All politics threads on HN are full of shillbots, and very likely bots now that GPT-3 is big.

It's not as well moderated or as complex as reddit w/r/t brigading, fighting obvious shills, or obfuscating upvotes, etc. Not that reddit is particularly immune to this, either.

I wouldn't trust jack shit about politics on HN. Details about the latest implementation of systemd? Absolutely. Anything not STEM? Dubious at best.

This is a very dangerous line of thinking because it lets you dismiss any comment you don't agree with as a bot or a shill without having to address the merits of it. Its an unfalsifiable yet plausible suspicion to have, which makes it dangerously tempting to believe on spurious evidence. The way I see it, our only good option is to engage with it anyway (provided they aren't gaming the platform itself via spam or similar). An argument isn't inherently bad or wrong because GPT3 wrote it, and the volume of these types of comments isn't insurmountable.

Maybe the guy I was replying to was a bot. Maybe he was a human who fails the Turing test. In either case it was worth the time I spent pantsing him for the benefit of the discourse.