-431

For posts following this one, see:

Today, we released a blog post from Stack Overflow CEO Prashanth Chandrasekar about AI and its intersection with communities. Because of the timely nature and the content of it, I’m reproducing it here for comments or questions from community members. You can find the original blog post here.

Senior members of the company staff will be reviewing everything here and will respond to questions, though it may be 24 hours or so before answers are ready, just because of the amount of interest that this is likely to get. Please leave any questions that you have below as individual answers.

Community is the future of AI

By Prashanth Chandrasekar

Throughout history, great thinkers have made predictions about how new technology would reshape the way in which humans work and live. With every paradigm shift, some jobs grow, some change, and some are lost. John Maynard Keynes wrote in 1930 that new technology meant humans would be working 30 hours a week or less, and that the main challenge would be what to do with all our free time. So far, predictions of this nature haven’t exactly come true. As new technology empowers us, we push ourselves to new heights and reach for previously unattainable goals.

Over nearly 15 years, Stack Overflow has built the largest online community for coders to exchange knowledge, a place where anyone with an internet connection can ask or answer questions, free of charge, and learn from their peers. Stack Overflow for Teams, our enterprise SaaS product, is trusted by over 15,000 organizations to serve as their internal knowledge bases. With the recent advent of dramatically improved artificial intelligence, many industries are wondering how technologies like ChatGPT will change their business. For software development, the answer seems more immediate than most. Even before the latest wave of AI, a third of the code being written on popular code repositories was authored by an AI assistant. 

Today, sophisticated chatbots, built on top of cutting edge large language models (LLM), can write functional code for a website based on nothing more than a photo of a rough sketch drawn on a napkin. They can answer complex queries about how to build apps, help users to debug errors, and translate between different languages and frameworks in minutes. At Stack Overflow, we’ve had to sit down and ask ourselves some hard questions. What role do we have in the software community when users can ask a chatbot for help as easily as they can another person? How can our business adapt so that we continue to empower technologists to learn, share, and grow?

It’s worth reflecting on an important property of technological progress. The Jevons Paradox shows us that, as innovation allows us to do more, we settle on a new normal, moving the goal posts for what we expect of people and organizations, then competing to see who can find new ways to pull ahead of the pack. For knowledge work, as the cost of an action diminishes, we often do more of it. Abstracting away repetitive or tedious tasks frees technologists up to make new discoveries or progress innovation.

If new AI systems make it possible to create software simply by chatting with a computer, my prediction is that, far from the job of programmer disappearing, we’ll end up with millions of new software developers, as workers from fields like finance, education, and art begin making use of AI-powered tools that were previously inaccessible to them. We are enthusiastic about welcoming this next generation of developers and technologists, providing them with a community and with solutions, just as we have for the last 15 years. We’ve got a dedicated team working on adding GenAI to Stack Overflow and Stack Overflow for Teams and will have some exciting news to share this summer.

I’ve heard similar sentiments expressed recently by Microsoft founder Bill Gates, by Geoff Hinton, the godfather of the neural network approach that produced today’s AI revolution, and by Stephen Wolfram, a pioneer across computer science and mathematics. Each sees in today’s AI the potential for the loss of certain jobs, yes, but also, if history is a guide, a future in which a great variety of more highly skilled work becomes available to an even larger group of people. Just as tractors made farmers more productive, we believe these new generative AI tools are something all developers will need to use if they want to remain competitive. Given that, we want to help democratize knowledge about these new AI technologies, ensuring that they are accessible to all, so that no developers are left behind. 

I talk to developers of varying experience levels all of the time, and I’ve been hearing anecdotes of novice programmers building simple web apps with the help of AI. Most of these stories, however, don’t begin and end with an AI prompt. Rather, the AI provides a starting point and some initial momentum, and the human does additional research and learning to finish the job. The AI can debug some errors, but is stymied by others. It can suggest a good backend service, but often can’t solve all the points of friction that arise when integrating different services. And of course, when a problem is the result not of instructions from a machine, but human error, the best answers come from other people who have experienced the same issues. 

For more experienced programmers, AI will be an amplifier of their existing skill, making them more ambitious in their projects. The result, as Jevons would predict, is that they spend more time with AI, but also more time creating new ideas, researching new topics, and asking new questions that had not occurred to them before. They feel empowered to reach farther beyond their traditional skillset and to push the boundaries in terms of the kind of work they want to take on.

