02:00 - 21:0021:00 - 00:00
9:01 PM
@ZoestandswithUkraine So they're either an abusive partner, or a child covering their ears, screaming "LALALALA!!!", with no humility whatsoever. — Andreas detests censorship 13 secs ago
@yagmoth555 It is mathematically and logically impossible for such a tool to exist, and work. It is computer-sciencely impossible to have an infallible detection algorithm with fewer than (the order of) eight-bytes-compressed minimal text units, since otherwise you could run it backwards and, by brute force, produce undetectable automated text. At this point I wouldn't completely put it past Stack Exchange to do something like you propose, but that'd be worse than even this policy. — wizzwizz4 31 secs ago
So ChatGPT is still banned but nobody gets suspended anymore for it? That would kind of defeat the ban. See also: How can we determine whether an answer used ChatGPT? — Trilarion 16 secs ago
"biases for or against residents of specific countries" I have difficulties understanding this. I can only identify if somebody speaks English as mother language or not and even that not with high confidence. AI generated content typically has top grammar and style. If anything the false positive rate might be higher for really well formulated content, but then don't we all strive for this? — Trilarion 16 secs ago
@Trilarion that really confsed me too, but see arxiv.org/pdf/2304.02819.pdf. A very limited study, on data (TOEFL essays all written by people using the same course, and speaking the same mother tongue) that are fundamentally different to SE content, so I don't know how relevant it is here, but that's an example of how detectors under certain circumstances may indeed be biased. — terdon 42 secs ago
Many one final comment, the false error rate may be controllable. One might be able to use a higher confidence threshold in these scanners and arrive at a lower, acceptable false positive suspension rate. Alternatively, wait a bit use the scanner on multiple answers from the same account and use an aggregate score. That should have a lower error rate. — Trilarion just now
Does this answer your question? We are graduating the updated button styling for vote arrows — CDR 24 secs ago
The whole premise is flawed because it starts from "regular users might use biased AI detectors" and the conclusion is "mods can't be trusted to judge from many kinds of signals whether a post is AI-written" which is the kind of fallacy that's hard to take as an honest mistake. Painting this all with "in the name of inclusivity" just drives it further home how cynical this all is. Not sure it's worth arguing about the face value of the "arguments" made by the company (for want of a better word). — Andras Deak -- Слава Україні 33 secs ago
I don't mind the circles to entice more voting. But the "darkened arrows" were such a big hep in letting me know which I've already up or down voted. — Chaplain Ray 13 secs ago
Not all questions are how to, some are also why or when or does exist X. Given only X you kind of lose a bit of information I think, but nothing that shouldn't be contained in the question body. — Trilarion 55 secs ago
2 and 3/5 are formally different. 3/5 could be answered by yes/no, 2 not. 2 and 4 are also different, 4 could maybe also be answered by not an SQL query, so I go clearly for 4 as the most general but still valid way of asking in this case, even though I think it might not make much of a difference for SEO. — Trilarion 37 secs ago
I see maybe to far, but handling chatgpt stuff was already a mess. Good luck to them now they block us to use such site and if they code nothing to detect chatgpt, all flag will be ignored in the end now — yagmoth555 46 secs ago
"It's tempting to think a change in leadership will help. But remember it's unlikely you'll find anyone who has just the right characteristics to pull off what Stack Exchange/Stack Overflow needs:" I volunteer... — Andreas detests censorship 23 secs ago
Something is always going to happen when it comes to the bottom line. Apple was the first one to discontinue the headphone jack, everyone followed suit. Twitter was the first one kind of to ban all third-party clients. Reddit and others are following suit. SE changed the voting arrows. it will get resistance probably for a week then everyone on the site will follow suit. It sucks but it happens. Not saying I like it, it's weird but it's reality. — Big Joe 33 secs ago
That image macro is begging to be redrawn by an ML model... — Andras Deak -- Слава Україні 56 secs ago
9:59 PM
I am constantly getting my attention drawn away from the user-posted content, and thinking "wait, did I already vote on this question/answer?". Having to look at and think about something that used to stay in the background isn't good. — JonathanZ supports MonicaC 48 secs ago
I am constantly getting my attention drawn away from the user-posted content, and thinking "wait, did I already vote on this question/answer?". Having to look at and think about something that used to stay in the background isn't good. — JonathanZ supports MonicaC 1 min ago
I agree with you, most of those using ChatGPT every day that claiming to ban it, one day, AI will be implemented here to do most of the jobs especially some forum routines. correcting language, etc,. — Akam 21 secs ago
