An interactive experience on how generative adversarial networks (GANs) can be used to produce artistic media. During the experience, visitors act as a curator for an art gallery that showcases work by both human artists and generative algorithms. As the curator, it's they both distinguish human-created from GAN-generated artworks and select up to three pieces (of any kind) for their personal collection. As more feedback is given by the curator, the generative algorithm improves its output and produce higher fidelity, if sometimes surprising, results.

Visual outputs are from two pre-trained models used to visualize GAN output across epochs: a custom CNN GAN model trained on a self-curated dataset from the WikiArt Project, and a pre-trained StyleGAN2 model developed by Baylies et al.