All of the following is my own perspective and how I currently see things. I could be wrong in some areas.

For more than two years, I have been using AI, and it has been absolutely amazing for developing features, MVPs, and POCs. I really enjoy it. It gives me the balance and the ability to do much more, both in life and at work, alhamdulillah.

But you know, we are becoming more like orchestrators than traditional coders. And honestly, sometimes I still feel a kind of insecurity when I don’t go through too much debugging, coding, searching on the internet, copy/pasting code from different places, trying to understand what we have and what we are doing, asking friends, and so on. This feeling was particularly strong around the coding part.

When I finish a task without touching the code, or barely touching it, I sometimes feel afraid, insecure, and maybe experience other negative feelings. Because at the end of the day, you need to keep your role, keep doing your work, keep your position, and continue to provide for yourself and your family. So naturally, you start asking yourself, Am I still developing as an engineer if AI is doing so much of the coding?

I don’t know how I will deal with this feeling in the future, but for now, I’m trying to approach it differently. 

I’m trying to focus more on software engineering itself. Understanding more than I used to know. Going deeper into backend engineering fundamentals, ports, proxies, protocols, DevOps, infrastructure, and many other areas beyond just coding and code style.

I’m also trying to understand more about the stack I’m currently working with, while at the same time implementing things using other stacks when I have the opportunity, because I want to understand the end-to-end flow of a feature rather than becoming attached to a specific framework or programming language.

I’m focusing more on understanding requirements, refining them, being responsible for the specifics, documentation, architecture decisions, and the overall engineering process. And when time allows, or when I’m working on a hobby project, I still try to write some code manually first and then use AI tools to refine it.

I think the goal isn’t to prove that I can write more/better code than AI. The goal is to make sure that when AI writes the code, I understand why it should exist, what problem it solves, where it belongs, what trade-offs it introduces, and whether it actually fits the system I work on. 

Maybe this is one of the adjustments we need to make as developers. AI can take away a lot of the struggle that used to make us feel like we were “doing engineering”. But engineering was never only about the struggle of writing the code.

And maybe now we have to become better at everything around the code, while still keeping our ability to understand and write it when we need to.

It’s a strange, hard and biiiiiiig transition.