I’m pushing 40 and recently started systematically learning about machine learning and large models. It kind of feels like going back 16 years, right after I graduated, when I had all these dev paths to choose from—backend, frontend, Android, iOS. I actually did about half a year as an Android dev intern, but I ended up picking iOS. Not for any fancy reason—just because I loved the feel of it. I loved iOS, I loved the iPhone, and I could stay up all night tweaking a single UI control and feel fine with just a few hours of sleep the next day.
This whole LLM thing gives me a similar buzz. Sure, I don’t have the same energy or curiosity I had in my 20s, but I can still feel that itch—that urge to really dig in and figure out how it works under the hood.
I don’t plan to spend much more energy on plain software development—Go, Ruby on Rails, that whole stack. With AI getting stronger, honestly, that old thrill of wrestling with a tricky problem, hunting for solutions, it’s just not there anymore. You loop with AI, it handles at least 90% of it, and I can cover the rest with experience. In 3 to 5 years, who knows, maybe that last 10% won’t even need me. So what’s the point of writing code then? Just for a paycheck? That was never why I got into this. And that’s exactly why I’m pivoting to AI—not because it’s trendy (if I were chasing hype, I’m already late), but because working on AI research and development feels like it could bring back that spark I had when I first started.
I’ve been going through some classic ML algorithms lately. They’re not that hard once you get what problem each one is trying to solve—then going back to the math makes a lot more sense. And yeah, gotta give credit to AI itself for explaining all those dense concepts in a way that actually clicks.
From now on, I won’t be posting dev-related stuff on my blog anymore—just personal updates or notes from my AI learning journey.