Entirety of math is too large. Are you trying to learn to learn math? Or are you trying to learn actual math? Which topic are you interested in? By that, I mean

Entirety of math is too large. Are you trying to learn to learn math? Or are you trying to learn actual math? Which topic are you interested in? By that, I mean something like say topology or algebra etc. Of course they have overlaps, but you have to start somewhere. Did you have even more specific ideas in mind? Maybe commutative rings or quasi groups are your jam. Once you have a vague idea of what you want topic wise, the best thing to do is sample as many books as possible. It seems that some people click with one book and other not so much. Take (baby) Rudin for example. Some people love it as their first analysis book. It was too dense for my liking and I needed something that did more spoon feeding then that. But I can see Rudin making for a good reference book if I wanted such a thing. Tangentially, the premise is that mentors aren't always available. Sure, but might not have access to them all the time but you'll probably want some one to talk to once in a while. Say you pick a highly recommended book X and you don't like it at all. What do you do? If you find someone who has not only read X but other books on the same topic, they can nudge you in the right direction. It can also help with discussing and understanding concepts. HN seems to have a decent number of mathematically aware people. There is also math stack exchange and friends (which I haven't used a lot). There also exists more niche solutions like the ##math channels in some irc servers (which i used extensively and was the best thing to happen personally). Lastly, a lot of the popular/most recommended style books never worked out that well for me. I ended up looking at lecture notes scattered across the internet by various professors on very narrow topic and also very niche youtube videos by random professors and such.