Used the app "Sora."
By welcomebach6 | Published | Fansmade-stuff
I just adjusted the text prompt based on the uploaded image.
Q: How did it generate the exact same picture?
A: It’s because models like Sora are trained on massive datasets scraped from the open internet—including Google Images, public websites, and media repositories. When you enter a prompt that closely matches common metadata or visual patterns from those sources, the model isn't pulling the image directly, but it’s reconstructing it from learned distributions. Basically, it's not copying—it's regenerating from statistical memory built during training.