Raw LLM Responses

No offense but you got so much wrong in this video. At 0:14 you say we have to understand how AI generates art. Yet you never explained how it actually does this. Nowhere do you talk about training keyword value pairs, nowhere do you talk about images generated from noise. (which is the key to AI generating images) You act as if its based on the work of artists alone. You say for AI to generate art there has to be an input in order for their to be an output. At 1:04 you say the images were collected without the permission of the property owner. My guess is that when singing up for these sites to display your work you never read the contract and you just clicked yes to it. You actually did consent for them to use your data how they want. Just because you didn't read the contract you agreed to doesn't mean they are in the wrong. The examples you are showing where the images look identical was done purposely by anti AI artists to make it look like the AI is creating copyrighted work. They do this by giving the AI a feeder image and telling it to use the majority of that image so very little is changed. Stable Diffusion is completely free to use. You act as if its just an evil corporation who did all of this to screw artists over. At around 3:44 you talk about how an artists name is used to create other work. This is only because that pushes the AI into that style of work. It's not actually copying anything from that artist. At around 6:40 you say AI takes data from the training and is able to make a perfect replica.....yet thats not how it works. Then you say a human looks at the work and uses it as inspiration and combines it with their other knowledge about the subjects.....but this is exactly what the AI is doing too.