If everyone had their dinosaurs fixed, there eventually wouldn’t be any!
ISO POLITICAL CONTENT
I posted this in The BBQ Pit, but I’m posting it here to see if anyone can help me. I use sora chatGPT because thats the one that worked the first time for me. [NB: I broke the link in this post so it would post. I used the intact link on sora.]
I tried to get sora chatGPT to replace Verruca’s face with Trump’s shouting face, using these prompts: ‘https://i mgur. com/Yh37yfj Replace ONLY the girl’s face with Trump’s angry face.’
It didn’t use the screenshot at all, and just put Trump’s face on a little girl. If anyone can make their AI do what I tried to do, retaining the caption on the original meme, please let me know. I’d love to have it.
I began to suspect what bothered Sora was the text of the image instead of the subject. I replaced the memed image of Veruca with a clean one and Sora was happy with it. I then added text back to the result.
OK, I found the image from WW&TCF, saved it to imgur, opened the image in a new tab, and copied the URL. I put the link into sora, and copied and pasted your text after it. It returned an image of a yelling Trump in a tan overcoat and red tie, standing in front of protesters with signs reading ‘STOP THE WAR’ and ‘NO OPPAN AND PEACE’.
As you know, I am new to this sort of thing. How do I tell sora, ‘No, that’s completely wrong. Replace the face in the image with Trump’s face, and don’t change anything else.’?
To be clear, the text on my version was something I added using a photo editing app, not text from Sora. Just about any image editor you have should be able to add text. (I use this one but I think you have an iPhone?)
I don’t use many apps on my iPhone, because I don’t know how to use them. I have Photoshop Elements on my Mac. I don’t know how to use that either, but I can add captions with it.
Maybe others have had better luck, but this is the first time I’ve sort of gotten actual vampire fangs. Could be more symmetrical though. It’s adventure time!
I was talking to ChatGPT about making a sword from the blood of one’s enemies per a meme, which led to it producing the first image for me, then I gave it a shot on my own getting increasingly abstract.
A Facebook AI group recently gave a challenge to create an image of someone licking their elbow. Obviously an anatomical impossibly for a normal human, But I tried for an elbow licker using a Rokurokubi. She would have no problem licking her elbow, but the ChatGPT renderer just doesn’t seem to understand what an elbow is.
This is two prompts in one: just change a single letter.
Summary
Emo girl sitting on park bench during extremely heavy rain. She’s trying to keep dry under a wet newspaper that she’s gripping with both hands. She has a distant, melancholic expression. There are misty trees and blurred lights in the background. Candid iPhone 15 photo with shallow dof and forced perspective.