We are excited about what we can bring to the fast moving arena of generative AI. One problem with modern LLM systems is that they will provide incorrect answers with the same confidence as correct ones, and will “hallucinate” facts and figures if they feel it fits the pattern of the answer a user seeks. Grounding our responses in the knowledge base of over 50 million asked and answered questions on Stack Overflow (and proprietary knowledge within Stack Overflow for Teams) helps users to understand the provenance of the code they hope to use. We want to help coders stay in the flow state, allowing them to create with the latest tools with the confidence that they will be able to document and understand the provenance, source, and context of the code being generated. 

Community and reputation will also continue to be core to our efforts. If AI models are powerful because they were trained on open source or publicly available code, we want to craft models that reward the users who contribute and keep the knowledge base we all rely on open and growing, ensuring we remain the top destination for knowledge on new technologies in the future.

AI systems are, at their core, built upon the vast wealth of human knowledge and experiences. They learn by training on data - for example open-source code and Stack Overflow Q&A. It is precisely this symbiotic relationship between humans and AI that ensures the ongoing relevance of community-driven platforms like Stack Overflow. Allowing AI models to train on the data developers have created over the years, but not sharing the data and learnings from those models with the public in return, would lead to a tragedy of the commons. It might be in the self-interest of each developer to simply turn to the AI for a quick answer, but unless we all continue contributing knowledge back to a shared, public platform, we risk a world in which knowledge is centralized inside the black box of AI models that require users to pay in order to access their services. 

As the AI landscape continues to evolve, the need for communities that can nurture, inform, and challenge these technologies becomes paramount. These platforms will not only offer the necessary guidance to refine AI algorithms and models but also serve as a space for healthy debate and exchange of ideas, fostering the spirit of innovation and pushing the boundaries of what AI can accomplish.

Our thesis on community as the center of a safe, productive, and open future for AI also offers some exciting prospects for our business. Stack Overflow for Teams, our enterprise, private version of Stack Overflow, helps to power a community-driven knowledge base inside of 15K+ organizations like Box, Microsoft, and Liberty Mutual. Decades of institutional knowledge, shaped and curated by subject matter experts and experienced teams, allows the employees at these organizations to more easily collaborate, improving productivity and trust. 

Incorporating generative AI technologies into the organizations using Stack Overflow for Teams will allow us to layer a conversational interface on top of this wealth of information. We believe this could lead to tremendous productivity gains: from new hires being able to onboard more quickly, to speed up developer workflows, as users are able to quickly ask questions and retrieve answers tapping into the company’s history, documentation and Q&A.

The example above is just one of many possible applications of GenAI to our Stack Overflow public platform and Stack Overflow for Teams, and they have energized everyone at our company. We’ll be working closely with our customers and community to find the right approach to this burgeoning new field and I’ve tasked a dedicated team to work full time on such GenAI applications. I’ll continue to share updates through channels such as my quarterly CEO blog, but I’ll be back in touch soon to announce something big on this topic. In the meantime, thank you to our community and customers for continuing to help us on our mission to empower the world to develop technology through collective knowledge.

