00:00 - 15:0015:00 - 00:00
00:01
Intentionally obstructing clear access to information, by increasing friction, as to make it harder to access information, is a form of censorship; however, I'm leaving that bit out of the answer, as some people seem to be against the notion of labeling things censorship. — Andreas detests censorship 35 secs ago
There's a network-wide search on Stack Exchange homepage, if that's what you're looking for. — Meta Andrew T. 45 secs ago
On a tangent note (not sure it's worth adding it to the answer itself) I'd be less in favor of saying "definitely not appropriate" if the policy post was constructed as it usually is: a support tag, a full lock on answers, and a link to a feedback post. Alas, it isn't, and the post got several detailed responses by the time of their deletion. — Oleg Valter is with Ukraine 31 secs ago
@MetaAndrewT. that seems to be the answer! I never go to the home page so had never seen it. Perhaps that's true for some others as well, so I think a short and accepted answer post would be great! (cf. stackexchange.com/search?q=%5Bcrystallography%5D+twinning) — uhoh just now
@FrédéricHamidi My brain apparently shut down, and I'm unable to process anything further today, so I'll have to revisit this tomorrow. — Andreas detests censorship 7 secs ago
Moderators were not, as far as I ever saw, provided an opportunity to give feedback or participate in discussions prior to the policy being crafted. — Thomas Owens 14 secs ago
Does this answer your question? Is there a way to search all sites at once? — Meta Andrew T. 28 secs ago
Also, as this affects all sites on the network AFAIK, can you please feature this? It directly impacts the moderation of all sites, so it seems worthy of being featured (unless you're explicitly not featuring it to hide it, which would be a separate issue) — cocomac 13 secs ago
I was just about to ask about why do some sites have full domain names while other have site.stackexchange.com. I would have been downvoted and duplicated. Thank god for searching. — Big Joe 56 secs ago
Yes, except that on most sites, the color change is so minor that you can't see it. Apparently, that's a known bug. So known that it was known back in October when this idea was first trialed. But not well enough known that it's been addressed yet. Supposedly, it is planned. — Cody Gray 41 secs ago
@Trilarion nothing is new. It's the exact same problems and unhappiness again. And quite a few sites have declined and become lesser than they had over time. Nothing unique about that either. — Journeyman Geek ♦ 59 secs ago
@Makoto well in a sense - maybe it's a blessing but also a symptom of the problems with the company, that it can't attract folks with deep community roots, nor values the ones who try to do it anyway. That's rather contrasting with the old days when a good chunk of staff were site/network regulars — Journeyman Geek ♦ 50 secs ago
@ThomasOwens Right, my mistake on the wording. It does seem more like an edict than a.... Proposal of a policy change. — Spevacus 12 secs ago
Your edit is very consistent with what I saw. Again, I'm not sure if any mods were approached privately or if there were things that were not notified. But yes, the only push notification to mods was an early preview of the policy change. — Thomas Owens 13 secs ago
There's also the impossible miracle- maybe SE will rethink this at some point in the future. At that point, it would be nice if we still have some communities that aren't drowned in a sea of crap. — Chris - Regenerate Response 50 secs ago
I don't think anyone was approached privately either, @ThomasOwens. They did that with the new CoC, but those who were approached later had no issue discussing that publicly, and I doubt anyone would remain silent about it now given the outrage. — terdon 15 secs ago
01:11
This is tagged both discussion and mod-agreement-policy. These tags are incompatible. You cannot have a discussion and expect to use this to set policy. Based on what I've seen, the company does not wish to engage in a discussion - with community-elected volunteer moderators nor the communities they support - on this topic. In addition, I also see that you opted to not apply featured to disseminate this broadly. — Thomas Owens 51 secs ago
Calling for a change in leadership generally ends up in annoyed people in power digging in. There's a reason it's a call for a change in direction. I'm not convinced we will get more than excuses but one of those will not be 'we had no idea' — Journeyman Geek ♦ 52 secs ago
I would also argue that the link to this discussion post should be in the Q part of the policy announcement. — Anton Menshov 40 secs ago
01:54
Stack Exchange prioritizes efficiency, expertise, consistency, legal considerations, and trust in leadership when not soliciting feedback from moderators. Quick decision-making, internal expertise, uniform policies, legal obligations, and community trust drive their approach. Regenerate response — Chris - Regenerate Response 1 min ago