It feels to me like there's really mixed/ unfocused messaging going on here. The question title and body focus on "moderat[ing] AI Generated content", but the policy answer is primarily about user suspensions, and gives little to no concrete guidance on the aforementioned moderation of AI-generated content. Can you update the title & question to reflect the actual policy you've posted? It's genuinely really confusing to me what your main aim is here because the answer and question title seem so disconnected. — zcoop98 39 secs ago
what does "Regenerate response" mean? Is it a ChatGPT "command" and your way of saying here that you want Phillipe to rewrite their answer post with a clear policy? (sorry if I'm dissecting humour and killing a frog here. I have never used ChatGPT before) — starball 50 secs ago
@starball It's a joke. It's at the bottom of ChatGPT when it's being used, and some users accidentally copy and paste it into their answers. Naturally, seeing it can make you pretty darn sure an answer is computer-generated. But apparently it's not even evidence we are allowed to consider anymore. Perhaps Phillipe thinks it's a hip phrase in some country we're discriminating against? — Chris - Regenerate Response 13 secs ago
this seems to substantially limit means to control quality of the posted content. Was company planning to compensate this somehow? — gnat 49 secs ago
@zcoop note that the question being linked is tagged mod-agreement-policy. I recognized that putting "agreement" in double-quotes is charged, but it was what I could think of when I wrote it. — starball 20 secs ago
@starball That's certainly fair enough, and I'm not at all knocking you for the wording, it's borne of justified frustration. I missed that "agreement" came from the tag, since it's not mentioned in the post anywhere else. — zcoop98 43 secs ago
friendly reminder that this question post is literally tagged "discussion"... — starball 29 secs ago
You’re censoring opinions and responses you don’t approve of. Disgusting — Andreas detests censorship 6 secs ago
To moderators: If possible, please move all comments from censored posts. — Andreas detests censorship 33 secs ago
Hi Aditya, questions on this site need to be about the sites or technology behind the Stack Exchange network, as defined in our help center. I'd direct you to check out a more appropriate network site, but this kind of recommendation question isn't really fit anywhere on the network, since we try to keep things more objective and focused around here, which recommendation questions don't do well. You might find better luck somewhere like Reddit if you're looking for a place to ask for open-ended recommendations. — zcoop98 30 secs ago
Please update your post. They now censored all other answers to their question, by deleting them. This is disgusting. — Andreas detests censorship 54 secs ago
The current lock notice states "It is still accepting answers, comments and other interactions" (emphasis mine). If that is not the case, please change the lock type to the appropriate one to prevent any new answers being posted, rather than just deleting them all afterwards. — John Omielan 1 min ago
They're burning through community goodwill like nuts right now... even if the deletion of feedback from the policy post truly is just a formality, e.g. allowing answers was a genuine mistake, it's still in suuuuuper bad taste, and the fact that staff/ CMs haven't already jumped in to clarify why they went missing is wild to me. It's a terrible look. Add that to the note on SE violating their own stated policy on amendments to the mod agreement, and you've got a pretty sizeable fire on your hands. — zcoop98 30 secs ago
@zcoop98 FWIW, CesarM left a comment on my deleted answer "Deleting this because we don't typically host answers on policy posts, if you wish you can create a new question on MSE for it." Not sure that's worth a lot, but maybe something. — Chris - Regenerate Response 36 secs ago
This post is manipulative. You have been told time and time again how you are wrong. Moderators have been very clear in the fact that you say one thing behind locked doors, and another in public. As such, I consider this post offensive. — Andreas detests censorship 48 secs ago
@Chris-RegenerateResponse Even that answer deletion explanation is questionable. I just jumped on your question post in relation to this (I hope you don't mind). — starball 53 secs ago
another weird thing: Your question post is written in a way that nobody would have really asked that question. How would anyone but mods know that this happened earlier this week? And why would mods be wondering about something that presumably you would/should have made clear to them? — starball 46 secs ago
They’re lying to us. They’re being dishonest. Please act accordingly. Don’t accept being lied to. — Andreas detests censorship 29 secs ago
This whole incident is - sadly - reminiscent of the Monica situation where SE staff keep posting things that make it clear they really don't care about the community. — Ward - Reinstate Monica 41 secs ago
Side note: could the chatgpt tag be added? It's locked so I can't, but... that is pretty directly connected to this post — cocomac 14 secs ago
This has happened before, with the policies regarding underage users and suicidal users - those were initially published without any sort of moderator consultation. They were recanted immediately and re-enacted after consulting moderators. — Sonic the Anonymous Hedgehog 42 secs ago
@SonictheAnonymousHedgehog I remember. There was the whole nonsense about how long moderator resignations could be featured or not, too. They tried to make that one secret, too. — Chris - Regenerate Response 5 secs ago
