72knots.ai · All Episodes
The Episodes
Two arcs. The Build (2026) asked whether an operator, alone, could take AI all the way to enterprise-grade systems. Ground 0 is the answer — one operator, four months, five systems in production, zero middleman. The record, as it happened.
Ground 0 · The record · Live now
Ground 0 — 4 Months, 5 Systems
The Build ended with a promise: "the real test starts now." I was running the test on myself. Each episode of Ground 0 is one piece of the record — where it started, what it takes, what broke, and what it proves.
Ground 0 — the episodes ahead, and what each one tells
EP 2I Already Knew the Job — describe, test, break, fix. The method was always mine; the developer was the interface.
EP 3For the New Starters — what I'd tell an operator opening a terminal tomorrow.
EP 4The Toolbox — the garage access control I built in my backyard, and why operators can do this.
EP 5Four Months, Five Systems — the scoreboard. Twelve years the old way vs. four months this way.
EP 6Break It Like You Mean It — testing AI-built software like it's real software. Because it is.
EP 7The Boring Proof — PCI, security, live payments, automated deployment. The receipts.
EP 8Get Out — software is a commodity, not a business. Parking belongs to parking people.
EP 9The Unthinkable — what comes after five systems.
The Build · 2026 · Complete — answered by Ground 0
The Build
Where it started: the operator adoption problem AI companies keep hitting, why the obvious solutions don't work — and the question Ground 0 now answers.
Can You Hear Me? The Hotel Room Version
A patent sat unused for seven years. Not because the idea was wrong. Because the middleman was brutal. Then came the couch, the Heat game, and two weeks that changed that.
Read episode →
The AI Finally Asked the Right Question
I showed ChatGPT my valet system without telling it who built it. It rated it 8/9 for operational power. Then it asked if I was an operator. Twenty years in the business. Zero lines of code written.
Read episode →
Can You Hear Me? When AI Tried to Teach Me My Own Business
I'm on the couch. Heat game roaring. AI jumps in with pricing strategy and go-to-market advice. I told it to shut up. Then the real work started.
Read episode →
Getting Rid of the Middleman
One month ago, half these tools didn't exist. Built apps before my coffee cooled. Couch. Heat game on TV. Eyes wide. This is Day One of building enterprise systems with AI.
Read episode →
5
Everyone Asked "What If You Get Hit by a Bus?" Then They Stopped.
When I started AVPMi, every prospect asked the same question. They weren't worried about the code breaking. They were worried about the business dying. AI is at that same stage right now.
Read episode →
4
The Problem They Didn't Hire Me to Solve
A hospital hired me to fix their valet chaos. That wasn't the real problem. I fixed what they asked for — then walked the floor and saw what was actually killing them.
Read episode →
3
Claude Built Something Great. Now Tell Us What It Does.
Anthropic dropped a genuinely impressive model. But they're speaking IT, not business. You won on product. You're losing on communication.
Read episode →
2
I Was the User Who Became the Builder
In 2004, I was a valet attendant using an automation system that didn't match operational reality. I knew exactly where it failed because I lived it every shift. So I built one that worked.
Read episode →
1
The 95% Problem
AI companies fight over 5% of the market — technical users who understand prompts, databases, deployment. Someone needs to solve for the other 95%. That's operators.
Read episode →