Preface: I originally published this post to my blog on June 8th 2022. I am posting it here in case some of the ideas are useful to other EAs. It has been lightly edited to fit the context of the EA Forum.
Epistemic Status: A personal reflection on my continued journey toward effective learning.
I enjoy the process of seeking out information and improving my knowledge. It is fun to "go down the rabbit hole" on any given topic, and I agree with Richard Feynman when he said that things become more interesting the more you learn about them. This is a consequence of having the trait of curiosity. I have met a few hundred people in my life, and while that gives me some sense as to how different people's perspectives are, I admit that I find it hard to imagine what it would be like to live without curiosity. Would it just be a matter of unquestioningly accepting how the universe is? For people without curiosity, did they experience wonder as kids, only to have it crushed out of them by the tragic events of their lives?
I don't know the answers to the above questions, and while my curiosity might lead me to want to investigate their answers, learning information is hard — and historically I haven't done much with the information I currently know.
I want to propose a simple, three-tiered framework for how my journey to better process new information has changed over time:
- Tier 1. Seeking Information
- Tier 2. Taking Smart Notes
- Tier 3. Learning Through Writing
Let's discuss the first tier:
Tier 1. Seeking Information: Locating sources, passively reading them, and updating my worldview — holding everything in memory
This is the obvious strategy. I knew that if I wanted more information, I had to find it, and I reasoned that the more information I had, the better my overall picture of the world would be.
This sounds great.
But the world is very complicated. It is full of divergent opinions. People have cognitive biases, they misrepresent data (intentionally or otherwise), and their questionable motives often lead to intentionally misleading information being published. Critical thinking and reasoning skills are a must when navigating this polluted information landscape. There are various resources which aim to improve these skills: LessWrong is a great start, and Julia Galef's excellent book, The Scout Mindset, is another.
I only recently encountered these resources. Before that, I was driven by my interest (and later education) in Science, because it is unparalleled in its effectiveness at cutting through the darkness of uncertainty. Science produces results. Although bias can exist, time and peer review typically (but not always) snuffs it out. Overall, the idea of testing our ideas against Nature is a powerful one — possibly Humanity’s most powerful. This makes information gained through good science very interesting to me. It is a ready-made treasure trove of ideas waiting for me to explore.
However, today I am not arguing for or against Science being valuable. Instead, I am interested in exploring how my method of learning effectively has changed.
At some point I realized that blindly following interesting threads of information to see where they lead is inefficient. There may be a more effective use of my limited time.
This leads us to...
Tier 2. Taking Smart Notes using a Knowledge Management Engine
Passively reading a text might be fun, but to recall information usefully, you have to take notes. However, like most people, I had to learn that I had been taking notes wrong my whole life. Luckily, there was a better way.
My friends mentioned the book How To Take Smart Notes, and how it was pivotal to their lives. I found the book so compelling that I even gave a speech about it at one of my Toastmasters clubs. From there, I learned about Obsidian.md, which is a software tool for managing notes. Using this tool, I started my first "Personal Knowledge Engine", using the Zettelkasten Method (German for 'slip box').
Ali Abdaal's video does a great job of summarizing the key ideas of Taking Smart Notes. Here are a few key excerpts:
- Capture Ideas Immediately, because "brains are for having ideas, not storing them." - Ali Abdaal
- Connect concepts by relation, not by categories. Your brain is a neural network — a web — and we can mimic its structure in our notes.
- Instead of thinking in terms of "fields'' of study, think in terms of ideas. It’s better to make a note about a raw idea, and link to it inside of other notes.
Here is what my note collection looked like after several months:
Taking notes using this method has drastically improved my ability to recall information. For the first time, I am actually able to remember what I read for longer than a month, and use it to generate new insights. The system works well, and it creates a natural kind of spaced repetition. I rediscover useful concepts as I create new notes, which helps my information recall even when physically away from my references.
When I read books, I at least try and take useful notes. Over time, this mindfulness improved my ability to sift through text for meaning. Because books are filled with argumentation, analysis, and sourcing, they are great for exploring all angles of an issue. However, this means that it takes longer to extract their key ideas. By creating smart notes, extracting pure ideas is now my primary reason for reading; summarizing is slowly becoming a practiced skill.
We have explored seeking information and taking good notes, but there is a further tier to reach. If I want to improve my recall, information parsing, and the usefulness of the entire enterprise, I would have to eventually turn my notes into something public.
Tier 3: "Ideas unshared are worthless": Information has to be published in order to be useful to others.
Recently, I realized that although I was making notes, I wasn't publishing anything. Information that felt obvious to me might still be to useful to others, especially if I wanted people to understand my thoughts. Therefore, my attention finally shifted toward how I could create something useful from what I was discovering.
On June 8th, I discovered Holden Karnofsky's brilliant post Learning by Writing. It advocates investigating new ideas through briefly researching them, then writing a tentative worldview on the topic, then updating it as new information is learned. Instead of trying to search an entire topic before creating a work, and thus never writing anything, you write an initial hypothesis based on available information. This hypothesis is then scrutinized in an attempt to refute or modify it. As the author’s thought process is transparent, the reader benefits from seeing how ideas are updated incrementally. This is a more realistic picture of how knowledge is refined, and lets the author start writing right away.
This seems to solve the problem of endlessly seeking in a Tier 1 pattern, without having anything to show for it. It also focuses the search toward finding information to answer specific questions, allowing the notes generated in Tier 2 to be employed as part of the writing process. Writing specifically to answer key questions requires obtaining a holistic picture of a topic, focusing on arguments for and against something. It mandates staying focused and seeking only relevant information.
What strikes me is that the method could be useful right away. Part of the reason I haven't written many posts is because I have ideas sitting around as drafts, but I don't feel they're good enough yet to publish. On the other hand, done is better than perfect. It certainly didn't occur to me that I could publish what I knew about something so far, while publicly acknowledging that my views were likely to update with more information — and that doing this was okay! Rather than holding myself to some unrealistic standard of perfection, I'll learn more quickly by trying out Holden's method.
Let this post mark my desire to switch into Tier 3. This post itself is me "writing my stream of consciousness, however imperfect", rather than aiming for perfection. I'd like to write posts more often, and have it reflect the growth of my ideas and knowledge over time. I want to share what I learn, and there is no better time to start than now.