Anybody else feel like the language we use with LLMs and neural nets in general is way too anthropomorphic? "It doesn't know X", " It understands Y". "it said Z". These things are glorified, n-dimensional curve fitting algorithms. They do not think. #asknostr