AI Changed Stack Overflow for the Better

AI Conversations Powered by ProsusGroup

2025 Developer Survey

Prashanth Chandrashekar [00:00:00]: The first answer will be a Stack Overflow link.

Demetrios [00:00:02]: You're growing and profitable.

Prashanth Chandrashekar [00:00:04]: Yes, we work with everybody, OpenAI and Google and all the cloud providers. At some point the AI taps out. That's why we want to be able to integrate into these tools. So we want to make sure we are where users are. The big portion of what we're doing now is expanding well beyond Q and A.

Demetrios [00:00:21]: All right, you just told me something very important. You, you're growing and profitable.

Prashanth Chandrashekar [00:00:27]: Yes, we are growing and profitable. Despite, you know, all the viral headlines around our. Specifically just our Q and A, which is an interesting stat to look at because obviously in the age of AI, you've got people asking a lot of simple questions on let's call it ChatGPT, any AI tool. Right. And that I think is like it's more, it's just a function of being a content site on the Internet. I think a lot of content sites now have that same point around who's creating the new knowledge. And so for us it's thankfully like the increment because of the experts on the platform, et cetera. And the standard has always been about let's have the bar really high and let's create this amazing corpus of cumulative information and knowledge that is going to be very useful, including for all the AI models who now work with us in an official capacity by licensing our data.

Demetrios [00:01:14]: Yeah, I was going to ask about that. Basically any coding helper is powered in one way, shape or form by models that are powered by the stack overflow answers.

Prashanth Chandrashekar [00:01:27]: That's right, that's right. So specifically, just to finish the earlier thought, so people come to our site to ask all the incrementally difficult questions. But all the simple questions now are also being asked on places like ChatGPT. But they're also now empowering people to ask those questions even on our own platform through multiple mechanisms which I'll talk to you about AI and non Q and A related. So we want to accept all questions but at the same time keep the Q and A corpus very, very sort of pristine and like incrementally kind of accurate. So it's very useful for these AI providers. So to answer your second question, yes, all the AI providers work with us officially. A few years ago when ChatGPT came out in late 2022, we noticed, you know, all the AI companies and LLM Labs or AI Labs had used our data to train or pre train their models.

Prashanth Chandrashekar [00:02:15]: So we soon put up a bunch of anti scrapers on our site and they all sort of wanted to work with us because we generate obviously quality data. We even worked with the process team to actually think about. We took an open source model and fine tuned it with our data and now we actually publish these results with something called Prolm AI where our stack data is used to rate all the latest foundational models.

Demetrios [00:02:41]: I love that that's so valuable, especially because the data isn't. It's like public, but not public. And so you have these models that as soon as they come out, you can get a quick gauge at how good they are on coding tasks.

Prashanth Chandrashekar [00:02:55]: Exactly right. So with that information we noticed, you know, something like a 30 to 40% improvement in accuracy of answers. And so either if people were using the data already, then they now realize they have to officially license it through us, which is now a requirement. Attribution is very important because that's part of our license. You need to attribute back to the sources so it recognizes our users, et cetera. And ultimately we're able to contribute back to the community by building new features which I'll talk to you about so that yes, we work with everybody, OpenAI and Google and all the cloud providers and the, and the, you know, all the AI labs and they officially license our data and they're more. Every day we're seeing new companies come into the, into the, you know, to the limelight. So we're happy to work with them.

Demetrios [00:03:38]: Yeah, there's something that is to be said about the experience of Stack Overflow versus just asking one of these answer engines.

Prashanth Chandrashekar [00:03:46]: Yeah.

Demetrios [00:03:47]: And it's along the lines of there's more than one way to skin a cat. And so when I go to Stack overflow, I get a much more robust answer because I can see various opinions on things and I can get the nuances a lot more versus oh, here's how it's done. Or oh, let me just go try and do that. If it's a coding agent and then it tries it and it doesn't work because some package isn't updated or something and then you have to go through that loop, et cetera, et cetera.

Prashanth Chandrashekar [00:04:18]: Right, right. No, exactly. I think that the depth of the answers where people over a 16, 17 year period have, you know, started by writing their first, I don't know, basic programming, you know, kind of like their first hello World program all the way to, you know, using more advanced languages over time. And then as things have abstracted higher and higher, I think people have, you see that journey. Right. And so that depth of what is the right answer to canonically write answer to a question and how that's changed over time. You know, the accurate answer from six years ago may have been very good, but now we've got a more elegant answer that gets, you know, completely evolves.