Improve this question
55
  • 109
    Anyhow, my personal feeling is that AI is a huge wave washing the world these days. And it's not a good thing. It's going to cause enormous damage in the long run, not end of the world as some say, but damage that will take lots of years to fix. So while I understand the desire to jump on the worldwide AI wagon, I'm sad SE will also take part in it. :(
    – Shadow Wizard
    Commented Apr 17, 2023 at 15:18
  • 111
    How will the LLM be trained to ignore all of the out of date and plain wrong answers on the SE network?
    – DavidPostill
    Commented Apr 17, 2023 at 15:33
  • 25
    If you have questions, to ensure that they are seen and responded to, please add them as an answer, not just a comment.
    – Philippe StaffMod
    Commented Apr 17, 2023 at 15:44
  • 46
    You've probably heard this a lot of times before, but you'll get fewer downvotes than people unhappy with this, due to this site having its own pool of reputation.
    – Andreas condemns Israel
    Commented Apr 17, 2023 at 16:40
  • 222
    So let me get this straight; we've deleted tens of thousands of CGPT answers just to have another CGPT clone introduced to the site itself? You could at least take us out to dinner before you screw us that hard. CGPT already exists; what's the point in introducing a technology we've so far distanced ourselves from precisely over quality concerns, and when either the same tech or a better tech exists elsewhere on the internet?
    – Zoe - Save the data dump
    Commented Apr 17, 2023 at 17:41
  • 38
    @JourneymanGeek Based on the wording, it doesn't really matter where I point it out. If I write an answer, I'm going to have to expand it because answers are more serious than comments. However, this isn't a feedback post. Note how the question explicitly asks for questions that'll be responded to; this isn't a feedback post, it just looks like one. This is an early announcement, and no amount of negative feedback will change anything, because this isn't a feedback post. I'm not going to spend more time on something that will almost guaranteed end up being ignored or at the very least not
    – Zoe - Save the data dump
    Commented Apr 18, 2023 at 0:06
  • 71
    This seems very much in the same vein as the cryptocurrency craze of 2020/2021, just with GPT: everyone is doing it, so let's do it too! Stack Exchange does have a massive Q&A count in its 50 million Q&A... the problem is that some 30 million (if not more) are... low quality or duplicates. Stack Exchange content is probably a great corpus of data for training big models, for sure, but I'd caution Prashanth (if he ever visits Q&A and reads comments) to not overestimate the content the company is working with here (lest we end up with yet another gen AI suffering from accuracy problems.
    – TylerH
    Commented Apr 18, 2023 at 16:00
  • 21
    I mean, i get that, but that seems a bit too aspirational to me, given what the company has promised and delivered previously. ¯\_(ツ)_/¯ It's more likely we'll just get clippy.
    – Kevin B
    Commented Apr 18, 2023 at 16:39
  • 9
    @Trilarion It shouldn't be so hard for a company like Stack Overflow to beat Google at making a better in-site search for its own content than what Google can do. SO has access to more data and better domain knowledge of the content, and should have no problem attracting enough talent to develop such a thing, given its overall reputation as a company.
    – TylerH
    Commented Apr 18, 2023 at 18:00
  • 9
    @Braiam It should fairly obvious how a company focused on broad internet searching has different concerns than a company focused on searching its own system. If Google wanted to dedicate a team to making a site search for questions on SO, for example, then yes they could probably out-compete even SO's own best concerted effort. But they aren't doing that, because it's not their job or concern, and much of what is useful for searching SO (all the parameters and keywords one can use for searching) are simply not supported in Google search because they only apply to SE sites.
    – TylerH
    Commented Apr 18, 2023 at 19:45
  • 19
    @Braiam Also, no, Google's primary goal has been ads, not search, since the acquisition of DoubleClick in 2007. You can tell that this is true given that search has remained fundamentally unchanged since around that time, except for the fact that ads and promoted content have continued to increase in presence in the Google search space. Google had the best product around that time, as a matter of hard effort, and has been able to keep and expand their domain presence by simply outspending any competition thanks to their ad money.
    – TylerH
    Commented Apr 18, 2023 at 19:46
  • 35
    10 years from now it's going to be really interesting to see how this post and the comments in reaction to it have aged.
    – Mentalist
    Commented Apr 19, 2023 at 4:20
  • 33
    Is anybody else a bit concerned that the CEO himself couldn't be bothered to post this with his own account? I mean I hope this gentelmen actually believe in the core product and understands how it works.
    – ouflak
    Commented Apr 21, 2023 at 15:28
  • 22
    @Philippe You can never win all of meta - but I do think tailoring the messaging to the audience helps. If I may be blunt - Prasanth writes very much like a management/business sort, and its almost a different language, as much as there's quirks to communication here for an 'outsider', or when technical sorts communicate and a non technical sort is trying to keep up.
    – Journeyman Geek Mod
    Commented Apr 22, 2023 at 3:17
  • 21
    The one piece of actionable information in this CEO's post was "I’ve tasked a dedicated team to work full time on such GenAI applications [and I'll back to you in 3 months]". Everything else was exactly the kind of ineffective mushy regurgitated keywords and calculated copycat know-it-all posing that could be effectively emulated by the current version of ChatGPT. --- Not even considering the question "how do ChatGPT and human "intelligence" differ? Intellectually as dead and un-curious as a doornail.
    – Craig Hicks
    Commented Apr 22, 2023 at 7:47

39 Answers 39

Reset to default
1
2
8

The way these LLM networks work is that they require a bunch of training data, and then produce a likely outcome.

But the network cannot think, it cannot perfom fact-checking, and it cannot reflect on anything.

People come to SO for answers to questions, but LLM's can at best repeat the answer someone else has already given, so they add no real value.

At worst, adding LLM's to SO will be really harmful to the site and the community built around it.

Please don't do this.

