Points of Contention in Social Media Discourse on Generative AI:Visualizing Value Systems through Utterance Analysis

Abstract

　In recent years, increasingly complex conflicts on social networking services (SNS) have become difficult to resolve due to the obscuration of key issues. To clarify such issues, attention was directed toward the underlying values held by individuals—an approach that has not been sufficiently addressed in prior research. The controversy surrounding "generative AI" was used as the topic of discussion, and a quantitative analysis was conducted on SNS utterances based on value system annotations. Utterances were classified into 23 distinct value categories using the KJ method. The co-occurrence of values across supportive and opposing positions was examined, and the centrality of each value was analyzed through network analysis. The findings revealed that the discourse consists of affirmative utterances centered around utilitarianism, which emphasizes the pursuit of benefit and technological advancement, and negative utterances centered primarily on copyright concerns, while incorporating a broader range of values. Notably, utilitarianism and copyright concerns appear difficult to reconcile, suggesting that the tension between these value orientations serves as a key point of contention.