ChatGPT generally solves stuff quicker and more precisely for me.

I have yet to receive a badge but could add that to the prompt or custom instructions as an alternative to its encouraging words when I'm nearly there.

It does now. If StackOverflow stopped then it would certainly not be as useful in several years time.

It will just learn directly from the code suggestions it gives you. Did you use the suggestion it gave you? Did you fix it manually? Did you generate or write a test case that passed for the suggestion. There are many ways for such a system to learn.

..until the next generation comes along who have only ever learned to code with AI doing most of the work....

the context ChatGPT provides around some of my error message inquiry is incredibly useful. on stack overflow, I have to read and discern the context.

this is the whole value of GPT so far, they can pivot the structure and align it to the context at point

It solves most of my issues. But I the problem for SO is that most helpful pages are from the past.

For finding answers, Stack Overflow may be useful, but for posting questions, it is not. I ceased posting questions a while ago due to the aggressive moderation, which often labeled my questions as duplicates or too common. I was very upset with the direction Stack Overflow had taken, as it seemed to embody the very criticisms it had once had about forums. The Stack Overflow community has lost its vibrancy and has become stagnant. I am pleased that platforms like ChatGPT are emerging as real competitors.

If they didn't moderate aggressively, they would probably end up with a bunch of duplicate questions. I really like the fact there's not a ton of dups on SO. ie, see other help sites. They also have a high standard for those questions, to the point where it will take 10+ minutes to post a good question. I don't ask questions either, mainly because I don't want to spend 30+ minutes writing a good question AND that's a good thing.

IMO, the aggressive de-duplication of questions means that people will keep updating the answers as things change, which is really nice as you don't have to worry about the responses to a question becoming out-of-date/stale.

I don't think people object to de-duping true duplicates. The problem is near duplicates, where the original question/answer didn't quite cover what you're asking. In those cases, it can be frustrating getting past the SO duplicate filter, which can sometimes be too aggressive.

Or, when searching for answers, you keep chasing your tail because the mods kept killing off the variants of the question you cared about, and you have to cobble together answers from the comments because those will often be all you get.

Or (and arguably worse), you'll find a relevant question but the answers are inundated with "like the real answer, but variation for situation X" because people are posting to the main question because they know the almost-but-actually-not dupes will all get closed out.

There still are a ton of duplicate questions. When questions get marked as duplicate they don't get merged or anything, they just lock the most recent ones (even if they are much better than the older one).

Their moderation doesn't prevent duplicate questions it just makes it more frustrating when the question you find has been marked as duplicate.

Is ChatGPT a competitor? I always viewed LLMs in this context as an interface for SO. They're not contributing new answers/code, right?

It bootstraps from the internet, including SO, but in fact they do have coders in house teaching it stuff: https://www.semafor.com/article/01/27/2023/openai-has-hired-...

Oh wow that's interesting, thanks for sharing! I thought that the value in today's LLMs comes from the wealth of "free" crowd-sourced answers online. It seems like OpenAI is trying to create more source material specifically for "basic coding", though I wonder how cost-effective that is in practice. Or effective in general, for that matter.

Suppose you are just learning C++ language. There is a syntax error or type mismatch. The error message is 50 lines long and it does not make any sense to you. You already spent 10 minutes on it. What do you do now?

If there is an experienced C++ developer around who is happy to answer questions, great. Otherwise you are stuck with having to figure it out by using trial and errors and Google-fu. If you post it to a forum, it could take a while for anyone to respond. If you post it to stackoverflow, very likely nobody wants to look at your horrible code (natural for a beginner, you know), and your question gets downvoted.

By contrast, ChatGPT can look at your code and explain very clearly what is wrong with it within seconds.

This is just one example. And it's not only for beginners. I have found that ChatGPT can answer high-level questions in programming very well. The alternative would be searching for the Internet and sift through all the noise to find the answers as well.

Isn’t it worth considering that the reason ChatGPT can do those things is that it was trained on data from platforms like Stack Overflow? This is hard to quantify but my guess would be that without the SO data it wouldn’t be as useful.

It is not worth considering that because that is irrelevant to the fact that ChatGPT is a competitor now.

You are correct.

This is why reddit and twitter both locked down their APIs, due to the data being very high quality and immensely valuable for training.

Too bad they all did it too late, no one saw ChatGPT coming. And since all the data was already scraped from Stackoverflow, it no longer has any value for OpenAI. Stackoverflow is rapidly declining in volume, so future data from it is irrelevant.

Counterpoint: I've never had a question moderated as duplicate, and I've been using StackOverflow for 15 years.

It can still happen. I've been actively using SO for 14 years and had my first question marked as a duplicate last week. The question was already 9 years old and had been thoroughly answered. Last week somebody decided to "optimize" it, which sent it to the review queue and finally led its closing.

Would be almost hilarious if it was not my next thing to search for

Your experience must be an outlier. After I posted I read other comments here and others had the same experience (just search "duplicate" here).

It does seem likely that I'm an outlier . OTOH it could be that just people that have had bad experiences are posting their experiences and the silent majority without bad experiences aren't posting. It's so hard to tell in a forum format.

I think the loudest people are the ones talking about their bad experience. I've been on SO for, I think, 13 years. I've rarely had questions closed as duplicates and if one was it truly was a duplicate and I found my answer on the referenced question.

I don't see how all the people complaining about moderation don't look at all the low quality questions and see why moderation is what it is.

Personally I don't remember having gotten questions closed. However, sometimes other people's good questions have gotten closed, making me annoyed that I didn't get to see any, or more, answers.

> I've been using StackOverflow for 15 years.

I guess you and any experienced user know the "meta game" of allowed questions. Survivorship bias, kinda.

Yes, seems a likely explanation. Being on the internet prior to the eternal September probably helps too.

Problem is ChatGPT is feeding on SO

Exactly. I've already hit the dreaded "knowledge cutoff" of 2021.

What happens when this training data is 10 years old? Where will we get new data?

In the meantime we'll get used to functioning with these models with assumption the "knowledge cutoff" will just be moved. But if there is little new data, how? Furthermore, how do we prevent feeding ML generated data to ML training in the first place?

> For finding answers, Stack Overflow may be useful, but for posting questions, it is not.

That is mostly as it should be. The primary original intent of StackOverflow is to curate a good Q&A, not to provide an interactive support/mentoring service. Questions are (supposed to be) evaluated by what they add to the repository of questions and answers.

That being said - there has certainly been a culture of harshness in relating to newbie question askers on large parts of the SO.

I love stack overflow, however I also hate it. I have asked some, I thought very reasonable question on there, and they got shot down and removed. The elitism, in this day & age of nearly infinite search and data storage capabilities makes me not feel all that sorry for them though. I do hate that we are doomed to AI generated trash answers on google now for at least a decade though.

I sometimes contribute. There's a lot to be done on the Python side, because Python and libraries change faster than the answers do.

"In theory" - I think at least - answers should be updated when new things become possible. A lot of answers used to be "you can't do that with X, do Y instead" and should now be updated to either, "You can do that with X" or "You have to use Z"..

Same here. I look for answers about 10 times a day I'd say and in 90% of the cases I found what I was looking for, in a few seconds, for free. I also post answers when I found old questions and the answers are outdated or not complete. I know it helps other people cause I usually get 1-10 upvotes / year for those answers on old questions. Mostly about python, sql, postgres, js, react, aws and docker, so popular topics. I rarely have to post questions on my own, usually cause it's already asked and answered, or cause I'm experienced enough to find the solutions on my own or somewhere else.

Anyway, have been using SO for 12 years and still think it's great.

I’m curious what you are doing that leads you to lookup stuff 10 times a day.

I'm curious what kind of programming job you could have where you look up stuff less than 10 times a day.

Something they are actually competent doing, I suppose.

That seems needlessly harsh.

But, generally, yes, I tend to know the details of what I'm doing already. If anything I might check some API docs for a library.

If much of your development is spent on implementing business logic in a language you are familiar with, and you have a decent IDE, I don't know what you use stackoverflow for at that frequency. Occasionally you may need to consult stackoverflow for a specific issue in string manipulation or API usage, but that's not something you run into a lot.

I rarely ever need to use SO on a daily basis. I've been programming on the .NET ecosystem for a while now and usually if I need anything I just directly go to the docs.

I used SO a lot when I was a junior engineer. Most of my issues now stem from biz logic or internal design that really can't be googled for.

StackOverfow is great for questions that are very basic for people who work in the corresponding domain but can take some time to put together based on the docs along.

