[slashdave]

It is exhausting, but so is the misconception that the output of an LLM can be cleanly divided into two categories.