Skip to main content
127 events
when toggle format what by license comment
Dec 23, 2025 at 22:59 comment added pilchard Did some one say ick
S Dec 23, 2025 at 10:19 history bounty ended Resistance Is Futile
S Dec 23, 2025 at 10:19 history notice removed Resistance Is Futile
Dec 19, 2025 at 18:14 history edited SashaStaffMod
edited tags
S Dec 16, 2025 at 9:59 history bounty started Resistance Is Futile
S Dec 16, 2025 at 9:59 history notice added Resistance Is Futile Reward existing answer
S Dec 16, 2025 at 9:59 history bounty ended Resistance Is Futile
S Dec 16, 2025 at 9:59 history notice removed Resistance Is Futile
Dec 11, 2025 at 20:05 comment added ch4rl1e97 @Lundin Thank you.
Dec 10, 2025 at 13:21 comment added ygtozc @AshZade AI Assist does not violate the genAI ban (yet), but it does motivate policy motivations and justifications provided on the SO Help Center: "Users who ask questions on Stack Overflow may have already sought answers elsewhere. Due to the ease of using generative artificial intelligence services, if a user wanted an answer from an artificial intelligence, they may already have sought one, and so it does not make sense to provide one here."
Dec 9, 2025 at 18:00 answer added Jane timeline score: 6
Dec 9, 2025 at 16:34 comment added Jason C Good bye site engagement.
Dec 9, 2025 at 12:33 answer added Jane timeline score: 4
Dec 9, 2025 at 10:09 comment added Lundin @ch4rl1e97 Wrong thread. This one is about you getting mislead to get incorrect information from an "AI" in various broken ways. This thread is the one about your CC by-SA data getting farmed by "AI" and sold for profit without attribution.
S Dec 9, 2025 at 8:47 history bounty started Resistance Is Futile
S Dec 9, 2025 at 8:47 history notice added Resistance Is Futile Reward existing answer
Dec 9, 2025 at 2:09 comment added ch4rl1e97 I have never knowingly consented to my Q&As to be farmed by some second-rate HAL and would never, if given the option, have given such consent. Nor do I wish to ever engage, see, or be forced to acknowledge the existence of LLMs. For the rare occasion over the years I've interacted here, I shall not be doing so again.
Dec 8, 2025 at 21:20 comment added NoDataDumpNoContribution @AshZade I don't know. Maybe it isn't. If we could do the experiment, we could see. If it's all the same, there shouldn't be any difference between all the copies.
Dec 8, 2025 at 19:58 comment added Ash Zade Staff @NoDataDumpNoContribution isn’t that what exists as of now? One SO but users can keep to Q&A if they want, use AI Assist, try unified posts, etc. While they have connections to each other, they can also operate independently.
Dec 8, 2025 at 19:56 comment added NoDataDumpNoContribution Sometimes I wish we would have made copies of SO and let them live against each other. Like one SO going full AI, one SO going anything goes, no downvotes, and one SO as it is. I would have liked to see the outcome.
Dec 8, 2025 at 19:54 comment added NoDataDumpNoContribution The culture, whatever of it is there, hardly affects the advice questions experiment (no downvotes, anything goes) or the AI Assist (prime spot, no user interaction). Cultural issues should not have any big influence on whether they fail or not.
Dec 8, 2025 at 15:58 answer added user400654 timeline score: 9
Dec 8, 2025 at 14:53 comment added Adamant @Alex - You are not wrong, but jumping on the bandwagon of LLMs and trying to turn the site into a poorer version of ChatGPT doesn't seem to be the solution. Issues with our community culture need to be fixed by changing our culture, not by eliminating it altogether in favor of chatbots.
Dec 8, 2025 at 3:45 answer added Weijun Zhou timeline score: 4
Dec 7, 2025 at 1:39 comment added ssokolow As a Canadian watching my prime minister rush to try to diversify us away from the U.S. economy before Mr. 51st tries to cancel USMCA/CUSMA as part of his economic warface to do to us what you did to Hawaii, I'm hoping this U.S. A.I. bubble crashes late enough to minimize the harm to Canada and early enough to stage an intervention for StackOverflow before it becomes unsalvageable.
