Bot farms rule the web. Bot-authored comments are everywhere, no less than 50%, regardless of site. Keep in mind, human-assisted bots have been used for the past 20 years but now the bots are a lot more productive thus the human part is largely quality control.
Fretting at the rare researchers who admit to their bot activity is like fretting at the criminals who admit guilt and praising those who don't.
Seems more likely to me. Reddit has a far broader userbase, whereas HN is primarily made up of the sort of people who would post with AI. And HN has a non rate-limited API (even though most people here probably use Algolia.)
And people already do paste comments from AI, it's becoming more and more common. At least they have the courtesy of pointing it out but I expect that will stop before the AI comments do.
While I do love seeing the comments starting with "I asked ChatGPT..." get absolutely eviscerated in the replies and negatively voted, it is probably sending the wrong signal.
Fretting at the rare researchers who admit to their bot activity is like fretting at the criminals who admit guilt and praising those who don't.