Reality has an incredible amount of detail. You can't fit it all in your head. Think of all the things you could learn. All the professions you could spend a lifetime in. Let’s imagine that you’re a biologist. You could spend your whole life studying the sight of the honey-bee. Or perhaps the 18,000 neurones of the california sea slug [0].

Much of learning is "getting used" to the detail of reality. Soaking it up by osmosis. Learning to abstract / model it at the right level. This is your S1 absorbing a crazy amount of information. [1] I've heard it said: “you don’t learn a language you get used to it”.

A friend of mine learns math through exposure. He reads a textbook iteratively. Jumps into a chapter where he doesn't really know what's going on. Then after re-reading previous chapters it starts to make sense. A crucial skill here is learning to be a little okay with most stuff not making sense. Not needing line by line correctness. “Well I got 60% of that” perhaps this is what Nate is talking about with his re-read advice

Learning maths occasionally feels like becoming comfortable with the concepts. Learning the ontology. At the start it’s all very confusing then you wake up a month or two down the road and you've been saying "surjective" or "manifold" like nobodies business. If you give it enough time you also normally forget what it was like to be so confused.


Here's a paper. Try reading it iteratively. set a timer.
2mins - just skim the abstract
5mins - the introduction & conclusion are also worth a look, can you feel the context “ahh this is what it’s about?”
10mins - Now everything from the top. Can you feel the nuance sliding into place. Be okay with confusion, clues, gently don’t expect knowledge. Follow your curiosity, what parts are interesting.

[0] Sea slugs are super pretty

[1] You think a system 2 knows how to walk?

The title is borrowed from [John Salvatier]


New Comment

New to LessWrong?