I agree there's a strong element of misplaced anthropomorphism. But the issue isn't even that LLMs are (themselves) a threat to humanity, or that we have any re

I agree there's a strong element of misplaced anthropomorphism. But the issue isn't even that LLMs are (themselves) a threat to humanity, or that we have any reason to think they are "conscious" (in a human, or even animal, sense). The reason for concern is that the emergent behavior we're seeing from a mere token-predictor should give us serious qualms about blindly pursuing the creation of alien intelligences, which are highly capable of unintended consequences precisely because they are neither human, nor conscious.