The Quiet Revolution Needs More Rooms Like This One

The Quiet Revolution

5:19

This morning I joined Ryan Moeller's LinkedIn Live — The Quiet Revolution: How the AI Community Is Building Its Own Launch Pad. I had 60 seconds and a flight to catch. But I've been thinking about that conversation ever since I left the room.

---

Here's what actually happened in that session that you won't get from a highlight reel.

Nino Jambalbo — former CTO, now focused on AI literacy — walked the group through a framework he calls the Game of Work. Level zero through level four. Zero is the person who hasn't touched AI at all. Four is someone whose systems are deeply integrated, agentic, running without them in the loop for most of it.

The thing that struck me wasn't the framework itself — it was how cleanly it mapped to what I've seen in the field. Most people aren't afraid of AI because they think it'll take their job. They're afraid because they can't picture what their life looks like on the other side of adopting it. They can't visualize the path from zero to four. Nino's framework gives them that map. That matters more than most people realize.

Then Jimmy Hayes — my partner at Data Power Supply — dropped something worth unpacking. He described a silent tax running through every simulation, every model, every computation at scale. Small rounding errors. Numerical drift. Compounding across billions of timesteps until your results are subtly — sometimes catastrophically — wrong. Nobody talks about it because it's invisible. You don't know you're getting bad math. You just trust the output. The hurricane analogy he used was clean: forty different models tracking a storm, each producing a slightly different cone. What EK-TORUS does is collapse that cone into a single accurate line. Same compute. Same hardware. Just corrected math. The API is live on AWS. Anyone who wants to test it can reach out directly.

I added one technical layer — this is a Hamiltonian corrector. It's not a patch on top of bad models. It's fixing the underlying integration drift that every numerical simulation accumulates over time. The results we've seen across A100, H100, H200, B200 architectures are real and they're repeatable. We don't make claims. We hand people API keys and let the math speak.

But here's the bigger point I want to make about what Ryan is building.

84% of the world hasn't seen AI yet. Not touched it. Not used it. Not even seen a demo. That number is from a piece I wrote a few months back — there's a chart with a tiny red square in the corner. That's the current builder class. That red square is who's reading this right now.

The AI Launch Pad isn't just a LinkedIn Live series. It's an attempt to build the on-ramp infrastructure for the people who haven't gotten there yet. Community-owned. Open forum. No gatekeeping. People joining from Cairo at 3am, from Miami before a DOE meeting, from wherever they happen to be — with whatever they happen to know — and finding a room where the conversation is real. That's rare. Most AI content is either deeply technical and inaccessible, or surface-level hype that doesn't actually help anyone build anything. Ryan is threading a different needle — creating the kind of open forum where a non-technical person in their 60s and a CTO who's run digital transformations can sit in the same conversation and both walk away with something useful.

We need more of that. Not less.

The conversation also surfaced something I keep coming back to: the pairing problem. Young builders — aggressive, fast, burning through tokens, no fear, no context. Older practitioners — deep domain expertise, instinctive pattern recognition, decades of scar tissue. Both are essential. Neither is sufficient alone. The veterans know what the problem actually is. The young builders know how to build fast. The unlock is getting them in the same room, pointed at the same problem, with AI as the connective tissue. That's not a metaphor. That's literally what happened this morning in Ryan's session. A former CTO and a CFO-turned-AI-community-builder and a team heading to a Department of Energy meeting in Miami all finding each other in an open forum on LinkedIn at 7:30 in the morning.

The revolution is quiet because it's not on the front page. It's happening in rooms like this one.

Thanks to Ryan Moeller for building the room and for the invite. If you're not attending these sessions, you should be. And if you want to test EK-TORUS — drop me a line. API keys are available. Run your own simulations. See what corrected math actually looks like. The numbers are real. The proof is in the output.