Midjourney is a machine learning system that can use keywords and phrases that you provide, find images that match, from a gigantic dataset, and stitch them together into some kind of coherent picture, using additional parameters to control the result. It does not understand language, and it doesn't have internal 3d models for the objects in the pictures. Instead it tries its best to synthesize a picture based on the request you give it, and the pictures and metadata it has been trained on. It's art for the non-artist, and it's both immensely useful and (potentially) deeply problematic for trained visual artists everywhere. But it's here and it's fun to play with.
Be the first to hear about deals, exciting new products and much more!