But then a human comes and selects one from a hundred images. Not to mention the human had to write the prompt, sometimes a very long and explicit one. I'd say that's enough human involvement to be able to use the image as his own.
I'd say that the largest amount of human involvement was on the original works of art that these models ingurgitated. What came after is only statistical interpolation.
It doesn't matter, even choosing a picture is a creative act that imparts copyright. Maybe if you hook up GPT-3 to generate prompts for DALL-E and then filter the images with another automated method, then the work is completely AI generated and deserves no copyright.
As for the authors of the training data, they still enjoy the copyrights over the works they created. Images generated by DALL-E only copy the style, something humans are also allowed to do. Anyway, you can't put the toothpaste back in the tube, now we have to live with image and text models. Next year we'll be shocked at video models that can play games, accomplish meaningful office work and also generate long videos.
But then who does the copyright belong to? Does it belong to every artist who's input was used for training? Was it the user who set up the conditions and chose the image? Or was it OpenAI who developed the model?
Given the court ruling I linked earlier OpenAI almost definitely does not have ownership over the output of the AI, which is what they claim.
