The Studio Ghibli Meme Gallery - Using OpenAI 4 to convert popular memes into weeb drawings

  • 🏰 The Fediverse is up. If you know, you know.
  • Want to keep track of this thread?
    Accounts can bookmark posts, watch threads for updates, and jump back to where you stopped reading.
    Create account
If anyone is looking for ideas, nobody has done king terry yet.

1743137283848.png 1743137309698.png
1743137421734.png
 
Are these all scrapped from elsewhere or are you guys making these using something. They look really cool and I really wanna make them (after I get a new computer)
 
Haloperidol said:
I’ll just admit it: I’m an idiot and if someone could tell me how the fuck to do this it would change my life.
edit—I know it’s Grok but it’ll only let me input descriptions, it’s aight I’ll figure it out *yawn*
Click to expand...
I discovered the horrible truth: gotta pay to win
2.png 3.png 6.png
8.png 7.png 9.png
 
Last edited:
im awake and i wanna get my $20 dollars worth so i am going to be doing some prompt requests. one thing i have observed is the degregation in image style. the way i do these is i start with the initial prompt and an image, after it finishes an image i put in another image and say "now do this one". but i find with subsequent images in the same chat instance the images start to change. for instance im ghibli-ing the 'jaks.
this was the inital prompt. looks good
10571428-c5a9-46bb-b0de-650c4f234d55.png

this one still looks good, no noticable difference in style
4d6d49ab-2639-434b-986e-c3689c71a5c7.png

this one still looks fine but with later hindsight the degregation has already kicked in
af63092d-9b19-4214-834e-822defa90cf5.png

from here it becomes noticable, the style no longer looks perfectly ghibli. its still there but its off, and the image is noticably beige
7630cf01-db8e-4861-9ff0-c3518a61acde.png

a witheredjak i did for a friend
51ca76cb-c6d4-4e05-b11e-f7a1bfe3d611.png
E7vD99wVoAEuERc.jpg

from here its completely fucked. the style is nothing like ghibli and it looks like its soaked in tea leaves
1b6d61c1-f8d7-4dc7-a919-2a16b84274ac.png

and after a quick reprompt of the origional it goes back to this
8002a41a-9199-4623-9c4c-c79b1f4009e6.png

so if you want to try this yourself the moral of the story is dont be lazy and just ask it to repeat itself, you gotta copy past the origional prompt every time
 
ive also just made another discovery. if you ask it to do multiple at once it mashes them all together
heres my prompt. just throwing a bunch of random images i have in there just to see if it can do mulitple at once
Screenshot 2025-03-28 170809.png

this was the result
image_2025-03-28_171055121.png

could you use this to do something? maybe. it seems random and up to the ai. theres probably something fun here
 
2f386d95-2e86-4fb9-8d17-ab7d26147798.jpg
bcd06f15-7b0c-4def-a4fa-98a47d33b466.jpgChatGPT Image Mar 28, 2025, 02_14_52 AM.jpgChatGPT Image Mar 28, 2025, 02_00_05 AM.jpgdf68a741-671e-43b2-8ae9-09d02691d6dc.jpgChatGPT Image Mar 28, 2025, 02_14_56 AM.jpgChatGPT Image Mar 28, 2025, 02_15_02 AM.jpg
Impressed on how it handled Eugenia's eyeshadow and Shawty's tattoos
Eugenia is too skelly to be Ghibli???
Bonus #1 Null voice enjoyer
ChatGPT Image Mar 28, 2025, 02_18_17 AM.jpg
 
You must log in or register to reply here.
Back
Top Bottom