Resumen

This article examines the ontological and economic mutations disrupting filmmaking in the era of generative artificial intelligence. The historical shift from material optical capture to algorithmic prompting shatters the classical indexical pact of the image, imposing a regime of hyperreal and probabilistic simulacra. Through a hybrid research-creation methodology, combining critical theory (Frankfurt School, French post-structuralism) and auto-ethnography from daily practice in a production agency, we demonstrate how diffusion models silently impose a “default aesthetic.” This visual standardization contributes to a massive industrialization of the globalized imagination. Faced with this brutal rationalization, which demotes the demiurge-director to a mere statistical operator and restructures their cognition, the sovereignty of the author is directly threatened. To counter this technical and intellectual alienation, this article theorizes the manifesto of “algorithmic sabotage.” By rejecting opaque conversational interfaces in favor of open nodal architectures, the creator deliberately reintroduces entropy, error, and intentionality into the heart of the latent space. This technical and aesthetic resistance restores the author-function and the aura of the image, redefining contemporary mise-en-scène as an active struggle against the statistical norm.