Stack Overflow simply bans folks who don't want their advice used to train AI Stack Overflow users are revolting against the Q&A site's partnership with OpenAI, announcing they'd rather remove their posts and sacrifice their reputation scores than have their submissions used to train ChatGPT. The outcry comes after it was announced earlier this week that OpenAI would soon be able to train AI models …
Same. Honestly just screw these guys, they seem to have forgotten that the entire Stack product is built on the back of countless hours of free work contributed by users. Giving those contributors the finger and claiming all the content as their own property, why would anyone contribute to that mess anymore.
I say: screw everyone. This is the only forum where you'll find me, the rest can DIAF. I know that all these websites try to own you once you join, and are totalitarian in their decisions about how you interface with them even if that decision is against their stated policies when you joined. They all treat themselves as little fiefdoms and, after just one or two exposures to them, I decided a forever "No, thank you!".
I'd rather be quietly anonymous than controlled like a puppet.
Tangentially related, 2 days ago, a comment I made on youtube was deleted. Thinking it was an error, I reposed. It appeared, then disappeared.
The comment was something trivial talking about someone interviewed on the Graham Norton show, and my comment was as innocent as "I agree. His performance was awful. He went on for far too long" (paraphrased)
Being the stubborn git that I am, I tried reposting various combinations. What worked? Replacing "awful" with "or full". I even tried using "awful" in another comment, and the same thing happened.
They seem to have a gradient naughty-step that I'm at the top of. Nothing I ever post on political channels with the words "Palestine", "Israel", "killing", "bombing", "Hamas" and many others EVER get through - despite them being used in the posts I'm replying to.
WTF? The contempt youtube has for users is terrible - deleting messages for no reason, or even worse - sometimes shadow-banning the comment. And all done quietly and invisibly.
And yet, you go look at comments for something like Daily Dose of Internet, a huge number of them are a bot posting the exact same thing over and over. You'd have thought it would be easy to implement "three identical messages = messages discarded" but clearly not.
On the topic of shadow banning, I've had comments appear in my notification bar that when I tap them and go to the app, they simply don't exist. Need to go to the website to find them (usually, sometimes they just don't appear until somebody else replies).
So while I'm quite sure there are certain trigger words that can bring in some sort of filter as you describe, I think a big part of the problem is that parts of it (the backend? the app?) are written by arsehats that don't know what they're doing.
Case in point, I just discovered that if I feed a .ics calendar file created by one of my programs to the Google Calendar app, all of the scheduled notifications (like five minutes before the event) are discarded, only the events themselves are imported. If I go to calendars.google.com, switch to the desktop version, and then import the exact same file, all of the notifications are present. Don'tcha just love the consistency?
> On the topic of shadow banning, I've had comments appear in my notification bar that when I tap them and go to the app, they simply don't exist. Need to go to the website to find them (usually, sometimes they just don't appear until somebody else replies)
Seen this on some Wordpress sites, mainly with my own contributions.
It’s almost like the notification feed is reading the submission feed, before the submission feed has been reviewed as this happens more frequently when moderation has been enabled and thus contributions need to pass through (human) review before being made public.
Ahhh,. I've also had notifications for "ghost" messages, but never thought to check via the website. I'll try that in future.
I do sometimes use the website in incognito mode to check if my posts got through, but mostly these days I don't bother.
And yes, it's the android app that I use.
So, you're probably right. Their programmers are too busy making the app shittier, removing config. options, and changing things pointlessly to care about DB integrity!
Worst for that, though, is the Play Store.... Adding extra hoops to get to your apps a year or so ago, to lately removing mouse scrolling in certain pages, to their god awful split-screen (with one half of the screen being irrelevant) change for desktop use.
And the "shorts" abomination... "Let's copy tiktok, including all the crappy things that it does.. So let's disable landscape, disable video seeking, and pump this crap onto users feeds whether they like it or not!"
And they wonder why people flee to third-party apps....
