Hacker News new | ask | show | jobs
by bowmessage 228 days ago
If they were careful, which I'm sure they were, the flawless English is the result of a round of LLM proofreading.
2 comments

That's what I am thinking.

I'd use a local LLM too to make sure the original prompt does not leak and can't be connected to the published output.

Yeah I was thinking about that, these days you just run whatever text you want to anonymise through an LLM with some instructions for style.