WoW

Blizzard spent 10 years building World of Warcraft. Even now, with AI that can write code, generate art, music, 3D models, lore, dialogue, everything, it’s still not possible to vibe code WoW. Not in a week, not in a month. AI can replace any individual piece. The code, the art, the music, the story. But WoW is the work of hundreds of people’s minds over a decade. Game designers, artists, writers, each bringing their own creativity and taste into thousands of interconnected decisions. One person can ask AI to produce any of it. But one person can’t direct all of it. There’s a limit to how much creative vision one brain can coordinate, even when AI does all the production.

On the other end, anyone can vibe code a nice-looking app in a weekend now. A todo app, a landing page, a small tool. But that’s the problem. If it can be built in a weekend, someone else can clone it next weekend. No moat.

So somewhere between “weekend app” and “WoW” there’s a line. And the question I keep coming back to is: what does a killer app look like for a solo builder or a small team in the vibe coding era?

I’ve been trying to figure out the formula. I think the first point is clear: interconnected complexity. Not just a big product, but a product where the parts talk to each other. Where changing one thing affects ten other things. That’s what makes WoW hard to reproduce. Not the volume of content, but the fact that everything is wired together. Classes balance against each other, the economy connects to crafting connects to raids connects to PvP. A product like that can’t be cloned by looking at the surface.

For a solo builder, the version of this is smaller but the same idea. A product with enough interconnected systems that someone can’t just look at it and rebuild it, that’s real protection. Not because the code is hard to write, but because the design decisions are hard to reverse-engineer.

That’s one point. I tried to find two more and I kept going in circles.

I thought about accumulation. Products that get more valuable over time through usage, user data, content. Sounds like a moat. But then I thought about paradigm shifts. Mobile wiped out desktop software’s accumulated advantage overnight. AI is doing the same thing right now. Accumulation is a moat until the ground shifts, and the ground shifts faster now than ever. So I’m not sure that holds.

I thought about iteration speed. Keep shipping, be on v5 while someone is cloning v1. But that’s not really about the product. That’s a tactic. And everyone can iterate fast now. It doesn’t make the product itself a killer app.

I thought about marketing. There’s a popular take right now that since building is easy, everyone should just focus on marketing. I get the logic but I think it’s incomplete. Marketing a shallow product just means more people show up, poke around, and leave. A product that can be cloned in a weekend, the marketing angle can be cloned too. It just becomes a marketing arms race with no product underneath it. Marketing a product with no depth is just speeding up the cycle of being found and then forgotten.

The thing I keep coming back to is: a killer app is a product people use deeply. Not Flappy Bird viral, that’s a lottery ticket. I mean products where the more it’s used, the more it gives back. Where users build their workflow around it and don’t want to switch. That kind of product markets itself because people talk about things they actually depend on.

Before vibe coding, building something like that was a team effort. Months of development just to get to v1. Now one person can get there in weeks. The ceiling of what a solo builder can produce moved way up. Things that used to require 10 people and 6 months, one person can now build while the idea is still fresh.

But I still don’t have a clean formula. I know one piece: interconnected complexity. The rest I’m still thinking about. Maybe the honest answer is that nobody knows yet. We’re in the middle of the shift and the rules are still forming. But I think asking “where on the spectrum between weekend app and WoW does my product sit” is a better starting question than most people are asking right now.