3 ms·
Not really, its trained on the previous object to recreate said object given outside changes... it very much is the same thing as warhol doing it with his brai
by jayzalowitz 3y ago
Not really, its trained on the previous object to recreate said object given outside changes...
it very much is the same thing as warhol doing it with his brain based NN recreating it with "do a prince portrait, but orange"
- davorak 3y ago> Not really, its trained on the previous object to recreate said object given outside changes... All generative AI packages, that I know of, are not going to use material only from one artist even when generating something in that artists style. In general it is more than for example tracing a photo and coloring it differently. Just like artists can copy the style and techniques of another artist(often several) and generate something original, current generative AI, often, seem to do something similar. I am not saying that the current generative AI can not produce copy righted work, but it is not obvious to me that if the original dataset is all copyrighted that all output retain those copyrights, though it seems likely select output could.
- usrbinbash 3y ago> to recreate said object given outside changes... Unless we're talking about something other than diffusion models, the new image is generated from initial random noise. It also doesn't "recreate" parts of it's training data, unless the model was overfitted to a particular work, or the prompt is deliberately designed to do so, or the user is running it in img2img mode against a preexisting work.