In other words, “Art by my own hand, on canvas/paper or Photoshop, versus Artificial Intelligence generated.”
From day one, watching SG-1 from the start, I’ve made digital wallpapers of characters and episodes. It’s hard when the studio doesn’t release photos for fans and I have to depend on screenshots. These days, not a problem (for other shows), but in 1997 to about 2000, the quality of the film they used was horribly grainy. We’ve all heard about builders whose structures collapse because they cut corners and used cheap shit. Well, the producers of SG-1 cut corners in their film production. It was an insult to fans, in my not so humble opinion (that needed writing out, not thrown in as the oft-used acronym.).
Point is, making fan art was a chore when the source material was cheaply made. But as time’s gone on, it’s improved. The old episodes haven’t because they’re not deemed worthy of the cost it would take to launch a remaster. They did it for the premiere pilot episode, but not the others. Damn shame.
I still make art, but I’ve switched to AI “text to image” or “image to image.” Most people find this hurts traditional artists. Maybe it does. But when carpal tunnel killed off my painting/drawing skills, I tried to compensate by manipulating images. That never ended well because the lighting in each image was at odds with the other images, and sizing was difficult. Creating erotic art was hideous, and I’ve never been able to Unsee that in other people’s art, however tasteful or hot. And even so, after hours on the computer, the carpal tunnel isn’t liking that either.
But. I can make art images now. Beautiful stuff, if just a tad pixelated because it ain’t vector–and an oil painting or watercolor can’t be vectorized. My problem is getting Midjourney or LeonardoAI or some other program to give me exactly what I asked for. I have to literally draw up a diagram and hope the AI gets what I mean. Most of the time, so far, it doesn’t for Midjourney. For LeonardoAI, it does, but it’s tricky. The AI art bot tech has a LONG way to go.
Of course, it’s all dependent on the prompt you write. If the AI can’t figure out your prompt the way you want it to, you end up spending hours trying to craft the perfect prompt and it’s Vastly Annoying. And it costs $$ because all AI programs charge by the image. I don’t think that’s fair, but they base this system on the GPU server time.
Somewhere out in the desert or in a warehouse district is a building with a shit-ton of servers that are **Always** being increased, upgraded, and expanded, and they pay for it by our subscriptions to their service. I pity those people so I never complain. Coding and re-coding and trawling through all those lines to locate errors isn’t easy. But, as the over-used saying goes, “It is what it is.”
So I’m trying to do some art the new way and incorporating old images from the TV show. It’s not up to par. I make pretty wallpapers, but as I said in another post, they’re BORING. To make better art, I need a new PC. And that’s gonna take a while. Sigh.
