"Why Study for Certifications When AI Exists?" — Returning to Basics

One night, with a textbook open on my desk, a small voice surfaced: “Is there any point in doing this myself?”

For the past several months, I have made AI-driven development the core of my personal projects. The impact has been bigger than I expected: learning costs drop to almost nothing, and productivity soars. An app takes shape in a few weeks even in a language I had never touched — exactly the experience I described in my piece comparing AI-driven development with the traditional process stage by stage.

But as I grew used to that convenience, two quiet worries took root.

Worry 1: In the end, the human’s knowledge is what matters

The AI does the implementing, but judgment is not its strength. As a rule, the human decides the direction, and the AI moves its hands and gathers information along that decision. That is exactly why the human needs the knowledge to steer the AI toward what is most likely correct.

You often hear that “apps built with AI have weak security.” I think that is a textbook example. Someone without the right knowledge says “it’s easy to build,” and orders an app that merely works from the AI. The hole isn’t in the AI’s ability — it’s in the asker’s underlying knowledge.

As I touched on in my article about using Claude as a partner for research and decision-making, the one who makes the final call is always the human. And the quality of that call rests on the depth and breadth of the knowledge you carry.

Worry 2: I start demanding fast answers

In Japanese or English, AI gathers information at a speed no human can match. Once you get used to it, you start expecting an instant answer to everything.

While studying for the Information Processing Engineer Exam, there were stretches where I thought, “AI could just look this up for me…” Reading a textbook one page at a time suddenly felt unbearably slow.

Apps build fast. I have the AI collect security information, I judge based on it, and I have the AI execute. Repeating that loop, I even reached the thought: “Maybe I don’t need to study for certifications anymore.”

And yet, what built today’s me was the accumulation

So I stopped and thought it through. I am where I am because of all the people and work I have engaged with. And the foundation under it was the motto I live by: “keep studying, keep learning, one step at a time.”

Move forward just 1% a day and a year later you are about 37 times further along — I have long believed in this “rule of 1.01.” Steadily shoring up my base with certifications has been part of that accumulation. Having an answer delivered by AI and having knowledge accumulate inside you look similar, but they are entirely different things.

Take the networking and security fundamentals the exam asks about: AI returns an answer in ten seconds. But only someone who has walked through that same ground in their own head can tell, in an instant, whether the answer holds. Borrowed knowledge slips through your fingers; knowledge you built yourself stays — as speed of judgment, and as the instinct to sense when something is off. Whether you can steer instead of swallowing AI’s reply whole comes down to whether that foundation is there.

My philosophy is “never assume you know it all; keep learning forever.” Ironically, the more AI answers instantly, the more easily we feel like we understand. That is exactly why slow, hands-on learning is something I need more of right now, not less.

Returning to basics

So I am returning to basics. I will face the Information Processing Engineer Exam head-on.

AI is the best partner I could ask for, and I intend to keep using it to the fullest. But the knowledge to steer that partner in the right direction — that, I want to keep building with my own hands. The intention to “keep learning,” which had nearly dissolved into convenience, I am taking firmly back into my grip.

The studying that looks like a detour is what pays off most in the age of AI. Believing that, I open the textbook again today.