The Grief of Letting Go: What Happens When AI Writes Your Code

There’s a particular feeling that’s creeping into the software engineering community—a sense of loss, of something valuable being taken away. Gergely Orosz, author of The Pragmatic Engineer, put it plainly: “I’m coming to terms with the high probability that AI will write most of my code which I ship to prod.”

This isn’t FUD. It’s a genuine emotional reckoning with a technological shift.

The Core Insight

The grief isn’t about competence or job security in the traditional sense. It’s about identity. For many engineers, coding isn’t just what we do—it’s who we are. The late nights debugging, the “aha” moments when code finally compiles, the flow state of solving complex problems—these aren’t just professional experiences; they’re deeply personal.

And now AI threatens to automate the very thing that made us feel competent.

Why This Matters

Consider what it took to get here. Years of learning C in university, feeling lost in a massive codebase, years of practice, reading books, studying blogs, making mistakes. The craft of coding was earned through struggle.

Now, AI does it faster. For languages and frameworks we’re less familiar with, AI often does a better job than we would. The validation that came from “I wrote this and it works” is being eroded.

But here’s the deeper question: What remains of the software engineer when the coding is done by AI?

The Love-Hate Relationship

Orosz describes coding as “a love-hate relationship”—the intensity of focus required, the time estimation conflicts, the way “time passes differently when you’re locked in and working on a hard problem.”

These experiences shaped us. The frustration was real, but so was the satisfaction. The “YES, it worked!” moment after compilation—that’s not just a checkmark; that’s dopamine, validation, proof of competence.

Looking Forward: A New Zone?

There’s a hopeful counter-argument: perhaps the “zone” doesn’t disappear—it transforms. Instead of typing code, we might find ourselves “in the zone” while instructing AI agents to write complex code. The work shifts from implementation to specification, from coding to reviewing, from building to architecting.

The best engineers have never just been coders anyway. They’ve been translators—converting business requirements into technical solutions, bridging gaps between stakeholders and implementation.

Key Takeaways

• AI already writes code faster and sometimes better than human engineers
• The grief is about identity and craft, not just jobs
• The satisfaction of “I built this” is being replaced by “I directed this to be built”
• The core engineering skills—problem decomposition, system design, quality judgment—remain valuable
• “Being in the zone” may shift from coding to directing and reviewing AI-generated code

Looking Ahead

The question isn’t whether AI will write most code—it clearly will. The question is whether we’ll find new sources of satisfaction in the craft, or whether we’ll mourn the loss of something that was always temporary anyway.

The tools change. The problems remain. And someone still needs to understand what’s being built.

Based on analysis of “The grief when AI writes most of the code” by Gergely Orosz