Members of the software developer community have reported deleting or altering their posts to prevent them from being used by OpenAI.
You must log in or register to comment.
People cheering on SOs demise don’t realize what we’re losing.
Support is moving to discord which sucks massive ass. Asking the same question over and over, hoping somone is around to help.
It sucks.
I never understood the move to synchronous communication for asynchronous questions. The ephemeral nature of discord is really a PITA. It’s like using IRC for a FAQ.
Discord is honestly the most awful way to create a helpful community.
It’s a great way to give the 20 most active members of the community someplace to trample on top of newbies trying to get questions answered.
It makes sense if the issues being discussed are time-sensitive. Sometimes people need a solution now, not to open a bug report and hope that it will get a response an unspecified amount of time later.
Not just that, it forces to create an account just to view!
It’s purely because of how easy it is to create and manage a community. Imagine if there were a way to make a Lemmy instance without any fees or knowledge of selfhosting, it would be an instant success.
They don’t operate well at scale needed, though.
And search engines are unable to index the questions and answers, so good luck finding the already answered question.
Exactly. I really hate when people use Discord for stuff like this.
Or slack
Pretty soon search engines won’t be able to return anything anymore. At which point we might be looking for communities where live people can help with our issue. And if that happens Discord won’t look that out of place anymore.
If you can go somewhere and have your problem solved do you really care that some schmuck later won’t be able to find the solution written somewhere and will have to go through the same process?
You’re forgetting about the schmuck on the other end of that equation that has to answer the same question a thousand times over.
A server I’m in actually has a bot that, whenever a user types a command, will respond to a question with an answer.
Simple example:
@\Johnny: “Hey, my browser says 404, what does that mean?”
@\Support: !\404 @\Johnny"
@\Bot (using Discord’s reply feature to reply to Johnny’s post): “‘404’ is a common error message that can be caused by a variety of issues. Here are steps to resolve the most common issues:…”
I didn’t feel like typing out steps to check if the website is down for everyone or to fix it if it’s just you, but you get the idea.
The server I’m in doesn’t typically handle questions about that, either. They’re more specialized to something more specific that that, but, again, you get the idea.
Nah, that person can just ignore/block you.
I didn’t say it was a perfect setup. 🙂
And tbf there’s lots of people who don’t read the FAQ even if you shove their face in it.
deleted by creator
Support is moving to discord which sucks massive ass.
It sucks but can you blame them? It’s a natural response when people see that the old method (public posting and indexing) is being corrupted and grows increasingly irrelevant.
We’re going to see more and more knowledge becoming insular and/or gated behind manual curation.
This doesn’t necessarily have to mean Discord, can be private forums of any kind but private nonetheless. Discord may be the wrong tool but the problem it’s being applied to is real.
it sucks but can you blame them?
For picking discord I very much can blame them, I figure it won’t be long until that goes down the drain too.
they did an ai, but it got shut down.
Discord could solve this particular issue by simply adding a wiki.
We’re going to see a lot of changes in online community tools and in the way people use those tools. Lemmy is not exactly revolutionary either, it’s modern but it’s still a forum at its core.
Discord could solve this particular issue by simply adding a wiki.
yeah, as Stack Overflow could. Discord is bound to the stakeholders, and is already getting enshittified.
and is already getting enshittified.
Not yet. They’re still trying to squeeze money out of people with their profile pic frames and stuff. They’ve yet to introduce ads into app (for now).
Not to mention the people answering the questions are liable to just start accusing you of being an idiot if you make any less progress with their solution than “it’s been fixed so hard that it gained five new functions I didn’t even write into it!” I wrote a 3Js project once and ISTG the people on that discord had all the patience of a three year old who suddenly has to go to the bathroom the red second you’ve merged onto the highway.
Once upon a time, they stepped forth from the forests of IRC, but back into those dark woods they then one day marched.
Discord is straight up unwelcome on corporate networks.
I give it a few months before the Community tier servers’ data is dumped and sold to an AI model company.
So four months til GPT starts calling everyone “kitten”
I just got a ping the other day from a Discord server that said they’d finished moving their support onto a forum on their website specifically because Discord’s forum feature is terrible.
deleted by creator
The period when dejanews just started to index newsgroups was a golden age for finding answers on the internet, IMO, and there’s a strong similarity to the fediverse. All we need is for it to be searchable… OK, I see your point now.
