Dancing through the latent space

A GAN(Generative Adversarial Network - an AI technique) animation trained on a dataset of sacred spaces and corridors. The dataset images come from Flickr, same.energy and some of my own, the AI creates with it a statistical average representation of temple interiors - an 'average' within my own taste, of course. I added an Isadora Duncan video, a modernist icon of spontaneity and delight with one's own body. Could the machine simulations become the background of our deeper desires and joys? I like to think of these AI hallucinations as visualizations of our collective intelligence, from all the people who created the original images to the long lineage of people working on mathematical representations/recreations of reality.

Special thanks to @thelemuet for allowing me to use his generative colors that appear for a few seconds!