Noah Simpson writes a technically correct but really dumb article reassuring people about LLMs.

Link. Yes, if we never couple LLMs to other knowledge models we’ll be ok. Except we did that last year. Who does he talk to?!