Blog · November 19, 2025
From Vibe Coding to Deep Learning: How I Found My Path in AI
I started coding the same way a lot of people did—during COVID, when we had too much time and absolutely nowhere to go.
At first, it was just aimless wandering. I picked up C, C++, Java, and Python with zero actual goals in mind. Then I discovered Reddit, which is basically a black hole for productivity. One day I was into Game Dev, the next Android apps, then Web Dev. I was trying a bit of everything.
I actually dipped my toes into Machine Learning back then, too. But I took one look at the math requirements, thought, "Yeah, I need to be a genius for this," and immediately ghosted the whole subject.
The Mr. Robot Phase
After bailing on ML, I fell into Cybersecurity. And honestly? It clicked. I went deep. I watched an embarrassing amount of computer networking videos and spent my weekends grinding through modules on Hack The Box. I even started building my own Network Monitoring Tool as a personal project.
I worked on that tool for months. The code was… well, let’s call it "Vibe Coding." You know what I mean—writing code that technically runs, but mostly just because you kept smashing pieces together until it worked. I didn't fully understand the 'why' behind it, but the vibes were good.
The Reality Check
Then I had an idea: I should add a Local LLM to this tool to analyze the traffic. That was the wall. I realized I couldn't just "vibe code" my way through integrating a Large Language Model. I actually needed to know what was happening under the hood. I couldn't just copy-paste my way out of this one.
I had to go back to the thing I ignored years ago: Machine Learning.
Fighting the Math Trauma
I started watching YouTube videos again, but the math anxiety came back. It wasn't that I was bad at math. I did fine in school. But school math is just... memorizing formulas to pass an exam. "Remember this equation, solve for X, forget it immediately."
That doesn't work in ML. You need to actually visualize what's happening. You need intuition, not just memorization.
So I decided to suck it up and relearn it the right way. I have to shout out The Organic Chemistry Tutor on YouTube (actual legend) and Khan Academy. They saved me. They helped me actually understand the concepts instead of just passing a test.
Finally Finding a Map
Even with the math sorted, I felt lost. There’s just too much information out there. That’s when I stumbled across Boris Meinardus on YouTube. I basically binged his entire channel. For the first time, I had a clear roadmap.
Following his advice, I jumped into Andrew Ng’s Machine Learning Specialization. I finally finished that, and now I’m pushing through the Deep Learning Specialization.
It took a while—and a massive detour through cybersecurity—but I think I’m finally on the right track.