Kiss/Crash: Using Diffusion Models to Explore Real Desire in the Shadow of Artificial Representations

This paper describes the art installation of Kiss/Crash: a multi-screen work exploring the subject of AI-imagery and representation as well as the autobiographical themes of loneliness, desire, and intimacy in the digital age. The installation consists of three individual works in a shared space, Kiss/Crash, Me Kissing Me, and Crash Me, Gently, which all play with augmenting, inverting, and negating the iconic image of the kiss using diffusion-based image translation. In the production of this work, several potentially novel video translation techniques were developed and refined to create high-quality results central to these pieces. Throughout our research, we reflect on the nature of images and place diffusion models within a history of image-production technologies that artists have been contending with and responding to for the past one hundred years. This paper aims to extend that artistic tradition with a provocative, original aesthetic and technique that reveals the logic of AI imagery and hints at how our relationship to reality will continue to be stretched and shaped by artificial representations at an accelerating pace.