We’ve had a breakneck first half of our year here at Stack Overflow, with the last two quarters marked by the announcement of major partnerships, the final availability of our OverflowAI features in Stack Overflow for Groups, and lots of extra milestones throughout our public platform. At a macro degree, we’ve moved properly previous the inflection level of the AI hype cycle. The purchasers, companions, and technologists I interact with every single day grapple with the very actual questions and challenges of how one can implement and use AI instruments to extend productiveness, enhance their developer expertise, construct higher merchandise, and drive worth. So whereas change is fixed—and AI has actually kicked off an excellent cycle in know-how innovation and funding—the widespread thread I see is that belief is the important thing to constructing thriving data ecosystems: on our inner groups, for our prospects, our companions, and in our broader group.
This previous July, we launched our annual Developer Survey, our flagship report on the state of the software program world, with over 65,000 builders from 185 international locations across the globe weighing in. One of the vital putting takeaways for us was the truth that the hole between the usage of AI and belief in its output continues to widen. In 2024 76% of all respondents use or plan to make use of AI instruments, up from 70% in 2023. On the identical time,AI’s favorability ranking decreased from 77% final yr to 72% in our newest survey. Moreover, solely 43% of builders say that they belief the accuracy of AI instruments, and 31% of builders stay skeptical. These outcomes spotlight the truth that, whereas AI is right here to remain, the standard of the outcomes that AI instruments produce remains to be very a lot in query by technologists.
Anecdotally, I continually hear this in my conversations with our prospects. If builders don’t fully belief the output of the AI instruments they use, the entire basis of the home you’re constructing is shaky. For instance, when you’re a backend engineer at a significant monetary providers establishment, are you keen to stake your skilled status on AI-generated code with out human evaluate? Most company prospects will not be keen to take that wager. For this reason validated sources of information are essential to make sure that the inspiration is safe, and why we hold returning to how we are able to bridge this belief hole.
Final month, Stack Overflow celebrated 16 years for the reason that founding of our public platform, and in that point the Open Internet has undergone huge adjustments. This previous summer time, I had the chance to take the stage for the second yr in a row on the WeAreDevelopers World Congress in Berlin the place I mirrored on the entrance seat Stack Overflow has had for a lot of of those transformations, from the rise of social media within the early aughts and into the 2010s, to the explosion of cloud computing, to the wave of societal change and adoption of distant work led to by the pandemic, and now the meteoric shift led to by AI. When you zoom out, GenAI is one other inflection level on this continuum of main technological advances. With greater than 58 million questions and solutions throughout our website, Stack Overflow is uniquely positioned to satisfy this second with our wealth of high quality, correct technical data and essentially the most reliable methodology to generate new data going ahead.
I additionally had the possibility to preview our concentrate on Knowledge-as-a-Service at WeAreDevelopers, or how we at Stack Overflow view the evolution of the community business model within the present info panorama. The methods we discover and share info are shifting dramatically with the rise of GenAI within the final two years, and there’s extra separation between sources of information on the web and the way customers work together with that data. Resulting from these tectonic adjustments, a brand new set of issues have emerged. How will the proliferating variety of giant language fashions proceed to enhance if there are not any new human generated sources of content material for them to coach on? What occurs when an LLM doesn’t have the suitable context or particular data to offer an acceptable reply? And naturally, what can we do once we generally use these AI instruments however can’t totally belief the data they current to us?
The corpus of information and content material on our public platform is continually rising, evolving, and enhancing from the contributions and suggestions from our customers. Mixed with the interior and particular data from our enterprise prospects, this expanded physique of information turns into knowledge-as-a-service, driving each innovation and sustainable group progress.
Selling and fascinating in a wholesome data ecosystem means exchanging concepts and collaborating with different gamers within the sandbox. To that finish, we introduced two milestone OverflowAPI partnerships with GoogleCloud and OpenAI this previous spring. Google Cloud and Stack Overflow brings generative AI to thousands and thousands of builders via Stack Overflow’s public platform, now powered by Google Cloud, and Gemini for Google Cloud supplies code and solutions to builders by way of OverflowAPI. Equally, our partnership with OpenAI delivers high-quality technical content material to strengthen the efficiency of the world’s hottest giant language fashions. These strategic companions share our dedication to advancing the period of socially accountable AI.
