Multimodal Generative AI and The Metamedium Condition

Abstract

Since the decline of medium-specific approaches characteristic of early 20th-century modernism, various art theorists have introduced critical frameworks to address the fluidity of contemporary artistic practices. This article explores the implications of multimodal generative AI as a metamedium in contemporary art, building on the concepts of the post-medium condition and the postconceptual condition. As traditional notions of medium specificity dissolve, AI introduces a new paradigm where all artistic mediums converge into a single metamedium—data. This metamedium is inherently transmedial, assemblage-like, and deeply embedded in the social knowledge reflected in AI training datasets. By enabling transformations across text, image, sound, and other modalities, AI fosters a condition of metamediality, which challenges established boundaries of art forms. The article examines how this metamedium condition redefines creativity, positioning it as a collaborative process between human and non-human agents. Ultimately, it argues that the contemporary artistic landscape is now marked by a fluid interplay of mediums and temporalities, reshaping the very act of creation as a communal and boundless phenomenon.