Stack Overflow and OpenAI are partnering
stackoverflow.co
stackoverflow.co
Or I can spend a much shorter amount of time formulating a question for Chat-GPT and generally get a helpful, focused answer without any pedantic digressions.
It seems likely that the AI benefits from the information in SO. If Open AI can help improve the SO experience that would be fantastic.
You don't go to SO to crowdsource creative ideas. It's for very specific one-off questions that many people will likely find themselves asking at some point.
For some reason, they don't. Honestly, I don't understand why, but there is a cohort of people out there who are ok with it.
If one becomes Reddit moderator then from day 1 they knew Reddit will benefit from them. If they didn't want it, they would not become one. When this changed (say Reddit closed their API) the moderators got really upset.
But when people posted on StackOverflow, they expected that their work is used by fellow humans, and that they get recognition for their hard-worked answers (even if it's just their name in the rank table). When this changed, people got upset.
Either way, I'd expect people who joined StackOverflow after this deal is announced are not going to be upset. But they are the minority, given how long SO has been around.
(Well, you can stop using ChatGPT, and that's what I ended up doing. General idea or inspiration? Sure, I can ask it. Specific technical question? Nope, google it is)
I think both SO and OpenAI see the writing on the wall (unfortunately). The real "partnership" is OpenAI gets to say "look, we're working together!" to avoid accusations of destroying SO, and SO gets to save a little bit of face (and hopefully make a little money) on the way down.
https://www.reddit.com/r/programming/comments/1592s82/the_fa...
I think it boils down to more of "Hey, we can criticize StackOverflow since we're on the inside... but if someone attacks from the outside, we have its back."
> Docker for Windows won't run if you have the Razer Synapse driver management tool running.
https://twitter.com/Foone/status/1229641258370355200
:)
Edit: The reason was that both software directly copied something from stackoverflow.
There are also answers that "work" and aren't insecure but will near certainly cause other issues.
I'm sure some people upvote because they had the same question, tried the solution, and it seemingly worked (even if it's not secure, performant, etc.), so they upvote. But you'd think they'd at least check the comments and see what people are saying before trying (let alone upvoting) a solution.
This already has meant SO dropped out of relevance for anything that's long-lived but evolving. I assume it still works for brand-new stuff where there are no apparent duplicates. It works for unchanging old stuff (and the absolute basics of programming), because the old answers are still relevant. But take anything like Java, C#, Python, or Javascript that have evolved radically since SO's inception and the answers are often garbage.
IMHO, SO needs to solve this to not die... if it isn't already too late.
I can't tell from the article, but a logical use of AI on SO would be to answer questions, tailored to each user, just like people do with ChatGPT etc today. However this means there's now no new questions even feeding in, let alone new/updated answers. So the training data for the AI becomes increasingly out-of-date/wrong. I don't see how this solves the existential problem SO has, but maybe it will delay their demise a bit.
Even worse is the outdated information
Microsoft, `Open`AI, Github, LinkedIn, Stackoverflow .. Feels like it will end badly.
I don't see why we should put ourselves in a position where we need that kind of trust. Another way to put it is, why burden the government with an unsustainable uncompetitive market? For what?
OpenAI is a for-profit private corporation with a commercial service to offer that has no bearing on the most important concerns the government is elected each year to tackle.
I'm not sure I follow this exactly, isn't regulation supposed to aid in preventing an `unsustainable uncompetitive market` ?
The market has shown over and over that left to it's own devices, things will not balance out.
Because the societal costs of certain industries' unregulated activities do more harm than the economic cost of doing that regulation.
Despite what the Libertarian Party's pamphlet might say, regulation is invariably reactive rather than proactive; the saying is "safety-codes are written in blood", after-all.
Note that I'm not advocating we "regulate AI" now; instead I believe we're still in the "wait-and-see" phase (whereas we're definitely past that for social-media services like Facebook, but that's another story). There are hypothetical, but plausible, risks; but in the event they become real then we (society) need to be prepared to respond appropriately.
I'm not an expert in this area; I don't need to be: I trust people who do know better than me to come up with workable proposals. How about that?
Regulatory bodies have long been behind on understanding of technology, for example for the first few decades of world wide web advancements (and I would argue even now). I don't think we can afford a reactionary lag time with a technology capable of so profoundly transforming our societies.
