Summarizer

There was definitely a bit of Pournelle's Iron Law of Bureaucracy [0] at work. I worked there during a lot of the drama you allude to, and... It sucked, for everyone. But also... For most of SO's history, the vast majority of visitors (and this questions, answers) came in via Google. Not "search engines"; Google. This was pretty much baked in right at the start, and it effectively served as the site's primary user interface for years. And it worked. It worked pretty well! Until it didn't. At some point, Google started surfacing fewer "tried and true" Q&A examples and more unanswered, poorly-answered or moderated examples. This broke the fundamental assumption that sat behind SO's moderation - that curating a smaller set of posts was preferable to encouraging more, and newer. Suddenly, Google wasn't a very good UI for SO anymore. ...and SO didn't really have a fallback. Heck, for a while during this period they actually stopped showing questions on their homepage unless you were already logged in; the core assumption remained that there was a never-ending spring of new people feeding the thing. LLMs weren't the start of the problem, they were the end - the final wake-up call. I don't know that a site like SO can exist without the old Google, the old Internet; it is a product of all that, in the same way that mass-market TV shows were a product of 20th-century broadcast technology, or trade paperbacks of a particular intersection of printing tech and reading habits. [0]: https://www.jerrypournelle.com/reports/jerryp/iron.html

There was some of that for sure; sites that were all but designed to be attractive nuisances and took near-heroic efforts to moderate at all, with little chance of not causing a lot of drama. OTOH, topic-specific sites like Mathematics, MathOverflow, Physics, even small ones like Home Improvement or Seasoned Advice... Managed to collect a lot of good stuff: common niche questions with good answers that have a good chance at staying relevant for a long time to come. In a sane world, a few relevant ads on these sites would be enough to fund them for decades. But that appears to be another area where Google kinda shit the bed.

It still seems a bit too simplistic… no one imagined that Google could behave less than 100% virtuously in the future? Really?

> This is why all the moderation pushes toward deleting duplicates of questions, and having a single accepted answer. Having duplicates of the question is precisely why people use LLMs instead of StackOverflow. The majority of all users lack the vocabulary to properly articulate their problems using the jargon of mathematicians and programmers. Prior to LLMs, my use case for StackOverflow was something like this: 30 minutes trying (and failing) to use the right search terms to articulate the problem (remember, there was no contextual understanding, so if you used a word with two meanings and one of those meanings was more popular, you’d have to omit it using the exclusion operator). 30 minutes reading through the threads I found (half of which will have been closed or answered by users who ignored some condition presented by the OP). 5 minutes on implementation. 2 minutes pounding my head on my desk because it shouldn’t have been that hard. With an LLM, if the problem has been documented at any point in the last 20 years, I can probably solve it using my initial prompt even as a layman. When you’d actually find an answer on StackOverflow, it was often only because you finally found a different way of phrasing your search so that a relevant result came up. Half the time the OP would describe the exact problem you were having only for the thread to be closed by moderators as a duplicate of another question that lacked one of your conditions.

> Having duplicates of the question is precisely why people use LLMs instead of StackOverflow. The majority of all users lack the vocabulary to properly articulate their problems using the jargon of mathematicians and programmers. Yes; so the idea is they fail to find the existing question, and ask it again, and get marked as a duplicate; and then everyone else with the same problem can search, possibly find the new duplicate version , and get automatically redirected to the main version with high quality answers.

> It's up to you to figure out why users are avoiding it. Hint: They've been telling you quite loudly. No, it is not up to me to figure that out. I have heard it said quite loudly many times, over a period of many years. What you are missing is: I. Do. Not. Care. The goal was never for the site to be "not dead". The goal was for the site to host useful information that is readily found. The site already has tons of useful information. But it's drowning in... much less useful information, and Google has become much worse (to some extent intentionally) at surfacing the good parts. > And that's exactly what happened to Stack Overflow, and you're one of those guys still trying to promote the use of Python 2.7 in 2026 This is a bizarre thing to say to me, of all people. I am always the one catching flak for telling people that 2.7 had to go, that the backwards-incompatible changes were vital, that the break wasn't radical enough, and that people were given way more time to switch over than they should have needed. But really, the feedback for Stack Overflow is trying to take it in the direction of places that existed long beforehand. If you want forums, you know where to find them. And now you can also find LLMs. Which, as commonly used by people seeking programming help, are basically a grizzled forum guy in a can. >Everyone has left, the lights are off in the empty debate hall and you're standing there at the podium telling a bunch of chairs and desks why everyone actually agrees with you. "Everyone actually agrees with [me]" is the polar opposite of what I actually believe and am actually saying. I am well aware that the model is unpopular. My point is that the popularity of the model is irrelevant to me. > Can I guess you had a lot of reputation points and you desperately don't want to believe they're worthless now? I have a lot of reputation points (the site still exists), far more than I ever felt I deserved, and I never really felt like they were worth anything. A huge percentage of them come from an answer to a terrible question (that was still terrible after heroic attempts at editing; this all happened long before there was a common understanding of the purpose of question closure or what would make good standards for questions) that, once I understood things properly, I closed and tried to get deleted. Over the last few years, with that new understanding, I have been trying to give away my superfluous reputation points in bounties, trying to get missing answers written for the few really good questions lacking good answers that I identify, always to no avail (the bounty system promptly became a honeypot for ChatGPT hallucinations as soon as ChatGPT became available). You do not know me or my motivations in the slightest.

