Images in the Cloud

Today’s image generation machine-learning models use artificial imagination techniques to learn to see shapes and textures where there is only random noise. In the same way we look at the clouds and find shapes like faces or animals. 

In this practical session, we’ll break down these techniques to learn how to intervene and makes spaces for experimentation. We’ll train an AI app to recognize faces on random textures such as clouds, marble or other textures that are packed with entropy. Once we have created this image generation engine with our photographs taken at the CCCB, we’ll put them to work with the free software SD-collage tool. The result will be a collective canvas that functions as an instant social network, a shared co-creation piece of digital art.