so two different massive Content Farms that depend heavily on the free labor of moderators have pushed their mods into a strike in the space of a week, one over AI nonsense and the other over trying to imitate Elon. Maximum 2023 behavior

(stack overflow and reddit)

@0xabad1dea any videos/podcasts explaining the stack overflow stuff? I missed that one
@redlilrascal the site owners told the moderators they're not allowed to remove posts for being obviously AI generated anymore
@0xabad1dea @redlilrascal hmmm, so they want now moderators to fact-check AI content which is a completely different effort than checking good ol’ users. What’s next? Moderators will discuss with an AI chatbot why their content could be factually wrong?
@0xabad1dea @redlilrascal Did they say why the AI posts need to stay up? Because no obvious justification is easily coming to mind for me! I disagree with Reddit’s motives but I do understand them af least
@sameoldstory @redlilrascal the best argument for tolerating the posts (IMO) is that ESL speakers may use AI to write more professionally and then have their posts removed.
@0xabad1dea @redlilrascal hmmm I’m not sure I buy this though. AI-generated text doesn’t come with a warning label or anything, you can only tell ChatGPT authored something if it basically wrote the whole thing — and IME it’s actually straight up inaccuracies in its writing that give it away (by this I mean when it writes something that sounds vaguely convincing but isn’t something anyone with subject matter knowledge would ever claim)
@0xabad1dea @sameoldstory wont esl users just use translation tools instead? I dont understand what AI has to do with non engish speakers...
@redlilrascal @0xabad1dea Translation tools often don’t spit out results that are nice to read (though I assume it’s only a matter of time before AI fixes that too). ChatGPT can be really help your text sound native speaker written — but I don’t think using ChatGPT to put some polish on your own writing should be noticeable to anybody else, and it should completely evade automatic detection tools (since in that case it’s just your own words with better grammar)
@sameoldstory @0xabad1dea agreed, im not a regular stack overflow user since the ai stuff happened but i imagine the vast majority of it is native english speakers who dont know what theyre talking about spamming the forums to feel smart, not ppl inputting original text to be translated into english
@0xabad1dea oh I'm out of the loop, what did SO do now?
@jorin the site owners told the moderators they're not allowed to remove posts for being obviously AI generated anymore
@0xabad1dea oh my fucking god how could they possibly believe that AI nonsense will improve the site
@0xabad1dea @jorin just because an answer is AI generated doesn't mean it's wrong. Sounds like they need to grow up over there (they def do)
@firewyre @jorin a) if the answer is noticeably AI generated, that virtually guarantees the person posting it doesn't know if it's right or wrong and it's not the moderators' job to know everything about every subject on earth. b) literally no-one enjoys reading the extremely distinctly obnoxious way GPT blathers on without saying anything meaningful.
@0xabad1dea the firewyre fella's post didn't federate for some reason and I looked it up and oh jeez am I glad it didn't
@0xabad1dea @jorin meh, that place became the domain of power-hungry douchebag moderators a long time ago (much like Wikipedia). I kinda hope it implodes.
@0xabad1dea Wait, what's going on with SO? Haven't heard anything about that
@udondan check the replies
@0xabad1dea Maybe that's a mastodon thingy but all replies *I* can see don't help at all

@udondan OH sorry I forgot that mastodon can be weird like that.

long story short: site told moderators that "being obviously AI generated" is not a reason to remove a comment

@0xabad1dea Heh, that's new. Last public announcement I'm aware of, was that GPT stuff ain't welcome.

But I feared worst, being mentioned together with Reddit.

Thanks for details!

@0xabad1dea I wish we lived in a world where the US Department of Labor concluded their investigation into the AOL Community Leader Program with a finding that unpaid moderation integral to a company's business is labor that must be paid.

https://en.wikipedia.org/wiki/AOL_Community_Leader_Program

AOL Community Leader Program - Wikipedia

@0xabad1dea Clearly we have not made enough of an object lesson out of Twitter. It's taking too long to go bankrupt. What can we do to accelerate the finding out?

@0xabad1dea Looks Web 2.0's structural problems are starting to backfire.

NGL, though, 5 years ago, I'd have never predicted things going down this way. Maximum 2023 behavior indeed

@0xabad1dea
EDIT: I've found it here: https://meta.stackexchange.com/questions/390106/moderation-strike-update-data-dumps-choosing-representatives-gpt-data-and-wh
Do you happen to have a link to the SO discussion? I only found the old meta discussion and the old gpt-policy which still forbids chat-gpt use (https://stackoverflow.com/help/gpt-policy)
Moderation Strike update: Data dumps, choosing representatives, GPT data, and where we’re holding

Update On August 2nd, 2023, negotiations between community representatives and representatives of the company concluded with an agreement being reached. The details of the agreement can be found at

Meta Stack Exchange