At A MAZE. Festival 2024 the installation titled Cre[AI]D was showcased, highlighting the intersection of traditional art and artificial intelligence. This installation continuously generates images using AI, trained on the drawings of artist Margarete Jahrmann.

Methodology

Cre[AI]D employs an AI system to produce its images. The first AI is dedicated to the visual output, ensuring that each image reflects the style and characteristics of the original drawings.

The process begins with the second AI, which generates prompts for the image creation. These prompts are derived from five randomly selected words taken from a book, introducing an element of randomness and speculative fiction into the generated art. Even the prompts that generate the images are now generated themselves.

Interactive Element

A significant feature of the Cre[AI]D installation is the large red buzzer present within the exhibit space. When pressed by a visitor, this buzzer initiates the image generation process. This interactive component allows visitors to engage directly with the installation, triggering the creation of new images with each press.

Continuous Image Generation

Cre[AI]D is designed to produce an infinite number of images, ensuring that the exhibit remains dynamic and ever-changing. The continuous generation of new art pieces means that each visitor experiences a unique set of visuals.

Playable 3D version

Explore our live online version directly in your browser, where you can generate images within a virtual representation of the “Postsparkasse”. Use the WSAD keys to navigate and your mouse to look around. Click on the one-armed bandit to generate a unique image.