe-guided by supporting evidence of parallel trends, a difference-in-difference (DID) analysis suggests the launching trigger an average 2.6% reduction of question-asking on Stack Overflow, confirming a lower-search-cost-enabled substitution. Our further analysis suggests that, this substitution effect has resulted in more longer, less readable and less cognitive and hence more sophisticated questions on average. Finally, the insignificant change in the score given by viewers per question suggests no improvement in the question quality and decreased platform-wide engagement. Our moderation analysis further ascertain the types of individuals who are more susceptible to ChatGPT. Taken together, our paper suggests LLMs may threaten the survival of user-generated knowledge-sharing communities, which may further threaten the sustainable learning and long-run improvement of LLMs.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
10-AI