AI is Creating Engineers Who Don’t Think

I am so jealous of modern web development. Back in the day, when we coded by hand with just a text editor. We solved problems by thinking. Trial and error, not just Googling. As the web became more in use and people started sharing, and we were lucky if we found a relevant Stack Overflow post. Something close enough to iterate on. It was a place where humans experimented, debated trade-offs, and considered speed vs. readability vs. maintainability.

Rise of jQuery and the Shortcut Culture

Then came jQuery (I preferred MooTools, RIP). As a seasoned JavaScript engineer, it was amazing. I no longer had to do boilerplate. All the cross-browser hacks were taken care of. I used it to augment my work, not depend on it. jQuery helped the web flourish ,but...

It also created a whole new breed of JavaScript engineers. Suddenly, you did not need basic fundamentals. You could import jQuery (all gazillion bytes of it) to hide a single element. The uninformed were amazed by how fast these engineers seemed. But did they actually know what they were doing? Not really.

The seasoned engineers used jQuery as a tool to enhance their work. The new engineers never questioned how something worked. Only whether the library did it for them.

Yes, the web became more interactive, more fun, and more user-friendly. But what we got were not JavaScript engineers. We got jQuery engineers. When the library failed, they failed.

The Great Reset: Node.js Changes the Game

Then Node.js came along. There was no more library, well initially. People had to learn to do things again. The jQuery engineers struggled in this world.

It was a wake-up call. Engineers who had relied entirely on jQuery suddenly had to actually understand JavaScript. They had to build things without shortcuts. It exposed who could actually code and who had just been moving fast with the help of a tool.

AI is the New jQuery

Now, we have AI. And the cycle repeats.

We are seeing seemingly brilliant engineers who can prompt their way to success but lack competence. Maybe some future technology will expose this, just like Node exposed jQuery engineers.

The problem is not AI itself. The problem is how we use it.

AI is the Ultimate Copy-Paste Culture

AI tools like ChatGPT, Copilot, and Cursor allow engineers to generate perfect-looking code instantly. The result?

• No deep knowledge of why their code works

• No debugging beyond copy-pasting from AI

• No exposure to real engineering discussions

This is not software engineering.

AI is not a substitute for thinking. It is a tool, just like jQuery was. But when engineers rely on it without first learning the fundamentals, they are outsourcing the most critical skill in engineering: problem-solving.

The best engineers are not the ones who can prompt AI the fastest, but the ones who understand what the AI is doing.

Trust But Verify

I use AI, and it is amazing. It saves me tons of time. But I always go back and read, test, and verify. I have seen AI fail at basic math or outright hallucinate. I do not trust it blindly. I trust but verify.

A Message to Junior Engineers

I urge you, think of AI as a tool, not a crutch. The best way to truly learn is to struggle through problems. Peer feedback, alternative approaches, asking “why” are what make you a better engineer.

Here is a thought. Could you explain your AI-generated code to another engineer?

If not, you do not understand it. And that is a problem.