I support the take over. People thinking that they can pick and choose who makes money and how are fooling themselves. The developer who wants all his IP protected so he can make money is upset that a larger entity is also making money is honestly tough fucking shit. Either go back to the origins of the internet napster days or shut the fuck up and live with what we created. There isn’t any middle ground. Its only going to get worse unless you make this place hostile to people building walls and stealing all the data for themselves.
I feel so bad for the long term contributors :/
The only good thing I could think off, is that someone is going to create a defederated stackoverflow alternative?
Or something similar, to bring back real human interaction…
If this wasn’t enough, This will probably raise war against corporated AI.
It won’t. Some people will scream bloody murder, most people will ignore it.
SO was in decline anyway. Most answers you’ll find are several years old and outdated, because some idiot thought the new ones are duplicates.
So now a few people will leave, the spamming idiots will keep spamming the platform with low effort nonsensical answers and its relevance will dwindle just a bit faster.
Look at Reddit. Last year there was a huge outrage and today it’s pretty much the same as before.
Most people don’t care. Most people feel so powerless, that they’ll accept every privacy scandal, every exploitive business strategy, every sellout of their platform.
Is Reddit pretty much the same? From my limited perspective, a lot of the genuine contributors left, quietly or otherwise. I’ve found it much more difficult to have an interesting discussion on there since the API debacle. Most of Reddit was already lurkers and bots, so all it took was a significant proportion of the tiny minority of quality contributors to take their time elsewhere for reddit to become a complete dumpster fire.
Anecdotally, pretty much every time I’m searching for information on reddit a number of comments are redacted or even the op is deleted. The only reason I didn’t purge my comments is in case someone might find them helpful.
I have all my deleted comments in a csv (with context links), which I plan on fine-tuning an LLM with just for fun. I guess if there’s a platform that’ll accept it, I’d be happy to upload it. Mostly I wanted to make sure the info remains free for everyone, including AI researchers.
Hugging Face is the usual platform for sharing datasets and models.
Subs that I go on that used to get hundreds or thousands of comments now are lucky to reach 50 or so.
Reddit is still pretty useful, but it will become less and less relevant as contributors leave, just like StackOverflow did. Side note: are contributors actually leaving Reddit? People keep saying that’s happening, but I don’t really see it…maybe it’s very slow? Might depend heavily on the subreddit too.
I don’t have numbers, but I did. Took Redact 9 hours to overwrite & delete the 17,000 comments on my 17-year-old account. But watching them scroll by, most weren’t really worth keeping. I saw several 15+ year-old active accounts do the same before I left.
Wait, how can it find all your past comments? I thought Reddit only have you a list of the most recent 1000 or something like that?
Through the API - well, it was through the API before they changed it. No idea if that’s possible now.
This is something I have tried to convey since I came here when people are fantasising about the death of Reddit, I just couldn’t put it as eloquent as yourself.
I take the approach that me not using Reddit, or Amazon or whatever else is a choice I make so I can live with myself, and not that I believe it will have an impact.
I have alluded to this in previous comments in the past, that many of the choices I make actually negatively impact me more than the company I’m avoiding. Example: Not using WhatsApp means I can’t join group chats with friends as they won’t use signal as the things I care about, are meaningless to them. Or that I can’t find some items to buy except from on Amazon so I just won’t buy them etc.
All we can do is stick to our own morals and let others do as they will as it’s futile to make people care about the things we think they should.
I feel this. I’m well aware that if my partner wasn’t on Facebook, we wouldn’t have a social life. I HATE that fact, but that, sadly, is where people put their events. I don’t think I’d join if she left, but I can’t deny that I benefit from her being on the platform.
She won’t leave until everybody else does, and they won’t leave until everybody else does, and so nobody leaves. It’s dystopian.
SO was in decline anyway. Most answers you’ll find are several years old and outdated, because some idiot thought the new ones are duplicates.
There’s been a new thing (for the past two years anyways) where some power tripping user would edit the highest rated answer, causing new users to fail to get recognition.
So a new user answers a old question with the latest way to do something based on new language specs… And they’d get 1-2 votes.
Why even contribute then?
To help others? That’s why I did it. Someone editing that into the highest upvoted answer is good because more people will see it.
Unless you care about karma, how is this bad?
The only good thing I could think off, is that someone is going to create a defederated stackoverflow alternative?
I’ve read of someone making an alternative to stackexchange federated.
Let me lookup for it and add a link here.Here it is:
https://lemmy.ml/post/15471686Interesting.
Someone there mentioned Codidact, which is available today.