@starball What joke? I'm just writing words. Writing more words is what this site is about, right? — curiousdannii 17 secs ago
I'll say this as many times as this kind of ironic joke is posted: please can you not? — starball 28 secs ago
02:22
"W/e just do it then" - I take it back. Holy F@%5(>? #%$7. That's terrible. Now you're just printing money and SE currency will lose value. I don't vote until I've read the whole page. If you came to my house, got on you hands and knees and asked me to vote, I'd be less put off, than by those circles begging me to vote before I even scroll down. (voting) Consistency is everything and you took it away. I ain't gonna BS you and add Leaving SE to my name; I'll go down with the ship, but there will be swearing. — Mazura 32 secs ago
99% of moderation is performed by regular users like me, but this post provides zero guidance for regular users like me. You haven’t even answered your own question, which was “What does this guidance include?” — Thomas Markov 7 secs ago
02:44
“AI detectors might be xenophobic, so we’ve decided to allow AI content.” — Thomas Markov 23 secs ago
"Though maybe a few more tweaks there wouldn't hurt" - actually I thnk more tweaks would hurt. I'd prefer no more tweaks please. Please. No, really... — but those new buttons though.. 22 secs ago
i mean, the policy post pretty effectively makes it clear that the the unguided flags should just be declined due to no evidence. despite what we can see with our own eyes. — Kevin B 32 secs ago
03:27
It's not clear what you are trying to ask or whom you are trying to reach, but this is the wrong place. This site is for discussing the Stack Exchange network of Q&A sites and the surrounding community and software which powers it. — tripleee 8 secs ago
03:39
Clearly much of the StackExchange community (including some moderators) have demonstrated animosity towards even the idea of kindness. This speaks volumes about the community and where its headed. — RockPaperLz- Mask it or Casket 46 secs ago
What would make sense as an alternative would be for the company to devise a better process for moderators to handle AI content, not a worse one. Off the cuff, some quick ideas; require two mods to agree on the verdict, or (gasp) have the company invest in hiring and training staff to handle these posts separately from the regular mod flow. This obviously only addresses the acute problem, but would show a way forward for the longer term. — tripleee 44 secs ago
@tripleee But that would require spending some of the CEO's paycheck on employees. Better just fire them, so the CEO paycheck can be raised. — Andreas detests censorship 53 secs ago
04:18
Echoing an earlier comment, why isn't this featured? Is it because it lacks the details necessary for meaningful discussion? — QHarr 6 secs ago
04:31
04:50
I noticed that the "after" new vote button image in your screenshots does not have a vote arrow button voted on, unlike the "before" image. This makes it a poor comparison. I suggest changing the "after" image to be like the "before" image, which is of an upvoted post. — galacticninja 34 secs ago
05:01
I tried your suggestions but I have couple issues that I cannot figure out. data.stackexchange.com/stackoverflow/query/… I don't know how to get the close table to join with the revisions. Moreover, I actually didn't understand the helper query that you originally had in the closed count table:
-- helper queries to build the main query, just copy paste Would you also elaborate on that? — M-- 29 secs ago05:11
@BakerStreet I agree that the atmosphere of a site can have a significant impact on the quality of participation, and that a negative atmosphere can lead to demotivation and discouragement. The snarky comments to this answer actually prove the point of the answer and your comment. — galacticninja 16 secs ago
05:21
05:33
@Ward well, what can we add to our names now? Maybe Shadow Wizard - Stop Lying? ;-) — Shadow The Spring Wizard 53 secs ago
I'm using Firefox on Android, and it doesn't look like I can use Stylus. Am I possibly wrong, or does anyone know any other user styling options? — JonathanZ supports MonicaC 10 secs ago
The deletion of answers looks not very well. Please state next time in the question body If community answers are accepted. This question is of low quality with regard to giving context and setting expectations. It should not be tagged discussion then. — Trilarion 52 secs ago
05:55
06:06
You can try uBlock Origin’s
:style(…) syntax, but it will be rather unwieldy. — user3840170 just nowI have the same problem - it is hard to see whether I have already voted or not. — Martin 17 secs ago
You can try uBlock Origin’s
:style(…) syntax, but it will be rather unwieldy. — user3840170 1 min ago06:25
@ShadowTheSpringWizard Hm.... now I know why I haven't changed my name back on Meta... it is never ending story... — Resistance Is Futile 24 secs ago
