Too many cooks...
Category Story / Comics
Species Hyena
Size 800 x 999px
File Size 316.2 kB
Some of the background hyenas are looking pretty Piccaso'd, lol. Was AI used to generate the background yeens? Sorta feels like it. Some of their hands are messed up too. I zoomed in to look at all of the details and that's when I noticed it. Misplaced eyes and muzzles, hand-blobs, and so on.
I'm lousy at drawing architecture--I can't draw a straight line even with rulers and computer assist--so I put a prompt through an image generator to see if it would make me a usable kitchen background. Most of them were not even remotely usable, but I liked the elevated angle and general busy-ness of this image. That it looked kind of Escher-esque was a bit of a problem, but it's really the least creepy-looking of the lot.
I started to use my tablet to draw over it, got as far as the Czhokulas, and decided, "Screw that, it's too much like work. " It was no different than any other background I might have plucked off the Internet, except that I made it with a prompt in an image generator.
I had to do a lot of repair work, like covering up a cook that was sprouting out of the countertop, adding missing legs to tables, fixing heads that were on backwards, and so forth. I didn't get them all, obviously.
"AI" is villified, but I see it as a toy and a tool. It's also grossly misnamed. There's no intelligence in it at all, otherwise hands would have the correct number of fingers and the laws of physics would be applied rather less randomly.
I started to use my tablet to draw over it, got as far as the Czhokulas, and decided, "Screw that, it's too much like work. " It was no different than any other background I might have plucked off the Internet, except that I made it with a prompt in an image generator.
I had to do a lot of repair work, like covering up a cook that was sprouting out of the countertop, adding missing legs to tables, fixing heads that were on backwards, and so forth. I didn't get them all, obviously.
"AI" is villified, but I see it as a toy and a tool. It's also grossly misnamed. There's no intelligence in it at all, otherwise hands would have the correct number of fingers and the laws of physics would be applied rather less randomly.
It was less a judgement and more a curiosity, I assure you! I only consider generative AI an artistic no-no when it's being used to produce art for a paying customer. If you want to use it for your own projects, be they personal or otherwise, that's entirely your prerogative, and I'm not about to give someone flak for it. I more thought it was funny that some of the yeens looked like eldritch abominations, lol.
That's the hallmark of computer-generated images (I refuse to call it "artificial intelligence." That gives it too much credit. It's just a cleverly written program.) They may look OK at first, but the more you look at it, the more nightmare-fuel you find in the background or in the inaccuracies of the figures. Hands are a particular give-away. One image I got looked terrific--until I noticed the six or seven fingers on one hand, and the other hand on backwards...
If you are trolling for inspiration then prompting a generator might help. I got some ideas for costume designs I would never have thought of myself by plugging in "female spotted hyenas in Greaco-Roman armor."
If you are after something very specific, forget it. You can prompt your brains out and wind up with forty failed pictures. If on the other (multi-fingered) hand you have just a nebulous idea, you might nail it in one prompt.
And since the images are unrepeatable, it's impossible to use them for any sort of comic strip. I think I might have even frustrated the program, as the results began looking less like hyenas and more like cheetahs each time I plugged the prompt through the system.
It can be fun to play with. I do have my concerns about the "AI" that is being used for things like research--without the ability to prove the results, which are accepted because "we used Artificial Intelligence!"--and especially the "deepfake" stuff.
If you are trolling for inspiration then prompting a generator might help. I got some ideas for costume designs I would never have thought of myself by plugging in "female spotted hyenas in Greaco-Roman armor."
If you are after something very specific, forget it. You can prompt your brains out and wind up with forty failed pictures. If on the other (multi-fingered) hand you have just a nebulous idea, you might nail it in one prompt.
And since the images are unrepeatable, it's impossible to use them for any sort of comic strip. I think I might have even frustrated the program, as the results began looking less like hyenas and more like cheetahs each time I plugged the prompt through the system.
It can be fun to play with. I do have my concerns about the "AI" that is being used for things like research--without the ability to prove the results, which are accepted because "we used Artificial Intelligence!"--and especially the "deepfake" stuff.
Yes, I too have used it for inspiration, and help with stories. At tmes I need to 'see' what I'm seeing in my mind, before I can actually write something. So it's a tool, nothing more.
(giggles) and once I tried to have it draw a deer doe. Worked well one nearly all, until I noticed that one doe had 4 ears. Whoops!
(giggles) and once I tried to have it draw a deer doe. Worked well one nearly all, until I noticed that one doe had 4 ears. Whoops!
Looking for pictures on Google Image Search works fine--for stuff in the real world. But if you want to see something like a steampunk crab-legged Winnebago on a tropical beach, you kinda need an image generator. It may not be perfect, but it helps with getting the perspective right.
Of course, when it puts actual crab legs on the Winnebago, instead of mechanical ones...
Of course, when it puts actual crab legs on the Winnebago, instead of mechanical ones...
FA+

Comments