Stack Overflow

if you use AI a lot to get your answers that means you're not ending up on a forum or on Stack Overflow where you previously would end up getting those answers.

After four amazing years at Stack Overflow, I'm excited to be starting a new chapter. My time at Stack has been a brilliant experience. I've learnt a huge amount, worked alongside truly fantastic people, and had the chance to partner with some incredible companies. Now... I'm absolutely thrilled to be joining daily.dev ! daily.dev are building something genuinely special - for developers, and for companies trying to reach them. I couldn't be more excited to be part of the team! More to come soon 🚀

84% um from the 2025 Stack Overflow service are planning to use AI development in their service in their process

Stack Overflow uh they lost all of their traffic cuz people can now ask LLMs, and obviously it an LLM can can reply to you in a way you understand it, right?

Most AI agents today are being fed junk food. 🗑️ They’re ingesting raw, unverified internal data — fragmented Slack threads, outdated wikis, tribal knowledge scattered across tools — and then expected to execute reliable agentic workflows. The result is high-confidence hallucinations at scale. At Stack Overflow , our 2022 "Code Red" created a clear leadership imperative: in the Agentic era, data access is table stakes. The real differentiator is verified, scored context. In this clip from HumanX 2026, I explain why the biggest bottleneck for AI ROI isn’t model capability — it’s the Trust Gap in enterprise knowledge systems. Stack Internal produces Gold-Standard Knowledge for AI Agents from raw context inside the enteprise. We do this by bringing our 18 years of battle-tested curation, voting, linking, and quality-scoring mechanisms inside the enterprise firewall — turning messy internal noise into structured, high-signal context that both agents and engineering teams can actually trust. The future of engineering leadership won’t be defined by who ships features the fastest. It will be defined by who builds the strongest Verification Engine for their organization’s knowledge. What’s the most expensive technical debt you’re seeing in your company’s internal knowledge systems as you scale Agentic AI? Would love your perspective below. 👇 #StackInternal #AgenticAI #EnterpriseAI #TechLeadership #KnowledgeInfrastructure #HumanX2026

The Context Gap: Why Most Enterprise AI Initiatives Are Underperforming Everyone’s pouring serious money into bigger LLMs and fancy RAG pipelines. The demos look impressive. The outputs sound confident. Then a senior engineer reads it… and does that slow, painful forehead-to-palm maneuver we all know too well. Here’s the uncomfortable truth: Documentation gives you the theory. Code gives you the execution. But neither carries the “why” - the messy architectural trade-offs, the buried technical debt decisions, the tribal knowledge, or the human judgment that actually defines ground truth in your company today. That’s not hallucination. It’s Ground Truth deficiency. And it’s exactly why most enterprise AI projects deliver more noise than value. Stack Internal (Stack Overflow's business solution) was built to fix this. Stack Internal's probably the only enterprise platform that delivers true Contextual + Human-Validated Truth Layer in an organizational Architectural stack . It captures the living knowledge that normally lives in long form documents, Slack threads and veteran engineers’ heads - then turn it into atomic, structured, versioned, and verifiable context your AI can trust. No other platform integrates this human-validated layer at the core of the stack. The result? AI that stops being a high-speed guessing engine and becomes a genuine high-fidelity institutional asset. The 1st phase of AI led development drove ~20% uplift in efficiency. We believe the next 30-40% will come from keeping Stack Internal in the loop - where contextually relevant human-validated truth continuously refines and grounds the AI in real time. The goal isn’t just more AI. It is the trust in AI's output, and doing it responsibly. #AiContext #EnterpriseAI #ArtificialIntelligence #KnowledgeManagement #RAG #InstitutionalKnowledge #StackInternal #DataStrategy #StackOverflow

“I thought you guys at Stack Overflow were just for public Q&A?” We heard this **constantly** at our booth during HumanX . These are some of my favorite moments as CEO — because that’s when we get to flip the script and show leaders the far more powerful enterprise side of Stack Overflow: **Stack Internal**. 🛡️ Most companies are sitting on a massive mountain of raw, siloed context — Slack threads, fragmented wikis, and stale docs. When they feed that into AI agents, they get “AI Slop” — confident hallucinations built on unreliable information. 🗑️ **Stack Internal produces gold-standard knowledge for AI Agents from raw context.** We do this by bringing our 18 years of proven curation logic, scoring, and guardrails inside the enterprise firewall — turning messy internal knowledge into verified, trustworthy context that both humans and agents can actually rely on. A lightbulb moment!💡 The future of AI isn’t just better models. It’s **better context**. 🚀 What’s the biggest source of “Context Debt” in your organization right now? Would love to hear your thoughts 👇 #EnterpriseAI #GenAI Stefan Weitz Jeff Bailey Scott Golden Nick Madison Jason Martin Carrie Koos Emma Bulajewski Johnathon Hawkinson Sander ⌨️ Van Vliet Jason Leach Andy Vadnais Britt Begley Chelsea Temkin Joshua Martinez Ranjit Singh Pabla Nick M. Ryan Donovan Michael Lefler

We had a "Code Red" moment at Stack Overflow in late 2022. 🚨 It forced us to transform our entire business model—moving from Ad-dependence to a high-stakes focus on the Enterprise context layer. I just sat down with Anna Tong of Forbes at HumanX to discuss this pivot and what we’ve learned about disruptive threats. We covered: ⚡️How our customers and users guided us on building our enterprise product focused on the context layer in the enterprise. ⚡️How Stack Internal, our knowledge intelligence platform, produces gold-standard context for humans and AI agents to leverage inside the enterprise. ⚡️How we evolved our Stack Public platform to be a high quality data asset with our AI data partnerships with the leading AI companies like OpenAI and Google and expanded opinion-based questions allowing for rich discussions for all types of users. ⚡️Advice to other tech companies and leaders that are faced with similar challenges in today's rapidly changing AI landscape. Stefan Weitz Andrew Blum