Imagine you have a giant boulder in front of you. This is the task you want to undertake. You know you need to push it to get it moving but you also recognise that the inertia you'll need to overcome at first is substantial. You unfortunately think that pushing the boulder will always take this much force. The truth is, if you can get the boulder moving, keeping it in motion only takes many small, infrequent pushes. The idea behind pushing a boulder is the same as fast systems incurring usage. Want to improve your coding? Code. Want to read more? Read. Action begets action.
When you try to acquire knowledge there is always a pool of unknown-unknowns whose size is, itself, infinite. We are constantly refining things from the unknown-unknown pool into known-unknowns while we learn. One way to acquire knowledge is to read. I recently joked on twitter about speed reading:
Here’s how you speed read: flip through a book and tell people you’ve read it.— Ryan James Spencer (@_justanotherdot) September 6, 2019
But in all fairness, this is a sensible approach. Some relate knowledge acquisition as panning for gold. I think that's a misguided analogy because panning for gold is slow. Yes, we want to find the gold and we're not quite sure where it is, but no modern mining outfit would pan for gold these days. They would use large machines that could tear, demolish, crush, rip, sift, sort, detect, and on and on in orders of magnitude time more quickly than someone holding a pan to a river.
There is a technique involving multiple passes of reading. I learned this first when I got into research papers. It's also the basic idea behind How to Read a Book. You read a first pass involving the abstract/introduction, the conclusion, and then you skim, noticing headlines, captions, diagrams, and all the other top-level items. You follow suit with other passes increasing in detail if you haven't already dropped interest in the material. Basically, actions get prioritised by cost, the lower cost stuff coming first. Don't pan for gold, mine for it with machines..
Take a book you want to get through. Try to just flip through it. Once you're done skimming, try to concentrate on the primary points you've just become aware of; your known-unknowns. In your future passes, try to hone in on the things you care about more. Reading books out of order is actually fine for a lot of material if the first pass didn't tell you enough of what you need to know. What this does is change your definition of done.
There is always signal-to-noise and it's largely why there are diminishing returns which is both true for what we produce as well as what we consume. Ruthlessly closing browser tabs to reduce distraction, keeping your output high and frequent, reading books in multiple pass, and on and on. Do things to put you in an uncomfortable state. Make tons of little changes and skim content frantically going back for more when you need it instead of sitting around dreaming or scheming. Be a giant machine. Tinker like a maniac. Push the boulder.