No, I Don't Use Generative AI for Images (or Words)

The featured image in the previous post (of Jesus communing with a lizard) was obviously generated by AI. When generative image-making software was first unleashed on the world, I was fascinated by it, specifically because of how bad it was. At the time, I didn't know exactly how thoroughly bad it was in all ways. I just thought it was so atrociously bad at interpreting instructions (and generating results), that the resulting images were fascinating in a surrealistic way.

I never tried creating anything serious with it, but I had a lot of fun generating bizarre images. The first 'project' I used it for was to illustrate a zine of my haiQu. It'll be faster to show you what I like about it, than to explain.

Background: haiQu is a little creation of mine that is a form of poetry flarf. Its key feature is the use of phrases that I would appropriate from my Twitter feed. Here's an example:

ant Jesus
tarpits and labyrinths
chum loves bait

When I was ready to publish my first haiQu zine, I thought it would be fun to use each poem as a prompt, and use the resulting image as an illustration. I count the project as successful. Here are a few of the pages:

There is one thing about this software that is impressive, despite the horrible nature of it: how fast it evolved. The Jesus/Lizard image was created just a few months after these, and while I liked the concept of the image, I found the increasingly pseudo-realistic nature of the images to be very creepy.

And I subsequently learned how these parasitic programs were part of the tech-bro fascist movement to replace human creativity with machine generated slop.

I am now firmly in the "I Hate AI" camp, and do not use it in any way. It was a blip. But I'm still fond of my Jesus/Lizard image.

Here's the full haiQu zine: