"Pinker’s reasoning on AI was so horrifically bad (as reasoning qua reasoning, not to mention as elementary scholarship; angry and dismissive and failing to consider the opposite or try steeling the imagined argument, as well as ignorant of the Bostrom book that even outsiders who’ve heard of the field have heard is the basic literature), that I’m disinclined to believe anything Pinker says about topics I don’t already know about, lest that just be Gell-Mann Amnesia on my own part. I frankly worry he’s gotten old and run out his supply of precision."