Lol bit of a funny one.

Fabel Five being shut down by the government was a major wake-up call for me. It reinforced my belief that open-source, self-hosted systems are going to play a major role in the future. When you self-host, you are not simply renting access to someone else’s platform. You can actually own the infrastructure, control the data, customize the system, and reduce the risk of your entire operation disappearing because a provider, corporation, or government makes a decision outside your control. I believe the ultimate opportunity will be building a company that creates these self-hosted systems, partners with hardware and technology suppliers, and delivers complete, ready-to-use devices and infrastructure directly to individuals and businesses. The future will not only be about using AI. It will be about owning the systems that run it.

I'd love to get to know you all more. Comment below what timezone you're in, day and time that works best for you. Let's catch up! :)

6 months ago I published a repo called "Business MCP Servers" on GitHub. 15 production-ready MCP server examples. Built with FastMCP. Open source. MIT license. Just a side project. Something I made because I couldn't find good examples anywhere. Yesterday I Googled "hamishfromatech MCP servers" and here's what I found: → Google AI wrote an entire summary calling me "a popular creator in the AI and developer space known for practical, production-ready tutorials and codebases focused on MCP servers." → My repo is listed on MCPMarket.com → It's on LobeHub (one of the biggest MCP directories) → There are 10 PAGES of results → My YouTube videos on MCP servers have collectively hit thousands of views Here's what nobody tells you about building in public: The moment you stop trying to "go viral" and just solve your own problem in the open — that's when the internet finds you. I didn't launch on Product Hunt. I didn't run ads. I didn't cold-email influencers. I just... built something useful and left the door open. And now there's an entire ecosystem of developers discovering MCP through my repo, my videos, my code. But here's the thing that's really messing with my head right now... I almost didn't publish it. I almost kept it private because "who cares about 15 example MCP servers?" The lesson? Ship the thing. Even if you think nobody's watching. Because the algorithm doesn't reward perfection. It rewards consistency in public. What's something you've been holding back from shipping? Drop it below 👇

Not because the tech was bad. The tech was great. But I was solving problems I thought people had, instead of the one problem that actually kept them awake at night. Here's the constraint I finally identified (and it's probably yours too): Most people don't need "AI." They need results without the rental fees. --- That's why I built the classroom the way I did. Not as a library of "cool AI stuff." As a step-by-step path from "I pay monthly for AI" to "I own the stack, the data, and the outcome." The core course — Be Practical — now has 14 playbook folders inside it. Each one is a complete business system you can run offline: Model Mogul, Autopilot Empire, The Silent Sales Engine, Fast Credit, Open-Oracle, The Prompt Millionaire, AI Arbitrageur, The AI-Assisted Agency, The Exit Strategy, The Data Dominance Playbook, Content Cascade, Custom Models & GPTs, Mastering Offline Voice AI, Building Applications with Ollama. Plus the full foundation stack: Private AI Foundations → AI Agents 101 → MCPs 101 → Automation 101 → Vibe Coding 101 → Making Money with Offline AI → AI for Business. --- If you haven't clicked through the classroom in a while, there's a lot more in there now. The question isn't "do you have time to learn this?" The question is: what's the ONE constraint in your business that, if removed, makes everything else easier? If that constraint is "I don't own my AI stack" — the path is already built. Drop a comment with your biggest bottleneck. I'll point you to the exact playbook that solves it.