The arXiv paper "AI-Experiment" describes the creation of a physical artwork through a collaboration among an artist, a computer scientist, and generative AI tools. According to the abstract, a latent diffusion model was fine-tuned on photographs of the artist's existing artwork, and the project was "performed in synergy" between the human participants and the AI.
The abstract does not provide details on how the fine-tuned model was used in the final piece, nor does it describe the artwork itself. The truncated announcement cuts off after mentioning the source photographs, so the full methodology and outcomes are not yet available from the abstract alone.
What is clear is that the project frames AI as a close collaborator in a physical art-making process, rather than as a tool for producing standalone digital images. This aligns with a broader trend of artists integrating generative models into tangible, real-world works.