This is because the real goal was SEO.

> They were primarily trying to make google searches more effective for the broader internet This is mostly how I engaged with SO for a long, long time. I think it’s a testament to SO’s curation of answers that I didn’t ask almost any questions for like 5+ years after starting programming

Sad? No. A good LLM is vastly better than SO ever was. An LLM won't close your question for being off-topic in the opinion of some people but not others. It won't flame you for failing to phrase your question optimally, or argue about exactly which site it should have been posted on. It won't "close as duplicate" because a vaguely-similar question was asked 10 years ago in a completely-different context (and never really got a great answer back then). Moreover, the LLM has access to all instances of similar problems, while a human can only read one SO page at a time. The question of what will replace SO in future models, though, is a valid one. People don't realize what a massive advantage Google has over everyone else in that regard. So many site owners go out of their way to try to block OpenAI's crawlers, while simultaneously trying to attract Google's.

Yeah, old news? It's how it is today with humans. You SHOULD be making things in a human/LLM-readable format nowadays anyway if you're in tech, it'll do you well with AIs resorting to citing what you write, and content aggregators - like search engines - giving it more preferential scores.

Too bad stack overflow didn't high-quality-LLM itself early. I assume it had the computer-related brainpower. with respect to the "moderation is the cause" thing... Although I also don't buy moderation as the cause, I wonder if any sort of friction from the "primary source of data" can cause acceleration. for example, when I'm doing an interenet search for the definition of a word like buggywhip, some search results from the "primary source" show: > buggy whip, n. meanings, etymology and more | Oxford English Dictionary > Factsheet What does the noun buggy whip mean? There is one meaning in OED's entry for the noun buggy whip. See 'Meaning & use' for definition, usage, and quotation evidence. which are non-answer to keep their traffic. but the AI answer is... the answer. If SO early on had had some clear AI answer + references, I think that would have kept people on their site.

Google also played a part. After a while, I noticed that for my programming related questions, almost no SO discussions showed up. When they did appear on the first page, they were usually abysmal and unusable for me. When it started all kinds of very clever people were present and helped even with very deep and complex questions and problems. A few years later these people disappeared. The moderation was ok in the beginning, then they started wooing away a lot of talented people. And then the mods started acting like nazis, killing discussions, proper questions on a whim. And then bots (?) or karma obsessed/farming people started to upvote batshit crazy, ridiculous answers, while the proper solution had like 5 upvotes and no green marker next to it. It was already a cesspool before AI took over and they sold all their data. Initial purpose achieved.

Because what you’re describing is the exception. Almost always with LLM’s I get a better solution, or helpful pointer in the direction of a solution, and I get it much faster. I honestly don’t understand anyone could prefer Google/SO, and in fact that the numbers show that they don’t. You’re in an extreme minority.

It seemed to me that pre-llm, google had stopped surfacing stackoverflow answers in search results.

My memory is there were a spate of SO scraping sites that google would surface above SO and google just would not zap. It would have been super trivial to fix but google didn’t. My pet theory was that google were getting doubleclick revenue from the scrapers so had incentives to let them scrape and to promote them in search results.

Seriously where will we get this info anymore? I’ve depended on it for decades. No matter how obscure, I could always find a community that was talking about something I needed solved. I feel like that’s getting harder and harder every year. The balkanization of the Internet + garbage AI slop blogs overwhelming the clearly declining Google is a huge problem.

Not to mention, it's not indexed by search engines. It's the "deep web".

The discoverability, both from the outside and within is absolute trash, but the closest I find of those old forums nowadays are Discord servers.

