Since I was starting my career in software, one rule kept coming back everywhere. Tutorials, courses, senior engineers, everyone repeated the same thing. Do not be a framework slave.
On paper, everyone agrees with this idea. Do not be a “framework expert” or a “tool expert”. Focus on fundamentals. But in reality, I noticed something strange. Many of the same people who gave me this advice ended up falling into that exact trap.
Our job as software engineers is exhausting. We spend eight or ten hours a day staring at a screen, reading and writing code. Most of the time, it is the same framework, the same patterns, sometimes even a company internal framework that only exists in one place. After years of this, it becomes very easy to confuse familiarity with real expertise.
I was always afraid of ending up there. Afraid of waking up one day and realizing that I only knew how to be productive inside a very narrow box. So I forced myself to do side projects. I contributed to open source. I took some freelance work when I could. Not for money, but as a way to protect myself from becoming just another engineer who is very good at one framework and lost everywhere else.
Honestly, it worked. In just a few months, I gained more insight and more perspective than I expected. I learned faster, made more mistakes, and understood why certain things existed in the first place. At some point, engineers with ten or more years of experience started asking for my feedback when they worked on new projects. Not because I was better than them, but because I had recently explored ideas and tools they were just starting to touch. Their experience still mattered a lot, especially in judgment and trade offs. I just happened to have fresh scars from areas they had not visited yet.
That felt good, but it also made me uncomfortable. Being good at many things comes with its own fear. The old saying always comes back. Jack of all trades, master of none. I kept asking myself if I was spreading myself too thin.
Then AI entered the picture. And before anything else, no, I am not a blind advocate for it.
I see colleagues jumping into new areas without a solid software engineering background, relying almost entirely on AI. What they produce is often pure AI slop. It looks convincing at first glance, but it is fragile, hard to maintain, and expensive to fix later. In that sense, having AI on your side can be a serious disadvantage if you do not know what you are doing.
At the same time, for me, AI changed everything in a good way. Tasks that used to take hours of explanations, calls, code reviews, and back and forth with juniors simply disappeared. I delegate a lot of that to AI now. It has been a life changing shift. I feel more productive, more focused, and strangely more creative.
I am not talking like an entrepreneur chasing some hype or trying to vibe code a startup. I am talking as a tech person who always wanted to learn. I could do these things before AI, but now the speed is different. Projects that used to take months can now be done in a much shorter time.
Still, there is a dark side. When I see people from data or machine learning backgrounds jumping into software engineering without understanding the basics, it becomes a real problem. The cost is not just bad code. It is refactoring time, engineering time, and eventually real money. A lot of it.
What frustrates me the most is that it is hard to explain how bad the output really is. My job often turns into cleaning up AI generated proof of concepts. At first, it can be fun. Over time, it becomes tiring. Even more so when the people who created the mess step away and see you as someone who is taking their work from them.
I still believe in that first rule I heard years ago. Do not be a framework slave. Today, I would add one more thing. Do not be an AI slave either.
Experience matters. Fundamentals matter even more. Tools change fast, and confidence can be faked for a while, but understanding always shows up in the long run. The goal was never to know everything. It was to stay curious, stay uncomfortable, and keep thinking for myself. That is still the job, with or without AI.