I do not consider the use of AI Art gotten through OpenAI to be any different than using references, but, that being said, I would not just copy someone else's art as part of my final image. Inspired by, yes, and perhaps a very clumsy trace with a light box, but I think that is more honest than simply doing a copy paste.
For a design hack, I would say that a lot of generated images could be roughly traced to produce a lot of quality keyframes for animation.
CGI Animation and Comics are off-putting. I hate the overly perfect effect. I want an organic effect at a high level of production without breaking the bank with commissions of sketches.
I love hiring artists for additions to my creative works, but there is a financial bottleneck slowing down the work.
jthrash
My conflicting opinions on AI art in a nutshell. Although until/unless the technology gets better at anatomy like hands and arms, I personally would continue to prefer to use photo references and good ol’ Google Images to make 100% percent sure the image isn’t leading me astray with bad anatomy (and even then, you have to account for camera distortion and things like that in the photo).
Mindblade16
I am currently using many different sources to build multiple design iterations. Layer after layer. Refine and repeat. The sources are not copied, but components in experimental compilation. I am paranoid about copyright law, and AI sources regarding ethics. The percentage of the original content used in the recipe has an ethical limit where it is no longer a copy.