Today, I’d like to share how I’ve been studying data science and AI recently.
This isn’t meant as a guide or a recommendation in a strong sense—just a record of what has been working for me. If any part of it is useful to someone else, that’s more than enough.

For several years, I’ve been studying data science online, on and off. I’m still doing that, but over the past year or two—especially since around 2025—I’ve noticed a shift. I’ve been enjoying AI engineering courses more, and I naturally find myself spending more time there.

That change made me reflect a little on how I’m learning, not just what I’m learning.

Online learning platforms I often use

The platforms I use most are Coursera and Udemy. I usually browse courses there and subscribe when something catches my interest.

One obvious advantage of online platforms is the price. Compared to traditional education, they’re affordable. But more importantly, many of the courses are surprisingly comprehensive—and very hands-on.

What I appreciate most is being able to see real code written by other people. Not just explanations, but actual examples:

• what tools they use
• how they structure their code
• how recent libraries are applied in practice

Even when I don’t fully understand everything at first, just seeing how things are done speeds up learning a lot. It feels very different from the old textbook-based approach, where you move step by step over weeks or months before seeing the bigger picture.

How my learning style has changed

In the past, learning often meant following a strict sequence:
textbook → exercises → assignments → finally, understanding.

Now, my approach is almost the opposite.

I often start by looking at code written by others. If it’s on GitHub, I clone the entire repository first. Cursor is one of the excellent tools for that. Then I ask in the chat simple questions like:

• “What is this part doing?”
• “Why is this tool used here?”
• “What’s the overall structure?”

After that, when I return to a course video or lecture, things click much faster. I already have a rough map in my head, so the explanations land more easily.

Coursera and Udemy

I used to enjoy Coursera a lot and even earned certificates there. The projects are solid, and the structure is well thought out. But to be honest, it can feel a bit rigid. Good—but sometimes too formal, and I find myself losing momentum.

Lately, I’ve been using Udemy more.

On Udemy, the atmosphere of a course depends heavily on the instructor. Their energy, tone, and way of speaking really shape the experience. Before subscribing, I usually watch the preview videos to see if the style feels right for me.

When it does, learning feels lighter. I often choose hands-on courses where I can watch someone code in real time—seeing how they think, how they use AI tools themselves. That alone can be very educational.

When energy is limited

Of course, actually setting up environments, managing versions, and fixing errors takes energy. Sometimes more than expected.

On days when that feels like too much, I simply watch the videos. Most are short enough. Even just observing the flow—what tools are used and in what order—can be surprisingly helpful.

Skipping steps isn’t failure. Sometimes it’s just choosing to keep going.

A small note at the end

This isn’t sponsored by anyone, and I have no connection to the instructors or platforms I mentioned. It’s simply how I’ve been studying lately.

That’s all for now.
Thank you for reading.