Framing Artificial General Intelligence in Academic Literature and Reddit Discussions: Governance Implications

Abstract

Artificial General Intelligence (AGI) is increasingly discussed in both academic and public contexts, despite the absence of a stable or widely accepted definition. This creates challenges for governance, as different groups interpret AGI in different ways. While prior research has examined AI governance and public perceptions separately, few studies compare how AGI is framed across expert and public discourse. This thesis addresses this gap by analyzing AGI framing in academic literature and Reddit discussions between 2020 and 2025. It aims to identify dominant frames, explain their divergence, and assess implications for governance and public understanding. The study uses qualitative framing analysis, combining thematic coding and comparative analysis. The dataset includes 75 academic publications and 215 Reddit comments. Frames are examined across six dimensions: conceptual definition, timeline, certainty, risk type, responsibility attribution, and governance orientation. Academic discourse frames AGI as a manageable governance challenge, emphasizing technical control and institutional responses. Reddit discourse focuses on economic disruption, corporate dominance, and definitional uncertainty. These differences are explained by four mechanisms. Uncertainty supports governance in academic discourse but limits it in Reddit discussions. Academic discourse treats risk as systemic, while Reddit emphasizes immediate concerns. Academic texts stabilize definitions, whereas Reddit remains focused on disagreement. Finally, academic discourse assumes institutional reform is possible, while Reddit often treats corporate power as dominant. These divergences weaken governance communication, reduce policy legitimacy, and limit public engagement.