It appears someone is using a LLM to make pro-Reddit (ie anti-subreddit-blackout) posts on r/programming. The accounts are all under 30 days old and the posts have that classic milquetoast ChatGPT vibe to them - utterly bland, inoffensive, “safe” posts. If you’ve used ChatGPT you’ll immediately recognize the style.