Dec 7, 2025 at 0:39 comment added Hashim Aziz @Alex 100%, the community really is the worst part of SO despite all their gripes about the company. These are people who might upvote one question a week if that, no matter how well-asked, but will kneejerk downvote and VTC at the slightest hint of every question that isn't perfectly asked. It's been a meme for a long time outside of SO that you don't go to SO if you want your question answered, and even as someone who's been here a while and knows how to ask a good question, I avoid it as much as possible nowadays because there are much less painful ways to get answers to most questions.
Dec 6, 2025 at 19:52 answer added drew010 timeline score: 9
Dec 6, 2025 at 13:33 comment added Alex SO has been dying for many years and is now flatlined. Its community has slowly condensed into the most extreme individuals, able to find meaning in this small bubble of rule-pedantry and an illusory image of “quality.” It’s telling to see this community recoil violently at anything that tries to spark life into this corpse, while instead loudly complaining about why the corpse hasn’t had a haircut or had its teeth brushed. It’s fascinating how such a dysfunctional community culture could have emerged, persisted without intervention, and eventually killed its host. Someone should write a book.
Dec 6, 2025 at 10:38 comment added kaya3 My Stack Overflow homepage, at full-screen, now has only three questions on it before scrolling, and they are at the bottom of the screen. Stack Overflow is no longer "a site for questions and answers". RIP.
Dec 6, 2025 at 6:09 answer added mOctave timeline score: 6
Dec 5, 2025 at 23:40 answer added izzyg timeline score: 12
Dec 5, 2025 at 15:37 comment added Bart Oh wow, this is absolutely dreadful. I'm not nearly as active as I used to be, but looking for some info just now I was presented with this AI block. "Surely there's a way to turn that off, right?", but it seem the answer is "nope"? What in the absolute clutter is going on with this site?
Dec 5, 2025 at 4:40 comment added You Old Fool Dumbfounding: stackoverflow.com/ai-assist/shared/…
Dec 5, 2025 at 4:23 comment added Laszlo Pav Isn't the whole point of SE that it's a place where you can go to ask other people about your problems?
Dec 5, 2025 at 0:52 answer added Mr. Squirrel.Downy timeline score: -4
Dec 4, 2025 at 17:38 comment added terdon @user400654 It is in the sidebar, as you said, and is therefore relevant to the entire network. Yes, the specific changes are SO-only, but sadly, the thing is everywhere so I want them to announce changes here so we can see them and not have them lost in SO.
Dec 4, 2025 at 17:32 comment added user400654 @terdon That is incorrect, in June it was embedded into all sites side bar, that hasn’t changed with this release, the only thing that changed was it was additionally embedded into SO’s homepage, all other sites still do not have this embedded into their homepage, outside of the menu that was done in June.
Dec 4, 2025 at 17:29 comment added terdon @user400654 no, for better or for worse, the thing is available and linked to from all sites. It isn't SO only, it's been embedded on all network sites' home pages. As far as I can tell, the only exception is this site, MSE.
Dec 4, 2025 at 16:56 comment added user400654 “we chose to post here because AI Assist works (can fetch content) from across the entire network, not just SO.” While that may have been relevant to the first iteration of this, given all you’ve done here was embed it into SO’s home page… the change here is solely a change for SO, not all other sites on the network. While I’m sure this will eventually be relevant to all sites on the network because stack cares more about selling stack internal than fostering a living breathing community, it isn’t today.
Dec 4, 2025 at 16:50 comment added Kijewski Its not often that you see someone so eagerly digging their own grave as SO does.
Dec 4, 2025 at 15:37 answer added Fattie timeline score: 27
Dec 4, 2025 at 14:53 comment added Ash Zade Staff @TylerH it's on the stackoverflow.com domain but accessible via the side nav on all networks and has access to all network content. We do send users to SO's question ask page but that's a trade off decision based on the % of queries fit for it. We will make that dynamic in the future though.
Dec 4, 2025 at 13:51 comment added TylerH @AshZade Except it's only available on stackoverflow.com, so it should be posted on that site Meta.