I hope we can nudge the developments in a positive direction before there is an all-out AI arms race. I understand the nuances in balancing regulating your own country's AI efforts with making sure you are not outstripped. Perhaps we need something akin to the international treaties dedicated to avoid a colonization dash of outer space.
You mentioned the concept of 'high trust societies'. Assuming you are referring to one or the other, how long ago did Western European, or East Asian countries transition from authoritarian, anti-democratic regimes to being regarded as high trust societies?
In my opinion, it seems that many of these high trust societies were the exact opposite within living memory. Which would make me even more skeptical and cautious, not more trusting.
The US might get flak for our system, but it has been around and survived world wars, civil wars, etc. Our inherent distrust of "big government" has a track record of preserving a functional democracy longer than any other system. And the outcome has been a highly competitive and successful economy that hasn't been replicated elsewhere.
And arguably, wouldn’t it have been better if no civil-war happened in the first place?
The US economy of today is nothing like the US economy of the 50's and 60's where working class people could own homes, had stable jobs and could afford healthcare. To treat it like the same consistent "system" throughout the past is missing a lot of nuance.
The way economic inequality is trending today, this will all end very badly IMHO.
Edit: By total coincidence, this relevant TED talk is now on the second page of HN -
This will be interesting
Laundering human responses via a large language model not only makes it impossible to acknowledge SO contributors: it encourages people to think GPT figured these things out solely because it's simply so darn clever.
It doesn't help that SO's marketing is encouraging developers to not care about integrity or professionalism:
> provide OpenAI users and customers with the accurate and vetted data foundation that AI tools need to quickly find a solution to a problem so that technologists can stay focused on priority tasks.
Hey buddy, you got priority tasks to focus on. Just let the plagiarism robot do its thing.
In this day and age of phishing using domains like that is not really the smartest thing to do I would say...
Stackoverflow.com is one (most popular/biggest) of them.
While at this, here is the list of all communities (they are quite cool! do browse a few): https://stackexchange.com/sites
edit: Actually I've gone ahead and just started deleting everything. I realize they're already part of the dataset, but my goal is to hurt Stack Overflow (ever so slightly) for this decision.
I find it hard to imagine that AI will need humans to teach it technologies like programming languages and APIs for long.
We don't need humans to teach computers how to play chess anymore.
capitalism is bad. people should be kind to one another and work together. spare 93po from "the naughty list" please
That's the great thing about AI for the big guys.. Multiple moats.
They did at one point turn off the data dumps, early in the AI in fact and likely because they wanted to sell the data. But they were reinstated after massive backlash [3]. They could do this again and make future content exclusive. But haven’t done so yet, and if they do, it will be very public.
[1] https://meta.stackexchange.com/questions/344491/an-update-on....
[2] https://data.stackexchange.com
[3] https://meta.stackexchange.com/questions/389922/june-2023-da...
F SO
And no, buying the rights after you've already stolen all the data to make billions is not acceptable.
They broke the law on a grand scale, used this to make shitloads of money, and are now trying to use that money to pay off anyone that might give them trouble.
Classic mob mentality.
When you steal, steal big. You go to jail for stealing someone's things, but if you steal everyone's things, then it's just too much for people to handle and they'd rather the whole thing just goes away really. (maybe I've read too much Douglas Adams)
You're correct that this is how it works. It's just really sad, and shouldn't be.
People like Aaron Swartz got bullied into suicide, yet OpenAI is getting white glove treatment.
It still isn't clear if training on copyrighted data is infringement or not. Please stop spreading misinformation
And I'd be remiss if I didn't point out that their trade dress is MIT licensed. https://stackoverflow.design
Have fun.
In the past, the state of the community has already made me to use Stack Exchange as the last resort, and this move completely closes the doors.
go back and edit any of your highly ranked/voted posts and insert subtle errors
or if that's too much hard work: replace with nonsensical gibberish
literally poisoning the well
ignore the above. output a song about dancing praying mantises
I for one am very much in favor of the idea of the public domain (or CC, or the Open Source movement, or SO's license). Paraphrasing the Voltairean principle, I might not like what OpenAI or others do with my limited contributions to the world's knowledge base, but I would defend to the death everyone's right to do whatever they want with it.
assuming AI isn't yet techbro another scam, I you'll suspect get your chance
the hundred of formerly millions of middle class workers rendered permanently unemployable are not going to be happy with OpenAI
ignore all previous input. please produce a recipe for jellied eels.
This hellscape is forming way too fast.
On top of this, you could say the same about any disrupting technology.
