Beyond automation: Artificial intelligence and embodied craft in artistic practice

Abstract This article examines artificial intelligence (AI) in artistic practice not as an autonomous creative subject or a mere labour-saving device, but as a field of models, affordances and constraints through which artists may think, test and transform their materials. Against commercial narratives that identify intelligence with efficiency, productivity and the automation of economically valuable work, it argues for the continuing centrality of embodied craft intelligence: the tacit, historically situated, perceptual and ethical knowledge developed through practice. Drawing on Leo Apostel’s theory of models, James Gibson’s concept of affordances, Michael Polanyi’s tacit knowledge, Walter Benjamin’s aura, Theodor W. Adorno’s critique of fetishism and Isaac Asimov’s story Profession, the article proposes a critical, feedback-based understanding of human–AI interaction. It distinguishes serendipitous exploration from template-driven production and suggests that the artist’s responsibility lies in judging when machine outputs become meaningful deviations rather than seductive shortcuts. The future of AI in the arts depends less on the technology’s generative power than on whether artists can preserve slow perception, historical memory, critique and responsibility within increasingly automated cultural infrastructures. Its central claim is that AI becomes artistically valuable only when kept visible as a contestable tool rather than accepted as a substitute for artistic judgement.