When people talk about building with AI, the focus is usually on individual productivity, how much faster one developer can move, how much code one person can generate.

My latest AI surprise wasn’t individual speed.

It was what happened when our entire team started building with AI.

At DBGorilla, we’re building an AI-powered DBA Assistant (hit me up if you want a demo) which means we use AI everywhere (like really for everything).

So with a very small team, we suddenly found ourselves producing the volume and complexity of code you’d normally expect much later in a company’s life.

At first, it was great! Features moved quickly. Ideas went from concept to production in days. The momentum was and is real.

But then we started hitting problems that usually belong to much larger organizations: bottlenecks, assumptions falling out of sync, and infrastructure straining under the load.

Nothing was obviously broken. There were no dramatic failures. But the system had quietly stopped scaling.

That was when it clicked for me:

AI doesn’t just scale developers. It introduces large company complexity into small teams.

So if your systems don’t evolve with that new scale, speed turns into slop.

In other words:

We didn’t need better Engineers.

We needed better systems.

More on that soon.

Question - If AI 100x’d your team’s output tomorrow, what would break first?