"Giving those contributors the finger and claiming all the content as their own property, why would anyone contribute to that mess anymore."
While I agree with you, those very same Ts&Cs, or words very much like them have been on public fora for about as long as I can remember. Both AOL and Geocities claimed ownership of everything posted or uploaded to their fora and hosting services. It's just that in most cases, until AI training came up, there often wasn't really any way to monetise the content, most of the effort going into monetising the users and the data through advertising or selling to advertisers.
researchgate dot net A-Time-series-of-weekly-posts-to-Stack-Overflow-since-early-2016-The-number-of-weekly_fig1_372404443
There are a multiple factors to unwrap. One is that the SE answer space was already getting saturated by 2016. The other is AI, improved copilot, etc.
I was surprised to see, last year, that CEO AK's salary was 500K - pretty low for a CEO.
Like most C levels at web companies, Stack succeeded due luck in timing, filling a void in the online space and a strong response from the community. I can think of zero genius or visionary managerial decisions they have made. Like Jimbo Wales, over hyped and over paid frauds.
The real mistake people keep making over and over is not owning a real stake in communities they put so much work into, and come to rely on. All those TikTok influencers may learn the same lesson shortly. In a more healthy word, we'd all spin up a co-op, clone the non-toxic parts of the Stack community and ad a migration tool to let people re-post their gems on a new community owned site.
Problem is half of you are flipping idiots and wouldn't move(because your still on Facebook, after decades of clear evidence your feeding a monster) and the other half wouldn't make it a week before getting butt hurt and trying to start your own version of it that's exactly the same but YOU get to decide who to kick off for being a moron.
Feel free to prove me wrong. Really. I've seldom wanted to be proven wrong this much. At least about the spinning up a Stack replacement. The CEO pay thing is at best a cautionary tale. Paying them what they are worth won't fix much.
The "you all" logical fallacy incorrectly assumes everyone but you is a hivemind capable of spontaneous unanimous decisions. And you, incapable of understanding why this imaginary "everyone but me" entity does the things that it does, causes you to project yourself on to this bogeyman and assume it's a singular entity incapable of making rational decisions.
No, in actuality, the social sciences and game theory behind why a herd animal acts the way it does is far more complex than you either can understand or want to understand. There's a million reasons why everyone does the things they do, but by and large, it's a chicken-and-egg problem. You can't go anywhere else because everywhere else doesn't have anyone, and nowhere else has anyone because they don't want to go where nobody is. Only by freak accidents and pioneering efforts does anywhere get populated. This is not new, this is how settlements form, this is how we humans survived and evolved beyond our Neanderthal brethren (who, while as intelligent if not moreso than us, did not want to join us in traversing the unknown seas in a possible suicide mission just to see what was out there). Your problem is that you yourself know you're just as afraid of colonizing as your friends are, and so you project the self-loathing you've developed from your dependency on your fellow man on to this "collective" you've dreamt up, believing if you can just say the right magic words to this imaginary friend you've made, then they'll all go to Nirvana with you and everything will be right with the world.
I really recommend you not be so emotionally invested in online forums.
"and nowhere else has anyone because they don't want to go where nobody is"
Case in point - Twitter is a massive steaming pile of excrement that is now actively excluding people like me who don't have an account (so yay for companies still posting announcements there, some of use just cannot read them), and yet...
...what happened to the Fediverse?
Elon taking the ruins [1] should have been the watershed moment, and Mastodon was a big thing for a few weeks, but it being too different, less centralised [2], and general apathy mean that, well, loads of people are going down with the sinking ship.
1 - I intended to swipe "reins" but my phone thought I meant "ruins". I like that option more so I left it.
2 - Yes, that's the point, but you try explaining that to a narcissistic brain farter.
