AI and Programming: A Weapon for Ninjas, a Trap for Beginners

I’ve been coding for 22 years. Ruby, Rails, APIs, frontends—you name it. And for the first time in my career, I feel like I have an army behind me.

When I use AI for coding, I move at lightning speed. I can focus on designing solutions while the AI fills in the boilerplate, scaffold tests, or drafts functions I’d otherwise lose hours on. For me, AI isn’t replacing my work—it’s multiplying it. It’s a dream come true.

But here’s the catch: I asked my team to do the same, and the results weren’t nearly as good.

AI in the Hands of Juniors

Most of my team is less experienced than I am. They haven’t had decades to master debugging, system design, and the subtle patterns of good code. So when they use AI, it often works against them.

The problem isn’t that AI writes bad code—it’s that they can’t always tell when it does. They copy-paste, run into errors, and get stuck. And unlike failing on their own code (which teaches you a lesson you’ll never forget), failing on AI’s code leaves them confused. They don’t know what went wrong, because they never really understood what was happening in the first place.

That’s dangerous.

My Rule as a CTO

I told my team this:

“If I give you a task and you do it yourself and fail, that’s fine. You’ll learn, try again, and eventually win.
But if you use AI and fail, that means I gave you a task and, even with the most advanced tool around, you couldn’t deliver. That’s a bigger problem.”

AI is not a shortcut. It’s a lever. And levers only work if you know where to place them.

The Fix: Training and Education

The solution isn’t banning AI—it’s doubling down on training. As leaders, we need to catch when juniors rely too much on AI, and redirect them to learn by doing, failing, and understanding.

AI will get better. One day it might deliver perfect solutions on its own. When that happens, our role as engineers will shift—we’ll spend most of our time designing systems and QA’ing both the output and the implementation.

That future isn’t here yet. But the skills you’ll need for it—critical thinking, system design, and quality assurance—can (and should) be developed right now.

Final Thought

AI is the deadliest weapon in the hands of an experienced ninja. In the hands of a beginner, it’s more like a trap.

If you’re leading a team, the answer isn’t to avoid AI—it’s to train your people to understand it, question it, and outgrow it.

Because when the time comes, those who can QA AI will be the ones who lead the next generation of software development.