One can only study one or two subjects at the same time in college, but with internet, anyone can approach any subject they are interested in without much hussle.
A standard way of looking at a subject is top-down: knowing its nature, its structure and then diving into details. Stanford philosophical encyclopedia is a nice resource to start for a wide range of subjects. Philosophers are responsible for answering a lot of fundamental questions, which never should have a definite anser. A question with an answer is just a piece of information. Knowledge is a jungle, our understanding of which is limited to the length we’ve gone through so far.
For dealing with details, AI can be handy. It can generate the standard learning process of a subject from top universities, and following that there are online playlists that deal with these sub-branches. If one really wish to dig into a field, a rigorous study is necessary, otherwise there could be patches that are left out if roaming within it without a method.
But roaming is fun. I roam all the time. The rule of thumb is that one should always be immersed in learning. If the systematical way fails to grab one’s attention, then it is better to leave it for a while, roam in the field, get some fun, and then come back. Forcing oneself to pay attention is miserable. We have suffered enough in life, and really should reduce pain whenever we can.
Problem-solving is essential, and should be done with quality instead of quantity. Wrong answers are treasures, from which one learns the most. If the answers are mostly correct, then it is only a waste of time. Mostly wrong answers are just due to different understandings of a concept from the official one. Understanding a concept in the same way as others is important in that people can only discuss the subject when they have a collection of vocabularies that defines the same meaning. Otherwise it is just one man’s game.
And if one hopes to learn something in order to do something, then it is quintessential to just do it. Think about it, try to do it. And if there is any problem, come back to online resources and look it up. After doing it a little bit, then use the time that’s spared for a systematic learning, so as to do it better. There is a hypothesis of vicious regress concerning know-how: if one needs pre-knowledge in any action, given that grabing the knowledge is also an action, there would be an infinite process before actually taking the action one wants to act at the first place. So just do it.