Dec 4, 2025 at 11:57 answer added tenfour timeline score: 39
Dec 4, 2025 at 11:22 answer added David Arenburg timeline score: 6
Dec 4, 2025 at 10:55 comment added Luc You forgot to mention another change, that is pushing this into everyone's notifications: Stop pushing product ads in notifications
Dec 4, 2025 at 10:50 answer added Mari-Lou A Слава Україні timeline score: 2
Dec 4, 2025 at 3:03 review Suggested edits
Dec 4, 2025 at 9:06
Dec 4, 2025 at 0:37 answer added Ken Y-N timeline score: 21
Dec 4, 2025 at 0:28 answer added Erin Anne timeline score: 7
Dec 3, 2025 at 23:33 answer added Chester Gillon timeline score: 2
Dec 3, 2025 at 22:25 answer added gerrit timeline score: 9
Dec 3, 2025 at 20:57 comment added Ash Zade Staff @Marsroverr we chose to post here because AI Assist works (can fetch content) from across the entire network, not just SO.
Dec 3, 2025 at 20:47 comment added Marsroverr Why is this announcement on meta.SE when the feature is only on SO? Shouldn't it be on meta.SO?
Dec 3, 2025 at 17:37 comment added dmh Looking for information about how to disable and delete all the AI chat features from my view of the service. I'm here to interact with other developers, not an LLM.
Dec 3, 2025 at 17:01 answer added Stephan Kolassa timeline score: 37
Dec 3, 2025 at 16:53 answer added tkruse timeline score: 5
Dec 3, 2025 at 16:46 answer added F1Krazy timeline score: 168
Dec 3, 2025 at 16:41 comment added Richard Why, precisely, are we trying to get users to ask fewer questions?
Dec 3, 2025 at 16:24 answer added anonymous timeline score: 8
Dec 3, 2025 at 16:03 answer added Snow timeline score: 20
Dec 3, 2025 at 15:55 answer added SPArcheon timeline score: 7
Dec 3, 2025 at 15:42 answer added Snow timeline score: 3
Dec 3, 2025 at 15:26 comment added Ash Zade Staff The screenshots are mockups. We fully test as different types of users. I think there's an assumption being made connecting the screenshots to our testing practices that isn't valid.
Dec 3, 2025 at 15:24 comment added Snow Yes, but "test as a user" is one of the fundamental tenets of software development. You can't, and shouldn't assume that it'll all look and work the same way with and without elevated permissions.
Dec 3, 2025 at 15:19 comment added Ash Zade Staff Good catch re: mod diamond, but the experience in the screenshots are representative of non-mod, authenticated users.
Dec 3, 2025 at 15:18 answer added 0stone0 timeline score: 4
Dec 3, 2025 at 15:18 comment added Snow When sharing screen captures of functionality, please please please show that you're testing the system as a normal user. It's painfully obvious that these screens have a mod diamond and mod-flags and therefore is being tested as an elevated user. So, please demonstrate that you're testing and viewing at the appropriate user level.
Dec 3, 2025 at 15:11 answer added 0stone0 timeline score: 7
Dec 3, 2025 at 15:06 answer added 0stone0 timeline score: 8
Dec 3, 2025 at 14:36 answer added 0stone0 timeline score: 6
Dec 3, 2025 at 13:28 answer added You Old Fool timeline score: 9
Dec 3, 2025 at 12:56 comment added Ash Zade Staff @GSerg I'll answer in earnest: since we launched the Alpha, +80% of AI Assist's usage has been technical questions. It's one of the strongest reasons we continued the project. I commit to sharing detailed data in the next few weeks as we collect it. I have mentioned this point in past meta posts: the sentiment here does not reflect the usage of AI Assist overall.
Dec 3, 2025 at 10:03 comment added GSerg Are there usage stats already? Is the most asked question from the AI box "How to remove the AI box from the main page of Stack Overflow"?
Dec 3, 2025 at 7:45 answer added Lundin timeline score: 58
Dec 3, 2025 at 7:42 answer added NoDataDumpNoContribution timeline score: 79
Dec 3, 2025 at 7:06 comment added talex "our mission to power learning and affirm the value of human community and collaboration" since when? Did I missed some announcements?
