Francesco D'Isa's AI-Generated Video Art Haunts Florence's Digital Refuge Bunker

Francesco D'Isa's solo exhibition 'Latent Rooms' at Rifugio Digitale in Florence presents unsettling video works created through a collaboration with generative AI. The artist, born in Florence in 1980, utilizes text-to-image models like Midjourney and text-to-video tools such as Seedance 2.0, refining outputs with his own previous AI-generated works. Screens lining the walls of the former air-raid shelter display short video art pieces that feature diffused light, domestic interiors, rural landscapes, and floral explosions. D'Isa describes the 'latent space' as the realm of all possible representations within an AI model, where commands create new points rather than uncovering pre-existing images. His aesthetic embraces glitches and errors as signature elements, diverging from attempts to mimic commercial productions like those from Pixar. This approach is influenced by Eastern traditions like Chán/Zen Buddhism and Taoism, treating AI as a disquieting machine that consumes and regurgitates images, opening onto a collective figurative unconscious. The works suggest a Baudrillardian hyper-reality where original objects are lost, leaving only representations of representations.