@thewebrecluse Having some experience with Large Language Model technology, I can best describe it as a lossy compression algorithm. There is no "understanding" or "intelligence" involved. It just returns what it was trained upon with some mixing and loss of fidelity. Some of the algorithms also incorporate entropy (randomness) to mix the output even more. Generative AI that was trained on people's copyrighted works is theft, plain and simple.