[Correction] I know it's likely going to happen anyways, so please be careful about it. AI as a search extension would probably be a good thing, as would grammar correction when posting questions.

But please use LLM's for what they are, and not just "jump on a wagon" becase Bill Gates et al. says so. The SO community contains the brightest minds in the entire world of IT, and their combined knowledge for how to drive this site forward will be the best recipe for success.

Just follow community feedback - you have an excellent system in place for how to use this productively. Easy!

Improve this answer
3
  • please see the responses by Phillipe here and here a lot of people have the same misunderstanding as you (about what the company plans to use AI for), but to their credit, they didn't have the information that you can now find my reading prior discussion. Please read prior discussion.
    – starball
    Commented May 9, 2023 at 17:27
  • 1
    "repeat the answer someone else has already given, so they add no real value" To be fair, I think that there is tremendous real value in being able to repeat over and over again what others have already said, especially for learning.
    – NoDataDumpNoContribution
    Commented May 10, 2023 at 16:31
  • @Trilarion I agree with that in general - it's how you learn to teach. But on a site like SO I think in general less is more, no? Best with only 1 source of truth, because most questions in computer science can be boiled down to "which is more efficient", and we tend to avoid questions or answers that are too subjective, like "which framework is better", unless we can determine "better" strictly from a perspective of performance or some other metric. In that sense I think LLM's are suitable for search engines, but not so much more - at least yet.
    – alexpanter
    Commented May 11, 2023 at 8:11
7

Perhaps GenAI would be useful if it a) cited all the sources used to pull its answer together, and b) every upvote or 'thanks that worked' tick on the GenAI text resulted in two upvotes on the sources cited.

Edit: Useful and correct answers are still the overarching goal. If AI can provide answers tailored to seekers individual circumstance, drawing from the library of experienced people's contributions, without invoking yet more effort from those people, well, that's a good situation. An awesome one even.

If however AI answers all float to the top, and people's efforts and contributions are quietly pushhed off to the background, that's a horrible situation.

This suggestion is only a way of incorporating generated text. The central idea is ensuring that people remain the forefront of the Stack Overflow community, that's the important part.

Improve this answer
13
  • 3
    why two upvotes?
    – starball
    Commented Apr 19, 2023 at 5:21
  • 2
    @starball I guess to give more credit to humans than to the AI (which would get a single upvote). However, I don't really see the GenAI boosting "Look I have 100Million rep and just because I plagiarized all the answers here."
    – NoDataDumpNoContribution
    Commented Apr 19, 2023 at 7:51
  • 1
    I'd only give at most 1 upvote. I can't understand why 2 upvotes would make sense.
    – starball
    Commented Apr 19, 2023 at 7:53
  • 4
    @starball the numbers I gave are just off the top of my head. They're place holders for the central idea of rewarding the people who gave the text generator something to work with.
    – matt wilkie
    Commented Apr 21, 2023 at 0:40
  • Would you also be in favor of banning human answers on StackOverflow that don't provide a reference? If not, why does it matter if an AI does that?
    – JonathanReez
    Commented Apr 25, 2023 at 4:51
  • 1
    @JonathanReez I think it's bad form for humans to not reference sources. Banning no cites from people? I wouldn't go that far. Downvote works fine for that. AI are not human, they shouldn't be treated the same.
    – matt wilkie
    Commented Apr 26, 2023 at 5:24
  • 1
    @mattwilkie but humans often can't reference the source for coding answers. I.e. you might not necessarily remember where you've read about a particular hash function implementation. Or you might not remember where you've learned about how to change the Windows settings to solve a problem on SuperUser.SE. Should all such answers be downvoted? You're making the assumption that humans always cite their sources but that's easy to invalidate with a 5 minute review of StackOverflow answers.
    – JonathanReez
    Commented Apr 26, 2023 at 14:22
  • @mattwilkie and I'm confused as to why AI needs to be held to a higher standard. Shouldn't we be blind to what the origin of the answer is?
    – JonathanReez
    Commented Apr 26, 2023 at 14:38
  • 1
    @JonathanReez it's because lack of citation is prevalent that I used 'bad form' rather than 'ban' or some other systemic approach. We should encourage a culture where referencing sources is normal, and incentivized. AI needs to be held to higher standard in this context because it can answer thousands of questions in the time it takes a person to merely read a single question. If the AI answers everything, as they very well might, then people will stop contributing, because there's nothing left in it for them. Upvotes and the associated community recognition are the only reward we have.
    – matt wilkie
    Commented Apr 26, 2023 at 19:31
  • The AI shouldn't get any upvotes at all. None at all.
    – Benjamin R
    Commented May 4, 2023 at 18:19
  • @BenjaminR I considered that first, but then realised that useful and correct answers are still the overarching goal. If AI can provide answers tailored to seekers individual circumstance, drawing from a library of experienced people, without invoking yet more effort from those people, well, that's a good situation. If however AI answers and the library of experts is ignored and disenfranchised, that's a horrible situation.
    – matt wilkie
    Commented May 4, 2023 at 22:06
  • @mattwilkie Since a GPT style AI in this context is just synthesising its response from actual SO/SE users, it shouldn't get any public reward. Internally, yes, it absolutely needs those metrics to improve the quality of its answers, but it should be seen for what it is: a facilitator and aide, not as a true "Answerer", because it is not. It is just synthesising the learning, expertise, wisdom, and voluntary help of actual humans who gave answers. This is the most correct way of viewing the situation.
    – Benjamin R
    Commented May 6, 2023 at 14:03
  • I think it'd be pretty important that it gets publicly visible rewards, actually. We have to consider that questions may get answers from both humans and AI. Consumers of Stack Exchange sites need a way to gauge what's the best answer, regardless of who or what wrote it. If an AI has the accepted answer (perhaps it answered first and was already accepted before humans got around to it) with 30 upvotes, but later a human comes along and answers and gets 200 upvotes, we'd want to see this. If the AI didn't publicly get reputation/votes, this comparison would be challenging.
    – Seth Falco
    Commented May 9, 2023 at 10:30
6

