Face Value: On Generative AI, Audiovisual Production and The Promise of Democratization
Exposé by Theresa Büchner
At the core of the project lies a scientific reflection on the historical, aesthetic, and socio-economic conditions of the development of generative AI, accompanied by two practical explorations: the print series „Corpus“ and an audiovisual piece. Both are complemented by the seminar „Schreiben und Schälen“. Together, these components form a constellation in which theory and artistic practice inform one another.
The history of media art is marked by recurring unease toward emerging technologies. When photography emerged, it seemed to threaten painting. When video became affordable, it was considered aesthetically inferior to celluloid film. Today, generative AI appears to automate and potentially displace the artistic act. The generating apparatus, located somewhere in the world, eludes direct experience. What we see and hear are visual and auditory representations of code, attempts by a program to interpret image and text instructions. At the same time, AI-generated images and sounds are derived from the compressed residue of training data, which technically no longer allows access to the original material. These technological shifts are accompanied by powerful narratives oscillating between utopian promise and dystopian anxiety, frequently infused with religious motifs and strategically mobilized by industry figures, media actors, and critics alike. Such narratives shape both the perception and commercialization of generative AI.
Against this backdrop, the project examines the socio-economic conditions of AI-assisted audiovisual production, focusing in particular on its promise of democratization. The notion that audiovisual production can be relocated into the domestic sphere, decoupled from large teams and expensive equipment, is critically situated in relation to artistic and feminist practices that have long operated within constrained, everyday environments.
Against this backdrop, the project examines the socio-economic conditions of AI-assisted audiovisual production, focusing in particular on its promise of democratization. The notion that audiovisual production can be relocated into the domestic sphere, decoupled from large teams and expensive equipment, is critically situated in relation to artistic and feminist practices that have long operated within constrained, everyday environments.
Cut up and reassembled, depictions of the human body in the series „Corpus“ emerge through an iterative working process in which digitally processed photographic images and found motifs are combined using generative AI. The medium appears marked by a distinctly morbid quality and an ambivalence towards or outright rejection of bodies, which finds expression in its military origins and current applications, the wasteful consumption of natural resources, as well as the boundless appropriation and subsequent pulverisation of cultural assets. Drawing on the concept of „Exocapitalism“ (Trillo/Poliks, 2025) the use of generative AI will be further explored in the generation of images and sound for an audiovisual piece, with particular attention to the use of photographic archive materials as prompts for node-based workflows.
In response to the rapid implementation of generative AI in everyday communication, media, and cultural production, the seminar „Schreiben und Schälen“ at HAW Hamburg positions writing practices against tendencies toward increasingly automated language production. Observing closely and refining one’s senses are understood as fundamental prerequisites for any form of staging and artistic creation. In other words, if you want to know what a generated potato should look and sound like, you must first peel a potato. This proposition culminates in a collective cooking situation inspired by a key scene of Chantal Akerman’s movie „Jeanne Dielman, 23 quai du Commerce, 1080 Bruxelles“ (1975). Accompanied by a screening of the film, participants peel potatoes and prepare potato salad. Selected texts that emerge over the course of the seminar are published in the format of a newspaper – a medium that may have lost its former authority in the production of truth, yet remains remarkably useful when peeling potatoes.
The project ultimately aims to critically assess the potential and contradictions of GenAI-based audiovisual production. In doing so, it takes the promotional promise of democratization at its word, probing what it actually means to enact and inhabit these claims in practice. It asks under what conditions a reflective appropriation of generative AI might be possible, given the apparent ecological, legal, and labor-related concerns, and understands it as embedded within existing power structures that shape both its outputs and the imaginaries it produces.
www.theresabuechner.de