While the content comes from other users, I think the company did a great job in creating the platform that made it fairly easy to ask and answer questions, way better than any "knowledgebase" system I've ever seen. It was clear from the start that you're assigning the copyright to the company but they're also keeping it public and I think most contributors are happy with this: if we wanted to assert copyright, we would. We're often pretty good, or even expert, in some areas and novices and idiots elsewhere. I've benefitted enormously from the feeling that I'm not the only one who can't see the wood for the trees and this is far more important than the gamification with "reputation". I have an ad-blocker and have never seen an ad but also never anything try to force me to drop it.
I will admit to growing grumpier over time to posters who clearly don't follow the rules and don't care about the documentation you may have spent considerable time writing. If these guys, often working for sweatshops in South Asia, get their answers from a chatbot in the future, then I'll be happy.
I totally disagree. The few times I've tried to post there (proper answers that correct answers with mistakes, some code fixes etc), I get "you don't have enough reputation points to comment"
I've since given up.
WTF? Are you expected to post lots of "fluffy" questions just to be able to offer valuable advice?
The thing is, any current content is only valuable to answer existing questions. Let us say that Stack overflow trains a robot capable of correctly answering every question in the existing database. It's only going to know how to deal with existing issues.
But how is it going to deal with a new set of issues a month after the existing community has been alienated to the point of leaving?
The only conclusion should be that value to humans is derived from being able to consult a community of knowledgeable thinking humans. If people wanted to read a dry computer generated statement of what something is supposed to do (rather than what it actually is doing) then they'd be reading the support documentation.
By now surely everyone remotely awake assumes that every word they tap into a web site is immediately and irrevocably out of their control.
Think before you type and, sadly, don't ever assume that web site owner will behave with honour.
Why, even at The Register there been cases wher.. x.&*/......
no carrier
Well if you had any residual trust in online services, this is a wake up call.
I remember when StackOverflow was in Beta, it was great.
Then there was that statement by one of the developers that they might open source it. The prevarication made it obvious, a lie.
Then you realised the game was rigged. You started an item, shortly changed your mind, couldn't delete it. Somebody had inserted a single irrelevant comment. Yep you realised, the design is pretty pure evil. Automated theft of your contribution.
After that there was a wave of "brownie point hunters" looking for upvotes irrespective that they had no idea what the question meant. (What kind of person is turned on by silly little badges? Dumb scores?)
Then came the organised gangs. I saw wrong answers top voted and right answers downvoted and "amateur lawyered" to death.
The huge numbers of outdated, and now wrong, answers to questions. The good questions unanswered. ...
Then it got sold out to enemies of us all.
Enshittification, upon enshittification. The course to the heat death of the Internet Universe. He who can't see it deserves what he'll get.
Automated theft of your contribution.
You have to sign up, it's not theft. You pro-actively have to submit your details to Stack Overflow and then pro-actively type in the text of the answers.
One of the lessons people won't learn from this is to read the T&Cs.
(Put it another way... very little, if anything, in this world is free.. if you're not paying with money, you're paying with something else. In Stack Overflows case it's your knowledge).
You would have gone through the same process before posting on this site. Would you be pleased if the Reg decided it owned your posts, and could do with them what they wanted?
You recently posted a funny jibe about range rover drivers (which I upvoted) - If used in certain contexts, that could be taken badly.
"You would have gone through the same process before posting on this site. Would you be pleased if the Reg decided it owned your posts, and could do with them what they wanted?"
I wouldn't really mind. They don't, by the way, because the operative part of the terms (specifically 8.2) is a lot tighter on El Reg than Stack Overflow's terms. However, I've already decided to write the posts I make here and give them away for free, to the extent that anyone values them anyway. My feelings on how much I should own them are very different from something I've written elsewhere. While I have defended and will continue to defend copyright owners against LLM companies beliefs that their business model supersedes the law, I am less concerned when a real legal agreement exists and is valid at all levels. Those who posted on Stack Overflow did so for a reason and knew they weren't going to be paid for making those posts. They can have and express any views they have about the decision, but I won't mind my few answers before I abandoned the site being used for this purpose.
