I built two things with AI last week. One feels great.

I built two things with AI last week.

One of them I'm proud of.

The other one I'm still figuring out how I feel about.

If you've been around here for a while, you might remember the newsletter where I talked about building a silly video game featuring my eighth-grade band on a quest to become famous. I'm a nerd at heart. And I've been completely mesmerized by what's possible right now with vibe coding — finding efficiencies in the business, creating new opportunities I never thought were even on the table given the resources I have.

A small humblebrag while I'm here: I've quietly released two pieces of software through my businesses that are generating actual recurring revenue. Nothing crazy yet — a few hundred bucks a month — but it's real money from products I built without a development team. A year ago I wouldn't have believed that was possible for someone like me.

That brings me to the first thing I built last week.

I've been working on a video strategy assessment as part of the book — basically a quiz that tells you whether you actually have a video strategy, and if you do, where it's leaking and what to fix first. It's nine pillars, twenty-seven questions, and you get a personalized report at the end.

Here's the thing most people don't realize: a real video strategy — the kind that's actually showing up on your website and your socials and doing the work for you — can save you a ridiculous amount of time and make you a lot of money. Most businesses don't have one. Most don't even know what one looks like.

I almost paid a SaaS company a monthly fee to host this thing. ScoreApp, Interact, Typeform — they all do it. They all charge for it. And for a long time I assumed that hundred bucks a month was just the cost of doing business.

Then I opened Claude Code.

I started prototyping the assessment on a Tuesday. By Thursday I had a working version, hosted for free, with all the scoring logic and the results page and the email capture working end to end.

I'm not a developer. I've never written a line of production code in my life. And I just shipped a quiz that would've cost me a hundred bucks a month and a whole setup process.

That felt like magic. The kind of magic where you start looking at every other line item in your business and asking what else doesn't have to cost what you've been paying for it.

If you've heard people talking about Claude Code or Claude Cowork and weren't sure what the fuss was about — this is the fuss. It's not just for engineers anymore. It's for the rest of us who have ideas and have been waiting for someone else to build them.

The second project felt different.

I'm wrapping a video for a software company right now. They're an underdog in their space — a much bigger competitor recently dropped over a million dollars on a real, honest-to-God documentary film. Big crew, real interviews, the whole production.

My client can't afford a million dollars for a video. Not even close.

So we're making a mockumentary. AI characters. AI voices. AI footage. We're using every tool in the box to create something that looks like a documentary, except none of the people in it are real.

And I'll tell you — it's stunning what we've been able to do for what they're paying. The quality is beyond anything I would've believed possible even a year ago. It looks real. It feels real. The story works.

The client is over the moon about it. They're already talking about doing more projects with the same AI characters after this one wraps. They think if we package this up and offer it as a service, it's going to be gangbusters. And honestly? I think they're right. There's real demand for this kind of work, and very few people can pull it off at this level.

But every time I sit down with my team to look at the edits, I notice this little knot in my stomach.

I've spent twenty years of my career finding real people and capturing real moments. That's the whole point of documentary. The whole point is that it actually happened. That this person actually said this thing. That you're watching something true.

And now I'm building something that looks true and isn't.

The justification I've been telling myself is that it's a mockumentary. It's not pretending to be real. We're not trying to trick anyone. The whole format is the joke. We're poking fun at the genre and we're poking fun at the giant competitor who spent a million bucks on the real thing. It's supposed to make people laugh, and it's supposed to make them question whether the polished version is actually any better.

That justification is real. I think it holds up.

But the knot is also real. And the fact that there's a viable business waiting on the other side of saying yes to more of this work? That makes the knot harder to ignore, not easier.

Here's what's strange.

Both projects use the same tools. Both projects let an underdog (me, my client) compete with budgets ten times bigger. Both projects are technically remarkable in ways that wouldn't have been possible eighteen months ago.

But one feels like leverage and the other feels like trespassing — at least a little.

I think the difference is what's underneath.

The assessment is my thinking, my framework, my voice — Claude just helped me ship it faster. The mockumentary is the appearance of human experience without any of the actual humans.

That's not a moral judgment on the client or the project. They're being smart with their money and they're solving a real problem in a clever way. But it's making me think harder about where I want to plant my own flag.

For the work that feels most meaningful — the brand films, the founder stories, the documentaries about real people doing real things — I don't think I'll ever use AI actors. The whole reason people hire me for that work is to find their actual story, with their actual face, in their actual voice. The minute I outsource any of that, I've sold the thing I'm actually good at.

But there's a whole other category of work where I think I will use AI actors — fun stuff, mockumentary stuff, satirical stuff, the kinds of videos I'd love to make but never had the bandwidth or budget to pull off with a real crew. That feels like a different game with different rules, and I'm genuinely excited about what's possible there.

I'm still figuring out exactly where my line is. But I think it has something to do with whether the thing being created is supposed to be real in the first place.

A quiz isn't a person. A framework isn't a story. A mockumentary isn't a documentary.

Maybe that's the line.

Maybe.

Until next week — go build something. Just make sure you know what kind of thing you're building.

Your story is the strategy,

Ryan

P.S. — The book is closer to done than it's ever been. I'm not running a waitlist or anything fancy yet — just keep an eye on the newsletter. When it's time, you'll hear it here first.

P.P.S. — My friend Mike Kim wrote two newsletters last month that are sitting right next to this one in my brain. One is called The Tabloidization of the Internet, the other is Visibility Fatigue. If you're trying to make sense of where content is going right now, his stuff is some of the clearest thinking out there. You can find them at mikekim.com/newsletter.