Dec 3, 2025 at 6:15 comment added NoDataDumpNoContribution It creeps and it creeps and it creeps. Why should human experts again spend their time answering questions here?
Dec 3, 2025 at 1:33 answer added Franck Dernoncourt timeline score: -10
Dec 2, 2025 at 23:56 comment added Brad Straight out of the "how can we make Stack Overflow crappier this quarter?" playbook. At least the Stack Exchange dictator is consistent.
Dec 2, 2025 at 22:59 answer added Franck Dernoncourt timeline score: 3
Dec 2, 2025 at 22:55 answer added Franck Dernoncourt timeline score: 3
Dec 2, 2025 at 22:40 answer added Franck Dernoncourt timeline score: -4
Dec 2, 2025 at 22:39 answer added Franck Dernoncourt timeline score: 10
Dec 2, 2025 at 22:32 answer added Franck Dernoncourt timeline score: 15
Dec 2, 2025 at 22:18 answer added Franck Dernoncourt timeline score: 4
Dec 2, 2025 at 22:15 answer added Franck Dernoncourt timeline score: 6
Dec 2, 2025 at 22:13 answer added Franck Dernoncourt timeline score: -10
Dec 2, 2025 at 21:50 answer added Franck Dernoncourt timeline score: 4
Dec 2, 2025 at 20:55 comment added DBS Obligatory "Here's a selector for hiding it" comment: #mainbar > div:has(#ask-stack-initial-question)
Dec 2, 2025 at 20:55 comment added GammaGames @Sloloem mass-deleting content is viewed as vandalism and will probably just get your account suspended. There have been some relevant discussions for ideas over the years though, if anonymization is too much
Dec 2, 2025 at 19:21 review Close votes
Dec 3, 2025 at 6:49
Dec 2, 2025 at 19:18 comment added Sloloem @Gewure Yeah but that just sounds like you're saying "Stop complaining". My answers were here for people, not to be incorrectly summarized by a half-baked LLM that can't attribute its sources worth a damn and I'd prefer not to be a party to it.
Dec 2, 2025 at 19:05 answer added cottontail timeline score: 151
Dec 2, 2025 at 18:57 comment added Gewure @sloloem I don't think deleting your answers makes a difference. Don't be Don Quixote, fighting Windmills.
Dec 2, 2025 at 17:55 comment added Ash Zade Staff @zcoop98 all great questions! The majority of users don't use the homepage (anon users are served the Questions listing page for example), but as you said, this is an issue with the homepage and part of the scope for the redesign: A First Look: Stack Overflow Redesign
Dec 2, 2025 at 17:54 answer added cocomac timeline score: 16
Dec 2, 2025 at 17:50 comment added zcoop98 @AshZade I don't think that's unreasonable, I just still don't like the overall message it seems to send, it doesn't sit well with me. I'd also posit that if the clickthrough on those questions is poor, then the homepage has an identity crisis that should be tackled wholistically... why is the clickthrough poor? Is it a relevancy problem? Would users prefer seeing something else other than questions? The AI box may very well be part of that solution, but I'd assert that it's not the whole picture. I haven't gotten the sense that Stack, or its users, really have a cohesive vision for that page.
Dec 2, 2025 at 17:39 comment added Ash Zade Staff @RandRandom by "user report", I mean if a user sees a shared chat that violates any trust & safety rules.
Dec 2, 2025 at 17:39 comment added Ash Zade Staff @computercarguy AI Assist doesn't post Q&A content so I'm not sure how it conflicts with the current genAI community policies.
Dec 2, 2025 at 17:36 comment added Ash Zade Staff @zcoop98 we looked at engagement data with those lists before implementing the widget - very little to no click-throughs to those questions. We made the trade-off decision based on that.
Dec 2, 2025 at 17:23 comment added zcoop98 Nah. I like AI tools, and I don't think the integrations are bad from the outset– but I feel that adding yet another widget to the top of the home page which pushes more human questions out of view is a terrible compromise that I feel is wholly unacceptable on a site that proports to value human contributions above AI ones. The focus just feels so wrong to me, and I think having no way to hide, relocate, or even minimize that pane is just ridiculous.
