Software Engineering After AI: Disruption, Not Displacement
It took me a while to admit that the craft I spent years building might no longer be the most relevant thing about me.
I’ve been a professional software engineer for eight years and have been coding for more than a decade. I’m confident in my tools: Node.js, web servers, REST APIs. I’ve built data pipelines, web apps, backend services, and integration systems by hand, and I took real pride in that work. I used to think of myself as a craftsman. The craft was reading requirements, reading technical documentation, translating all of it into code a machine could run, and then debugging how the machine behaved.
Looking honestly at that list, most of the work is information processing. Reading a thousand-line document used to cost me half a day. An AI model can do it in minutes. AI now handles both sides of the loop: it takes in documentation and it produces code. My daily model isn’t even the most advanced one available, and at high effort, the code it generates is very accurate. Not perfect, but it does a lot of the heavy lifting.
Do I regret learning to code? Absolutely not. Eight years of writing code taught me judgment and intuition, and that’s what I’m still paid for. Coding was the gateway to rooms I would have reached much more slowly otherwise. I don’t write most of my code by hand anymore, but I still produce it, and the code an agent generates under my approval is still my code to manage.
Something unexpected has happened: I feel more ownership over my code now than ever. I didn’t curate every line, so I feel a strong responsibility to make sure what I approved meets the standard I’d hold myself to, or a higher one. I demand more from the agent than I ever demanded from myself. The time it frees up goes into the work I never had room for when I was shipping features every day: more unit tests, security review, performance optimization, and design decisions.
The anxiety is real. The industry feels it too. State of Dev survey 2026 paints a discouraging picture of how developers feel about their future. Seeing a single-page app build in seconds and deploy to the cloud through a one-line script is unsettling, and I won’t pretend otherwise.
We need to start thinking about what comes next: what the next abstraction layer looks like and where engineers add the most value. I don’t have a full answer yet. But I’m not waiting to find out. I’m already in the rooms where requirements get clarified, design decisions get made, and questions about adoption and security get asked. AI has removed the bottleneck of writing code, so the time to sit back and wait is over.
Doom, gloom, and why I’m still optimistic. I don’t believe software engineering is dying. I expect disruption, not displacement. Coding by hand may stop being the job, but building production-grade software was never only about coding.
Some people ask why anyone should still learn to code. I recently started reading Fluent Python and spend my free time writing code for fun, because I love it. As long as people enjoy tinkering with machines, people will write code. Giving up the ability to understand the instructions machines run, at a time when we depend on them more than ever, doesn’t seem wise. Typewriters and computers didn’t kill the joy of handwriting.
On jobs: layoffs have always been part of the industry, and so has fear-mongering. Each wave of change (cloud, no-code, now AI) has come with predictions of doom, and each has reshaped the work more than it eliminated it. I don’t know if that will hold for AI, but the one skill I’m confident matters most is adaptability. Learning is more accessible now than it has ever been, and that’s a real advantage.
Where I land. I won’t pretend the job is as predictable as it was. If your job is taking Jira tickets, writing some code, and opening a PR, rinse and repeat, that predictability is gone. Is that good? I honestly don’t know. I’m equal parts worried and excited. For now, I’m focused on the next two years rather than the next two decades, because the world is moving fast and we have to move with it.
AI disclosure: I wrote this post myself. I used AI tools only to check grammar and sentence structure