With the new thing being AI art generation, I couldn’t help but give it a shot. After spending a few days trying to get Stable Diffusion to work with my GPU, I finally was in luck. Seeing I have no real place to post some of these besides here and in a few smaller communities (remember I don’t do social media), here’s some examples.
First thing off, was realizing you have to order your words exactly in a specific order or unexpected results occur. I thought bacon on a skillet would be easy… WRONG!
Having bacon after the skillet caused it to be part of the campfire. I got a massive laugh from it. Fourth try got exactly what I was looking for.
I tried to feed some ideas from my book writing times into it and see what it gave me. If it worked great, I could possibly not bother with trying pay an artist commission for book covers and a few web banners in the near future. Not really what I was looking for but still funny.
AI art generators, both stable diffusion and craiyon seem to have issues with real faces. Not a really big deal. It handled anime girls quite well though and paintings as well.
Now for some crazy stuff, because why not.
If I was told just a few years ago you could have art generated by a computer program by just text, I’d be skeptical that it could occur that quickly. I do have some concerns that as this stuff becomes more prolific, there will be restrictions of some kind restricting people from entering specific words or using it in a certain fashion. Where I am okay with metadata being embedded in the images saying they were generated is okay, that’s about as far as I go. Sadly, it’s not the stage yet where I think I can use it in some manner for here yet unless I want everything to look like anime girls, not that I have a problem with that. I’ll be following it for a while though.
Note: I’ve laughed my ass off while generating as many images as I could. This has been quite enjoyable.