You just reiterate what other anti-ML extremists have said like a sad little parrot. No, LLMs don’t just copy. They network information and associations and can output entirely new combinations of them. To do this, they make use of neural networks, which are computational concepts analogous to the way your brain works. If, according to you, LLMs just copy, then that’s all that you do as well.

Maybe because that is more dangerous than any other use?

Can you read?