Fair enough, and actually the above examples of use wouldn't bother me. I was just replying to the comment that it's fine that they can do what the hell they like with our posts.
One difference, though, to The Register - you probably used stack overflow with your real name..
> One of the lessons people won't learn from this is to read the T&Cs.
One tiny issue with the T&C's and EULA's if you read and understood them all fully, including the all updated revisions, it would probably take 2.5 lifetimes just to do that and that alone.
Unfortunately many, many human beings.
Years ago I was on the evaluation committee for some open positions in a local university. To facilitate the evaluation, we do an interview and ask for proof of the highlights of the CVs of the candidates.
One of those had a screenshot of a technical website's forum showing his user name as one of the top posters as if it was a scientific award or something.
And one of the committee members said it was impressive....
That depends on what you were doing with it. If it was something where they were answering people's questions, and you were hiring for a teaching position, that could prove that they have a passion for teaching and, if you read the posts, you could get an idea of how well they teach. As jobs go, a teaching position is probably one of the ones where SO posts are most relevant. That doesn't mean that it's necessarily useful enough to consider, but including it may not be as daft as you imply.
End of an era, I guess.
To be honest, SO was never good. It was only good for C# (the C# responses are invaluable and extremely in-depth, written by actual geniuses), every other language's community is either horrible and completely unhelpful, (Python answers are all written by children, every JS answer is "Just Use JQuery™", C++ users are violent animals who would rather burn you at the stake for misunderstanding their glorious language, etc.) or you usually flat-out don't need them (like how virtually everything in Rust is uber-documented, therefor SO questions/answers are mostly redundant and just say what the docs say). Maybe AI is going to raise a generation of people who deliberately do the wrong thing out of misinformation and hallucinations, but SO was never too helpful to begin with, either.
And there are people who do it for fun, fame, humanity, or pure evil.
Where are you standing yourself? Are you throwing plastics into the ocean, or rather trying to reduce your ecological footprint?
This is why I am worried about Linus Torwalds and other project owners. Once not with us, a new owner might have completely different motivation or vision. It is still puzzling that a new AlphaFold article has been published behind the Nature paywall instead of sharing it free. So much great knowledge is collecting dust (and money) instead of benefiting us all.
Maybe Torwalds and others should create own AI-clones, with their personal moral values embedded, then driving their projects forever.
Honestly, Linux is owned by corporations at this point. Primarily Red Hat, but lots of companies, including Microsoft, and especially Google (read: Android), are maintaining it now. Linus is a figure head and gatekeeps some dumb ideas, but it's pretty much become it's own living thing by this point that nobody can rightly control. I assume after Linus passes, Red Hat & Friends are going to keep maintaining it, people will maintain forks of it that cherry-pick patches (like Linux-libre but more arbitrary forks for the sake of it), and of course other OSes will start to rise to prominence to replace the old software (if you think Linux is going to be around forever, I applaud your reverence to it's legacy, but I do feel you're being a bit naive). Redox OS is a good contender, but, whatever the next big thing really is, is just something we'll have to see.
Outside of that, replacements are surprisingly easy to come by. I imagine there was a time when everyone thought NGINX would be *the* web server, here forever to serve all your serving needs. And now Rust exists and you can literally just make your very own server out of a couple crates which outperform it. I'm sure something will replace SQLite once it's owner passes. I mean, there are *tons* of KVS databases out there, and the concept of SQL itself is starting to become outdated (if not a security vulnerability).
Things change, people invent new solutions and move on. We're not trapped in a morass of technical debt. If anything, each generation of technology brings with it more standards, more intention, more paths to upgrade. Things are more replaceable than ever.
Presumably https://www.nature.com/articles/s41586-024-07487-w. That's the paper announcing AlphaFold 3. Haven't read it myself, just seen some of the summaries.
I generally prefer journals associated with professional societies over commercial publishers. Further, I'm not a fan of Nature (for quite range of reasons), but they are a business, and so like to make money. They are entitled to charge for access, if that is their business model.
