Accusations of 'AI Slop' Don't Screen for AI Text (arxiv.org)

Recent research has revealed a sharp rise in the use of the term "AI slop" as a pejorative label for what users perceive as inauthentic comments in online discourse, particularly on platforms like Hacker News and Reddit. An analysis of 25 million comments from 2023 to 2026 showed that accusations targeting AI-generated content have more than doubled, constituting 94% of negative mentions. Interestingly, a matched-control test indicated that the features distinguishing AI from human-written text did not correspond to which human comments were flagged, suggesting that current accusations serve more as a form of social gatekeeping rather than a legitimate detection method. This shift is significant for the AI/ML community as it highlights a new social dynamic around the perceived authenticity of online communication. Instead of accurately screening for AI-generated text, users are using these accusations to signal membership in a perceived in-group and to reinforce standards of authenticity. The findings imply that emerging technologies for AI detection may struggle to address this evolving discourse, raising questions about how society interprets and responds to the increased integration of generative AI in content creation. This underscores the need for a deeper understanding of the social implications of AI in writing, separate from its production capabilities.