Raw LLM Responses

I find it hilarious that OpenAI is upset at DeepSeek for 'stealing their work', given how AI models gather their data to begin with. The hypocrisy is hilarious. Also, regarding the Tesla 'self-driving' thing, when a lawsuit was made against them over the tech not actually being viable for quite some time yet, the courts ruled it was 'puffery', not fraud. For something to legally be puffery, it has to be 'so absurd and exaggerated that no reasonable person could possibly believe it to be literally true'. A BS ruling IMO, but that's the courts for you. I'll admit I'm not completely against AI in limited circumstances, primarily for personal non-public non-profit use. Making a visual pun or something that you wanted to share, and it going no further than a Discord DM. I'd personally say the absolute limit of what I'd consider 'acceptable' is for something like a fanfiction doing it for illustrative purposes, though I am biased in that having a visual approximation of something helps in a way a description just doesn't. Also, while I wouldn't exactly call it art, I do think there's a bit of a skill in prompt engineering; that is, basically wrangling a generator to give the output you want without overloading it. Seeing a consistent sequence is something that does mildly impress me. Still, even that is like one step below finding a sequence of photos and captioning them with a story. Honestly, not sure where I"m going with this, or even if it'll be read. Ultimately, I'm not completely for or against it. My strongest stance on the subject is probably that the models should not be allowed to freely scrape data off the internet without some kind of appropriate recompense. To the creators, I mean, not the ones hosting the works.