Pratiche visive e immaginazione algoritmica/ Visual Practices and Algorithmic Imagination

Abstract

This paper articulates a theoretical and methodological framework for understanding the visual apparatus of ViaggIAccademici as a locus of negotiation between human cognition and algorithmic generativity. Rather than treating generative AI as a merely operative tool, the study conceptualises prompting as a form of visual dramaturgy through which the theatrical text is reconfigured into a sequence of iconogenic units. The integration of LLM-mediated descriptions, hand-drawn studies and diffusion-based image synthesis reveals how artistic intentionality is modulated, refracted and at times destabilised by the latent structures governing machine-learning models. By incorporating a customised painterly dataset, the research examines stylistic drift, the epistemic status of the glitch and the emergence of visual indeterminacy as constitutive aesthetic conditions rather than technical contingencies. The resulting corpus of images and videos operates not as representational support but as a critical dispositif that expands the dramaturgical field, interrogating the boundaries between authorship, perception and algorithmic inference. The paper ultimately argues that contemporary artistic practice in the context of generative AI demands a shift toward a paradigm of co-emergent imagination, in which human and computational agencies converge to produce novel regimes of visual sense-making.