Decided to unwind at the end of the day by playing around with Diffusion Bee, on of the hot new AI image generators. I’d spent the day planning out next years publishing schedule (war, famine and plague notwithstanding) and thought it might be fun to see if these things were up to the job of generating some background imagery.
Meh.
Not so much. Not Diffusion Bee anyway. A couple of images got halfway there. I already have cover art for Sleeper Agent, for instance, but I figured generating some new stuff with a few prompts like “spy, running down a wet street at night in the style of film noir” might spin up something cool.
These were the best of a mixed bag
I like the composition of the one on the left, but that guy isn’t running from anything. And his mate on the right looks like he’s ready to break into a show tune and tap dance.
This one, looked nice as a super basic mock up for a graphic novel cover, but a real artist would have nailed it.
And this one, Donald Trump as a GTA character? Bitch, please.
I’ll be setting aside some of the folding stuff to pay for real human artists for a while yet, I reckon.
i reckon it will come down to paying people who can spin just the right words to get the right image. It will become an art in itself
A whole new career as an "AI Whisperer" could be born
clicking the like button wont work, but just imagine the bullshit titles you could come up with. "i'm an A(i)RT facilitator"
oh - seems it does but i need to refresh the page. oops