A.I. MATERIALITY
CAST GLASS, SLIP CAST CERAMICS
This project explores the materiality and physicality of GenAI models and the forms that they generate. To create the images that the forms are based on, I built a dataset of images of classical sculptures and trained a StyleGan model to create new ones. These images were then used with different image to 3D models to see what materiality they produced.
The first itteration of artfects were created with an appropriated ML model designed to generate human forms. The model would create beautiful glitchy messes when asked to create certain forms. This worked well with the glass casting process as it allows for unlimited undercuts and complex shapes.
The second used a much more recent, polished model, that was more suited to ceramics casting due to the simplicity and simitry of the forms. These artefacts are uch more easily replicatable but are less detailed and organic.
© Billy Dixon All Rights Reserved