Demetrios [00:04:56]: That's so true. And the right answers bubble up because somebody tries something and it doesn't work. And then they come back and they put in the thread, hey, this, this doesn't work anymore.

Prashanth Chandrashekar [00:05:06]: That is right. That's right.

Demetrios [00:05:07]: And so you have that almost like up to date answer.

Prashanth Chandrashekar [00:05:11]: Exactly. And your point around, you know, the depth of the answer is very useful because what we do notice is that. And the reason why I think we're very excited to work with all the AI companies is when AI, let's say, does a lot of the, let's call it the grunt work and the kind of the, you know, the craft. And that's good because it basically is all about abstraction. Let's remove like the kind of the menial tasks. That's all good. However, it does have a complexity cliff issue. At some point the AI taps out or it doesn't have the answer to a question because it's very specific and contextually relevant.

Prashanth Chandrashekar [00:05:43]: It doesn't have it based on the fact that nobody's actually come across that issue and actually produced a viable answer. Which all shows up on Stack Overflow. So that's why we want to be able to integrate into these tools and what we call knowledge as a service. Go where the users are. So previously the interface was go to Google, type in a question, how do I spin up a Kubernetes cluster? The first answer will be a Stack Overflow link. You click on that, you come to our page and so on. Now people use Google, but they also use a bunch of AI tools to go and search for answers to your earlier point. And so we want to make sure we are where users are.

Prashanth Chandrashekar [00:06:18]: So by many of our partnerships are about also not only licensing our data, but also being able to empower users to ask questions straight into Stack over from those places. Right. Or and in addition also attribute back to Stack overflow from their AI, their chatbots. Because that way people come and look at the more detailed answers to your earlier point also around here, just exactly how you do it, why we're suggesting this sort of solution. So that's the full loop that we're trying to do kind of Enable plus obviously on our own website being able to basically bring people and we just launched this, this Week, which is StackOverflow AI on the site where people can bring their LLM chat straight into this and drop it in there, it imports it and then it looks at the corpus of 50 million questions and answers that we have and then provides a relevant answer. Or if none of that works, then you can post a question to the community either in one of our now newly formed live chat rooms with humans, human experts, or in our traditional Q and A or even a discussion board nowadays. Yeah.

Demetrios [00:07:21]: So I think it's a good moment to just mention the. What everyone thinks of as they look at Stack Overflow now and they're like, oh, well, the questions are going down because of this you rightly so kind of talk about. Yeah, probably the basic questions are getting answered by the answer engines. The more nuanced discussions are still happening.

Prashanth Chandrashekar [00:07:44]: Yeah.

Demetrios [00:07:44]: Can you explain what you've been seeing as far as questions being asked?

Prashanth Chandrashekar [00:07:49]: Yeah.

Demetrios [00:07:49]: Is it more in depth questions? Is it always more topical questions that are just not in the corpus of data and so you don't have them and they're encountering them?

Prashanth Chandrashekar [00:08:00]: They're very. If you take a step back, we've historically only had Q and A. And the Q and A again, the mission and purpose of Q and A was what is the. Almost like the encyclopedia, the Library of Congress for all things software programming. Right. So it literally has that.

Demetrios [00:08:15]: Super valuable too, by the way, I was going to tell you before, the most valuable form for developers by far, but yes, exactly.

Prashanth Chandrashekar [00:08:23]: And that's why it's been useful for all the AI companies to leverage. So with that Library of Congress, it's like, you know, if you have the same book like six times, like it's not useful. Like what we're really trying to do is like say, what is the next book? You know, and so on. And so we want to keep adding new books into that. So previously. So 1, 1 out of every 2 questions were closed on Stack Overflow in just the Q and A. So the reason for that is because they were all duplicate questions. Right.

Prashanth Chandrashekar [00:08:48]: So even my own first experience using Stack Overflow when I, even before I joined the company was like, hey, this is question's been answered already. Like, you know, slap on the wrist. Right. And while that can be harsh, the whole. That was very deliberate by the original founders of the company to say, let's create a Library of Congress that's truly like the cumulative effects. Cumulative effect of all the works in the world around this topic. So that, yes, it is now all just, you know, what is the something that's never been answered before. I'm a, you know, an expert on Python.

