SYNTHETIC CINEMA: GENERATIVE OR DEGENERATIVE? NOVELTY OR CULTURAL AUTOPHAGY?
SYNTHETIC CINEMA: GENERATIVE OR DEGENERATIVE? NOVELTY OR CULTURAL AUTOPHAGY?

Summary/Abstract: This paper argues that synthetic cinema can achieve artistic novelty through an author's singularity, that is, the novelty of a personal mode of production, not through the same automated process of production. Otherwise, with the advent of generative AI, the delegated automation of audiovisual production contributes to a linguistic and aesthetic standardization of cinematic language through predictability. The artistic potential of generative AI thus presupposes an ethos on the part of filmmakers, such that these technologies ensure an ecology of production rather than an autophagy of creation. The syntheses produced by large models degrade over time, a process through which the output becomes the input for further syntheses and future models. Conversely, the use of small models, trained on limited datasets, such as personal or collective archives, offers filmmakers novel ways to explore cinema critically, through self-reflexivity. Much like analog degradation or digital glitch art, generative degradation can become an aesthetic, a form of synthetic expressionism. By creating their own models, filmmakers select the contents used for training qualitatively, whereas large commercial models quantitatively appropriate random data from the infosphere.

• Details
• Contents