People need a place where their problems can be solved using original, sophisticated thinking.

Current cutting edge large language models (LLM), used by technologies like ChatGPT, are able to copy-paste and recycle what has already been created by humans, but do not actually create new ideas, do not think. Perhaps, one day we might have LLM that actually create new ideas, think.

The major danger I see in CEO Prashanth Chandrasekar's proposal, assuming I understand well the proposal, is that it risks undermining actual innovation and creativity. Humans are capable of surprising, unpredictable levels of innovation and creativity, and we might lose that with unwise use of LLM.

There is also accountability that humans have, but LLM do not.

The individuals that CEO Prashanth Chandrasekar's blog post mentions to support some of their points might have conflicts of interest due to possible associations with the companies behind technologies like ChatGPT, or something related.

Improve this answer
1
  • 3
    Hi ChatGPT, nice to know "your" views :-P
    – David Roberts
    Commented May 16, 2023 at 7:48
3

I am here to get answers to my problems. I am here to give answers to others problems. I am here to gain knowledge. I am here for a bunch of other things too. I am not here to give knowledge for an AI to misuse to spread misinformation. I am not here to see AI generated nonsense all the time. If you want to build an AI, there are other places to do that. Stack Exchange has never been a place to do so, and I am not sure I want to be here if that is what it becomes.

Improve this answer
3
  • 2
    you have little to no control over how your content will be applied by people outside of SE if they follow the CC-BY-SA license. People can use what you write to make weapons of mass destruction. When you made an account, you implicitly agreed to the ToS, which also gives SE the right to commercially exploit your content. Not that that brings me great joy, but it's a fact, and you need to accept it.
    – starball
    Commented May 31, 2023 at 21:45
  • 2
    I am aware of that. But what I am saying is I dont want to post more content to be fed to an AI. I know they can, but if I dont like what they do with it, I dont have to keep posting @starball
    – Starship
    Commented May 31, 2023 at 21:47
  • While true and fair, anything you write publicly is going to be fed, at some point, to AI. That's the sad reality. So not posting anything just in SE is quite pointless, IMO.
    – Shadow Wizard
    Commented Jun 1, 2023 at 7:28
3

When I read this, I got something completely different out of it.

After filtering out all the marketing fluff and buzzword salad, this passage seemed the most relevant to me:

Incorporating generative AI technologies into the organizations using Stack Overflow for Teams will allow us to layer a conversational interface on top of this wealth of information.

To me, that's saying that he's wanting to build an AI answer-bot trained on the existing Stack Overflow content, and then sell that capability to people who use SE for Teams. Building that sort of functionality would be extremely expensive, and offering it as a paid service makes far more sense. Adding it to the public site would reduce the number of real, accurate answers, which means your bot training data gets worse and the system spirals downward. It'd be hard to bring in enough new users off of that for the additional ad revenue to cover the costs of the project. Instead, let the public site continue generating quality content by actual intelligent humans, and then keep your AI-generated stuff on paywalled private systems.

