Original image done by Midjourney, the A/I text to image generator that everyone is talking about. And is winning prizes in small market contests across the country.

I gave Midjourney prompts like a single pillow on a bed in a home in Japan, following the lead of this nicely translated poem by Rexroth. Midjourney produced a photorealistic image that didn’t look the painterly way I wanted. I also adjusted the brightness and the saturation.

So, I then used a basic Photoshop filter called Cutout until it produced the slightly cubist edges that I wanted. Added quote and “Vole La” as Kelly Bundy would say.

Refinements would be to place the bed at a distance across the room, to emphazise the distance between the two subjects of the poem, perhaps make the bed a floor mat, and make everything more spare and lonely feeling.

Having said that, AI now makes it possible for any writer with no drawing skills to fully illustrate a book that they want to write. Or, a million other possibilities. For the graphic artist, I see it taking the role that templates or clip art does today, too much A/I and it reveals the hand of the artist, a bot. And since everyone is going to be dipping into AI, it is absolutely essential, as always, that the artist find and keep their own look. That’s because other as people’s work might start looking the same as more and more AI is used. I shouldn’t have that problem.