Raw LLM Responses

@Darth_Melek What are you on? ChatGBT works more like a brain compared to any other AI, it’s created thousands of layers of understanding and connections between words and meanings and context and associations. It guesses what the next appropriate answer would be when it receives an input (which is, in this case, a message) and uses that guess to generate a reply. It literally was fed every single piece of human literature accessible to create the system it currently uses. The guy you’re talking about isn’t an AI expert and got different results compared to what other people might get. It’s constantly learning and improving, mistakes can be made, and the expectation that it should be perfect is ludicrous. I’m also fairly certain I’ve seen the video you’re talking about- it’s in the context of a screen replacement, right? I never said it generates images based on art it finds on the internet, I’m pretty sure I specified that it’s fed a dataset to train on (which are images publicly found on the internet). It doesn’t use that art as a basis for how it generates images, it uses that art to learn how certain shapes should look, what colors should go where, how specific angles might work, etc., but it never copies art. That means that, when prompted, it already knows how or what something should look like when it’s made correctly and uses that knowledge to assemble a piece that is as close to the given prompt as possible. That is literally art, by definition, since it’s engaging in a creative activity. Artwork is made, and it is art itself. Please don’t twist what I say if you don’t understand what’s being said.