**Some general steps for better [[Metalearning]]** Ask yourself (in order) why, what, and how. 1. Why do you want to learn the thing? 1. is it for _instrumental_ or _intrinsic_ reasons? 2. if instrumental, tailor your learning to the thing you need 3. if intrinsic, just focus on what's most interesting or fun 2. What do you need to know about the thing to have learned it? 1. How does the thing break down? 2. What are the 3. How are you going to learn those things? 1. Are there benchmark techniques you can mine for the subject? 2. Can you find experts to talk with? You should spend about 10% of the total time you have to learn a thing doing metalearning, establishing the roadmap for how you'll learn it. Heavily front-loaded, but also return to it periodically, especially when progress is slow. **** # More ## Source - [[Ultralearning]] ## Related - [[Metalearning]] - [[First Idea ≠ Best Idea]] - don't settle for the first resource you find - [[Decomposition Understanding]] - [[Black Boxing as Step One to Learning]]