Agreed, it’s the discoverability that’s the real problem here at the end of it all. All the veterans are pulling up the drawbridges to protect their communities from trolls, greedy companies, AI scraping, etc. which means new people can’t find them. Which then means these communities eventually whither and stop being helpful resources for us all.

> - Success: all the basic questions were answered, and the complex questions are hard to ask. I think this is one major factor that is not getting enough consideration in this comment thread. By 2018-2020, it felt like the number of times that someone else had already asked the question had increased to the point that there was no reason to bother asking it. Google also continued to do a better and better job of surfacing the right StackOverflow thread, even if the SO search didn't. In 2012 you might search Google, not find what you needed, go to StackOverflow, search and have no better luck, then make a post (and get flamed for it being a frequently-asked question but you were phrasing yours in a different / incorrect way and didn't find the "real" answer). In 2017, you would search Google and the relevant StackOverflow thread would be in the top few results, so you wouldn't need to post and ask. In 2020, Google's "rich snippets" were showing you the quick answers in the screen real estate that is now used by the AI Overview answers, and those often times had surfaced some info taken from StackOverflow. And then, at the very end of 2022, ChatGPT came along and effectively acted as the StackOverflow search that you always wanted - you could phrase your question as poorly as you want, no one would flame you, and you'd get some semblance of the correct answer (at least for simple questions). I think StackOverflow was ultimately a victim of it's own success. Most of the questions that would be asked by your normal "question asker" type of user were eventually "solved" and it was just a matter of how easy it was to find them. Google, ChatGPT, "AI Overviews", Claude Code, etc have simply made finding those long-answered questions much easier, as well as answering all of the "new" questions that could be posed - and without all of the drama and hassle of dealing with a human-moderated site.

One factor I haven't seen mentioned is the catastrophic decline in quality of Google search. That started pre-llm and now the site is almost unusable to search web. You can access something you know exists and you know where it exists, but to actually search..? Most SO users are passive readers who land there using search, but these readers are also the feed of new active users. Cut off the influx, and the existing ones will be in decline (the moderation just accelerates it).

Are there any publicly available options to actually interact with real people about software development anymore? There doesn't seem to be anywhere that's accessible with something like a google search... Sure there are derelict IRC/Discord/$language forums, but of the handful I've been part of they aren't active or in the case of discord, weirdly disjointed. AI is great and all, but somewhere with a little bit of an opinion and push back to carelessly thrown out questions would be nice (as a thrower of careless questions). SO obviously went off the toxic deep end, but has that culture of shared problem solving just died completely online?

I think the biggest issue, what lead to the toxicity, came down to the question/answer format not suiting the problem it was trying to solve — The answer could only be as good as the original question, and the platform gave little leeway to "get to the bottom" of the problem. Getting to a high-quality question/response required a back-and-forth that the platform made difficult by burying the discovery/definition work in comments and edits instead of a clear discussion mechanism. All of this meant the learning-curve on how to participate was high, and this spurred gate-keeping and over-zealous moderation. High-quality but out-of-date information was preferred over lower-quality but more recent updates. When combined with the rapid shifts brought on with mobile development and web frameworks, the answers could easily get out-of-date months after being answered. I remember a time when StackOverflow dominated every search query. Now we're seeing searches take you to a dedicated forum/discussion board, which feels more appropriate for the current state of the industry.

I'm glad I learned how to program when you could coax useful answers from Google searches. Whenever a Stack Overflow result comes up now the answer is years old and wrong, you might as well search archive.org.

I suspect a lot of the traffic shift is from Google replacing the top search result, which used to be Stack Overflow for programming questions, with a Gemini answer.

Has AI summarization led to people either getting their answer from a search engine directly, and failing that, just giving up?

Probably similar for google. My first line of search is always chatgpt

I wonder if google search saw a similar hit

My personal bet is that traditional search engines face a -70% usage drop at the moment.

I doubt it. If I want to ask AI a simple question I type it into Google now.

anecdotally, i personally stopped using google a lot in the last few years

It makes sense to see the number of questions decline over time as people google questions and get results. It would be interesting to look at the number of comments and views of questions over time to see if that has declined as LLMs have driven declining engagement and discussion.

Good riddance to bad rubbish (TLDR: Questions are now almost never being asked on Stack overflow). The most annoying example I can think of (but can’t link to, alas) is when I Googled for an answer to a technical question, and got an annoying Stack Overflow answer which didn’t answer the question, telling the person to just Google the answer.