Raw LLM Responses

The point people are trying to make is that there are no new truly original ideas out there. Everything is based on something else that comes before it, which then either is a tweaked version of that, or a mix of other ideas combined with it. Though it usually still contains some aspect of originality that your bring in yourself. The difference between this and AI though, is that AI needs to be trained directly on other people's work to generate anything that isn't just noise. With LLMs you train it to mimic something else, and encode parts of other's work into it so that it can be copied verbatim. With image generating models you train it to identify images, and then use that to get it to modify images (random noise at the start) to look more similar to something else. AI is trying to copy something else, humans use other work as inspiration to make something new. Also, I'm not a luddite either. I literally work as a programmer... sometimes even dealing with AI related things. (Though I'm skeptical AI and it's capabilities, as people claim that it can do way more than it actually can, and ignore all the technical and ethical issues/limitations that it has. It's mostly hype and lies, rather than it being good. Sure it's a cool tech demo, but beyond that it isn't much. Also, people tend to heavily misunderstand AI. It isn't something that magically can solve all issues. They have to be trained to solve very specific problems, and are used when trying to manually write out code to solve it would be too difficult.)