I have drook to thank for first calling my attention to “midjourney,” a text-prompt image generator. How I wish I’d had access to this technology back when I was a one-man art department! In those days I had to bolt my work product together variously from purchased stock photos/artwork, from scanned images, or from pixels snagged from the internets, usage rights be damned (most of my work product was for in-house consumption—I was more careful with the public stuff). With midjourney I could have generated bespoke art.

Thus far—I’ve only got around to playing with it this month—my output has been uneven, and not a patch on the work I’ve seen from seasoned veterans. I need to learn what goes into a successful prompt. Particularly when i’ve done human faces, some of these visages could be mistaken for photographs and others stare at me right out of the Uncanny Valley.

I’ve played with inputting scenes and characters from my semi-moribund novel-in-progress, and have compiled some of the results here.

cordially,