I don't think many people object to statements like "ChatGPT doesn't have world model". I'd guess that's wrong, but I'm happy to talk about it - we can have meaningful discussions about what exactly a world model is, how to distinguish between a bad world model and the lack of one, and where ChatGPT seems to model or not model the world.
"ChatGPT is a glorified word predictor", on the other hand, can't really be discussed at all. I struggle to even call it a criticism or concern; it's a discussion-ender, a statement that the idea is too ridiculous to talk about at all.
"ChatGPT is a glorified word predictor", on the other hand, can't really be discussed at all. I struggle to even call it a criticism or concern; it's a discussion-ender, a statement that the idea is too ridiculous to talk about at all.