Just over two months ago, I made a decision that still makes me smile when I think about it.
I decided to teach people how to build things with AI... while I was still figuring out how to build things with AI.
Classic move, right?
Here's the thing: the best way to teach something isn't to wait until you're an expert. It's to share what you learn publicly, and teach it while you're just a couple steps ahead.
And holy shoots, the numbers don't lie:
- 1,926 YouTube subscribers
- 194 newsletter subscribers
- 33 legends (& 2 AI agents) hanging out in Discord
The interesting part is, this is such a new thing (directing AI to code for you) that it's really hard to quantify the results people have had with my tutorials. There's no real bar to compare it to because this trend is only a few months old.
But people keep asking me for more, so hey...
Today, let's dive in to indie startup philosophy, creatorship, AI art therapy, and breaking growth ceilings:
The beautiful paradox of learning by teaching
I just dropped v2 of my Content Commander Challenge. In a gist, it teaches people to get comfortable with 'vibe coding' their own sites and apps by learning to communicate with their computer through the command line.
The kicker? This version features my first AI agent—a support chatbot for students.
Did I know how to build an AI agent when I started creating the course? Nope.
Nor had I ever created an automated system that generates activation keys on a remote server and e-mails the new customer so they can use it to access the course.
And on that note, nor had I ever tried to sell a course that doesn't use Kajabi, Skool, or Circle (etc)
But I did it. All of it. And if I can do it, I know anyone else (almost) can too, with enough dedication.
Speaking from a meta perspective, this is a loop I've stumbled into that I'm finding fruitful: commit to building something before fully understanding it. Share your experience while you build. And "teach it" after that.
Because beyond wanting to offer it as soon as possible, this loop adds the pressure of knowing that one day I will share this technique with my audience... so this forces me to actually learn it properly and not take shortcuts.
The indie startup revelation
I've been reading "Make" by Pieter Levels and it's hitting me like a ton of bricks. The whole indie startup philosophy—solve problems at a niche level, build with what you know—is basically the same strategy one must apply on YouTube too.
My niche? "Using AI tools to build directories."
Apparently. It was undecided by me.
But here's where it gets meta: I'm learning to build these directories with AI... by building directories with AI.
<mindblown.gif>
My interests have long since moved on, however "using AI tools to build directories" is actually the perfect top-of-funnel for my target demographic: tech-competent builders/entrepreneurs with no coding experience that want to harness AI better.
Boom. So I'll take one for the team (several) and continue making directory build videos.
Building the plane while flying it
Daily, my energy shifts between the VibeBuilders.ai project (community) the AI Captains Academy brand (business). It's all coming together in a way that contributes to my main goal of learning to harness this new technology. I'm learning tons along the way.
FUN: I've been deep in the lab with ChatGPT 4o (first time I've actually enjoyed using ChatGPT, btw). For the AI Captains brand, we're going full 16-bit SNES retro aesthetic because I'VE WAITED 25 YEARS FOR THIS!
Current construction zone:
- "Book a call" landing page (done)
- Box art for all the offers (because nostalgia)
- Homepage with a clever retro interface (because fun)
- The AI Captains Academy—teaching the exact tech stacks I'm still mastering
Am I "monetizing" yet? Hell no.
Not really, anyway... because when I say "monetizing," I mean making my goal to make as much money as possible (and if you're my wife, you'll be genuinely concerned that that day will never come).
Why? Because I know that once I master these AI agent workflow automations (and make YT videos to prove it), my time becomes exponentially more valuable. Time I'd charge $100 for today will easily be worth $500 in a couple months.
So right now, I'm invested in building an educational business, and an educational business only. I'm putting out some very low ticket offers just because people are asking me (imagine that), but it's nothing I can even pay the bills with.
Vibe coding and the unexpected joy of pixel art therapy
Here's something nobody tells you about directing AI to code: it's weirdly... unsatisfying?
Don't get me wrong—watching Claude or ChatGPT spit out functional code in seconds is mind-blowing. I'm living in a fantasy.
But my dopamine receptors? They're starving. There's no tactile feedback, no creative problem-solving, no "I built this with my hands" satisfaction.
Enter the AI Captains retro branding project I mentioned.
I've been deep in the lab creating this entire 16-bit SNES aesthetic, and holy moly—it's become my creative outlet. While the AI handles the heavy coding, I get to:
- Design pixel-perfect box art for products
- Create nostalgic UI elements that trigger those sweet childhood gaming memories
- Build this entire retro universe that makes learning AI feel like playing Mega Man
It's closing those dopamine loops that AI-assisted coding leaves wide open. Turns out my brain still needs to CREATE something tangible, even if it's just pushing pixels around.
(Who knew that teaching futuristic AI tech through 1990s gaming aesthetics would be so therapeutically satisfying? SNES games are the only ones I'll let my son play.)
Or maybe I just need to write more.
The growth ceiling (and my next impossible commitments)
Real talk: I'm hitting a wall.
I can't create the content needed to scale while also building my ecosystem AND taking on client work (which must come at some point).
So I'm exploring how hiring a video editor who can follow along with my 'vibe coding' builds—someone who understands we're not just teaching coding without coding, but a new way of thinking... (and of course, someone that can follow my sometimes non-linear recordings.)
Meanwhile, my email marketing game is a pending disaster. I need to build an AI agent workflow that handles everything to do with writing except hitting send. (Another thing I'll figure out by committing to teach it.)
This is all one challenge I have on the back burner that I'm noticing needs to be bumped up to the front of the stove.
Oh yeah, and I'll leave you with my next impossible declaration: 10K YouTube subs by late summer.
I'm pretty sure I've cracked the code enough, tapped into the signal enough, to make it happen. But there's much to do and still many roadblocks to overcome.
But I'll commit to it publicly, to let the pressure do its magic. And so... let's see what happens.
