Python for Engineers Who Never Wrote Code in School

I never went through a formal computer science program. What I had instead was a lot of restless curiosity through university, and a habit of teaching myself whatever seemed interesting at the time. I tried circuit design. I spent a stretch messing around with Kali Linux, trying to understand the basics of cybersecurity. I picked up MySQL for database work, then Java, then .NET. None of it stuck long enough to call it a specialty. I was testing water after water, mostly because I could, and partly because nothing had forced me to commit yet.
Python is what actually stayed. Toward the end of my first degree, I made a decision that if I had a problem, I'd reach for Python first and figure out the rest as I went. Data to manipulate, I'd use Python to see how it behaved. Curious about building something simple like a countdown timer, I'd figure out how to do that in Python too. It wasn't structured learning. It was closer to using Python as the tool I picked up whenever something needed solving.
Then I ended up in downstream oil and gas, and the restlessness finally had a direction. I decided I wanted to build tools for this industry specifically, with Python, and that I'd actually commit to it properly instead of moving on to the next curiosity.

What That Path Actually Teaches You

Bouncing between circuit design, Kali Linux, MySQL, Java, and .NET before landing on Python sounds unfocused, and for a while, it was. But it also means I never learned any of it purely from a textbook. Everything I picked up, I picked up because I was trying to make something specific work. That habit carried straight into how I use Python now. I don't sit down to learn a concept in the abstract. I have a real problem in front of me, usually something in downstream operations, and I learn exactly what I need to solve that piece of it.

Where AI Fits Into This

Time is the one thing I don't have a surplus of, especially juggling this alongside everything else I'm building. So I use AI tools to generate boilerplate. There's no real value in typing out the same Flask app factory pattern from memory for the tenth time, or reconstructing a form class I've written variations of before. AI gets me to a working starting point fast. What still has to come from me is knowing whether that starting point is actually right for the problem, and being able to explain why it works once it's in place. That's the part I don't outsource.

Why I'm Writing This Down

I think there's a real gap in how programming content gets written for people like the version of me at university. Most of it assumes either a completely blank beginner or someone already deep into software as a career. There's not much for someone with genuine curiosity and no fixed path yet, testing multiple things before one of them earns their full commitment. If that's where you are right now, the years spent trying different things before it clicked aren't wasted. They're usually what makes the thing that finally sticks, stick properly.
If you went through something similar, several things tried before one actually took, I'd like to hear what that looked like for you.