The brilliant Professor Andrew Ng teaching his Stanford class on how to teach themselves AI. Really useful meta-guide on an often neglected topic - useful teaching yourself anything.
How to read an academic paper: don't try to read it from start to finish. Make multiple passes. Start with title/abstract/figures. Then intro/conclusion/figures. Then read it all but skip the maths. Then try the whole thing, but skip parts that don't make sense.
Prof. Ng is a great teacher. I did his deep learning course on Coursera a couple of years ago and it was excellent. He has a knack for making you feel OK about not understanding things yet, and seems to have real empathy for his students. This is a really useful guide for anybody trying to get to grips with a new technical subject.