"Community is the future of AI" => "[our] Community is [the training data for] the future of AI"

If that's the case, then the damage to the public site should be minimal. All the auto-generated garbage will be locked away where only a few people can see it. After a while those SE for Teams customers will realize that chatbot output can't be trusted and will quit subscribing to that service. The main losses will be the resources diverted to the project, plus the lower quality of the software produced by any company that's trusting a chatbot to write production code.

One other thought for the CEO (and anyone else who's thinking about doing something like this). There's a lot of content on this site about programming Linux kernel drivers and other systems that are licensed under the GPL or similar. When I include part of that code in a question/answer, that's a derivative work and thus also GPLed. Courts haven't made rulings about AI-generated software yet, but what is your plan when they inevitably rule that an AI's output is considered a derivative work? Your customers, who have been using the output of your bot in their commercial products, now get surprised that your product has forced them to involuntarily inherit the GPL and publicly release the code for their proprietary software. That sounds like a fantastic way for the company to get sued out of existence. You could of course put waivers in your contracts preventing lawsuits over that, but then you'd be tacitly admitting that there are serious unresolved legal risks in using your product. Who's going to be dumb enough to buy it then? Even if you escaped the lawsuits, it'd be the end of the product and you'd have wasted massive amounts of money.

Improve this answer
5
  • 1
    I doubt this. for one thing, I assume the content and value of SOfT instances is based on business-internal knowledge. I don't see how data from SO/SE is useful to that in any way that's more useful that a generalized LLM model like in ChatGPT.
    – starball
    Commented Jun 2, 2023 at 4:42
  • "There's a lot of content on this site [...] licensed under the GPL or similar. When I include part of that code in a question/answer, that's a derivative work and thus also GPLed" I'm not a lawyer, but I thought you weren't supposed to do that? Content on SE is CC-BY-SA, so wouldn't you be breaking the code's license by publishing it under a different, incompatible license? Or are they actually compatible? I'm only aware of adapting from CC-BY-SA 4.0 content and relicensing under GPL3 to be ok (link)
    – starball
    Commented Jun 2, 2023 at 4:46
  • also related: Is Stack Exchange's CC-BY-SA v3.0 content compatible with the GPL? (about the older CC-BY-SA 3.0 license)
    – starball
    Commented Jun 2, 2023 at 4:49
  • "wanting to build an AI answer-bot". And Here It Is
    – Shadow Wizard
    Commented Dec 5 at 21:41
  • People post GPLed content here, regardless of what the rules say. That doesn't change the license. Also, for many years, the site let you select the license of content when you posted it, and GPL was one of the options. SE force-migrated everything on the site to the CC license at one point, but there have always been serious legal doubts as to whether that move was legal. There's a significant amount of content here that is most likely still subject to the GPL.
    – bta
    Commented Dec 6 at 20:05
2

I think like so many alluring tech advances, trained language model interpreters (AKA "AI") are here to stay, so we need to deal with them. This means including them in a safe way, and not falling into the trap of putting our heads in the sand... and that includes being ethical in using SE to train an "AI", based on the free, hard work of SE users/reviewers.

"AI" for use as a gateway to collating existing knowledge is a great refinement to existing search tools... but as we all know, such tools require critical interpretation. I don't agree with flippant comments that pump hype, without clarifying the superficial analogy, such as:

Today, sophisticated chatbots, built on top of cutting edge large language models (LLM), can write functional code for a website based on nothing more than a photo of a rough sketch drawn on a napkin.

Gaining peer-reviewed knowledge is hardly ever the product of just brilliance, or an "AI" reviewing existing data. It takes hard work, blood, sweat and tears. You will see this often in the progression of answers on SE.

I think the use of an "AI" interface to aid with searching on SE through existing answers, advising on duplicates, and generally serving as a support to the humans using the site, is a great idea. But the belief that an "AI" tool will somehow be able to create answers is flawed - the very nature of the beast is that it is not aware, cannot critique it's work, and suffers from the bias of its training models. Do not anthropomorphise "AI"! It's an information interface, not a ship mind from the Culture [1].

[1] Iain M Banks - Culture series.

Improve this answer
1

I am against all of this. I'd hate to see AI integrated as a part of SO. SO is for learning and helping other people, not for using an AI to help you.