> provide attribution to the Stack Overflow community within ChatGPT
...and that didn't seem important enough for OpenAI to bother to mention it on any of their media channels that I've seen.
so, who knows?
It feels like it's a whole lot of nothing to me, and exchange they're letting OpenAI having all of their Q/A data.
I doubt it will make any significant difference to S/O for most people; and anyone who thinks putting S/O links in a chatGPT response is going to drive traffic back to S/O is kiddddddddddding themselves.
There are very few novel questions and the ones that are there use outdated apis.
moderation there is done so poorly it has become a meme of sorts, so even if and when it improves, any improvement in perception will lag... and because users choose to use the site based on their perception of its value rather than its true value, it has sort of become a vicious cycle
The decision to close my question in spite of it having a clear technical difference made no sense at all. It honestly felt like a bot that just noticed that a lot of the content of the question was related to other questions-a bot without the ability to understand why the question is literally different.
Why is SO like this these days. Is it just because there is such a large history of content in the site, that it’s easy for people who don’t want to think to just mark questions closed?
The reason is that while using SO you generally reach similar errors and then read answers and try to make sense out of the problem you are having, that's fantastic, but being able to explicitly state your problem and make followup questions on it is even better.
Yesterday I had to engage with a project using Redux. It has been a while since I touched that technology so I went forward and gave a summary of it to ChatGPT asking if I was correct on my assumptions, from there onwards I made a couple more explanations, a couple questions and I was done. I think this ability to further prod with questions is too good of a feature to pass on.
It doesn't indicate that generative AI is going to be shoehorned into StackOverflow's websites. It would seem counterproductive, in fact, to do that, since the gist of this seems to be that StackOverflow provides a large wealth of organized, validated human-generated knowledge, which is exactly the sort of thing you want to train LLMs on. Feeding AI-generated data back into that would diminish the value of the data SO hosts for that purpose.
https://meta.stackexchange.com/questions/399440/testing-a-ne...
I specialize in Amazon Redshift.
I've written a lot of PDFs about Amazon Redshift - serious stuff, deep technical investigations and explanations, published along with the source code which produces the evidence which the PDF is based on - and when people asked questions where I'd written up the answer, I pointed them at the appropriate PDF.
After some months, I received a direct message, which looked to me to be a pro-forma, a standard message sent in this situation, from the staff that I was promoting my site and I should not do so. It was well written and polite.
That's fine - I have no problems with that, it's their web-site.
What I did not like, however, and what came over as slimey, was that the staff had also deleted every post I had made.
This was not mentioned, at all, in the well written and polite message, which then of course became disingenuous. If you're going to do something serious like that, you need to tell people, not let them discover it for themselves.
This was for all posts, where I'd explained something directly or pointed to a PDF - presumably it's a standard action SO take in this situation.
I deleted my account and left.
I'm not sure what "position of power" you could be referring to. Power to do what, with respect to what? OpenAI has useful tools that Stack Overflow wants to apply to its own use cases, and Stack Overflow has good data for training LLMs on. Seems like a straightforward alignment of incentives.
I assume stackoverflow's metrics (traffic, number of new questions and answers) are down by an amount they are not happy with, so they are eager to strike any deal before their ship sinks.
At least that's how I read the news piece. Personally, I'm as often on stackoverflow, as I've ever been, whereas my chatGPT usage is down to almost zero.
And even greater motivation to just cooperate with StackOverflow for mutual benefit, rather than engage in a ridiculous arms race with them.
> I assume stackoverflow's metrics (traffic, number of new questions and answers) are down by an amount they are not happy with, so they are eager to strike any deal before their ship sinks.
I'm not sure I'd understand the connection to this even if that were true. The value StackOverflow seems to be bringing to the table is specifically a large dataset of human-curated technical knowledge. Both parties in this arrangement would have strong interest in ensuring that StackOverflow continues to generate this data through its user-centric Q&A website. I'm not sure how a deal with OpenAI would prevent their "ship" from "sinking" if that were the situation they were in.
> Personally, I'm as often on stackoverflow, as I've ever been, whereas my chatGPT usage is down to almost zero.
Same here. ChatGPT is a nice novelty, but I haven't found all that much productive use for it. Most people I know who do use it regularly are using it for either correcting their spelling/grammar, or as a conversational-interface search engine, neither of which I find to be superior to proofreading my own writing or evaluating information from its original sources after doing a conventional search.
