Public attitudes and sentiments toward ChatGPT in China: A text mining analysis based on social media

Abstract

ChatGPT, an innovative artificial intelligence language model, is attracted significant attention around the world, sparking both enthusiasm and controversy, but identifying its societal impact and addressing its potential concerns necessitate an understanding of the prevailing public's attitudes toward the tool. In this study, we leverage text mining techniques to analyze the sentiments and themes prevalent among Chinese social media discussions of ChatGPT. In total, 96,435 comment data and 55,186 repost data were used, and the results show that public discussions mainly focused on ChatGPT's technical support, AI-related effectiveness, impact on human work, and effects on education and technology. Concerns were related to disinformation risks, technological unemployment, and the human–computer relationship. In addition, we found that social media played a prominent role in information dissemination, while official media and government units demonstrated a limited influence. The insights obtained through this study can inform policymakers, industry stakeholders, and the public of the public's prevailing attitude toward AI technologies, and they can facilitate informed decision-making.