The whole point of SO was to ask questions, and get answers. To help other people with your knowledge of coding, or to get help with their knowledge of coding.

Also, taking a quick look at the home page/front page of SE: wow.

Guess the CEO needs to make space for the AI.

I hate how when one company does something and it gets super popular, all the other companies have to copy it. e.g. Google Bard AI copies ChatGPT, YouTube Shorts copies TikTok. Every company has to copy the more popular one and it's stupid. It's a shame that that's how the world works though.

So, SO should not have AI.

Like many other people have said, AI can be inaccurate some times and it doesn't know when it is.

If the accuracy of AI was better, maybe I might reconsider my opinion. But, sadly, it isn't so I am against this.

Adding AI to SO is even worse than Discord removing discriminators.

Feel free to downvote this, but at least explain why you downvoted.

Improve this answer
1
  • 4
    I didn't vote, but the answer doesn't really add anything that wasn't said already in other answers.
    – Shadow Wizard
    Commented May 18, 2023 at 7:30
-25

People worrying that AI code generation is moving us towards some existential risk ridiculously misinterpret what ChatGPT does. The software simply recycles online content, filters it, and formats it in a readable format using a language model. If we called this software, Search 3.0, there would be a chorus of approving voices just as there were when Google came along and made the internet more consumable.

Fear of AI becoming AGI is as ridiculously overblown as the prophecy that the Y2K bug would cause social collapse. This software is simply a data recycling tool, and if you get behind it, it makes online resources like SO more useful.

AI cannot replace human developers. It just streamlines many laborious repetitive coding tasks that have been explained to death online, because so many people repeat those tasks. If AI could ever write new code for us going forward, it would imply there is a sufficient corpus of existing code from which to write remaining software. This is an astronomical miscalculation that fails to comprehend the vast complexity of program variations.

How sufficient is the existing body of code, compared to the code we might possibly choose to write? We can enumerate possible programs as sets of input-output pairs. So an example program might only accept a 0 input to be converted into a 1 output, and so be defined as ((0,1)). Another might be defined as ((0,1),(123,456)) and so on. How many possible trivial programs are there that have a single ASCII character input and output? A lower bound on the number of variations is the powerset, so 2**128 variations for trivial single character input-output processes. How many possible programs involve character pairs? 2**16384. Typical programs have complex inputs and outputs. These are numbers that make all the programs written to date look infinitesimal.

'AI' cannot write our code for us. What we call AI today is nothing like 'AGI'. AI is a system that recycles our existing ridiculously tiny body of software. It cannot possibly extrapolate what we might want to write. It is not at all in the realm of possibility. When people claim 'AI wrote my app' then either they are using components that have been written 1000 times before, in which case it is a somewhat trivial invention, or it’s fake click bait.

