The AI Acceleration Era: Making Two Games in a Day and Starting a Web Agency

Everything Changed (2024-2025)

Look at these numbers:

That’s not a typo. Q3 2025: 2,189 commits. More than my entire output from 2020-2023 combined.

AI changed everything.

The StickerApp: Building for My Kids

Let me start with something personal.

I have three kids, all six and under. If you’ve ever tried to manage a reward system with young kids, you know the chaos. We’d promise stickers for good behavior, but then:

• We’d be out somewhere without the sticker chart
• My wife and I would lose track of who promised what
• The physical chart would get messy or lost

So I built StickerApp.

A digital sticker board. When the kids do something good, they get a sticker. Each sticker is worth about $0.50, so we use it as money too. Want to buy something at the store? Count your stickers.

I remember one time at Walgreens checkout, my son saw a candy bar he wanted. It was about $0.50, so I told him it was 7 stickers. He thought about it, checked his sticker count mentally, and said: “Oh, never mind.”

That moment made every hour of development worth it. Teaching kids about delayed gratification and financial trade-offs through a silly app I built on evenings and weekends.

The NodeBuster Project

Q4 2024 shows 268 commits on NodeBuster. This was fun.

The game NodeBuster was a puzzle game written in Godot. I reverse-engineered it, got it running, and then started modifying it to be co-op. The original was designed for single player, but I wanted to play it with friends and family.

It ran on phones. It ran on Linux. The netcode actually worked. I had fun with it for a while, then (like most projects) moved on when something shinier appeared.

The AI Thesis

Here’s the thought process that changed my 2025:

AI is accelerating. Rapidly. If the trajectory continues, it’s going to displace a lot of white-collar jobs. Programmers included.

I have two choices:

1. Stay an employee and hope I’m not one of the displaced
2. Ride the wave as a business owner, leveraging AI as a multiplier

I saw a guy on Reddit posting about making multiple six figures just building websites for small businesses. Nothing complicated, just solid, functional websites. And I realized: I could do that. I already knew how to build things. AI would just make me faster.

So I started Web Town Hero.

The Web Agency Journey

Summer 2025. I launched the agency.

The first several months? Nothing. Crickets. I was learning sales, which is a completely different skill from engineering. Building the product is easy. Finding people who want to pay for it? That’s the hard part.

But I kept at it. Cold outreach. Networking. Building my portfolio.

And then clients started coming.

My model is different from the typical agency. Instead of charging a big lump sum upfront, I do monthly subscriptions. $200/month, one-year minimum contract. Most projects I complete 95% of the work in that first month, so it’s definitely an undersell for what I deliver.

But here’s the thing: recurring revenue compounds. If I keep clients happy, they stay. And every month, the base grows.

As of late 2025, I’m at almost $1,000/month in subscriptions. That’s not life-changing money, but it’s not nothing. More importantly, it’s a safety net. If something happened to my day job, I have something to fall back on.

The Commit Explosion

Look at Q2-Q4 2025:

• Web-Town-Hero/lead-generation-admin-web: 312 commits
• sdg9/html5-coop-digital-card-game: 828 commits
• Web-Town-Hero/client-murren-properties: 274 commits
• sdg9/white-label-app: 500 commits

That’s just the top repos. The total commit volume is staggering.

AI didn’t just make me faster at building things. It made me faster at starting things. Ideas that would’ve taken weeks to prototype now take hours. The friction of “is this worth building?” dropped to near zero because the answer is usually “let’s find out.”

Making Games with My Kids

December 2025. Naptime for my youngest. My other two kids are awake.

“Let’s make a game,” I say.

We made a Pokémon-themed Vampire Survivors clone. Together. In front of them.

The three of us were playing it on controllers within hours. You could pick from four different Pokémon, each with distinct attacks, fighting waves of enemy Weedles that would spawn and come at you. Vampire Survivors meets Brotato gameplay, but with their favorite characters.

Later that same day, I made a mobile toilet-themed football game. You’re a pixel-art football player trying to run as far as you can. Every first down or special zone gives you an upgrade. Roguelike elements on a simple endless runner.

Two playable games in one day.

That would’ve been impossible two years ago. Not because I couldn’t code, but because the time required would’ve been prohibitive. AI collapsed weeks of work into hours.

The Philosophy Remains

Here’s what’s interesting: the AI acceleration didn’t change what I build. It just changed how much I can build.

I’m still a tinkerer. Programming is still digital Legos. My goal isn’t to build a business empire or ship AAA games. It’s to scratch itches, solve problems, and occasionally stumble into something useful.

The StickerApp helps my kids learn about money. The web agency provides a safety net. The games I make with my kids create memories. The client websites help small business owners get online.

None of it needs to be the next billion-dollar startup. It just needs to be useful and interesting.

Looking Forward

The web agency keeps growing. I’m hoping it can become substantial enough to be a real option if something happens with my day job. Not because I want to leave (I don’t) but because having options is valuable.

The game development continues. CardCoalition/html5-coop-digital-card-game is still evolving. I still want to play a co-op Slay the Spire with my friends.

And the random projects will keep coming. That’s just who I am.

Twenty Years of Digital Legos

Looking back at this retrospective, a few things stand out:

The motivation was never money. The Payday 2 success happened because I built something I wanted, not because I chased profit. The tax software fizzled when the Rich Dad motivation faded. The projects that stuck were the ones I found intrinsically interesting.

Most projects don’t ship. And that’s okay. Every abandoned repo taught me something. Every unfinished game made me a better developer. The journey matters more than the destination.

Side projects feed the day job. React Native at my company? That came from my side projects. Mezzo at work? Side project. The infrastructure knowledge from Kubernetes game servers? Side project. The skills transfer, even when the projects don’t.

Life has seasons. The game dev intensity of 2019-2021 couldn’t last. The quiet of 2021-2022 was necessary. The AI explosion of 2025 is happening because the foundation was already there.

Family makes it better. The Walgreens sticker moment. Making games with my kids. My wife suggesting the Android app that paid for our pizza. The best side project memories involve the people I love.

The Next Twenty Years

I don’t know what I’ll be building in 2035. The technology landscape will be unrecognizable. AI will be doing things we can’t imagine. My kids will be teenagers with their own interests.

But I know I’ll still be tinkering. Still scratching itches. Still treating programming like digital Legos.

That’s the constant through all of this. The tools change. The projects change. The motivation stays the same.

Build interesting things. Learn constantly. Share when it helps others.

Everything else is just details.

Thanks for reading this five-part retrospective of my side project journey. If you’re a fellow tinkerer, I hope seeing the messy reality (the abandoned projects, the scattered focus, the bursts and lulls) is encouraging. You don’t need a polished success story. You just need to keep building.