I keep saying it, but there's this weird narrative tic I keep seeing about AI where someone will lay out all the reasons it's bad, then conclude with "but there's some good usecases" or "it's just a tool" or some other hedge.

And like, I get it. Hedging against a lack of knowledge is good, actually! If folks who know more than you about a subject say thing $x, and you don't have the expertise to evaluate that claim, it can make sense to hedge and assume that the claim is at least partly true.