Prashanth Chandrashekar [00:09:18]: And here's an additional book, you know, Python. A lot of questions have been answered in the world, but I would say let's say something newer, you know, maybe it's Transformers, maybe it's even something even, even more, more current. And one of the latest technologies in AI and so that those additional books keep getting added to the library. But that's the Q and A form factor. But a big portion of what we're doing now is expanding well beyond Q and A. So back in March, we had an AMA with our community members and we talked. We jointly created this now vision, which is to be the most vital source for technologists, period. Right? So the source is how we want to do that is to make sure that we cultivate community, we want to power learning and unlocking growth in people's careers, that their work to be productive and so on.

Prashanth Chandrashekar [00:10:01]: And the way we want to do that is, especially with what's going to happen in the world, there's going to be a lot of job displacement, etc. We want to make sure that you, any question actually also is fine on Stack Overflow and maybe not in Q and A, which is why we've opened up, you know, hundreds of chat rooms on the site which are live human chat. Now that's not exactly novel, but the fact that you've got experts in these places because you again, remember people have been building expertise on the side, they're, they've got reputations on the site. I have a million points on Python. Okay, so you know what you're talking about because you know, you're in the school of hard knocks and for you to be in the Python chat room, to be able to help another developer out when they have an issue, it could be one of the newer AI topics. It could be an AI specific chat room. How do I prompt on this? Why is my prompt not working, whatever it may be? So chat discussions, coding challenges, because coding challenges like how do we help a young developer? Because you know, those folks are going to struggle in this environment, most likely because companies, et cetera, unlikely to dive straight in and hire a bunch of junior developers. Only because a lot of the AI agents are pretty capable to do the, let's say, the simpler, you know, workflows.

Prashanth Chandrashekar [00:11:10]: So all these ways that I just described, we're broadening the site to be much more multidimensional, to realize our mission and to help our developer community and technologists community. So yes, questions in Q and A are incremental, but we are also now opening up all these Other avenues where any questions should be fine. You ask a question, it either shows up in Q and A discussions or chat. And all that will be helpful through different mechanisms for the community.

Demetrios [00:11:35]: Yeah, I really like the idea of ask an expert because that is something that with the mlops community we've had in mind. And it's almost the way that I feel like the MLOps community was valuable or is valuable to folks because you, you have different experts from different walks of life. And we used to joke back in the day that the MLOps community Slack was like slack overflow. And it was almost like real time slack overflow. And what I've seen when it comes to the questions that folks are asking in the MLOps community SL is that those basic questions that you can really answer easily, they're not being asked as much anymore. What you do have are more nuanced questions and things that you really want to talk to a human about and you want to get a human opinion on. It's not like what's the, what's the answer to X.

Prashanth Chandrashekar [00:12:26]: Right.

Demetrios [00:12:27]: You want this more nuanced discussion. You want to hear opinions, you don't want to hear an answer.

Prashanth Chandrashekar [00:12:32]: Correct. Exactly right. Which is why we opened up something called discussions on the site. So they're a lot less stricter in terms of the Q and A is meant to be the canonical answer to a question. This is more of like, look, you can ask whatever you want. What is. What is the discussion around the topic? You know, should I be using XYZ Vibe coding tool versus the other wipe coding tool? What are the pros and cons of X versus Y? Whatever you want. Right.

Prashanth Chandrashekar [00:12:57]: So that, so I think it's less about. It's an. It's a conversation versus our discussion versus here's the accurate answer to the question.

Demetrios [00:13:03]: Exactly. There's so much gray area that you have to cover and you have to hear all the different opinions from folks and you can't just say this is how you do it.

Prashanth Chandrashekar [00:13:12]: That is correct. Yeah.

Demetrios [00:13:13]: Because if you're looking for the this is how you do it, you have it there, it's closed. Or maybe it's getting updated as accepted.

Prashanth Chandrashekar [00:13:19]: Answer or you know, and so on. Right in the Q and A area. Exactly.

Demetrios [00:13:23]: I think this is probably a good moment to talk about what types of questions you've seen.

