Nintil suggests that you should massive input most things to build intuition/tacit knowledge and use spaced repetition for things that you really want to remember.

I talked to Jacob GW about this, and he posits that massive input works best in domains where there’s immediate feedback like coding or rubix cubing. In other domains, spaced repetition provides the immediate feedback. This tracks with my ideas around Environments that are good for deliberate practice.

This makes me wonder about Making Massive Input Better.

Massive Input vs Spaced Repetition Research

I want to research into whether there’s even better heuristics for when to use massive input and when to use spaced repetition for coming up with better ideas/getting to the edge of your field. I want to pioneer a new approach that combines the two into a system that can be used effectively for whatever you need.

A few research questions and directions:

  1. Is there evidence that one is better than the other in certain scenarios, and if not, can you do research towards that?
    1. What differentiates language learning from physics or something like that? What are the best ways to learn languages?
    2. How do researchers get new ideas, and do they use one of these over others traditionally?
  2. How can we structure massive input slightly more to make it more effective? (This probably comes from answering the question what is massive input trying to optimize for) Most likely this is going to look like a lot of the work Linus Lee has done with making skimmers and easily searchable personal knowledge databases.