Public Perceptions of Generative AI: Insights from Social Q&A Platforms
摘要
Generative AI (GenAI) discussions on social Q&A platforms reveal diverse public perceptions and user engagement shaped by platform-specific characteristics. This study analyzed discussions on Zhihu, Quora, and Reddit using topic modeling, topic similarity analysis and sentiment analysis, uncovering distinct user priorities and emotional responses. Zhihu focused on content creation, ethics, and personalized learning, with polarized sentiments. Quora emphasized business strategies, customer support, and technical learning, showing predominantly positive emotions. Reddit explored a wide range of topics, including communication tools and software development, with balanced sentiments. Cross-platform analysis revealed shared interests in communication, translation, and education, with notable overlaps in personalized learning and technical applications. These insights highlight how user demographics influence GenAI discourse, offering practical guidance for tailoring tools to regional needs. The comparative results have both practical and theoretical implications, offering valuable insights for stakeholders to tailor applications and strategies for different user groups related to GenAI.