Generative AI Echoes Historical Shifts in Art & Authorship

Researchers frame Generative AI not as a new break from the past, but as the latest stage in a long history of artistic and technological development. The paper traces the evolution of generative art, arguing that current questions about authorship and creativity have historical parallels. They propose a system for categorizing generative AI based on whether it functions as a medium, artwork, or instrument, asserting that these classifications are determined by editorial choice rather than inherent qualities.

The study identifies potential risks of creative automation, including cognitive decline and “Model Collapse,” and suggests “environmental enrichment” as a possible solution.

Ultimately, the authors posit that the role of the artist is shifting from creator of objects to a designer of systems and curator of outputs. They introduce the concept of “Manifestation” as a third state for generative works, beyond simply original or copy, and suggest that older, less stable generative models offer aesthetic possibilities lost in contemporary aligned systems.