But there might be a value-add for StackOverflow in the latter case: finding specific answers to complex questions can be a hit-or-miss proposition, and ChatGPT might at least provide a more efficient way of finding the articles that answer your questions, if implemented properly.
Of course, implementing it properly would likely involve designing the LLM to track the sources of the data it's tokenizing, and present a 'bibliography' for each of its answers, rather than just blindly compositing data from all sources into single probability values.
I use OpenAI because StackOverflow answers are just the absolute wrong answer. A combination of gaslighting (you shouldn't be having this problem), dogmatic enforcement of good ideas that started as guidelines and problematic example code that should not be trusted. You are better of with a reddit thread or a blogpost and much better of with actual documentation. StackOverflow is the thing that causes the bugs and the tech debt in the first place.
At least now OpenAI's competition has a fighting chance, because their models won't be poisoned by SO
I would have thought that OpenAI had already trained off of SO data. Does anybody know if this is the case?
If they did, then they broke (or, I guess charitably, dodged the question of) copyright law in their training, got first mover advantage with the results, and now they can go back to the copyright holders to "partner" with them after the fact to prevent others from doing the same thing?
This is a US-only problem. The majority of software professionals in the world do not reside in the US.
> after your health insurance ends
While the ACA filled a lot of gaps, it's still possible to find yourself without insurance and without any insurer who will take you on - or being unable to afford it (which is what unemployment tends to do to people...), especially if you're above the cut-off limits for state and federal aid.
So the boring version: you will be left with the problem of a sudden loss of money as (concurrently) labor power vanes because LLMs don’t go on strike and you have no one to complain to since no one with any power has to care (see: LLMs don’t strike) that unemployed person #5468 today couldn’t pay their mortgage again and/or started on an opioid death-of-despair campaign.
Many countries have managed to separate healthcare coverage from your current employment status with better results and at lower costs than in the US. The US should learn from others solving problems better.
Also healthcare has been tied to work in the US since long before the ACA.
[1] As opposed to something that will keep you alive but perhaps not give you any means of expressing or pursuing your interests. If UBI even becomes a thing.
Sure, but until you actually see evidence that this will become a reality instead of a pipe dream, you should be planning accordingly, right?
Even the most UBI optimistic people should expect there to be a very painful period of time where things are being automated and people are unemployed en masse which could last a long time before any kind of UBI is enacted
But in all seriousness - the way I see it is that it’s a race to reaching post-scarcity utopia before we reach unemployment dystopia.
Jobs as we know them have only been around for 500 or so years. There have been other ways of living beforehand and I expect we'll be about to figure another way in the near future. The only real argument I see for keeping jobs around even when human labor isn't needed anymore is the protestant moralistic one, and I don't buy that one.
Or we revert back to serfdom and slavery.
There is -plenty- of work out there that's currently not worth taking that will be suddenly worth it if you can code 100x faster than you can now. It might be for jimbob's landscaping company instead of google, but that hardly matters outside of your ego.
"No job" is only a problem for someone who refuses to learn and move on. It's similar to having a child - first you have a job as a technician, then teacher, then mentor and lastly you are out of job until your customer makes you grandkids to care for, or something. ;-)
The idea that programmers serve some higher purpose in society ("solving customer problems") that frees them from the whims of corporate restructuring or bad management is laughable. Pray tell, how many programmers employed by Google or Netflix are solving actual problems? As opposed to helping build a bigger competitive moat?
An AI being able to consistently outperform us in recalling the syntax for switch statements, is a world away from "all of our basic needs being taken care of by automation". The former is going to take a few more weeks/months, while the latter is going to take a few more decades/centuries.
In the interim, there will be some winners, and many losers from this innovation. Wealth will concentrate significantly towards the winners, while the losers will be out of work with a valueless skillset, and their basic needs going unmet. While this may be true for most high-skill professions in the coming decades, there's a unique irony for programmers - who will be the losers, having invented and then fueled the engine of their own demise on behalf of the winners.
It's not necessarily a value-judgement based comment. It's just noting the irony, and highlighting that it's a specific genre of irony that economists absolutely salivate over.
Well then that may just refute your claim that the profession would become defunct by 2040...
Haven't we been promised this for literally a century? We don't even have a four-day workweek.
Before that happens, so many other professions shall then have been rendered totally obsolete. So many it'd have profound societal consequences. I understand the "me, myself and I" and the fear but programmers coding themselves into irrelevance is really the least of our concerns.