Do you mean federated? And what would federation solve?
The only way you’re separating humans from LLM will be by asking for government ID but that would eliminate anonymity. And even so people could sneak in LLMs under their credentials.
is that someone is going to create a defederated stackoverflow alternative?
God I would love that. But as simple as the UI is, it would be some real work.
What do you think the hardest challenge would be?
I’m a software developer but I don’t know too much about working with de-federated services, but I would be interested in working on a Stack for us, if it was feasible and maybe we got a few more devs on board.
I’m working on a P2P Reddit alternative, but Reddit is basically a more complicated StackExchange. Here’s the mapping:
- stack exchange site -> community/subreddit
- question -> post
- answer -> top level comment
The missing bits:
- selected answer - easy metadata to add
- comment on question
- bounty - is it needed?
- tags - again, easy metadata to add
Maybe I’ll consider forking my project once it’s ready and turn it into a QA site. The hardest part will be useful search due to its distributed nature.
Thanks for this. I’m intrigued by your P2P Reddit too. Do you have a public GitHub where I can follow?
Not yet, but I’ll post it somewhere here soon. I want to make sure the basics work properly first so people have a good experience on the first try. It’s pretty rough right now.
What do you think the hardest challenge would be?
I honestly wouldn’t know where to begin. Hardest to start would be learning enough of the problem space to identify p problems, but that’s just what’s in front of me
That actually sounds like a good idea. Like Lemmy you have communities of common or popular languages like java or python which you can join and everyone there assists with questions. As it grows you might see a node for spring or flask get created for more niche discussions.
Maybe someone should create Lemmy. Oh wait…
Why need another federated protocol, Lemmy should be more than plenty
why would they want to move to a defederated SO, it’s going to be public anyways too.
And they smell bad too!
What?
It’s an old, bad, joke on the two definitions of revolting.
I think it would’ve been funny if the title was “StackOverflow contributors are revolting” and the comment was “a little more than usual.”
But hey, gotta get whatever amount of humor in while you can.
He’s punning on the multiple meanings of “revolting”.
Stack Overflow Users Are Revolting Against an OpenAI Deal
Stack Overflow Users Are Revolting
https://en.wiktionary.org/wiki/revolting
Adjective
revolting (comparative more revolting, superlative most revolting)
repulsive, disgusting
The most revolting smell was coming from the drains.
Looks like downvoters don’t get the joke. I first seen it on the cover of Horrible History.
Smelly neeerds
I hate 2024
And so the enshitification continues…
I bet its related to the continuous need to liquidate resources to support the proxy war effort. You can’t print money withour offering some value.
Stack overflow still have users? These days it rarely shows up on my search results and when it does the answers are always outdated by several years.
Interesting, I still see it pretty consistently in the first few results in my experience and usually with a pretty recent one too
What options are there to use instead? I think they’re still often having the best results, and are usually near the top.
I guess I’ve moved to the part of my career where Stack Overflow isn’t that helpful, but I’ve found a lot of utility in searching issues on GitHub and Reddit posts.
I’m right there with you. It’s nice to know it’s been there if I needed it. I don’t find myself there very often anymore and when I do it’s often to compare official docs to other ways to approach something or because the getting started section of the official docs felt weird or wrong.
I’m grateful that Stackoverflow is a valid alternative to reddit for many things. Fuck reddit.
But are they losing the first positions because they’re losing relevance, or is it due to other sites abusing seo and search engines abusing from advertising results?
I’d say the former.
Many queries don’t find relevant questions, and the relevant questions are often not answered properly. I often find the exact same problem I’m having, but the answers are just a bunch of those CV padders that post completely irrelevant answers based on a buzzword they saw while skimming the question.
When I do find relevant answers, lately they’re all so old that they no longer work, or rely on now deprecated functionality of a library or system.
Finding code snippets for interfacing with Azure through PowerShell is a crapshoot because Microsoft keeps deprecating different PowerShell modules for it.
It’s time for a federated version. How about “OutOfMemory”? It would fit because I always return to the same topics on StackOverflow.
Stack Overflow, technically a neutral term. Idk though whether the name in such a context would violate any trademark laws even if it’s a non-profit platform.
Snack Overflow
Nullpointer Exception
Access Violation
Syntax Error
Page Not Found
Random Access
Guru meditation
:D
Segmentation fault
RubberDuckHaven
Time to edit all my answers :)
This title doesn’t mention it, but it was reported earlier that users editing their past posts against this move get banned for it.
