Not necessarily? We have ai that can de-age people created without CSAM so it stands to reason that it’s possible. I still don’t think it should be accessible though. It still requires children and that’s uncomfortable enough for me I think
Something like stable diffusion can absolutely make things it’s never been trained on. For example if puppies are in its training set, and poodles are in its training set, but not poodle puppies. It can generate a poodle puppy.
Yes but fron what i know every image AI is technically able to make “cp” you don’t need cp to get a ai to make it. The training data is definitely actual children but not necessarily from CSAM
Yes, but, well, the neural networks need to be trained on something. The smoke is better than the fire… But you can’t have one without the other
Not necessarily? We have ai that can de-age people created without CSAM so it stands to reason that it’s possible. I still don’t think it should be accessible though. It still requires children and that’s uncomfortable enough for me I think
That’s not true and that’s not how it works.
Something like stable diffusion can absolutely make things it’s never been trained on. For example if puppies are in its training set, and poodles are in its training set, but not poodle puppies. It can generate a poodle puppy.
Yes but fron what i know every image AI is technically able to make “cp” you don’t need cp to get a ai to make it. The training data is definitely actual children but not necessarily from CSAM