Coding with AI, an endeavor
For years, coding has been about building things brick by brick.
Developers love to dive deep into logic, shape ideas, and watch them come to life through code.
But something is changing.
With AI tools like Cursor, coding feels faster than ever. It’s like having a supercharged assistant that can turn hours of work into minutes.
You describe the problem. It writes the code.
And if it gets things wrong, you nudge it in the right direction. Give it more context. Refine the prompt. Eventually, it delivers.
At first, it feels magical.
But after weeks of this, something unexpected happened.
I started feeling less like a developer and more like a prompter.
The excitement of debugging, optimizing, and piecing logic together was fading.
Instead, I was spending my time fine-tuning instructions for the AI.
It’s not that prompting isn’t rewarding. It’s satisfying in its own way, getting the AI to produce working features can feel like an achievement.
Yet, there’s something missing.
I wasn’t writing the code anymore.
Developers care about the details.
We obsess over clean code, elegant variable names, and consistent indentation. We fine-tune parameters. We pick just the right compiler version.
AI tools don’t share this obsession. They deliver functional code, but the soul of craftsmanship often gets lost.
And with that, a subtle but important part of programming: You name it “control”, slips away.
I still believe developers are needed.
We bring creativity and systems knowledge to the table. We know how to structure complex systems and design scalable architectures.
AI doesn’t replace that.
But it’s changing how we interact with code. Instead of solving problems ourselves, we’re asking AI to solve them for us.
And that shift might cost us the joy of building.
I’m not sure where this leaves us.
AI-driven development is here to stay, and it’s undeniably powerful. But as we move forward, we need to ask: how do we keep the joy alive?
How do we stay connected to the craft?
I’d love to hear your thoughts.