He allegedly used Stable Diffusion, a text-to-image generative AI model, to create “thousands of realistic images of prepubescent minors,” prosecutors said.
How could the AI know what that looks like, unless it was specifically fed with images of pregnant Voldemort? There’s no way it’s just satisfying multiple independent prompts, with a human being choosing the ones that look plausible and ignoring the ones that are anatomical nonsense.
Here is an image of Voldemort while pregnant.
How could the AI know what that looks like, unless it was specifically fed with images of pregnant Voldemort? There’s no way it’s just satisfying multiple independent prompts, with a human being choosing the ones that look plausible and ignoring the ones that are anatomical nonsense.