Raw LLM Responses

AI art is not copyrightable, and it shouldn't be. As said in the video, anyone can do it. The barrier for entry is a grasp of the English language and a cell phone, laptop, or PC with an internet connection. The only part of AI image generation that a user should every possibly be able to copyright would, I think, be the prompt/negative prompt, and specific settings to include the combination of model/checkpoint and LoRAs used. That is, as long as the prompt isn't something that uses copyrighted material like "Luke Skywalker" and is something is obviously unique. The prompt is the only "man-made" part of the equation; as easy as it is to use the various programs and to make nice looking images, there is such a thing as a "skilled prompter" which basically comes with experience as you learn the technical limitations, how to work around them, and the advanced techniques of prompting like additional plugins for regional prompting, utilizing commands to end a chunk and start a new one, and understanding the best way to structure a prompt for the LLM to have the best adherence to your input. Can that take a lot of work? Of course. Doesn't mean it should be copyrightable. A prompt and the settings are the only thing the user actually creates. And I say all this as someone who makes a living working for an image generation and chat bot company (no, you can't bang the bots. we're not _that_ kind of company). And Midjourney sucks balls. (So does DALL-E 3)