But the problem presented in GP still exists right? Not everyone starts out with a good understanding of style and composition from the start. They need time to master those skills but also need a line of income to survive till then. If mediocre level work is all automated someone starting out might not get the time to reliably skill up
Learning composition and style is a different thing than developing technical skill though. You can learn the principles of composition/framing and get a good survey of art styles in months, compared to technical skills which frequently take years and years to develop. With this tech, you could start out as an enthusiast generating your own art, then get hired as an assistant of sorts to do low level prompt and input "exploration" for a head artist in a sort of apprenticeship.
The general public are not going to be the ones using these tools for the most part. Sure some people will try it for fun and some will become hobbyists but the ones actually producing salable work will be in the industry or attempting to get into it. So now instead of drawing a bunch of crude pictures and selling them for cheap they’ll be generating those images with an AI and selling them cheap.
