Raw LLM Responses

as a law student who is currently taking copyright law and has recently covered this case in class (briefly), I just want to say this case not nearly as "delusional" or "dumb" as people think. In fact the case presents a very interesting legal question that I can't wait for an appellate court to review. Currently there are no appellate cases on the copyrightability of AI art, meaning there's no controlling law on this topic in any federal circuit. Too many people, including Charlie here, seem misinformed or ignorant of copyright law in the US, so please let me explain: The threshold for something to be copyrightable in the US is EXTREMELY low. You do not, and I repeat, DO NOT need to be an artist, let alone a "talented" one at that, to have your creation protected by copyright law. You could spend 2 seconds taking a picture with your phone--that's most likely copyrightable. You could also spend 20 seconds drawing an ugly ass cat that looks like a 5 year old drew it--that's also most likely copyrightable. Charlie's drawing that he showed at 2:38 , believe it or not, is (definitely) copyrightable. This dude in question, Jason Allen, spent a lot of time on this AI picture: inputting more than 600 prompts to carefully adjust the output of MidJourney, then using photoshop to remove flaws, to finally create the picture that he wanted. For those mocking him saying that AI did all the work or that it took no effort, that's clearly not the case here. As I said previously, a cat that you spend 20 seconds drawing is most likely copyrightable, and that cat probably required far less effort and "talent" to create. The point here is that effort does not always matter (I'll come back to this). The main issue in this case is not effort but "authorship." Current law demands that a HUMAN must be the author of a copyrightable work. So here's another question: a picture taken by a camera, or music created by a synthesizer are nonetheless considered work to be created by humans. What's the difference? Well, the Copyright Office explained that AI image generators (specifically MidJourney in this case) are way too random, and simply repeating the random process again and again until it finally produces something you are satisfied with does not show that the HUMAN exerted enough control over the process. With all this said, the copyright office did NOT say no AI art can ever be copyrighted. The office's main point is that to satisfy the "human authorship" requirement, the input/prompt must exert enough control over the output, which they don't believe Jason Allen did. I think this may be a scenario where effort DOES matter. If you just put in a sentence into the AI and the AI gives you a picture, that's most definitely not copyrightable (because it lacks human control and thus human authorship). However, what if you wrote an entire essay describing every detail of the drawing, and the AI faithfully recreates the image according to the description? I think then it is possible to argue that it is copyrightable. But we will have to see. My point is that this is not as "stupid" or "delusional" as people seem to think.