Like lots of other people, I've been playing around with AI generated art (text to image). In my case, I've been using Stable Diffusion UI. Stable Diffusion UI is an easy to install variant of Stable Diffusion that also includes a graphical user interface. I was able to get it running on both Windows and Linux without too much effort. While the included GUI is convenient, the easy install was what appealed to me the most. Some versions can be a pain to install and setup.
To get reasonable performance you need a decent GPU with lots of VRAM. However, a GPU is not a requirement to use Stable Diffusion UI as long as you are not in a hurry. If you don't have a GPU and use only the CPU, each image will take much longer to produce. It can take thirty minutes or more depending on your hardware and the settings you use. However, this is open source software so you can generate as many images as you want without having to worry about fees or limitations. The end result is that even with limited hardware, you can still produce unlimited images as long as you are willing to wait for them.
In yet another steampunk mashup, I was trying to generate an image of a steampunk witch when I generated this image. I'm including it here not because it turned out like I wanted. It's just the best example I have so far of how disturbing some of the images that Stable Diffusion generates can be. She looks more like a victim of a steampunk with than one herself. Or maybe that's just how she wants to look.