About

This project explores the potential of queer expanded cinema to challenge and transcend heteronormative bias in AI video systems. Generative AI video is enabling novel forms of artistic expression and experimentation. However, these AI models may also perpetuate normative symbols, icons, and fantasies from cinematic history, limiting the medium’s expressive potential. This project aims to expose and subvert these limitations through a practice-based study that combines critical media theory, advanced AI video research, and expanded cinema arts.

Drawing inspiration from 20th-century radical filmmakers like Kenneth Anger, John Waters, and Carolee Schneeman, this project aims to develop subversive strategies—incorporating queer and expanded cinema tactics—that demystify and transgress AI video pipelines. These strategies include queer methods of appropriation, pastiche, and pleasure (Rich, 2013), and expanded cinema techniques of multi-screen projection, interactivity, and performance (Youngblood, 1970). Drawing on these methods and the recent literature on AI aesthetics, this project will produce and evaluate new audio-visual compositions of AI cinema. This practice-based study will demonstrate whether such techniques can disrupt the heteronormative fantasies embedded in AI video systems and generate more diverse, poetic, and sensual alternatives.

The expected outcomes of this three-and-a-half-year project include a written thesis that critically examines the theoretical and practical implications of queer AI filmmaking and a new body of work that showcases the aesthetic potential of this approach. Through an interdisciplinary methodology, this project seeks to contribute original knowledge to the fields of queer media studies, AI engineering, and the visual arts. This project also addresses an urgent need in the creative industries: to ethically redefine AI aesthetics and challenge the normativity of AI content proliferating across our media landscape.