feel free to post this as an answer post in my question. I think the ideas are related. Or you can post your own question post when this gets deleted. — starball 12 secs ago
I mean... that's precisely how we handle FAQ's on meta. Why should company policy statements be any different? They aren't there for discussion, they're there to present the policy. — Kevin B 23 secs ago
@KevinB well... because it's neither faq nor announcement in the first place? It's explicitly tagged discussion, which states as follows: "This tag is used for questions that ask users to discuss an idea, concept, or problem." — Oleg Valter is with Ukraine 1 min ago
I like to think if enough of us do it, the sheer amount of work needed to micromanage us and make sure we're not taking action against computer-generated content will turn them off from doing it. Or at least doing it effectively. — Chris - Regenerate Response 28 secs ago
@OlegValteriswithUkraine While also being tagged mod-agreement-policy which states "This tag indicates posts that are official Stack Exchange policy that moderators are expected to abide by. Answers to these posts are generally locked so only staff members can edit." I can quote tag descriptions too! — Kevin B 40 secs ago
@KevinB note how nothing says they have to be deleted :) And also "generally". I am honestly surprised that you chose to defend the company actions on the technicality given all that's going on. — Oleg Valter is with Ukraine 33 secs ago
Honestly, I am confused. Are we allowed to delete AI-generated answers or not? The new policy is not very clear to me. The teams-post said different things than MSE announcement, and the things I heard in chat seem to indicate we should still delete AI-generated content. — Dharman 19 secs ago
I'm most certainly not defending this policy, but i do find it quite silly to try to make an argument on a post that they're going to purposely bury rather than making a new one that gets to HMP and stays high scored. — Kevin B 39 secs ago
@Dharman The messaging is super unclear. One more reason not to change anything about the way we currently moderate over it. — Chris - Regenerate Response 15 secs ago
It would be censorship if they went on and deleted every post we make about the policy, that hasn't happened yet. — Kevin B 26 secs ago
Excuse me, but why did I get a downvote literally 0.5 seconds after posting? Please read the post before voting. — Andreas detests censorship 35 secs ago
The second downvote was me and I downvoted as soon as I read "this is effectively censorship". It isn't censorship when you're simply told to post a new question. It might be wrong, that's up for discussion, but let's stop calling everything censorship. — terdon 23 secs ago
One of you retracted their downvote. Please be aware that the sentence is still there. I moved it to the end, and change its wording slightly. — Andreas detests censorship 22 secs ago
@BigJoe They could remove all the moderators (despite multiple promises that they won't over this sort of thing). Would that kill the community any faster than telling moderators we aren't allowed to use our judgment to act on things that everyone agrees are hurting our communities? — Chris - Regenerate Response 56 secs ago
"We’ve heard your feedback from the previous posts about this change, and are not moving forward until we address them"... BAHAHAHAHA, good one! 🤣 — Phil 18 secs ago
Oh crap, yes indeed. Thank you, that was very fair minded of you. I have now downvoted again. — terdon 2 mins ago
There is no precedent of this. No precedent of "dissenting" answers being removed? You, sir, must be new here I'm afraid. — Frédéric Hamidi 2 mins ago
@FrédéricHamidi I have edited to clarify. I'm definitely not new here. I know they have deleted lots of content they don't agree with, in the past. Is it clear now, or do I need to clarify it further? — Andreas detests censorship 1 min ago
@Andreas, well, it makes it clearer that you consider the introduction of the
[policy] tag as a kind of game-changer, which I... don't understand why exactly. SE-the-company has always reacted in the same way to dissenting answers under their announcements ^W edicts in the past, and I personally see no reason for this to change, now or in the future. — Frédéric Hamidi 23 secs agoquoting my own comment: "If you want to put an announcement where it won't get downvoted into oblivion in the answer section, put the policy in the question post.". See also my post here — starball 30 secs ago
@starball yeah, had the policy been posted as a question, there would be far less justification for deleting answers. On the other hand, I admit that the one thing I like about that horror of a post is that it was actually posted as a proper Q&A. Yes, I know I am clutching at straws. But since it was posted that way, I can see how answers aren't really appropriate. — terdon 5 secs ago
I agree with this. The major issue here is a failure to read the room, but that is not the basis of a sound policy. It is reasonable to separate the feedback from the official announcement. As was noted by Kevin B, this is how we handle questions tagged faq on per-site metas, so whether there's the precedent for it here on MSE with staff announcements that was claimed or not, it is a reasonable policy, notwithstanding the emotion of this particular situation. — Cody Gray 50 secs ago
02:00 - 21:0021:00 - 00:00
« first day (6 days earlier) ← previous day next day → last day (356 days later) »
Transcript for
May29
May '2330
May31
♦ Meta Stack Exchange Comment Archive
A comprehensive archive of all Meta Stack Exchange Comments. C...