As we laid out earlier this yr, our framework round socially accountable AI signifies that we and our product companions who interact with our OverflowAPI product decide to offering attribution to the subject material specialists and group who contributed content material and that the very best high quality content material is pushed and vetted by people. Technical communities and AI might be higher collectively, with suggestions from group customers fueling innovation and elevating the solutions from AI in a virtuous cycle. We just lately added Jeff Bailey to our senior management workforce as Chief Income Officer (CRO) as a part of our dedication to this vital a part of our enterprise. A veteran know-how chief with greater than 30 years of expertise throughout the SaaS and API landscapes, he can be liable for main the income group with a robust concentrate on increasing upon the OverflowAPI enterprise and Information-as-a-Service enterprise product line.
Whether or not it’s at an API degree or in offering sources for his or her careers, we all the time try to be of service to builders. To that finish, it felt like a pure becoming a member of of forces to companion with Indeed on our newest iteration of Stack Overflow Jobs, which now brings hundreds of extremely related technical jobs to builders in each stage of their profession, from seasoned professionals to these simply beginning out within the business.
Keep tuned for updates on extra partnerships, coming quickly!
Stack Overflow for Groups, our market-leading data sharing platform utilized by 20,000 organizations throughout the globe together with Bloomberg, Microsoft, and Nationwide Devices, continues to be the go-to useful resource for builders to seek out the data they want, seek the advice of subject material specialists on their groups, and collaborate asynchronously. Within the spirit of all the time being the place the builders are, we launched OverflowAI in Could to basic availability, a significant step in integrating our GenAI choices to raise the developer expertise even additional. Whether or not it’s of their Stack Overflow for Groups data group, their built-in growth atmosphere with the Stack Overflow for VSCode extension, or in chat apps similar to Slack and Microsoft Groups with our Auto Reply App, OverflowAI represents group and AI coming collectively to ship a world-class expertise to builders and make their lives simpler.
Our groups right here at Stack Overflow are continually iterating, soliciting suggestions from our customers, and making adjustments to make sure we meet the wants of our prospects. We additionally have been proud to ship many new options and upgrades to Stack Overflow for Groups via the spring and summer time with a number of Enterprise releases, together with a reimagined Weekly Dashboard Report to raised perceive and analyze insights out of your data group, enhancements to enhance subject material professional visibility and engagement, and the power to look and filter by way of our Communities characteristic.
At Stack Overflow, one in all our core values is ‘Maintain Group At Our Heart’, and we take that ethos to coronary heart. Our Product workforce right here at Stack is continually discussing methods we can provide again to our international group after the belief and time they’ve put into contributing to our public platform for 16 years. One in every of these latest investments was the launch of Staging Floor to basic availability after a number of profitable beta launches. Staging Floor acts as a sandbox atmosphere for brand spanking new customers on our public platform to attach with extra skilled customers, search steering on the query asking course of, and be taught concerning the norms of the platform. We’ve since seen enhancements in query high quality, and last month we launched a brand new stats module that demonstrates and incentivizes evaluate exercise and affect.
We’re additionally exploring methods during which AI and ML can help group customers in writing drafts for questions and might scale back the burden of labor on our mods and curators by aiding in content material evaluate and flagging old-fashioned content material. Different group initiatives we’ve been engaged on embody lowering sign-up friction to make the method of becoming a member of our group simpler, and we additionally just lately held a week-long Group Asks Dash geared toward resolving small characteristic requests and basic high quality of life enhancements for our customers.
We additionally know that given the shifting info panorama and the rise of AI brokers, lots of our group customers are involved about privateness and accessibility. Stack Trade has a protracted historical past of publishing Q&A knowledge, which has been used for a wide range of purposes similar to constructing group instruments and tutorial analysis. To that finish, we introduced the internet hosting of our quarterly information dump in-house. You possibly can view the info dump for every community website through the user profile. The thought behind that is that these quarterly information dumps are a vital useful resource for our group, however we wish to put safeguards round who’s allowed to obtain the technical content material on our website, and incentivize current and potential companions to speculate again into our group and be a part of us in our mission of constructing the period of socially accountable AI.
At Stack Overflow, we’re dedicated to our group, whether or not that be our public platform customers, our prospects, companions, and Stackers, however we additionally deeply imagine in constructing a future the place our bigger data communities can thrive within the period of AI. A vital a part of that mission is preserving the belief that our group has positioned in us for 16 years, and which means persevering with to speculate again into our data ecosystem. One of many causes I joined Stack Overflow was the chance to make an affect on a group that has the worldwide scale this one does, and I’m impressed often by the various of you who give their time, effort, and data to others. I’m grateful for the belief this group locations in us to proceed to hold ahead our mission of empowering the following era to develop know-how via collective data.
