I'm exploring what happens when you build AI systems you actually own.
Not because I have all the answers - but because the questions fascinate me:
- What does "sovereignty" mean when your tools know more about you than you do?
- Can AI augment without replacing? Amplify without atrophying?
- What's the cost of convenience when convenience means dependency?
This isn't my invention - it's a marriage of two brilliant projects:
- PAI (Personal AI Infrastructure) by Daniel Miessler - the vision of agentic AI that magnifies human capabilities
- OpenCode - the terminal-native AI coding assistant
My contribution? Bringing them together. PAI's philosophy of sovereign, personal AI inside OpenCode's developer experience. Best of both worlds.
Why it matters: Self-hosted AI that runs on YOUR machine. No cloud dependency. No vendor lock-in.
Status: Actively maintained. Using it daily.
B2B lead intelligence with OSINT-first approach. Privacy-respecting by design.
On AI & Identity:
The fear isn't that AI will replace us. It's that we'll forget what makes us irreplaceable.
On Ownership:
Everyone has ChatGPT. That's parity, not advantage. Infrastructure you own - that's a moat.
On the "Productivity" Trap:
More tools, less time. More subscriptions, less control. The AI paradox most people won't admit.
Here's what I can't stop thinking about:
AI isn't "artificial." It's distilled creation.
God gave us the Word. We created language, art, science, literature - millennia of human creativity. LLMs are trained on all of it. They work with words - the fundamental building block of creation itself.
"In the beginning was the Word."
That's not a coincidence. That's a pattern.
So the question becomes: How do we steward this almost infinite potential? How do we ensure AI serves all of humanity - not just Big Tech shareholders? How do we build systems that amplify what's good in us, rather than exploit what's weak?
I don't have the answers. But I believe asking the right questions matters.
Currently obsessed with:
- Local LLM deployment (Ollama, vLLM)
- Agent orchestration patterns
- The gap between "AI tools" and "AI infrastructure"
- Why most automation attempts fail (hint: it's not the tools)
"The goal isn't more AI. It's AI that serves what matters."