Stack Overflow has just signed a deal with OpenAI to provide them with their data to get included in OpenAI's models. This is really interesting to me because there was a ton of drama earlier this year. Essentially Stack Overflow, their company valuation tank, I think the big thing that tanked was the usage metric. So you can track web traffic for sites and Stack Overflow took a massive hit.
It's a developer website where you can upload different comments and questions about code and people are responding. And it took a massive hit because ChatCHPT essentially could summarize an entire thread and just give you the bit of code you need. You don't need to go read through the whole thread. And it was, I don't know why, but it was just an area that very quickly got cannibalized, a site that I think like a 30% decrease in web traffic.
Like the month following ChatCHPT coming out. Because of this, Ac Overflow, you know, shut down access to their site. They made it really hard to tell everyone you're not allowed to scrape our site for data because I think what actually happened was OpeningAI scraped their whole site and included it in their models to get data. Now, I think because of this, Stack Overflow has went through a bunch of different challenges.
I think it kind of felt bad probably for OpeningAI to feel like, look, we scraped their whole site, took all their data, and now their company is literally suffering because of it. I don't know if they're kind or they want to look good in the public eye or what the optics are or maybe they really just want the data. But they've now signed a deal where the Stack Overflow and all of their developer software and tools are going to be, and content is going to be embedded in the next version that ChatCHPT is training. So you could argue like, oh, but didn't they already grab it?
I think you like one use is like, there's a bunch of new data. So maybe the new data is valuable. Again, maybe they want to just say like, look, we're licensing it. So they don't get sued by Stack Overflow or another company.
And the big idea here though, is that with Stack Overflow really being added in, it's interesting. Stack Overflow when this whole thing first happened, they actually banned people from using ChatCHPT responses on the platform because they're worried they was going to make spamming like responses. So you couldn't use ChatCHPT to answer questions on Stack Overflow. But I think they began experimenting with a bunch of different generative features last April.
They said that they were going to build a model that rewards developers who can contribute knowledge to the platform. So trying to come up with some sort of monetization, right? They didn't want to be the person where they just scraped everyone's data and made money off of it, I think, necessarily. So in July, they actually launched a conversational search tool that let users, you could ask questions, you got answers, all that kind of stuff.
And I think they have like a data set right now. Their database on Stack Overflow is over 58 million questions and answers. So this is obviously a huge data set that is very useful. So I think that the developers kind of rebelled against some of the changes allegedly.
They're saying that some of the information wasn't perfect and all this different stuff. And also they said that the data privacy for individual contributors on the platform, there was issues there. I think there were some basis for those concerns. There was an analyst done with that analyzed, I think, 150 million lines of code which were committed to project repos over the last several years by GitClear.
And they actually did find that when it comes to like generative AI and developer tools and all that kind of stuff, they resulted in more mistaken code which was pushed to code bases. So I think some security researchers also kind of weighed in on this and they warned that using those types of tools could, if you have like a bug or a security issue and a software product and it gets like sucked into one of these AI models, now all of a sudden it's going to be amplified. So there, people have all the sorts of concerns. At the end of the day, this is an amazing tool that's really valuable for coding and developers.
And so I think this is going to get rolled out. So Stack Overflow is doing some of their own things. And I think inevitably they realize like you got to kind of go where the stuff is popular, where the users are and that is JWT. So they've now made this deal with OpenAI.
They haven't disclosed like the actual financial terms. But all of this is coming after Stack Overflow also partnered with Google to help Google's Gemini model. It feels like all these Q&A sites are turning into essentially data sites. So you can see like Cora who's very closely integrated with OpenAI and their CEO is on the board of directors for OpenAI.
So like it seems like a lot of these data sets are very valuable, but they want to be very integrated with these AI models because really these AI models can can analyze or kill them in the same way that they're kind of cannibalizing, killing Google search or search has been very disrupted. So Prashanav Kedashkar, who's the CEO of Stack Overflow, previously said that 10% of the platform's nearly 600 staff was focusing on its AI strategy. And he also described a bunch of different avenues and ways that they're going to make money pretty much from making sure that Stack Overflow can keep the best quality information. Like this is really, this is the value of the company that was like they had the best responses.
And so if they can keep the best quality responses better than what else, better than whatever you get from necessarily chat GPT, like they'll be viable, at least as a data set on chat GPT. So he also said in a resuppressed release, Stack Overflow is the world's largest developer community through our industry leading partnership with OpenAI. We strive to redefine the developer experience fostering efficiency and collaboration through the power of community, best in class data and AI experiences. Our goal with OpenFlow API, so this is what essentially they're allowing OpenAI to use is their API.
And our work to advance era of socially responsible AI is to set new standards with vetted, trusted and accurate data that will be the foundation on which technology solutions are built and delivered to users. Sounds like that was all written by chat GPT. I'm sure they had a person in their PR department who did that or maybe they used a little bit of both. It's hard to get away from the technology.
It's good stuff. So all in all, I think this is fabulous. This is the way that these data sites are going to go. This is really interesting to see how this rolls out and if this can save the company and their finances essentially.
If you love the episode, if you enjoyed it, please write us a review. Let us know. Write a comment. Subscribe if you get this on YouTube or follow us on Apple and Spotify where we get your podcast.
I'm thrilled to have you on board and I hope that you have an amazing rest of your day.