My published thinking on AI and the future of knowledge work—structured so your AI assistant can load it.
Every content format we have—newsletters, blogs, podcasts, video—was designed for humans to consume. brianmadden.ai is designed for an AI to load on your behalf.
It's all of my published content—blog posts, LinkedIn articles, speech transcripts, interviews—plus files that tell any AI how everything connects: which ideas are authoritative versus still developing, how a new post extends or challenges an earlier argument, what I'm currently thinking about topics I've already written about, and a knowledge hierarchy that tells the AI what to prioritize and where to be careful.
That's what makes this a knowledge system rather than a content dump. Your AI doesn't visit a website. It loads a structured knowledge source and draws on it during your conversations.
Because it lives on GitHub (toomanybrians/brianmadden-ai), every change is tracked. Unlike a newsletter which tells you what someone thought last week, brianmadden.ai lets your AI discuss what changed in my actual thinking, when, and why—and your AI can read it natively.
brianmadden.ai is an MCP server your AI connects to:
brianmadden.ai/mcp as a connector in your AI tool's settings.You're not chatting with brianmadden.ai. You're chatting with your own AI, which now has access to everything I've published. Nothing leaves your conversation, I can't see your chat, and you get my thinking synthesized through your own AI's lens—not a canned chatbot response.
Hi! I'm a futurist at Citrix. I've spent 32 years in enterprise technology, focused on end-user computing (EUC) and the digital workplace. I've written 6 books, 2,000+ articles, and given over 1,000 talks globally. Originally from Ohio (Go Browns!), I now live in Paris. 🇫🇷
At Citrix, I explore how AI is reshaping knowledge work. My core thesis: the real AI transformation is happening worker-by-worker, not top-down. The invisible 80% of knowledge work—the judgment, reasoning, and tacit expertise that lives in workers' heads—is where the action is. Enterprise AI automates the scaffolding. A second brain amplifies the cognition.
Supporting essays:
brianmadden.ai is me publishing my second brain like I outlined in that second essay above.
Is brianmadden.ai a chatbot? No. It's a data source your AI connects to. You talk to your AI, and it draws on my knowledge when relevant. It's more like an always-updated context & knowledge source for your chatbot.
Is brianmadden.ai a digital twin? No. A digital twin simulates a person. This makes published thinking accessible. It's a living reference library, not a simulation.
Where does brianmadden.ai come from? I maintain a personal AI-powered knowledge system (sometimes called a "second brain") which is how I use AI on a daily basis. That system has everything I need to work: content, ideas, my todo list, thinking, meeting notes, documents, plans, etc. brianmadden.ai is a subset of that private system, updated daily. Content flows from that private system to the brianmadden.ai public repo using explicit publishing principles.
How current is brianmadden.ai? The "current thinking" file updates frequently. (Click here to see the latst updates.) The synthesis updates when new posts are published. The AI flags stale content rather than presenting old thinking as current.
Can I fork brianmadden.ai? Yes. The GitHub repo is the source of truth. Fork it, build on it, use the frameworks in your own work.