Prashanth Chandrashekar [00:13:28]: Yeah.

Demetrios [00:13:29]: As far as like the amount of AI questions or the way that questions have changed over time.

Prashanth Chandrashekar [00:13:36]: Yeah, yeah. I think, I think we sort of covered in that they're more specific and they're More you can call them like, you know, incrementally sort of very complex relative to the standard sort of questions that now that, that will also change with what I described because you could ask that repeat question in chat and discussions. So we're sort of in between both those. So specifically in Q and A we see the more complex questions. We see increasingly number of AI related conversations happening on that site or questions around certain AI technologies. So we have everything from Nvidia CUDA tags to TensorFlow to any of the people who are here at this conference. They all have content on the site. I would say there's obviously an increasing number of developers learning about those technologies because they realize that's where the world's headed and so that is naturally the place where that happens.

Prashanth Chandrashekar [00:14:27]: So and then, but even for the person who is early in their programming journey or developer journey, they're learning the fundamentals which we very much encourage because it's like, you know, you know, you know, you want to know when you are using a calculator that 2 plus 2 is not 6 right. At some point. So like you want to learn the fundamentals of math. So in the. Similarly we encourage people to learn all that and do it on Stack because you will learn through, you know, going through this journey of learning on Stack. So the questions are complex on Q and A, but again simpler on chat and discussions and, and yes, they're increasingly AI oriented on the Q and A side.

Demetrios [00:15:03]: I've got to know because I asked you before we hit record, why don't you build a cursor or a wind surf with all of this information?

Prashanth Chandrashekar [00:15:13]: Yeah. So I think we are, it's, I don't think there's actually now A. It's kind of the same question I sometimes get around. Why haven't we built our own LLM? Right. So why can't you just go do that? And so the answer is that I think we have certain points of view. For example, I believe the LM game is fairly commoditized. So over time they're all going to sort of get to the same level. Maybe one will get to AGI and all those sort of other predictions that people have.

Prashanth Chandrashekar [00:15:38]: And there's also the open source topic is like when this knowledge just becomes ubiquitous and then the question is, you know, you spent let's say a few hundred million dollars training these models and the whole ecosystem is running on that. So the question is like, you know, how do we. One, we want to just serve our community the best way we can. We have A great knowledge base. The user interface seems to have changed. So we want to sort of play this substrate layer at the moment where we are powering all these AI, LLM providers, AI labs, even the AI agents and I can. We should talk about our enterprise scenario, which is like a very fascinating thing to see how enterprises are using Stack Overflow inside their companies, which has given us a very sort of key insight. So all this suggests that for AI to work you need a very trustworthy foundation of knowledge from experts.

Prashanth Chandrashekar [00:16:30]: And it comes down to when we talk to our developers, every year we do a developer survey, as you're aware. And thanks, love that survey.

Demetrios [00:16:38]: Yes, I'm a big fan of that because you get so much information from folks and there's so many diverse opinions. Again, that's. I think the beauty of Stack overflow is the diverse opinions, the willingness to share.

Prashanth Chandrashekar [00:16:52]: That's right. So, so the, the key insight that, to answer your question, the key insight from our this year's developer survey that just closed, that we have not yet released, but I'll share it with you, is that not surprisingly, the AI adoption for coding related tasks has gone up. So it's gone up roughly over the past three years from let's call it 60% of our users up to about 80% of our users. So not surprising everybody's using AI tools at least trying or interested to try it. But what is very surprising is the trust level around what's coming out of these AI tools. I would have expected to go up, but actually it has gone down. So for the past three years it has been 40% of the users trust what's coming out of AI tools to 40% year two, which is last year, they will trust AI tools to 29% this past year.

Demetrios [00:17:42]: This year spaghetti code is coming to roost now finally three years later.

Prashanth Chandrashekar [00:17:45]: So it is quite counterintuitive but at the same time quite fascinating. That was the one data point that stood out to us. Like the trust factor is really, really key and you know, in a world where you've got a lot of AI slop and if you go to the Internet, you, you know, you're going to see all sorts of AI generated content. So there's going to be very few places where you actually go to get very credible, authentic human curated knowledge. Even if you use AI to sort of do things, but you need the human to make sure it's accurate and validated and you know, it's actually trustworthy ultimately. So that's what we glommed onto saying that's the problem that speaks to our strengths. Our strengths are people trust us, they trust our brand, they trust our community. It's in fact, it's really hard in some ways to engage with our community because it's actually really, really, the standards are very high.