For example, "how to insert multiple multi-columns rows dynamically using psycopg2" It is trivial if you are writing sql everyday but may take longer to find the right postgres docs. StackOverflow gives you the answer in 5 minutes: several variants with different tradeoffs (readability, performance, security).

Every time SO comes up on HN, I’m wondering what universe I live in where it’s such a useful site with strict guidelines that actually make it useful.

Maybe it’s the bias of being there since the early days and really getting how to ask and answer questions.

Exactly. The biggest issue for me is duplicate questions, which is why mods are tough on new questions. 99.9% of the time the question has been asked. And yes, I also racked up over 60 necromancer badges doing what you do. The tricky part is deciding which of the duplicate questions to submit my answer to.

For Android development SO is pretty useless. I often find answers on there from <=2015. That's when Android 6 came out, its way too old to be relevant. But if you try to re-ask the same question they mark it as a duplicate.

I'm skeptical that if you ask a good question and the "duplicate"is so old that it's not relevant anymore that your question would be closed.

I am not. The mods there absolutely love closing questions as duplicate. It has earned that reputation for a reason.

Can you give one example? I keep reading the negative comments about SO, how the moderators are bad, and there's never one example confirming it.

I've found that SO varies a lot by tech/language. In my experience, the C++ and JS subworlds are pretty brutal. C# is middle-of-the-road when it comes to culture. Clojure is very friendly and welcoming, but obviously more niche. This may explain why folks have different experiences on the same site.

I agree. Rust is the worse by far in my experience. There's a couple of really prolific people on there (Shepmaster and Stargateur) who treat it as their personal fiefdom. Shepmaster does provide a ton of high quality answers but I don't think that gives them the right to be so unfriendly.

Kind of ironic considering how the Rust community prides itself so much on being welcoming.

Sure, here's an example:

https://stackoverflow.com/q/11635/265521

It's the top result for "case insensitive string comparison in C++". It has 372 votes and a ton of useful high quality answers. But it was closed 5 years ago for being "opinion based". Presumably because the author made the foolish mistake of asking for the best way to do it, rather than just a way to do it.

I don't see how anyone could defend that.

That's not cherry picked, and actually the situation is worse than it might seem to a casual observer because when you do occasionally manage to reopen a question (quite difficult) it leaves no evidence that it was ever closed.

That's one of common failure modes for SO. When people ask for "best", they actually mean tell me the options, and the reasons to choose them, and their advantages and problems. It's hugely helpful. But moderators see "best", and automatically close it as "opinion" because there's no objective best. But that isn't the point of it!

So the author made a foolish mistake of not reading the site's rules? I know, I know, who does that? I might be one of very few… Just like I might be one of very few who actually take their time when asking a question, showing respect to anyone who might spend some time answering…

The question doesn't show research effort, it's just two sentences. Sometimes that's all it takes, but if you add a requirement "without converting to lower/upper case", I think it should come with an explanation? I often encounter questions like that "How do I drive my car without using a steering wheel?" - and upon a confrontation I usually hear "well just for fun" or a similar answer - SO is a bad place to ask questions like that. If you're e.g. interested in performance, describe your situation, paste your current code and explain how you're concerned about the efficiency of your current approach.

Also originally the post wasn't not even properly written… Just a low quality post, so why bother improving a question (editing away the "best") that is so bad to begin with, especially as that doesn't guarantee you will avoid the critique (instead of critique of bad moderators who close a question, I would be reading about bad moderators that modify a question changing its meaning).

As I hinted at the beginning, we might be from different worlds: it's beyond my understanding how you could think of pointing out the example is not cherry picked, as if it was some beautiful post, with images, thoroughly explaining the issue, perhaps even addressing (preemptively or as an edit) the "opinion-based" flag, and yet was unfairly closed. Meanwhile it's shit. The revision history doesn't show any real OP's effort to fix the question, other than that meta exists to discuss such things, and if you're not satisfied with answers, you can write a real question describing your case.

So we might be from different worlds, and I'm happy SO/SE exist, because I find them, and their policies - for the most part - useful. You don't, and it's fine, you can use the alternatives.

> when you do occasionally manage to reopen a question (quite difficult) it leaves no evidence that it was ever closed.

you can see it in revision history: https://stackoverflow.com/posts/11635/revisions

The "closed" revision won't disappear when you reopen a question.

Of course not. It's still my first recourse if GPT-4 can't answer my question adequately.