Generative AI as a probabilistic medium: a systems-theoretical analysis of the art system
摘要
This article examines generative AI as a probabilistic medium that reshapes communicative selection under conditions of structural opacity. Drawing on systems theory, it argues that generative AI alters how communication selects, attributes, and stabilizes meaning. The analysis focuses on the art system in order to show how probabilistic media disrupt established programs of authorship and authenticity. As AI-generated outputs cannot be securely tied to singular human origins, attribution shifts toward externally verifiable signals such as process documentation, provenance, and cryptographic certification. The broader significance of generative AI, the article argues, lies less in its adoption as a tool than in its reorganization of the conditions under which social systems process novelty, legitimacy, and selection.