Improve this answer
20
  • 41
    I mean... i don't see anyone here arguing "AI BAD IT'S GONNA REPLACE US"
    – Kevin B
    Commented Apr 17, 2023 at 20:58
  • 4
    -38 for OP's comment supporting AI? Why? There's fear of AI and of AGI on the internet and I want to address that.
    – Rian Rizvi
    Commented Apr 17, 2023 at 21:13
  • 20
    there's certainly more reasons to not want this change, as expressed in the several other answers, that aren't related to some ridiculous "existential risk"
    – Kevin B
    Commented Apr 17, 2023 at 21:15
  • 5
    There is also fear of snakes. I don't see you addressing that. Is it because nobody has actually expressed such fear?
    – VLAZ
    Commented Apr 17, 2023 at 21:16
  • 25
    Y2K failed to cause the collapse of society because people recognized that it was an issue and put in a lot of hard work leading up to 2000-01-01 to fix their software. Similarly, ChatGPT and related models could cause major societal issues if left unchecked indefinitely. Right now, we should have the discussion of what we need to do to avoid that.
    – The Guy with The Hat
    Commented Apr 17, 2023 at 21:17
  • 2
    You know, it would kinda be nice if AI could replace us. It would be very useful, and we’d get a lot more time for fun stuff. Yet I posted a very unhappy answer to this statement from the CEO. I’m not afraid of going out of business.
    – Andreas condemns Israel
    Commented Apr 17, 2023 at 22:54
  • 1
    Re "many laborious repetitive coding tasks": What are some examples of those tasks?
    – This_is_NOT_a_forum
    Commented Apr 17, 2023 at 23:15
  • 2
    @RiazRizvi Teach them how to use a search engine, not how to eat bullshit from ChatGPT.
    – Andreas condemns Israel
    Commented Apr 18, 2023 at 1:31
  • 1
    @Andreasdetestscensorship correct me if I'm wrong, but I think your concern is a feedback cycle where AI is recycling content posted by humans on SO and then humans are further posting output from that AI because they can do it prolifically and boost their own status on the sites, but in so doing, they degrade the status of people like me who spent years manually submitting content? That to me is a problem of content moderation in the face of a new technology that provides an astroturfing capability. I do support SO enhancements to prevent astroturfing and content degradation.
    – Rian Rizvi
    Commented Apr 18, 2023 at 1:39
  • 22
    The first two paragraphs of this answer miss the mark completely. SO is not some sort of tin hat haven, and does not harbor any sort of conspiracy views on AI. The transformer algorithm, specifically the popular one which is pre trained and aimed at generative language, is horrific in its falsehoods. While it may be useful for looking up existing knowledge (#plagairism), it is absolutely worthless when it comes to creating new works (aka generating). This is the reason for the pushback, regardless of what the name is or what public talking points seem to be around the subject or concept.
    – Travis J
    Commented Apr 18, 2023 at 2:05
  • 2
    Huh?? A system that efficiently looks up existing knowledge can obviously be used to create new works, that are composed of elements from those lookups. Think Lego. So no it is not 'absolutely worthless'. The whole premise of SO is to provide existing knowledge for others. 'plagiarism' is loaded language and implies you're in conflict with yourself as a user/contributor of SO.
    – Rian Rizvi
    Commented Apr 18, 2023 at 3:11
  • 9
    I didn't say it cannot be used to create new works, I said it cannot create new works, there is a very strong difference there whereby in your out of context version the person is creating versus my in context version where the algorithm is creating. SO is premised on providing knowledge... non hallucinated, factual, helpful, knowledge in the form of adaptation for use. Plagiarism is not loaded language, the algorithm will directly reproduce existing works. I am not in conflict with myself. Please let me know if you have any other thoughts, aside from Lego's....
    – Travis J
    Commented Apr 18, 2023 at 3:35
  • 2
    'Please let me know if you have any other thoughts, aside from Lego's..' that was good, it gave me a chuckle. I think we can agree AI doesn't create.
    – Rian Rizvi
    Commented Apr 18, 2023 at 3:45
  • 4
    "The cottage industry of Y2K consultants helped society by paying a few bills for some people. No social collapse crisis was averted." Well, now I can ignore everything you say.
    – JonathanZ
    Commented Apr 21, 2023 at 3:41
  • 1
    @RiazRizvi don't take a huge downvote personally - remember people are downvoting the answer, not you. Second, sometimes just want to Edward Macaroni Fork a post that is already scoring badly.
    – Criggie
    Commented May 8, 2023 at 13:13
-33

The announcement is vague, but integrating an LLM to help write SE answers would make sense (instead of banning LLMs). However, we may want to first analyze how often LLMs give an incorrect answer to an SE question to see whether it is good enough to be used more systematically. Otherwise, one could simply let users decide when to use LLMs, since the LLM accuracy depends a lot on the type of question.

SE may want to add some mechanism to mark an answer as incorrect that would complement the current voting system (which could also be beneficial for non-LLMs answers too), at least for the answers mostly generated by LLMs.

Lastly, it's still a bit unclear how a language model can keep track of the provenance of the main knowledge/sources used to generate a given output, and in some cases LLMs may go beyond the fuzzy boundaries of fair use: aside from the legal aspect, unreferenced answers sometimes aren't that satisfying (e.g., hard to check the sources and can't read the additional information that the sources may have provided). But it's rather rare, esp. for non-expert questions, and humans are known to plagiarize too anyway. If that's an issue, SE could also integrate some plagiarism detection tool.

Improve this answer
2
  • 5
    Quick note - since we keep getting flagged and its causing a bit of friction. We've reviewed things and really, its fine to try to link back your own meta posts to try to support something you'd want SE to review. I do it all the time myself. And I think its pretty obvious its the OP linking their own posts
    – Journeyman Geek Mod
    Commented Apr 20, 2023 at 4:39
  • 3
    @JourneymanGeek I guess some people prefer unsupported statements. Not to mention the linked posts quote other people's papers or have posts from other SE users.
    – Franck Dernoncourt
    Commented Apr 20, 2023 at 12:54
1
2

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .