You didn't touch on the "who will have them" part. Enforcing copyright on training data won't keep image, music and story generators from being built. It will

You didn't touch on the "who will have them" part. Enforcing copyright on training data won't keep image, music and story generators from being built. It will only stop the open-source ones. OpenAI and Google can afford to license the datasets. It'll just add to their moat! I can respect wanting to ban image generators in their entirety. I can't respect allowing image generators to be monopolized by the three megacorps who signed a billion-dollar deal with Getty Images.