Hacker News Reader

> if you edit those [LLM tropes] out, or ignore them, you can’t really argue that the writing produced by these models is objectively ‘bad’ any more.

Well, it isn't good either and that's the problem. It's just average.

I think that's a valid take, just one I disagree with in the article. I think a lot of it is average or bad, but a lot of it is good too.

And some of it is good, but has the annoying tropes, so removing them is enough.