Early in my first CS Foundations course, I put a simple three-line swap algorithm on the board and asked students to take a couple of minutes and think about it. Not solve it — just look at it. What do you notice? What do you think it’s doing?
In my peripheral vision, I saw a student typing on his keyboard. A moment later, his hand went up. I called on him. He gave me a clear, accurate answer.
I asked him directly: “Did you just look that up?”
“Yes,” he said, without embarrassment.
I redirected the class — the point of the exercise was to work through it themselves, to see what they could figure out before the answer was delivered to them. I asked them to honor that process.
The same student raised his hand again. Good manners — he waited to be called on. Then he asked the question I haven’t been able to stop thinking about since:
“Why should I think when I can just get the answer online?”
I’ve been teaching in the public school system for six years. I spent more than twenty years before that in the technology industry — writing code, managing engineering teams, sitting in the room when hiring decisions got made. I have never heard a question that more precisely captures the problem we are in right now, or the cost of getting the answer wrong.
The student wasn’t being defiant. He was being logical. He had identified an efficiency. Why spend five minutes puzzling through something when you can have the answer in fifteen seconds? In a world that rewards speed, that instinct makes sense.
It’s also the instinct that will quietly disqualify him from the most valuable work of the next decade.
Here’s what’s actually happening when a student outsources the thinking: they’re not just skipping a step. They’re skipping the only step that matters. The confusion, the wrong turn, the moment of not knowing and having to sit with it — that’s not the obstacle to understanding. That is the understanding. The struggle is where the judgment gets built.
And judgment — the ability to evaluate a solution, recognize when something’s wrong, decide what question to ask next — is precisely what the AI-driven workplace needs more of, not less. AI can produce code. It can generate an answer. What it cannot do is tell you whether the answer is right, whether the problem was worth solving, or whether the question you asked was the question you should have asked. That’s the human’s job. That will always be the human’s job, and it is getting more valuable every year, not less.
Students who use AI to bypass the hard thinking aren’t learning to work with a powerful tool. They’re training themselves out of the skill the tool can’t replace.
I want to be careful here, because this isn’t an argument against AI or getting help. In my classroom, I encourage both — students collaborating with each other, and yes, using AI. What I don’t allow is copying. There’s a meaningful difference, and it matters more than most students realize.
My rule is simple: try it yourself first. Sit with the problem for five or ten minutes. If you’re genuinely stuck, get help — from a classmate, from AI, from wherever. But get the insight, not the answer. Understand what you were missing, then close the solved code, go back to a blank screen, and see if you can work it out with your new understanding. The aha moment only counts if it’s yours. If someone hands you the solution and you copy it, you haven’t learned anything — you’ve just filled in a box.
This sounds like a small distinction. It isn’t. The student who struggles, gets stuck, seeks understanding, and then works it through on their own is building something. The student who pastes in working code and moves on is building nothing — and is often unaware that this is what’s happening.
The hiring managers I know — and I’ve been one — are not fooled by this for long. Competency built on shortcuts tends to collapse quickly under real conditions. The junior engineer who can’t debug code they didn’t write, who can’t explain the logic behind a solution, who freezes when the AI produces something plausible but wrong — that person is a liability, not an asset, in an AI-assisted team. The ones who thrive are the ones who can think independently and use AI to go further. You can’t get to that second part without the first.
So what can students and parents do right now, practically?
First: make the AI visible, not invisible. The shortcut is most damaging when it’s hidden — when the student doesn’t even notice they’ve stopped thinking. Name it. If you’re a student, ask yourself honestly: am I using this tool to go further, or to avoid going at all? If you’re a parent, ask your kid to walk you through how they solved something. If they can’t explain it, they didn’t learn it.
Second: value the struggle explicitly. Not in a vague “hard work builds character” way — in a specific, observable way. What did you get wrong first? Where did you get stuck? What did you figure out that you didn’t know before? These are the questions that reveal actual learning, and they’re the questions worth asking.
Third — and this is the one that matters most — understand what the job market is actually selecting for. The students who will thrive in the AI-driven workforce aren’t the ones who can produce the fastest answer. They’re the ones who can evaluate the answer, improve it, and explain it to a room full of people who can’t. No AI supplies that. It has to be built, the hard way, one confused moment at a time.
The student who asked me why he should think when he could just get the answer online wasn’t wrong that the answer was available. He was wrong about what the answer was for.