Most likely you should complain to the authors (and/or whoever they work for), who declined to pay the "open access" fee, and so left it behind the "Nature paywall". And even if they were unable to pay the arguably substantial fee charged by Nature, then there were alternative avenues for open publication.
Of course, it is hard to say much more, since it is unclear which article is being referred to (I had a look, and there are several AlphaFold-themed articles in the Nature family of journals that might be considered "new"). See:
https://www.nature.com/search?q=alphafold&journal=
Maybe Torwalds [sic] and others should create own AI-clones, with their personal moral values embedded, then driving their projects forever.
An AI with "personal moral values", eh? Is it just me or is that the dumbest goddammed idea I've heard all week?
I just can't wait for the inevitable AI with the "personal moral values" of t'pineapple....
I don't know what other ideas you've heard, but there's a significant branch of alignment research looking at embedding human values in models.
I don't hold out much hope for it, since 1) we have no good proposals yet for eliciting such values, 2) we already know they're highly contested and inconsistent, and 3) millennia of philosophical thought on ethics has yet to produce definitions of them that are even very useful, much less pragmatically instrumental. But that isn't stopping people from trying.
See e.g. Singh.
This post has been deleted by its author
People who don't want their answers to be SOLD by Stack Overflow to become part of some AI's training, have a right to do what they want with their answers and reputations. If Stack Overflow allowed people to opt out of that, or delete their content, this wouldn't be an issue.
Whether a mostly right answer is better or worse than one that no longer exists because it is deleted is left up to the reader. I feel that someone who is using it to learn will have an opportunity to learn and figure out what was wrong with the answer they found. Someone who is using it to cheat at homework or avoid having to do real work on their job will get their just desserts.
"You agree that any and all content … that you provide … is perpetually and irrevocably licensed to Stack Overflow on a worldwide, royalty-free, non-exclusive basis"
Many decades ago, while a student, I was an anonymous extra in a crowd for some movie or other (I forget what). But I do remember that I had to sign a release form entitling the movie company to use my persona for any purpose "in perpetuity throughout the universe."
That's what happens when lawyers turn up.
From this very site: "You retain all your ownership, copyright and other interests and rights in your comments but by posting any comments on our Website you grant us a non-exclusive irrevocable and royalty free worldwide licence to use, modify, alter, edit copy, reproduce, display, make compilations of and distribute such comments throughout our Website."
On two occasions I had a perplexing programming problem. I asked everybody I knew for help. Nobody had an answer. I went on StackOverflow and, following their advice, first searched their database for a solution. None.
So both times I posted my problem on their site. Both times I was ridiculed for posting such a problem. They said the answer was readily available.
Nice bunch of chaps, those guys at StackOverflow!
It's even better than that... anything slightly esoteric that's had this treatment will bubble to the top of search results, so anyone else with the same issue will find your post and all the unhelpful replies, but never actually find a link to the answer.
Starts out starry-eyed and full of hope, only to be crushed by the weight of gamified stats and overloaded moderators.
The oldest posts are generally good, gradually reducing in utility until it becomes a sea of slime.
Except of course, it's worse for StackOverflow because even old answers become wrong.
One of my most highly voted answers included "at the time of writing, this is not documented". About five years later someone replied "No, it's in the documentation" - using a rather ruder form of words.
Well yes. That'd be because I got the documentation corrected.
While that's a trivial case, there are many old answers that have now become very, even dangerously wrong.
Instead of leaving the platform altogether, how about crafting answers deliberately to be meaningful only to humans with a sense of context, and poison AI systems?
A human being will have an easier time disentangling instructions to request persistent storage in a mobile browser from titles of songs by 1980s bands than a machine .....
Could the right to be forgotten be implemented by making technical responses anonymous instead of deleting the entire bit of knowledge? Of course, the credibility might be decreased, but at least the knowledge might be retained? (The credibility might be implemented by a poster credibility score totally replacing the name of the poster?)
