When a non-fiction author writes a book, there is a certain quantity of information that he or she is trying to convey. Call that quantity "X". We as readers, try to transfer as much of X into our mind as possible. What is the percentage of X that ends up in our mind? I'm fairly pessimistic on this number. I think it's probably around 10%, if we are lucky. This percentage slowly atrophies over time until we are left with just a few, small percentage points.

This is where my interest in knowledge-management systems started. I want a way to capture and organize the knowledge in the information that I read.

I've been reading a lot about note-taking strategies. There is a subset of these strategies that I would call "connected notes". This entails writing small, atomic notes and linking them into other notes in a graph-like structure. Nicholas Luhmann famously had his Zettelkasten system, which is an implementation of connected notes. Luhmann's system has gained a lot of recent attention and is probably where I'll start. One of the difficulties is that most discussions about Luhmann's systems are by bloggers that are using his system to write blog posts. I'd like to see more examples, especially around people in STEM that are using this sort of system.

There are many books that discuss note-taking strategies. Some of the ones I've read or are about to read include:

• How to Write a Thesis by Umberto Eco
• How to Take Smart Notes by Sönke Ahrens
• Duly Noted by Jose Arango
• Digital Zettelkasten by David Kadavy
• Antinet Zettelkasten by Scott Scheper
• The Notebook: A History of Thinking on Paper by Roland Allen
• Effective Notetaking by Fiona McPherson

Enough reading, it's time to get down to it and start building.

I want to build a knowledge-management system that captures and links 500 notes. Initially, this system will be built using index-cards, but I want to keep an eye on whether a digital system would be better. I've read pros and cons of each. It will be easier to digitize an analog system than the other way around, however.