06:36
@Phil They kind of addressed them. It's not as if they have written nothing about them, they kind of acknowledged them. They could have delivered more reasoning why they decided to push through this change nevertheless. But a promise to address something surely doesn't include any liability for taking a specific action. They were still free to decide however they want to decide. It's their site, it's their rules in the end. — Trilarion 41 secs ago
@rene Not entirely. Remember that in 2018 SE added an arbitration agreement to the Terms of Service, meaning that unless a mod opted out they can't take SE to court. Whether they'll consider taking them to arbitration is another matter... — Sonic the Anonymous Hedgehog 17 secs ago
I'm unable to find this funny at this time, or even appropriate, given what's going on. Sorry. — blackgreen 5 secs ago
@ZoestandswithUkraine "..That said, they ran it on old questions.." That's a reasonable thing to do. I proposed to do that already in December 2022. However not telling about the exact results is kind of not inspiring confidence. And there doesn't seem to be enough confidence in the company left. However, everyone else could do that same thing, i.e. run scanners on content created before 2022. — Trilarion 28 secs ago
"Over time, they might have earned a reputation for making informed and fair decisions that benefit the community as a whole." See, this is the proof that AI generates demonstrably false bullshit. — Cody Gray 9 secs ago
"quite a few sites have declined" Maybe one could quantify that a bit in order to show that the changes from the company side basically didn't seem to have a strikingly positive effect. Moderators should know the company even better than normal users. They should know exactly what to expect. Maybe this is the proverbial straw that broke the camel's back or maybe not. I mostly care for the quality of the content, so if you moderators think that a strike is the best course of action in order to protect against too restrictive use of AI detection technology lowering the quality then I support it. — Trilarion 58 secs ago
06:58
It rather demonstrates that LLMs could in principle run a company as well as they are run by humans currently. Upvote for producing ontopic content. — Trilarion 58 secs ago
I really would like to have some stats, even stats from the moderators about how many people got suspended for posting allegedly AI generated content. The company and everyone else is discussing wild stuff, but nobody presents any numbers at all. So what should I believe? Maybe I am an AI??? — Trilarion 1 min ago
It's very much an on the ground observation - from the sites I am active on. Lots of previously thriving communities are a lot quieter than they used to be. — Journeyman Geek ♦ 40 secs ago
Obviously it's a difficult topic. Last year the network banned ChatGPT without having a good answer on how to actually enforce this ban. Of course there will be collateral damage. Now the discussion is how much damage one is willing to accept. This cannot be solved easily or simply by words. I would argue that we need more insight into the actual numbers. How many posts have been deleted? How many users complained about it? Can we estimate the false positive rate somehow (even that is probably quite difficult)? — Trilarion 21 secs ago
Just as a small remark: Even if it would be difficult to determine if a user was wrongfully suspended he/she still might been exactly that. The conclusion should not be that it doesn't happen if it cannot be proven. For example I could imagine that users may in general not protest in all cases where they would be eligible for a protest. If it's difficult to determine if a post was AI generated it's probably also difficult to determine if a post wasn't. For example, how would I convince anyone beyond doubt that this comment has not been generated by an AI? — Trilarion 53 secs ago
The communities decided in most places "we don't want AI content" and you say "We allow it..." - that still means the communities don't want it — Trish 48 secs ago
What strikes me as curious, given that SO supposedly want to monetise access fro AI companies, is that as more and more AI-generated content is added to the web non-AI-generated content is going to become valuable like pre-WW2 steel. Why throw away such a selling point? (I know, I know, number must go up before next investor call..). — snakecharmerb just now
"Are LLMs the iceberg to SEs Titanic" Nope, that would be letting AI generated content in. — E_net4 24 secs ago
07:43
@snakecharmerb Agreed. It will be "garbage-in-garbage-out" situation. I have no idea what they are thinking. Most likely there is no thinking at all. — Resistance Is Futile 40 secs ago
@rene some more dup targets I'd suggest: meta.stackexchange.com/q/384355/997587 and meta.stackexchange.com/q/386983/997587 — starball 49 secs ago
