How I learn new technologies faster — skip the course, build the real thing

I have a confession.

I almost never finish courses.

I've started probably 40 courses across various technologies in the last 3 years. I've finished maybe 4.

For a long time I thought this was a discipline problem. Something to fix.

Now I think it's the right approach — as long as what replaces the course is better.

Here's what I do instead, and why I think it works:

The method:

1. Pick something I actually want to exist — not a clone, not a tutorial project, something I personally need or want.

2. Choose a technology I want to learn as the implementation.

3. Start building. No tutorial. Just documentation and search.

4. When I get stuck (immediately), figure it out in the context of my actual problem.

5. Ship it when the core works. Not when it's complete — when it does the thing.

Why this works better than courses:

Courses teach a curated path. You learn what the creator thought was important, in the order they thought was logical.

Real problem building teaches you the actual path through the technology that your specific use case requires.

Those are different paths.

The parts of a technology that matter for a CRUD web app are different from the parts that matter for a real-time collaboration tool. A course teaches you a generalised version of both.

Real building teaches you exactly what you need, deeply, because you needed it.

The other thing this does:

It produces real things.

Some of the tools I use in my daily workflow came out of "let me learn X by building something I want."

Not portfolio pieces. Not tutorial projects. Actually useful things.

The course produces knowledge you may or may not use.
The real project produces knowledge you're definitely using — plus a thing that exists.

What to do when you get badly stuck:

This is the part people worry about. What if I can't figure it out?

My approach: 30 minutes of genuine effort on the specific blocker. Then search specifically for that blocker (not a general tutorial). Then ask in a community or forum if search doesn't resolve it.

This cycle — attempt, search, ask — teaches you both the technology and how to learn in it. Both are valuable.

The stuck moments are the actual learning. The flowing parts where everything works? Those you already knew.

One caveat:

This approach requires having a real problem to start with.

If you don't have one — talk to people around you. What do they do repeatedly that seems like it should be easier? What do you do in a week that's annoying?

The problem is often closer than you think.

What's your approach to learning new technologies? Do you finish courses?