The Artificial Literature Laboratory explores large language models (LLMs) not as services, assistants, or all-knowing agents, but as tools for creative writing. A specially developed interface enables interaction with the LLM through the physical hardware that makes the generative process possible in the first place — inviting visitors to collaboratively write a short text.

Instead of the optimization-driven logic that dominates current applications of AI, this installation centers on an experimental interaction process rooted in the traditions of electronic literature, constrained writing, and chance-based poetics. Collaborative writing here is not a simple prompt-response-output routine. It offers visitors the opportunity to explore working with LLMs as a creative tool in a non-commercial way — while critically reflecting on their underlying logic of optimization.

The installation engages with the generative logic of LLMs through the act of human writing — and vice versa — inviting reflection on what it truly means to create something new.

Credits: 

Artists: Eugénie Desmedt, Paul Kloker, Marton Zalka, Christine Haupt 

Production Assistance: Jonas Martschin

With thanks to Martina Mara and the Robopsychology Lab, Paul Janisch and Naoki Matsuyama

Funded by the State of Upper Austria