Ever since Turing’s definition of artificial intelligence as an indistinguishable simulation of humans, it has been characterized as a fiction. Art itself, moreover, is based on fiction, as Pliny’s account of Zeusi and Parrasius describes well: the painter Parrasius wins the competition with Zeusi because he manages to fool the man by simulating a veil over his painting. We are facing a deepfake ante litteram. The so-called trompe l’oeil, literally “deceives the eye,” is a dominant feature of painting, generating the illusion of reality from Roman art to the present. AI generates synthetic worlds that by their nature are artificial, not real. It should be identified as fiction, which distinguishes it from photography, which has been used for more than a century as proof of existence. An AI system can never generate an actual image of a battle; a photograph can. Unlike “technologies of the real” (photography, video, X-ray…), AI and the metaverse propose “fake people in fake worlds.” This raises ethical and cultural issues, in journalism for example, while in art, which does not claim to adhere to reality, this “falsification” is not problematic, and can be exercised with complete freedom.
Italiano