While you can assign the economic rights in your posts to Stack Overflow, you can't assign your moral rights, and the main one here is the right to attribution, the right to be identified as the author of the work.
In the past, Stack Overflow have complied with that by putting your username next to the text in question, and for many people, that is what motivated them to contribute it in the first place.
That how I would attack if I had ever contributed to Stack Overflow.
I'm assuming people contribute to this website to increase the sum of human knowledge. Adding it to AIs does this. Are they pissed because the AI companies are making money off it? People make money off information posted on the net all the time. I submit papers to Academia in the hope that they will be helpful. I'm OK with people making money off them, incidental to building good tech.
We have hypertext tags to prevent search engines scraping content (making it harder to find, which is general pain in the arse). We can do the same with a 'no AI' tag in anything we post or on any web page. Ultimately, it is a bit daft. If you post something online you presumably want people to benefit from it. Does it matter if they benefit from it by reading it or by their AI regurgitating it for them?
So maybe everyone needs to calm down a bit, read the T's and C's, and promote a universal 'no AI' tag for those who only want to share their thoughts a bit, rather than as much as possible.
I wonder what El Reg's T's & C's are with regard to all this. Do any of you check before posting?
"and promote a universal 'no AI' tag"
What, like the "do not track me" header that most places simply ignore?
Scum are scum and they'll do whatever they want until the judicial penalty is enough to cause a rethink.
Oh, and you have it flat out wrong by the way. We should be promoting an "AI okay" tag and treating all existing material as off limits until the original author says otherwise. But, then, how could these AI hypsters make their megabucks ripping off everything in sight if they had to, you know, respect the same sorts of copyright rules that mean I can't quote one single line of an old song in something I write without jumping through a lot of hoops...
Better, then, to just blatantly steal and then try to bullshit us that it's not really stealing and anyway AI is the hot new thing and if it can't rip off everything in sight then this place will become a tech backwater, blah blah blah.
on the parts of SE/SO I used to frequent, there was one egregiously officious "person" (very hard to stay polite but I made it, yaay!) called Schroder (sp?) whose behaviour was so high-handed that -- after a few incidents, each of them minor and inconsequential by itself but not when taken as a pattern -- I deleted my account, deleted as many as I could of my posts and comments, and never looked back.
Perhaps because if somebody goes and reads your missive, they will read what you wrote, which assuming you're being helpful will be a relevant and useful contribution to resolve a problem.
As opposed to something spat out by a machine that "sort of" knows the context of the question and "sort of" has access to an answer, and if it doesn't (on either count) then it is prone to simply making shit up and passing it off as correct.
I have also banned people, that go deleting their posts to ruin context. That's a problem with butthurt forum users. That's what edit time limits are for and disallowing deletion. I'd let users edit their posts indefinitely, but I can always restore edit history that way if they are doing it maliciously or thinking they are going to make others look stupid by removing the posts they are responding to.
I wonder: in some countries such as mine, France, authors have a "moral right" («droit moral») on their creation. This right cannot be revoked, and allow them to have a say in how it's used, up to removing it, no matter how it's been licensed.
Could it apply for those users in such jurisdictions?
Unfortunately when you sign up, you agree to the exclusive legal jurisdiction of Fuckyoustan. That's probably not even remotely legitimate in other countries, but they don't care...
Quote: Both you and Stack Overflow hereby irrevocably agree to the sole and exclusive personal jurisdiction of the Courts of the State of New York with respect to any action, suit, or proceeding brought by it or against it by the other party in connection with the Network or Services.
I contributed knowing very well - never having read the small print - there was a dark side to that site: if you found a solution to a problem elsewhere on the net, SO wasn't happy with a link to that solution. Always someone reminded you you needed to copy (!) the solution into SO. Never mind any copyright or that the www's intention was to link, and not to duplicate.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