Prashanth Chandrashekar [00:18:36]: So, you know, we're not going to be bought by commercial interests. You know, it's like, it's one of these things. It's like a very independent, like a Switzerland type of place. So that I think is one of our primary strengths. So, yes, could we go and build a cursor or, you know, something like that? Perhaps. But you know, it's not exactly. We could, you know, we never say we wouldn't do it at some point, but at the same time it's not something that we are focused on now because even I think there is. It's very early in that world of trying to figure out, is it a, you know, is it an IDE tool, is a command line tool, is it a, you know, is it an agent? Yeah.

Prashanth Chandrashekar [00:19:09]: And for what use cases? So we are very. So I don't think that world's been sorted yet. And I think there's still, you know, there's a lot of movement between multiple, you know, people have very low switching costs between tools. Devs just, you know, move from one to the other. Yeah.

Demetrios [00:19:23]: A lot of folks use two or three.

Prashanth Chandrashekar [00:19:24]: Right. So the question, yeah, so I think we always will be open to it. And our AI functionality on our public site now is maybe a step in that direction to say, look, get the best of the human's expertise on the site, plus leveraging AI search and discovery and all the things that, you know, are good about that. But also over time, if we're able to capture the user's context, you know, be able to perform tasks over time. But, you know, that's not exactly a current focus, but the idea is to power the AI ecosystem through that trustworthy knowledge intelligence layer, if you will. And we should talk about the enterprise, which is another very key data point.

Demetrios [00:20:00]: I want to get to that. I do think that it's smart to almost like be patient here because you're right. Two years ago we were having the same conversation about why aren't you building an LLM? You have all this data, you should make your own. Now we realize that would have been a horrible decision.

Prashanth Chandrashekar [00:20:17]: Right, Totally. Yeah.

Demetrios [00:20:18]: And that is not where you're going to have the delta, so. Or the alpha, I guess I should say.

Prashanth Chandrashekar [00:20:25]: Yes. Yeah, in, in financial terms.

Demetrios [00:20:27]: Yeah, sure. Yeah, yeah, exactly, sure. You can see where I play. I, I don't necessarily do the CEO stuff. I don't have to go to the earnings calls. Yeah, like, like you, I, I like that you took that approach, like, hey, let's be patient now, when you're looking at ways to make the company more interesting, you've put a big bet on the enterprise stuff. Can you talk to me about what you're doing there and how that even looks?

Prashanth Chandrashekar [00:20:57]: Yes, totally. So the two sides of our company. One is obviously our public site, which we've spent a bunch of time on the way in which we obviously were building all these new features, et cetera, that I explained. And the way our business model that supports that is our data licensing business, which is 12 big AI labs as we talked about, which is a very fast growing area for us, where they get access to the knowledge for a fee, a recurring fee, and then advertising, obviously. So that's one side. The second part of our company is the enterprise version of Stack Overflow that's used in literally thousands of companies around the world. And that's actually for some reason less known in the ecosystem because our brand is so prevalent on the public platform that people just sort of, they look at. That's why, you know, I joke about this, like one chart that shows up the number of questions and it's like, okay, that's, oh my goodness, it's the end of the world.

Prashanth Chandrashekar [00:21:43]: Not really, because that is a component, it's what we're well known for. But our company is like multidimensional in nature. You know, we've got this B2C or B2D rather business developer, you know, business developer. And then we have this B2B business which is all around a private knowledge intelligence layer inside your company that allows you to have human curated knowledge that will always stay up to date and updated by your subject matter experts inside the company. And now, most recently, what we discovered is that all our largest customers are using this very, very accurate knowledge base, which is Stack Overflow Internal, to power their AI agents. So what we started doing is that we noticed our APIs of all our customers were red hot. And when we talked to our customers, we said, what are you doing exactly? And they're like surfacing it in search results inside the company. They're surfacing it, using it in AI assistance.

