Post #590348

@thomasfuchs I decline to state whether "artificial general intelligence" is possible or not. I do not believe that we understand the problem adequately to frame the question, much less answer it. But I can say with great confidence that "large language models" are quite irrelevant to the whole topic. They pose, however, a very special kind of epistemic risk, in that they corrupt the automated, high-speed information indexing and retrieval systems which modern society depends upon.