Need to make sure the diff is small enough. A tiny change that creates a bug or makes the answer effectively useless is much worse than sweeping changes
But that leaves a lot of good code. The bad parts are very unlikely to appear in the AI results due to the amount of good code in the training set.
I got banned as expected. For a month. After I’m unbanned, I’ll edit a few characters every day. Not giving up. As unlawful it sounds, apparently you are not entitled to ask them to remove all your data.
I’m pretty sure they keep the revision history, so there’s no point in it
Correct. They banned people and then undid the changes.
I did on Friday and within 5 mins they suspended me and reverted them all. I knew they would so I didn’t care - I just did it so they’d see as many unhappy users as possible.
I then deleted my account of over 10 years with over 50k reputation. Fuck stackoverflow.
I don’t understand why you and others are so mad about this. Stackoverflow is a great resource that takes significant time and money to maintain. I don’t have a problem with the maintainers making money by selling access to train AI on the data.
Having Stackoverflow as an alternative to reddit is important so that people aren’t stuck using reddit.
Yeah I don’t get this either. As long as they keep the main SO sites up and available, why should anyone care if they train AI on it? Deleting or defacing content is like cutting off your nose to spite your face.
Same reason for me. Take some of their time and say fuck you to them. But I will do crippling edits by single characters in one month time. I have time.
I’d assume they would provide a backup from before the announcement for the training.
Oh, look, it’s Reddit all over again.
(Yes, yes, different reason. Same user response, though.)
If it’s free online, you are the product
Not necessarily, no. There are a plethora of non-profit services and pages on the net.
What platform do you think you used to write that commeent?
Regarding platforms of for-profit companies, you are more right than wrong though.
Oh I was referring to for profit. Should have specified. I’m used to using that when meta screws over their users.
Alright, I’ll bite, what’s the FOSS endgame?
Lately it’s been companies using liberal licenses to build their product up really fast and then doing a rig pull bcz “oh no we can’t have competition!”.
But it’s also from the user perspective… Freedom to not pay a lonely dev that’s been coding on the project for “fun” or “for the community” for 15 years.
I love free software, but if y:all want it to succeed you need to support the devs!
God forbid someone get a JavaScript snippet that uses var instead of let or const.
Me over here still using none of them, just straight raw-dogging my variables
Total power move. If they’re global, your colleagues will know you’re the Beyonce of the band and can do whatever you want.
gasps
time for users to shred their comments and posts
Okay but why? It’s not like it’s personal Data or something. I don’t get why people are mad ._.
EDIT: Ofcourse you can downvote me but I’d really like an answer, tho. The article is not very clear about this.
Because it’s original work they contributed for free. Lending others that kind of expertise and time, just that it get’s used by a machine learning algorithm, which aims to reproduce this, without giving it back to them or the community in a similar free manner, feels violating.
Apart from that, creators feel ownership over their content and it feels wrong not to be asked what happens to it. (Although those probably wouldn’t – or shouldn’t – use SO anyway, as their content gets commercialised anyway by giving it SO for free.)Aaaaah okay that is somehting I can understand, thank you!
I get the feelings, but once they added the content in SO, it was no longer their content.
it was no longer their content.
all posts were supposedly licensed under creative commons
Do they have the non-commercial/no-derivative clause?
It gives worse answers and hallucinates a lot, problems specific to the model’s design and not how much training data it gets. Even if it did work, it would be taking jobs away from actual programmers. It’s a total net negative, users have nothing to benefit from this. Plus, since this is a community of programmers, they’re all very much aware of these limitations and the lack of ethics of OpenAI.
Except from the “taking Jobs part” I can understand you, thank you :D
Do people still use stack overflow? I feel like their days are numbered.
Well, when good documentation is available it isn’t necessary but good documentation is not always available. And it can still be helpful with niche issues 🤷
PowerShell Microsoft Graph Module. Most of the official doc pages have at least one section of “TODO” in them.
After they officially deprecated the previous modules with similar functionality.
I mean, good documentation + LLM = Stack overflow is 100% obsolete
Why else would they speed run enshitification?
Nope, Documentation + LLM misses the big chunk what we call human errors, commonly x y problems, LLM are good until you need to get deep then they start giving surface level answer it’s possible to point them toward the right direction by refining but at that point i would prefer reading Documentation.
To make money. If user engagement drops, revenue drops. The existing content is the only real asset this company has.
Irc: I ain’t dead yet bitch! Discord is fucking awful for dev and search.
