saturday’s thoughts on ai-generated art

A couple of weeks ago I saw several artists spewing vitriol on an acquaintance’s post because it was illustrated with an AI-generated image. As an artist, I can understand their frustration. But those artists-against-AI-art rants leave me shrugging and shaking my head.

To specify, I’m writing about situations when AI-generated art is used by folks like me: bloggers or social media posters who are looking for a quick way to illustrate an idea. We’re not trying to generate income directly from the images.

I have different feelings and opinions about situations where the AI-generated art is offered for sale. That belongs in another post, and I’m still forming my opinions there.

When I use a MidJourney image to illustrate a piece of writing, I’m not cutting into any artist’s slice of the pie.

First, I don’t have the budget to buy someone else’s art, even on Fiverr.

Second, this saves me a helluva lotta time.

No, I probably couldn’t produce something that looked like this AI art in less than 2 weeks. But if I didn’t have MidJourney, I would manage to create some kind of original art that would suffice.

I’d like to mention this image from a recent post.

I created it in Photoshop by layering images I prompted into MidJourney AI. Thank goodness all the key facial features are aligned in all three images. They were all generated from the same seed file; I just tweaked the prompts. It’s a happy coincidence that this girl looks very much like I did at this age. Go figure.

Now let’s look at a grief image that MidJourney AI created for me: “the heart-shaped void left behind.” I could not have created these images by myself. They express the loss, desolation, and bewilderment that I felt when my Mom died in November – in ways that I would not have had the nerve or the talent to convey.

And using only six words that I typed.

Sure, I understand that these images were generated as they were because someone uploaded a ton of images created by artists who probably labored long and hard over their work.

But there’s also some magic involved. A secret (and not well-understood) algorithm that excises slivers and grinds away pixels into whatever it (who?) finds sufficient to spurt out onto my screen.

And I’ve had enough nauseatingly botched results to conclude this: AI-prompt-writing is a craft unto itself.

I’m old enough to remember when digital art emerged as an acceptable medium. It offended many smock-wearing artists.

I’m old enough to remember feeling indignant that young architects could not hand-letter legibly on sketches or formal drawings.

I’m old enough to remember needing an electric eraser when the rapidograph pen slipped on the mylar drafting sheet. And the delicate touch it took to remove the errant ink without destroying the tooth on the film. And the astonishment of French students watching me plug that eraser into an outlet, while they witnessed an alternative to scraping the ink off with a razor blade. Every convenience comes at a cost, and ushers in feelings of loss.

So what?

So, I’m old, and I carry memories? Is that it?

Is that a valid reason to reject new tools, especially if they save me time and effort?

On the other hand… when I touch my paint-laden brush to canvas there is no indestructible digital evidence trail of what I was thinking, or of the emotions behind those strokes. The mysteries of motive and creation may be kept private.

Every time I type a prompt into my AI interface, I wonder where my words might resurface. And I acknowledge that I have no control over how they may be used by others who stumble across them in the future.

I can take that as a hint to practice self-restraint.

Or as a reminder that somewhere, in some form, gods are watching.