alt Hacker NewsI'd also SERIOUSLY ask for an "LLM generated" flag in addition to normal flags - on both comments and submissions, that is.
Obvious slop comments thankfully already get flagged or downvoted in the hell they belong, but the submissions, oh dear god I do not wish to engage with shit written by AI any more. It's questionable enough if you let LLMs do your coding work for you, but if you can't even be arsed to write a proper README.md on your own, why the fuck should I be forced to waste my time reading it?
Replies
Unfortunately most of what gets posted to HN or could conceivably be posted here is already LLM generated to some degree. Because content of interest to HN tends to be written by the same people who increasingly prefer to use an LLM for everything they do, machine-generated submissions are only going to increase over time. Tagging submissions as LLM generated is ultimately futile, as one might as well tag water as wet.
Also HN's design ethos of not having "silos" precludes having tags or filters or blocklists, since they want everyone to see the same content at the same time. This is the kind of thing you're going to have to use with a browser plugin or third party service if you want it at all.
That should be somewhat feasible to be done objectively, now that we have LLM text fingerprinting regulations in place.