I've been impressed with the speed at which Stable Diffusion's release has spawned a community and economy around AI-generated art.
Browsing for a few minutes inspires me to exactly the opinion Simon outlines in Stable Diffusion is a really big deal. The ability to generate art with a text prompt (and in some cases a seed drawing) makes high quality art much more available than ever before.
You can play with stable diffusion yourself, free, over at Hugging Face. There's nothing about this technology that is limited to images. We have AI's that generate text, generate software, and generate images. This same technique can be extended to video and other, more complex, artistic designs.
Regardless of the quality of the output, all of these AIs pose a very complex question regarding copyright. It will be very interesting to watch the law catch up with technology in this particular instance, I think.