> Nowadays you can learn nearly anything for free with just an internet connection and some dedication
It's no different from the situation before the internet, since books are cheap (compared to your time) and usually are still a better source of information on such structured subjects anyway.
In fact, when I was in college, I mostly read textbooks anyway instead of attending lectures (it's easier for me to learn this way).
This however is not the whole story. Even if you are motivated ("I'd like to learn math"), you are probably not motivated enough for some feats of cognition. If you know people who stayed home and studied math on their own for a few years, full-time, at college level (I don't), they are probably very rare.
No, the situation is very different now with the internet since you can now study subjects in any arbitrary order, as long as you diligently identify and follow the dependencies (i.e. prerequisites). Prescribed curricula and structured courses were much more important in the pre-internet days since it was much harder to use the dependency-order learning approach.
I've learned a decent amount of mathematical stuff on my own without traditional books, and I'm pretty sure this will become more and more common over time. I'm a member of the earliest generation (b. 1985) for which this was really possible, and I just fell into it by accident, but in the coming decades I expect we'll see a lot more people who are almost entirely self-educated through the internet. I'm sure it'll still be a relatively uncommon phenomenon, but at least the people who learn best in this way now have access to the information and tools they need.
Learning math from Wikipedia is like learning foreign languages from a dictionary. It seems that you are intent on believing what you chose to believe, though, so good luck with that.
I'd be seriously impressed by anyone who could actually learn math or physics from wikipedia. Those articles are so poorly written and jargon heavy that I rarely learn anything from them.
Happily the computer science articles tend to make at least a little sense. But then I already have a degree in the field.
The articles are in fact extremely well written, but they are encyclopedaic. They are not written as a textbook, but as a reference work. Most articles assume you're nearly there and need the details.
There are people who can learn from encyclopedias, but most people need well-paced, linearly(ish) arranged material.
Wikipedia is a great way to get introduced to topics. Like any other encyclopedia (not dictionary) it serves as a great way to get a high-level overview of a topic. There's plenty of other resources on the internet for more in-depth information -- including free PDF versions of books.
It's no different from the situation before the internet, since books are cheap (compared to your time) and usually are still a better source of information on such structured subjects anyway.
In fact, when I was in college, I mostly read textbooks anyway instead of attending lectures (it's easier for me to learn this way).
This however is not the whole story. Even if you are motivated ("I'd like to learn math"), you are probably not motivated enough for some feats of cognition. If you know people who stayed home and studied math on their own for a few years, full-time, at college level (I don't), they are probably very rare.