Generative adversarial networks (GANs) are capable of far more than generating deepfake videos.

Researchers are partnering with artists and musicians to generate entirely new forms of creative expression. From synthesizing African tribal masks to building fantastical, fictional galaxies, A.I. is being used to explore new ideas.

Last year, Nvidia launched GauGAN (named after post-Impressionist painter Paul Gauguin), a generative adversarial A.I. system that lets users create lifelike landscape images that never existed.

The National Institute of Informatics in Tokyo built an A.I. lyricist, while Amazon released its DeepComposer system, which composes music “automagically.” These A.I.s aren’t intended to replace artists, but rather to enhance their creative process.