Prashanth Chandrashekar [00:22:35]: So as an example, Uber built something called Uber Genie, which one of our customers, Uber. And Uber Genie leverages the thousands of questions and answers within Uber's internal version of Stack Overflow and surfaces it into that, sorry, powers their AI assistant, which then goes into their Slack channels and answers a whole bunch of questions from users automatically that are all like, you know, questions about the same stuff. How do I reset my password at Uber? And so on. And so it drives a tremendous amount of productivity inside the company because it's tapping into this very accurate knowledge base. And knowledge bases are underestimated because you know, you know, you have a lot of knowledge inside the company. But either it's like hard to find or it's out of date. Right, dude.

Demetrios [00:23:19]: And we all can sympathize with the fact that maybe there's some stuff in Jira, maybe there's some stuff in Confluent, maybe then somebody came and they have a Notion database and then there's. Oh yeah, all that stuff. No, that's actually on Gitbook and you've got this sprawl across all these different areas. And I imagine that that type of thing, building an abstraction on it and having the ability to pull from wherever the answer is. If it's in Stack Overflow or if it's in Notion, whatever it may be, you have that and you can get that answer. That's a huge unlock because I can't tell you how many times I've spent searching for something and I'm like, I.

Prashanth Chandrashekar [00:23:58]: Swear it was somewhere.

Demetrios [00:24:00]: Name was like around this. And you're typing in keywords and you're not finding it.

Prashanth Chandrashekar [00:24:04]: Yeah, totally. It's like, it's a very, very common frustration. And that's why customers since 2019. I came on as a CEO in 2019, so. And partly it was to solve this problem. So inside companies, they had this huge issue around developer productivity. Employees like, you know, spending a lot of time just asking the same stuff and tapping on the shoulder of the subject matter expert always. Exactly like you said this.

Prashanth Chandrashekar [00:24:29]: Now, Stack Overflow internal ingests information from other knowledge sources that are all out of date or getting out of date and creates this area on Stack Overflow through the subject matter experts that we automatically identify and automatically make sure that the information is constantly up to date, much like our public platform has done for 16 years. So we have that secret sauce of what makes the curation engine really work, right, with reputation points and all the things. And then that when you plug it into an MCP server, you can do all sorts of amazing things because you're getting this really accurate knowledge stream that's powering whatever could be your agent to perform an action, could be assistant to provide information, whatever it may be.

Demetrios [00:25:11]: Yeah, we were talking to Ben Young yesterday, who is the CTO of Sourcecraft, and he was mentioning one of the biggest unlocks for their agent.

Prashanth Chandrashekar [00:25:20]: Yeah.

Demetrios [00:25:20]: Amp is giving it a little bit of a nudge on where to look.

Prashanth Chandrashekar [00:25:25]: Right.

Demetrios [00:25:25]: And so I imagine that's why people are, that's why you saw the APIs are red hot, because the agents then if you just say, oh yeah, check stack overflow for that answer.

Prashanth Chandrashekar [00:25:35]: Yeah.

Demetrios [00:25:36]: Or get the context from stack overflow, that could be a reason that all of a sudden these agents are performing better and they're able to just start looking around stack overflow for a lot of different stuff.

Prashanth Chandrashekar [00:25:49]: That is right. And we, you know, and it's very apparent because a lot of our customers are in financial services and healthcare and retail and tech companies globally. Right. And so they all have realized that this is one of the few credible sources that their experts have spent time making sure it's accurate. And so absolutely trust it goes back to trust. And it's the same way we help AI agents on the public platform train off of our data now with the commercially using partnerships. Similarly here those AI agents inside companies in the enterprise are leveraging the platform's data to be truly value added inside companies. Because, you know, ultimately these agents, AI agents and assistants need to show roi.

Prashanth Chandrashekar [00:26:35]: And there's going to be a tremendous amount of pressure to do that. I think we're probably second or third now in terms of phases around people's attempts to deploy some AI related, let's call it technology inside companies. So and now they're like, look, okay, we figured out the dependencies, our privacy and great data and it has to be secure and, you know, and so on. And we don't want our stuff getting out of our company to other company. Our competitors. Like all the things that people have been debating over the past three years are finally, I think people have figured out, okay, you need like a very accurate knowledge base and you want that to be very seamlessly accessible through an agent that can do a lot of great things. So I think it's just like it's timing, it's saying patience, especially when so much is happening in a kind of a hype cycle. You just have to know if you've been through it a few times, you will know the trough of disillusionment.