Dec 2, 2025 at 17:13 comment added V2Blast StaffMod @RandRandom: As stated in the Share popover: Anyone with the link will be able to view a read-only version of the conversation. So anyone you share the link with will be able to view that specific chat. (If you don't share it, folks won't be able to view it, but you count as a user who could report the chat too.)
Dec 2, 2025 at 17:12 history edited Joachim CC BY-SA 4.0
improved phrasing
Dec 2, 2025 at 17:01 answer added Joachim timeline score: 40
Dec 2, 2025 at 16:55 comment added computercarguy So, how does this work if there's a ban on AI in questions and answers? Is it a ban on everyone else's AI, but SO AI is "fine"? That just sounds like more corporate hypocritical BS to me.
Dec 2, 2025 at 16:49 comment added Rand Random @AshZade what does ...or users report can be removed. mean? a different user can report my chats? so a user can see my chats?
Dec 2, 2025 at 16:46 answer added 0stone0 timeline score: 68
Dec 2, 2025 at 16:10 answer added DBS timeline score: 21
Dec 2, 2025 at 16:00 answer added Sinatr timeline score: 13
Dec 2, 2025 at 15:58 answer added Sinatr timeline score: 17
Dec 2, 2025 at 15:57 comment added wizzwizz4 @AshZade Will deleting chats also revoke SE's license to use the content of the chats?
Dec 2, 2025 at 15:43 comment added Ash Zade Staff @Lundin we're going to add the ability for users to delete chats. For any violations, we have a process for content deletion for the entire site already. Anything our monitoring catches or users report can be removed.
Dec 2, 2025 at 15:43 comment added Lundin Is it supposed to archive previous chats for all eternity? What if a user writes something in AI prompt that is clearly a CoC violation. Is it still going to sit there with no way for anyone to delete it?
Dec 2, 2025 at 15:39 comment added Sloloem I couldn't be less happy SE is getting in on the LLM game. Is there a way to opt out of being used for the hallucination engine or do I just need to delete all my answers?
Dec 2, 2025 at 15:35 answer added esqew timeline score: 255
Dec 2, 2025 at 15:33 comment added Sinatr Where is a post, where someone asking for this? Oh, AI hype, I got it, we must have it here because.. well, no reasons. I don't use "random" assistants there and there. I have preferred one, which is way better than any website can offer me for free. Thanks.
Dec 2, 2025 at 15:24 answer added ColleenV timeline score: 11
Dec 2, 2025 at 15:23 comment added Lundin Is it the same bug as "Oops! Something Bad Happened!" or is that a different one? I keep getting it randomly when clicking "AI Assist" in the SO left panel.
Dec 2, 2025 at 15:15 answer added wizzwizz4 timeline score: 22
Dec 2, 2025 at 15:14 comment added Ash Zade Staff @l4mpi, Lundin, thank you for reporting that annoying issue. We're taking a look.
Dec 2, 2025 at 15:12 comment added Lundin Is it supposed to crash, drop your last prompt, have you "rejoin" with the prompt lost in cyberspace? Because it didn't do that previously.
Dec 2, 2025 at 15:11 comment added l4mpi At the moment it mainly seems to fail hard. I'm getting a popup with a " Rejoin failed... trying again in n seconds" message counting down from 5 and then repeating. But I would call that an improvement over a hallucinating chatbot that is by nature unable to provide suitable answers for most questions, does not provide any attribution for the "ai" parts, and has various other issues. I wanted to test if it can still print offensive words on your screen and only delete them after printing, but I can't confirm that's still the case as it's not working at all.
Dec 2, 2025 at 14:51 comment added Mad Scientist The screenshot for sharing conversations is misleading, it suggests that the mechanism could be vulnerable to enumeration. The actual mechanism uses a UUID though, not a plain, incrementing number.
Dec 2, 2025 at 14:43 history edited Dan Getz CC BY-SA 4.0
markdown fix
Dec 2, 2025 at 14:42 history edited V2BlastStaffMod
edited tags
Dec 2, 2025 at 14:38 history asked Ash ZadeStaff CC BY-SA 4.0