one problem with this question is that it lacks focus. It's asking multiple questions. strip out the questions that have already been asked / points that have already been covered in other Q&A posts. — starball 55 secs ago
@SyedM.Sannan I'm sorry, but I don't understand the first sentence of your comment. You then assert that "more votes normally equate to a better quality experience" without giving evidence; furthermore, I'm not talking about the quality of the experience, but quality assurance of content. Finally, it may be true that there are more people not voting than voting ill-informedly, but that simply does not mean that the former is a bigger problem. My answer is terse; if you need more details, perhaps see the final section of meta.stackexchange.com/a/389458/205264. — Keelan 6 secs ago
The most amazing thing is that SE is going to utterly ignore this discussion, we won't hear even a beep from any staff member. During Monica case there was Yaakov to serve as bridge, but pretty sure he got tired of this and won't do that again. — Shadow The Spring Wizard 53 secs ago
I genuinely don't believe the pixel pushers coming up with these changes actually use any of the SE sites. — MMM 48 secs ago
@RockPaperLz-MaskitorCasket If that's (including some mods being against kindness) what you took away from my comments, read them again. I'm not against kindness, that's why you got my downvote and a comment explaining why. I could've been 'mean' and not leave a comment, like so many users complain downvoters do. What I'm against is any suggestion of using votes for any other purpose than quality control, including 'kindness'. — Tinkeringbell ♦ 58 secs ago
Hi Keleigh Berenger, welcome to the Stack Exchange Network Meta site! I'm not sure which search brought you here but the problem you describe will not be answered on this specific site. To get an expert's answer for the topic of your question you'll have to find and then re-post on the proper site. Check How do I ask a good question and What is on topic on the target site to make sure your post is in good shape. Your question is definitely off-topic on Meta and is better deleted here. — rene 36 secs ago
Great work. I've made the following changes for myself: Added
font-weight: unset !important; to the .js-vote-count, to get rid of the bold text, setting the --fs-subheading variable to 1.9em on the container to restore the font size, I've overridden --black-400 to #babfc4 to restore the gray color and finally I've set the arrow sizes to 34px. — MMM 31 secs ago
@ThomasMarkov moderators have not received any more guidance than this. In both cases it is just a "don't do this" message. — wimi 27 secs ago
08:28
Actually, I went through archive.org and discovered that the original font size was
1.61538462rem, so I've set it to that wonderful number instead :) — MMM 48 secs agoOne more step to making Stack Overflow a first Q&A platform where answers are one hundred percent AI-generated. — Augusto Vasques just now
One more step to making Stack Overflow a first Q&A platform where answers are one hundred percent AI-generated. — Augusto Vasques 1 min ago
Additionally, Stack Exchange has an official channel for advertising, so the question arises: Why not use that paid service instead of trying to push for free advertising? Does the advertised company lack the funds to promote its brand? — Augusto Vasques 19 secs ago
This advertising piece aims to attract or repel the audience from the platform. Engaging in this kind of unsolicited intervention only devalues your brand, as the community where the advertisement was placed has specific rules regarding the type of content that can be published, and advertising is not one of them, which implies that your brand is automatically associated with non-compliance with rules. — Augusto Vasques 59 secs ago
08:45
But that is part of the problem: the experience with the voting buttons changes severely based on which browser extensions were used, which site in the SE network is being visited, or which theme was picked on the site itself. — E_net4 8 secs ago
08:56
Is a user who has never or hardly ever used capitalization and punctuation, has through the months or years of their activity written answers full of typos and "try this" in lieu of explanations of their solutions, and suddenly learned to write with perfect grammar and spelling, full-sentenced, plausible-sounding but conceptually entirely wrong answers not evidence enough? If not, then what is? — CodeCaster 30 secs ago
@Cerbrus ChatGPT actually stated good tools this time. They were actually useful. — PlaceReporter99 59 secs ago
Point 2 is useless, we can't know how fast something was written. 4 is just not an indication. 5 is also a slippery slope. So no, not useful. — Cerbrus 42 secs ago
