@stuckdeep
Use A. I. to generate stuck pictures
-
@sirstuck said in Use A. I. to generate stuck pictures:
@spintire said in Use A. I. to generate stuck pictures:
there is Bing AI too: https://www.bing.com/images/create?FORM=GENILP
That's a different interface to the same processor
But Bing gives us four results instead of just one and, in my experience, is less fussy about "terms of use" etc.
It would help if we knew exactly what request generated some of the pictures here. I'm trying to get a left rear view of the driver looking backwards out the open door but nothing has worked so far. If you've obtained something sensible, please let us know how!
-
@Florian I’ve been having the same issue. My other problem is it fills the cab of the truck with mud or snow.
I’ve found that it doesn’t like complex sentence structure, so if you say “a guy with his truck stuck in the mud” it takes the subject of the sentence, “guy”, and will show him stuck in the mud. I have the most luck when I think back to sentence diagraming in school and structure the sentences around placing the subject in the right spot. -
@gabe i will second this. commas separating features definitely help the AI discern what you are looking for, especially if its extremely specific. i've been having some success with extremely specific features of the person in question in a scene
i can't wait for a video generator for the AI. i'll have a banger of a fun time with that
-
I'm okay with a majority of the AI pics! Won't replace the real thing...BUT....some pretty sexy pics from what I've seen!
-
-
-
If you find phrasing that gives good results, then please share it here.
I copy the text I use to generate an image, and then paste that into the downloaded file's name. Then I can recall what worked particularly well. For example:
"the Wienermobile, stuck deep, in thick red clay mud, strong young men pushing the rear, on a country road, in Alabama, in 1969, rear side view"
-
@gabe I know woman stuck is almost taboo here, but I don't intend on posting them here but, I've been having the same issues, no matter how it's phrased or described, or even just using commas to separate the various criteria... I get the terms of use violation warnings.
I've been able to get some results but not the desired results or it's as some say, the girl is like part of the car or something stupid like that.
I personally think it's because the terms of use is written in a sexist manner and describing a woman in a stuck situation is objectifying her and it triggers the warning. Just my take on it
Might try the Bing AI generator -
@HardStuck I really LOVE this one... Imagine him smoking a few cigarettes first, then trying again to get out without help for 15mins or so, then starting smoking again before trying to get out again... Yummy!!!!!
-
@stuckfanjay said in Use A. I. to generate stuck pictures:
I've been able to get some results but not the desired results
Right now it is a guessing game as to how descriptions work. I've observed a few things:
- The exact same description re-submitted will produce different responses. There is some random factor here.
- Sometimes the process will take a while as if it is generating an image, then when it is almost through it will stop with a violation. What may be happening is that the picture has been finished and then a post-process content checker found something objectionable. Per my first observation, sometimes re-submitting gets a violation when there was none the first time.
I believe that the AI companies are providing these free services in order to exercise and improve their products. In other words, we are the alpha- and beta-testers.
