Public Discourses on AI Companions in Indonesia: A Semantic Network Analysis of Twitter Conversations

Abstract

The advancement of artificial intelligence (AI) has transformed how humans communicate and make sense of technology. One notable example is ChatGPT, an AI-based conversational system that has become embedded in everyday digital life, functioning not only as a tool but also as a communicative partner. This study aims to map public discourse about ChatGPT on X (Twitter) using a semantic network analysis approach. Data were collected from user posts mentioning the keyword “ChatGPT” during August 2025. The dataset was processed and analyzed in Gephi to calculate degree centrality, betweenness centrality, and modularity, and then qualitatively interpreted through Papacharissi’s (2015) Affective Publics framework. The results reveal two dominant clusters: a functional cluster that emphasizes ChatGPT’s role as a productivity and learning assistant, and an emotional cluster that reflects users’ affective attachment to ChatGPT as a conversational companion that provides empathy and emotional validation. The “ChatGPT” node achieved the highest betweenness centrality, signifying its bridging position between rational and affective dimensions of human–machine communication. Drawing on the Affective Publics perspective, the study concludes that digital public spheres are shaped through technologically mediated emotional resonance and collective storytelling, producing new forms of shared meaning and social connectedness within Indonesia’s digital society.