Raw LLM Responses

1:12 "the product matters more than the person creating the art." Brandon Sanderson would say, in terms of art, there's actually an argument for the artist themselves being the piece of art or even the product. Asmongold would say, "do you really think that people care about how it is made? Or do you think that they just care about the result. I think they just care about the result." These are not contradictions. Both are correct. Sanderson is 100% correct in limited "layers" of practicality and Asmongold is generally correct. There are multiple layers to the problem since everyone is different and Sanderson is only correct in some layers, however I think that Asmongold is most correct since most people exist in the layers where they don't care about big "A" Art (as in, the Arts) and only care about the end product. It may be difficult to avoid this bias towards art being important since anyone who cares about art (your entire audience) exists on different layers than your average Walmart American. I say this as someone whose primary income comes from the Arts, and as someone who is at risk of being replaced by AI if it is good enough. (Yes, I realize that AI Art cannot exist except on the shoulders of real artists and, in practice, data theft. I do not like the AI industry in its current state.) 1:18 "AI will be better than artists and will replace artists." Anyone who says this, including but not limited to Asmongold, is speculating. They might be wrong and they might be right. As an artist, it is best to assume the worst, but as someone who has also tried getting into ML and even trained their own SLM, it's pretty clear what the limitations to our best architecture ideas are, and figuring a way around those limitations is incredibly elusive. AI generation looks, sounds, and reads really nicely at first glance, but it gets worse the longer you experience it, and it is really cheapened when you realize there isn't any humanity in it, so Brandon Sanderson was definitely 100% right, but so many people will never realize it. 8:00 you have too much faith in humanity if you think everyone is as thoughtful as you, and that would be your only mistake, because you are right. Unfortunately, the person you are trying to contradict is also right and probably more generally correct. 11:10 I like that you actually call Asmongold wrong, though you feed his ego by talking around it saying "for that to be correct..." because his rhetoric is that "they say x and y about me but they never say I'm wrong." As far as if Asmongold is actually wrong or not, my intuition tells me that he is wrong. I've heard a ton of anecdotes of people saying that they are genuinely entertained by AI books, AI images, AI music, etc. but the fact that AI seems to be in a bubble would seem to say that people don't actually value it and are entertained by it because it is virtually free, though scammers are lying and selling AI Art as if it were made by a human, so someone is profiting off of AI, though the fact that they have to lie (sometimes by omission and sometimes they explicitly lie and say "no AI") is telling. 17:12 To be fair to Asmongold, the full context of his streams includes a theme that "it just keeps getting better." So his audience will deny your take since you didn't address that. I sense that you are a thoughtful person, but, to someone with his audience's context, your response comes off as unnecessarily aggro and stupid. I personally think that the quality of the AI Art Product is presently plateauing, but who knows what the future holds. 21:30 I agree with your premise—I kinda went over this already, but you and your audience (including me) exist in different layers of society than the majority of humanity. So you are also mostly wrong about your proof. How to help clean up this issue: 1. Push for legislation to require licenses for training data used in the optimization (training/learning/backpropagation/whatever you want to call it) processes for machine learning architectures. As a bonus you can only require this for companies who sell the generative output of inference (inference: the forward propagation of the model) since the counter argument without this bonus is "Ah, so you don't want us to cure cancer and do all of the other incredible stuff that AI has already achieved to improve the lives of humans in the field of medicine.' 2. Show people how cool artists behind good art are! Do not tell them that "actually humans like other humans making stuff" because that is easy to dismiss. Do show them the process and what is skillful and cool about it, and let them draw their own conclusions. 3. Show that AI "artists" constantly lie and deceive and scam people who want to pay for art. Show that they aren't transparent because even THEY know that people prefer people. Be fair to the AI bros because some "artists" are 100% transparent and people still want to pay them, but just don't downplay that and show how many more scammers there are than transparent AI "artists".