IA generativa: o quinto paradigma da imagem

Abstract

This research investigates the emergence of generative text-to-image artificial intelligence in contemporary visual creation, analyzing whether image production with AI can be characterized as a new paradigm distinct from the four paradigms described by Lucia Santaella. The study begins with a theoretical review of Santaella's paradigms of the image and the technologies that shaped them, including artisanal, photographic, post-photographic, and hybrid production. Subsequently, it delves into the technical and conceptual analysis of generative artificial intelligences, such as DALL-E, MidJourney, and Stable Diffusion, examining their technological, productive, and creative foundations. Using a comparative methodology, the research evaluates the similarities and ruptures between previous image production models and generative AI, considering aspects such as the role of the producer, storage, transmission, and reception mechanisms. The study concludes that generative AI inaugurates a paradigmatic moment in the history of images, introducing a model of human-machine co-creativity that reconfigures traditional image-making practices and significantly expands the possibilities of creation and access to visual production. The impact of this transformation suggests an ontological reformulation of images and their cultural interactions, consolidating generative AI as a new evolutionary stage in visual production