> often even their comments will all be fully LLM-generated
replyThis really bothers me, coming here asking for human feedback (basically: strangers spending time on their behalf) then dumping it into the slop generator pretending it is even slightly appreciated. It wouldn't even be that much more work to prompt the LLM to hide its tone (https://news.ycombinator.com/item?id=46393992#46396486) but even that is too much.
> I promise it's not that hard.
replyHow many non-native English speakers are on HN? If it's more than 30%, why should they have to use a whole new language if they can just let an LLM do it in a natural sounding way.
You could write in the language you are most eloquent in and prompt a bot to do a faithful translation.
replyPost both versions
I’d rather read broken english than LLM output.
replyYes, that's what we tell people too.
replyhttps://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...
Ditto
replydeleted
replydeleted
reply