You might not have heard, but there's a new AI out there called DALL-E from Open AI that will take any query you give it, like "space ship landing in Monet water lillies pond" and create an image.
As an AI researcher myself (though not on that project), I can tell you it is indeed pretty amazing. The results still need curating (that, is not all the resulting images are awesome), but there's a little randomness in the system so you if you generate a dozen images for the same query, often one is pretty great.
Switching to my hat as Art Director here at Artifact Puzzles, we had to ask, "Can AI make great jigsaw puzzles?"
Our co-owner Jim Muller investigated, and you can read his report and see DALL-E's AI images on the Didero substack blog.
Permission to use DALL-E is by invite only at this point, and not for commercial use, so we won't have any AI generated image puzzles quite yet.
The obvious ominous question is "would this replace real artists"?
My answer is: No, but it has the potential to change their workflow, much like any other awesome tool. Like the emergence of calculators, which meant some computing jobs were lost, but opened up new jobs and possibilities.
I think what we'd see with DALL-E technology is artists using it as a tool like Jim did - envisioning images, then crafting queries, editing with the AI, and also editing with more traditional (ha!) tools like Photoshop. I'd hope that would open up art to more people and more creativity, and a more visual and fun world, much like we've gotten with everyone having a camera phone in their pocket.