Raw LLM Responses

Very nice video. Regarding the issue concerning the illegal use of scrapped data for commercial gains, I completely agree. As per the point of "devised work" as generated by AI as opposed to the original artistic human work, one could ask the following: when you create a new work of art, and you sell it or profit from it, are you compensating each and every artist whose work served as an inspiration for your own work? I am not an artist, I come from a social sciences background, but to me, art is a subjective apprehension of what is in front of you. Just because an AI generation tools used millions of different refferences to diffuse and denoise existing pieces of art in order to create something that is corresponding to the user's prompt, does not automatically make it a copy of millions of other already existings pieces of art. If an artist were a tabula rasa and creates something, I could accept the argument that this is a trully original work. But in the real word, artist take inspiration from others, read books, on art view other pieces of arts, study art, etc. How exactly is that different when compared with an AI taking inspiration from other pieces of art? I think the correct answer to this questions depends from person to person, and there is no universal answer to this. After all, we humans process information, AI processes information. Artificial or not, there is neural activity to be observed, in some cases already far more complex than what we humans do on an average basis. So, I object that notion (the AI art is no art but devised work), as presented in your video as well as in the claim action against StableDiffusion, Midjourney and DevianArt. Again, as per the training data issue, I fully agree. This has been an issue for the past 10-15 years when certain classes of people, namely, authors, started objecting Google scanning their books without any consent. Nobody paid any attention really, and Google still does it. Then taxidrivers started objecting Uber and Lyft pretending not to be taxis. Nobody cared and welcomed this "distruption" with open arms. Now artist are objecting this "creative disruption". And guess what... nobody cares :) All I can say to you is: welcome to the club :D, and good luck in fighting a fight that is already lost. What people need to understand is quite simply. The only way to ensure certain rights are followed is to regulate it. In this case, you have to have a regulation in place that ensures that each prompted image contains a reference statying cleary that this is an imagine created by this-and-this engine. Each engine providers needs to have a clear reference regarding the data base it is utilizing. Each data base utilized for image generation (or any other type of AI synthesis) needs to be obtained in a legal manner (ie, respecting authors' rights, etc.). In order to enforce this, you need to have a legal framework in place. Law will not regulated whether something is a piece of art or not, but it will regulate the provision of services. Sorry for a long reply, but I found your video very good foundation for an extended debate over this issue.