What I believe
about building.
For decades, building software needed a gatekeeper: a degree, a title, an engineer willing to take your idea seriously. I never had any of that. I build with AI anyway, and I have run enough experiments to know the workflows, tools, prompts, and skills that help PMs and founders build AI-natively.
I am handing over everything I have figured out, and everything I figure out next. Take it so you skip the trial and error it took me to get here.
- 01
The gate is gone.
No computer science degree. No engineering title. No one signing off on whether you are allowed to build. If you can think clearly about a problem, you can build the thing that solves it. I never took a coding class and I ship products anyway.
- 02
AI runs through every step.
From the first problem sketch to the last go-to-market call, AI sits in the loop. It is not a feature I add at the end. The whole way of working shifts once you build with it from minute one.
- 03
Judgment is the job now.
AI made the building cheap, so the hard part moved. Picking the right problem, and knowing when the work is actually good, is where the real effort goes. Taste still separates people.
- 04
Ship the mess.
A working version beats a perfect plan, and it beats a polished thread by a mile. I would rather put something real in your hands than sand it down for another month.
- 05
Show the work, not the highlight reel.
You get the real workflows, the exact prompts, and the calls that went sideways along with what I did next. People learn from the mess, not the trophy.
- 06
Steal it.
Everything here is yours to take and use. Knowledge compounds when people pass it around. Hoarding it helps no one, me least of all.