When I was first learning how to code, I spent tens - maybe hundreds - of hours glued to online courses. Video after video. Tutorial after tutorial.

Once I “made it,” though, I found myself more drawn to books. Maybe it was because my employer stopped paying for PluralSight. Maybe because I suddenly had an O’Reilly subscription. Or maybe because I realized video pacing never matches what I need - either way too slow (wasting hours) or too fast (forcing me to rewind endlessly, or drop the playback speed to 0.5x).

These days, I don’t pay for either. AWS won’t reimburse for O’Reilly (Seattle residents get it free through their library - unfortunately, I’m not one of them). And while we do have access to AWS SkillBuilder, it feels more like marketing fluff or entry-level intros - stuff I can figure out on my own just by clicking around.

To be honest, I haven’t read a technical book or followed a structured course in over two years. I just read the docs. And when I need clarity? I ask ChatGPT.

Which brings me to the real question: Are tech tutorials even necessary anymore?

---

Experimenting with AI Learning

Last week, I thought: maybe it’s worth learning about AI through structured courses, since we’re already using it so much. So I tried “ChatGPT Prompt Engineering for Developers” and went through AI Coding University on Cline’s site.

And what happened? I realized I didn’t learn much that I didn’t already know. Because I’d already learned the hard way - by doing.

Still, there was a part of me that liked the structure. Even if I didn’t immediately need every feature, at least I came away knowing those features exist. Sometimes, awareness is its own kind of learning.

---

When Structure Still Matters

So no, I’m not done with courses or books forever. But I think they serve a narrower purpose now:

1. Starting something new - When you have no idea where to begin, a course can guide your first steps.
2. Expanding your breadth - When you want to know what you don’t know, structure helps you discover blind spots.

Outside of those cases? Documentation, experimentation, and AI explanations usually get you there faster.

---

The Bigger Question

Maybe the real question isn’t whether we should keep learning - that answer is obvious. We should.

The question is: What’s worth learning in a structured way - and what’s better learned by doing?

So I’ll turn it back to you: What are your go-to sources for learning these days?

Cheers!

Evgeny Urubkov (@codevev)