Prashanth Chandrashekar [00:27:25]: And when people actually understand the reality of what it means to actually deploy.

Demetrios [00:27:29]: Things, it also is something that avoids the bus problem. If somebody gets hit by a bus and they have that information in their head, that's a big problem. But if it's there on the internal stack overflow.

Prashanth Chandrashekar [00:27:41]: Exactly right, right, exactly right. That's a very, very common use case why people bring us in in the enterprises, let's make sure. Because you know, people are constantly moving in and out of companies, people are bringing in new people, you know, people are changing teams. So when all that happens, Demetrios is like answers, you know, on a certain topic that he's an expert on are codified and they're built on over time by other experts in the company. But the fact is that you're the expert on the topic that's your intelligence is now in the company. Companies like Knowledge Repository and irrespective of what you end up doing with your own career, that the company continues to benefit from investing in you. Right. By investing in you.

Prashanth Chandrashekar [00:28:22]: So that's. So the company should benefit because that's just the right trade because you're actually ultimately it's all about people and what they bring to the story.

Demetrios [00:28:28]: So yeah, last thing that I wanted to ask around the questions was do you see a lot of very obscure languages getting more questions? Because these things like the not Python or Typescript type of questions. Yeah, I imagine they don't have the same amount of data and training Data that the LLMs can't answer those or they probably hallucinate more when they're talking about newly languages, all that stuff.

Prashanth Chandrashekar [00:29:01]: 100%. 100%. And you'll see some of that in our developer survey that'll come out in a few weeks and we'll share that publicly. But yes, absolutely, that's always been the stack overflow. We know what's happening in the ecosystem even before we all realize it because it's already the number of questions and certain topics have been increasing. So the new languages a couple years ago was about Rust and Svelte and all these things. So as they keep sort of progressing, it happens on the platform first because somebody has to figure out how to use this in the spirit of abstracting and going up the stack, so to speak, how do you make it more elegant, et cetera, as you're aware. So those things absolutely are the incremental sort of questions, the net new etc.

Prashanth Chandrashekar [00:29:43]: Because the world's questions on let's say JavaScript have, let's say vastly been answered. Right. So what's more to be answered is like the more the incremental things and also the dynamic of who is asking these questions is also quite fascinating. Maybe I'll leave you with that is that one of the I mentioned earlier the trust score has gone down to 29% from AI related topics, even though adoption of AI has gone up. And one of the fascinating things we noticed the demographic is that the more experience the programmer, the less they seem to trust AI, which is interesting, right? But that is what we notice also it's another sort of quick tidbit from our developer survey is that the experience level age group has gone up and the trust level has gone down. I suspect the younger developers are all like rushing straight to kind of the easy button of like let's use the agent to do a bunch of things, which I think is good. But at the same time again, we would encourage people to do both, learn the fundamentals and do that. And so the expert programmers or the, let's say the more experienced ones are the ones who are spending time thinking about what is the latest programming language and what should I ask, how do I really sort of fully understand how does this hook up into the rest of, you know what libraries I should be using? All the things you need to do and why is this more elegant than the previous version of the same problem?

Demetrios [00:31:05]: Yeah, awesome, dude. I think that's it. Is there anything else that you wanted to touch on that we didn't hit?

Prashanth Chandrashekar [00:31:11]: I think we covered everything. I think maybe one thing would just be, I think we are in the spirit of realizing our vision to be the most vital source for technologists and all those dimensions that I mentioned, whether that's creating community in an age of AI, which is going to be very disruptive and a high change environment, whether that's powering learning for new developers who are looking to learn on the site through our challenges and we'll do all sorts of other things like hackathons potentially in the future, plus for them to build their reputations and then for that to unlock people's careers or even their productivity inside companies through our enterprise products. That's what we're building this with the community, we want to do this with them. So I think my ask would be to your audience to give us a lot of suggestions on how they would like us, how they'd like us to evolve based on their own needs. Right. Because we are moving from just being a Q and A site to much more of a multi dimensional company. And yes, Q and A has changed, the form factor has changed relative to AI. But for us to really work with the community to build a future of stack overload.

Prashanth Chandrashekar [00:32:17]: So we're excited about that. It.