It has, to its credit, pretty accurately described a number of signals of AI content (4 is even correct, sometimes, when multiple people try to post ChatGPT answers on a bountied question). But your suggestion that "ChatGPT will be the best at detecting the answers they made" is demonstrably not true. If asked whether it wrote a given answer, ChatGPT will either hallucinate a plausible-sounding response, or say that it can't. — Ryan M 10 secs ago
09:13
09:28
@BigJoe well, if there was something that was better and and changed something worse, and if you remember this then you will get a face of what the .. is this? I still, prefer the old question left side.... Yes, there was a problem with mobile usage, current this not solved the issue and make it worse, they only need more padding on the mobile! — kelalaka 37 secs ago
@Velvel That's not what I read when the post states "explain why upvoting is a kind thing to do". SE should never go around explaining the need for upvotes as 'a kind thing to do'. Luckily, they don't have to. Voting is important for quality control, getting users to help with quality control is important, users need votes to be able to help with quality control. No need to explain upvotes as 'kind' and downvotes as 'unkind'. — Tinkeringbell ♦ 20 secs ago
Yes, my main concern is about SE financial outlook. If I was an SE product manager negotiating with potential and existing paying customers of SE products and looked at these charts meta.stackexchange.com/questions/387278, I might feel pessimistic. — tkruse 1 min ago
@CodeCaster Last week or so I came across an answer with an initial version that was sloppy in style and had a lot of grammar mistakes, and in an edit it was converted to the signature chatgpt style complete with needless recap paragraph at the end. It was clearly done by chatgpt. But the technical part of the answer was unchanged. While the final state was technically written by chatgpt, I think this was fine in the spirit of the gpt ban: the issue is when gpt writes the technical part. So no, style and AI detection is not enough. But flat out banning AI banning is the wrong conclusion. — Andras Deak -- Слава Україні 43 secs ago
@Catija yeah, destroy the working one ( that was only need more padding on the mobile applications where it is hard to click correctly) and try to solve the new issues. Do you think that circle button solves the issue, then continue to bookmark, activity, Share, Edit, Follow, Flag, etc. they are all waiting — kelalaka 59 secs ago
I know it's meant as joke, or sarcasm, but.... sorry, not funny. Too soon. — Shadow The Spring Wizard 40 secs ago
Although your comment includes "conceptually entirely wrong" (may have been added as an edit?), in which case I agree. And these are exactly the kinds of circumstances our mods can use and have been using. — Andras Deak -- Слава Україні 29 secs ago
I wonder what will happen first, humans leaving SE or SE replacing humans with AI that generates high-noise low-signal engagement. — Akixkisu 51 secs ago
@Andras AI to fix your writing style has existed for decades; it's called spell check. — CodeCaster 39 secs ago
@E_net4 I believe that what you've mentioned also applies to the old vote button styling. IMHO, the vote button experience is worse in the old vote button styling, especially in certain site themes. The area covered by the contrasting color of the voted arrow button was too small. The contrast is much easier to see in the new button styling, regardless of the site theme, due to the larger area covered by the contrasting color. I can't comment much on non-dark mode since I never browse sites in non-dark mode. — galacticninja 8 secs ago
@IanCampbell SE Inc.'s business "model" has been built on "infinite growth promised to investors", and disallowing AI-generated crap and removing users who spam it is a clear and obvious threat to that "model". — Ian Kemp 28 secs ago
@Laurel you did not fix grammar, you made a major change in my answer. Are moderators still allowed to delete answers, and just not suspend the users? If so, should clarify, if not, should roll back the change. — Shadow The Spring Wizard 47 secs ago
@Laurel but why did you change it in the first place? If it was correct, why you removed it? — Shadow The Spring Wizard just now
The buttons may cover a larger area, but most of that is using a dimmer color, or using stronger, contrasting colors in the outlines. These perceptions may indeed be biased by the use of browser extensions. — E_net4 27 secs ago
We can delete "low quality" answers. That's pretty standard procedure on a lot of sites, going back before AI. — Laurel 59 secs ago
10:08
@BigJoe BTW, "Stackers" is the term that Stack Exchange staff use to refer to themselves, not the community members. — PM 2Ring 10 secs ago
What if we let AI both post questions, answer questions and moderate, all at once? The main advantage of this is that we don't need any users and not even online visibility. We could let the site run in a server hall on its own without ever getting published online. This has the following benefits: 1) using AI 2) using AI some more 3) and then some more AI. The only down sides are all users and customers leaving, the network reputation getting blasted to pieces and the shareholders stock getting reduced to the value of an AI-generated haiku. But hey... it's AI, so lets do it! — Lundin 43 secs ago
Sure but it would be ignoring it, and feedback that was given, not things getting missed cause silence on both sides — Journeyman Geek ♦ 53 secs ago
10:24
I'm afraid you've posted this in the wrong place. This is Meta.SE, for questions about the Stack Exchange network itself. You need to post this on Stack Overflow. — F1Krazy 53 secs ago
I wonder if any current AI detector can determine that the words of the CEO aren't generated by a marketing bot... — PM 2Ring 5 secs ago
@Tinkeringbell You seem to believe RockPaperLz-MaskitorCasket wants to undermine quality control. But I don't think that's the case. ("Upvotes literally cost nothing, yet so many people are incredibly unwilling to be kind enough to upvote helpful posts.") They're simply stating that helpful posts tend to attract less voting that flawed ones. But maybe I'm interpreting my own thoughts... — Velvel 49 secs ago
I see with extreme pleasure that once again SE has resorted to the same pathetic trick: posting a policy as an ANSWER so that the only way to ANSWER directly is as a comment. Once again Philipe I am telling you this (don't take it personally, I know you are just the one that got the shorter straw and had to post this) - next time post this as your "question" so that people can post actual answers instead of commenting.... — SPArcheon 44 secs ago
Even the claim of detectors' having high false positive rates for S.E. content is perplexing. The detectors return a percentage score purporting to index the likelihood of the supplied text's having been written by ChatGPT; a moderator could then choose a threshold over which to classify a post as having been written by ChatGPT & take action accordingly. Even if the company were under the mistaken impression that this is what moderators have been doing, they'd need to know the threshold to calculate a relevant false positive rate. — Scortchi - Reinstate Monica 11 secs ago
At any rate, they have shared a false positive rate from their study (on the Stack Moderators site) but not the threshold for which they've calculated it. (You can obtain as high a false positive rate as you like by setting the threshold low enough.) — Scortchi - Reinstate Monica 58 secs ago
In my experience modding another site, any post gets indiscriminate votes (off-topic sob stories get upvotes, out of kindness, I guess?) and these users involved in quality control are sorely needed to counteract the damage done by these indiscriminate votes, sadly, that makes them and their vote ratios, according to this post, 'unkind'. — Tinkeringbell ♦ just now
@Velvel that quote is again equating upvotes to kindness: "unwilling to be kind enough to" before even mentioning quality control (helpful). I don't see any data in this post to support the interpretation that 'helpful posts get less voting than flawed ones' either. Just data that says users most involved in quality control are voting more on flawed posts, and thus have higher downvote-to-upvote ratios. — Tinkeringbell ♦ 7 secs ago
11:12
It is also not unusual to post both the policy (in question form) ánd a discussion post, and link the two: meta.stackexchange.com/q/336364/369802 — Tinkeringbell ♦ 59 secs ago
@VolkerSiegel "You can expect better than human answers in many cases." Not sure if better than human in many cases. I would rather say surprisingly good almost as good as human in many cases. But what is mostly evident is a lack of specificity. The general problem area is well explained, but when it comes down to grasping the core of the problem there is often still a lack. It can be helpful, but you always have to take it with a grain of salt. — Trilarion 53 secs ago
11:25
You don't want it locked, as that would prevent voting. Currently, there is no tool that would lock a question from receiving answers but not other interactions such as voting. (Well, there's closing, but... that'd be even weirder.) Ah, I guess you cover that, though, with the "new type of lock". Did you know there's an emoji specifically to convey this sentiment? — Cody Gray 1 min ago
11:39
@Someone If I want to judge based on the content, and I don't currently have the ability not to let the score affect my judgement. — wizzwizz4 6 secs ago
@Mithical the announcements tag wiki says "For questions about Stack Exchange announcements" and this question is about announcements, why did you remove the tag? — Shadow The Spring Wizard 25 secs ago
@CodyGray while I didn't realize that locking the post would also prevent voting, I see that as a feature. If the whole idea is to show that "we don't care about your voice or feedback", voting is useless too. — SPArcheon 34 secs ago
@MMM I've incorporated these changes into the main script. Hopefully it didn't break anything :p — Ginger 30 secs ago
My impression is that "not interested in hearing from you" seems to be the current company's policy. Along with "sometimes we pretend we want to listen, but all feedback will be ignored and we'll do whatever we had already planned" — hkotsubo just now
They can do what they want with their platform (in the limits of whatever legal contract they had before). But we are reaching the point where we should realize that we all are also free to stop provide them content. — SPArcheon 45 secs ago
see my tangential post here. It should be made clear what is just an announcement and what is open to feedback — SPArcheon 7 secs ago
The button should say "convert to comment", not "edit". (It does say that for any mod on an English-based site.) — Laurel 27 secs ago
@hkotsubo I've been saying the later point for long long time, even almost got suspended for that. Now can finally say "Hah! Told you so!" — Shadow The Spring Wizard 9 secs ago
I am quite divided on this answer. On one side, it looks as a very badly aged joke/troll. On the other side, it seems a quite sharp bash at enthusiasms for "AI Suggestions" that the company demonstrated recently. Did you really mean that if SE thinks that AI generated suggestions are so great they are kinda hypocrite by not following them? — SPArcheon just now
Also, the are interested to hear from us, but they just take what they want from it, and never responding to things that make them uncomfortable. — Shadow The Spring Wizard 18 secs ago
12:30
The problem with the current scenario is that this policy was introduced without prior introduction or requesting feedback, which is what many moderators and other users are annoyed with. Yes, low quality content is still our enemy, and high quality content our friend, but that is not at all what this is about (any longer). This is about current policy-making, transparency, bad assumptions, and a (seemingly?) capricious attitude of the company. Besides, your answer doesn't add anything to either side of the equation. — Joachim 44 secs ago
Example of ChatGPT being terrible at detecting itself. We should not rely on it. — Seggan 57 secs ago
Thus, why not concentrate on getting rid of low-quality content instead of identifying AI-generated content? Because ML content is worse than low quality, it looks good quality, but is actually a steaming pile of garbage. — van dench 13 secs ago
@Seggan very funny, but I’m asking ChatGPT for its methods, and not whether something is written by it. — PlaceReporter99 32 secs ago
@Joachim That is why I added that this discussion is about the addressed subject. The company policy is a different subject discussed elsewhere. If any discussion turns into generalized meta-arguing, conflicts will never get solved. — Philippos 30 secs ago
Just wanting to let you know that for sake of coherence I expect that this answer has to be deleted too, and the fact that it is "pro" policy won't be the reason someone will forget to notice that another answer was posted. Don't worry, I don't blame you, but if you want this post to live, it is better to post it on the "open discussion" thread instead — SPArcheon 10 secs ago
@vandench Thank you for giving an example of one-sided comments that don't help. I meet enough user-generated content that looks like good quality, but is not. And ont each AI-generated content is just a steaming pile of garbage. — Philippos 54 secs ago
@Joachim clearly Atlas my fellow interloper. // 16 16 16 //. Jokes aside, thanks for noticing. — SPArcheon 59 secs ago
@Sha - Good point, I was remembering when the announcements tag was special cased — Mithical 29 secs ago
MaxBodySize can be increased from 30000 to 65536, as was done in '15. i.stack.imgur.com/w1eBm.png — Mast 1 min ago
If users got suspended for contributing good AI-generated content, I hope they get unsuspended. If they got suspended for flooding the site with junk, they simply should stay suspended because of that; it doesn't matter how the junk was generated. — Philippos 19 secs ago
This answer would be clearer if you were to specify that a "policy post" is indicated by the 'mod-agreement-policy' tag. Or, if that is not (necessarily) the case, to specify how else it can be told from a post. That way there is an objective method to tell "policy posts" from "posts that contain policies". — Joachim 52 secs ago
I like the old design much more. Sadly, the people behind this have definitely seen the very negative feedback, and most probably will keep the change because they have the power to do so. This is just like YouTube making dislike counts private – they never undid it. While I think a design change isn't that huge of a problem, it does show that Stack Exchange ignores our community (and does so loudly). The choice is up to all of us: keep supporting the company, or leave and find someplace else. Personally, I'll wait it out first to see what happens. — TheStacker- or rather AStacker 23 secs ago
As a matter of fact, if you don't you will be incited to by other users, and even by the site, which promotes accepting answers. The result is basically the same. — Albert 11 secs ago
I'd recommend checking out meta.stackoverflow.com/questions/405302/… - attempts have been made to try to reduce the influence of answer acceptance on SO. — van dench 11 secs ago
No, @CDR, accepting an answer does more than simply signaling a post as not needing more replies. — Albert 44 secs ago
Actually, the mere possibility of accepting an anser, even if it is not mandatory, carries over all the above problems. — Albert 16 secs ago
Please don't do this, curiousdannii. We get enough AI crap as it is, and it will only get worse, let's not add to it just to make a point. This is just making me more sad, and I don't see how it can do anything useful. — terdon 11 secs ago
I say that they opened themselves to a Catch 22 scenario that can get abused on both sides. It could be used to circumvent the problem (after all you are trying to remove low quality content, not AI generated content. Who cares about the source if the content is good?) and yet is blurry enough to allow for demonizing a new scapegoat unlucky mod should that be the case — SPArcheon 55 secs ago
@Trilarion That makes the impression that you are talking about answers without much prompting to steer the answer in the direction you want? It should be easy to make it specific, like by literally including "Be specific" in the prompt or the system prompt. Do you have an example query, so I can try to prompt it for very specific and very unspecific answers? (Try to include "Be terse." in the prompt for illustration) — Volker Siegel 15 secs ago
@AugustoVasques Well, once people find out nobody's allowed to moderate AI content, people will take care of making Stack Overflow like that next week. — doppelgreener 13 secs ago
@JourneymanGeek I'm not surprised. I agree that pull/vet would be even better, but that's no longer a small change. That's not just modifying a simple check, but introducing a new meachanism that requires additional database structures and so on. Sometimes I prefer the simple second-best solution over the best you won't ever get. Another year passed and many will follow while we wait for the big deal. — Philippos 27 secs ago
Its a big change sure but it would scale a lot better with as many sites as we have — Journeyman Geek ♦ 33 secs ago
It looks to me like they see bad traffic numbers in their future and they’re positioning mods as meatshields to buy time. Also consider they may be having conversations like “Could we use AI to mod SO?” You’re all right to be angry. — Mr. JavaScript 37 secs ago
My post is not about how to "convert to edit" works, it's about the state of the button dependent on the page where that button placed. Added to the Q to clarify. — αλεχολυτ 45 secs ago
@Trilarion landing path for total votes, bookmark, activity, share, edit, follow, flag, upvote for comment, flag for comment?... — kelalaka 30 secs ago
Thank you for replying with something else than "you don't need to accept an answer". I did not thought of your second point, which indeed is a good one. — Albert 45 secs ago
If future content is heavily AI-generated, there will be no market for SE to license the content to AI generators for training... The only way for SE content to be worth anything is for it to NOT be AI generated material. — Jon Custer 56 secs ago
Even SE needs to realize some day that in the long run, only "better" leads to "more", so let's assume we are on the same side. — Philippos 21 secs ago
14:16
There's plenty of money to be made with "low quality / high volume". — JonathanZ supports MonicaC 53 secs ago
At least this one is blue. On meta.so, it's grey, and a very light gray so that you have to peer closely at it to figure out which one has been selected. — aynber 26 secs ago
14:39
For those wondering what the duck is for, it's a powerful-if-rather-tenuous metaphor, reflective of the fact that not everything in an SE post can be considered to have any weight. — CDR 25 secs ago
In case anyone would like to use custom icons instead of triangles, I posted a variation of this script at tex.meta.stackexchange.com/q/10080/36296 — samcarter_is_at_topanswers.xyz 36 secs ago
@αλεχολυτ it's very unclear then. You should change it to reflect that you want the behavior of the button to change. Currently you're asking "Why I can't convert to comment", and this answer address this. — Shadow The Spring Wizard 53 secs ago
00:00 - 15:0015:00 - 00:00
« first day (7 days earlier) ← previous day next day → last day (480 days later) »
Transcript for
May30
May '2331
Jun1
♦ Meta Stack Exchange Comment Archive
A comprehensive archive of all Meta Stack Exchange Comments. C...
