Learning Theory
Oct 22, 2024
This is a live essay, I'm updating it almost daily with my thoughts and findings on the subject of what I'm calling "Learning Theory" — please don't judge it too harshly for the time being!
Thinking and writing about this extensively, will publish more here. Also making a film about this. See the bottom of this article for my latest "plan of attack"
Core questions (and my current answers to them)
Why should we all want to learn more? Learning is one of the greatest joys and it leads to progress.
Are we learning less? It sure seems like it.
Why aren't more people, learning more things, more quickly in the age of the internet (and now AI)? It's a cultural and self regulation problem.
Are we experiencing an "attention span crisis"? Yes.
Can you "know it all"? I hope so.
Can we "psyop" people into learning more? I hope so.
Currently reading the works of:
Barbara Oakley — "Don't just follow your passions, broaden your passions."
How are you learning without a memory system?
During the COVID-19 pandemic one of my most joyous home imprisonment activities was going deep on the rabbit hole of Roam Research. I did a book club with the Roam community and the author Sönke Ahrens on How to Take Smart Notes. It's a deceptively self-help-y title, but looking into the pages it's absolutely fascinating for a variety of reasons:
It argues the benefits of the Zettelkasten Method, created by Niklas Luhmann, which proposes that you (1) mostly spend your time reading and researching, (2) using a Zettelkasten system (in modern days Roam or Obsidian), (3) then when you're ready to publish go through your notes, compile them, draw conclusions, and focus on prose.
The author himself writes in this style, which I found exhilarating and fresh.
The core argument: your brain doesn't think in a "filing cabinet structure," or with folders and notes — you think inter-connectedly, ideas inspired by totally disparate ideas at random. The benefit of a connection-based system like Roam or Obsidian is apparent.
I used Roam a bit too much like a productivity software (tracking todos, tasks, and journaling) which clogged up my intellectual experience with the tool and found the lack of a solid mobile app frustrating — some of my best ideas come on walks!
So I abandoned Roam, and my Zettelkasten system, around 2022 with intermittent spurts of usage. However in 2023, I began to rebuild my system as I embarked on more technical projects in Obsidian. I've mostly very enjoyed this experience, however I miss Roam's node-based system, the ability to copy the link to a specific node/block of text was truly special.
"Timelapse animation" of my new Obsidian graph.
Now that I'm back, with lessons learned of how not to build this, I'm back to being in love with the process of learning with a connection-based system, and I wonder:
How the hell does anyone learn without one of these?
I imagine, and wish, I had this tool (with the proper guidance) when I was a kid in school. I fear I would have learned much more, faster, and deeper.
What should you learn?
Caution, very incomplete section
This is a very complex question, right now I'm exploring the frame work of learning specifically for mathematics and physics. Learning math and physics is fundamental because the core ideas haven't changed for centuries (which is really good news because the basics are actually easy to learn if done properly).
Math and physics is potentially better as a foundational layer for learning, even better than STEM, as Arun Rao argues in his new AMP program.
Rocket science, history, and math
The cultural virus against learning in the 2000s-2012s.
Rocket science
Can you recall the first time you heard the phrase, "It's not rocket science?" It's an incredible phrase because it implies quite a few things:
The thing you are talking about (allegedly) shouldn't be that complicated
That rocket science is in fact complicated
Because it's complicated and colloquially represents the peak of human innovation, it's beyond you to understand
Rocket science is in fact very simple.
That's it, that's rocket science. Rather, that's Tsiolkovsky's rocket equation, which is pretty simple.
Obviously, making an orbital rocket and making it do incredible things like deliver tonnes of cargo to orbit or re-land itself is very hard. But even that is understandable!
Why have we as a general society adopted an attitude of ignorance towards hard things when they are knowable and easier to understand than ever before?
The greatest generation & history class
They won World War II, they put man on the moon, and they built the great skylines of America — their sacrifice and work gave us what we have today, a launchpad to dwarf their accomplishments with even greater feats of human ingenuity.
Today, in history classes around the country, we are taught that the greatest achievements of American innovation are not monuments of inspiration for future generations (in this case the students learning in history class) but instead relics of a long gone past.
History should seek to teach learners about the past to build a better future. Modern history, especially modern history of a given society, should incubate feelings of curiosity, inspiration, and bravery. Instead, the Apollo era is commonly taught as an anomaly of innovation rather than an ode to possibility.
Being "gifted" at math
Unless you were of the "naturally gifted" math students, it's my observation that the majority of American math students (like myself) may hit a wall, or really a slow decline, anywhere from 8th-11th grade in learning mathematics. Modern majority society has seemingly adopted an ideology that "not everyone is gifted at math" which is incredibly dangerous and untrue.
The modern crisis
Instant gratification, short attention span, and comfort = a very scary future
The 1st principles of learning
Or, our current best mechanisms that neuroscience seems to agree with for learning.
Recall is everything
Idea chunking is good
Practice makes plasticity, spaced repetition works great
Sleep is very important, exercise is important
The more you learn, the easier you learn — transfer
The frontiers of learning
We need to know a whole lot more about neuroscience.
Tools for learning
There seem to be very few, and very few innovative ones
Obsidian, Roam — connection-based note taking software
LLMs (Llama, Claude, ChatGPT, Grok) — for explaining complex topics & teaching
Interactive portions of Khan Academy
Anki, Quizlet
(Learning + application) * speed = progress
I would make the argument that learning in a vacuum, while wonderful, is wasted without application. It's multiplied by speed, or intensity, and results in progress.
Tricking interest
We know that students today like to go really deep on things they're interested in — like people on Reddit going deep on tying 18th century knots or people building complex things in Minecraft. It's clear that we must learn to trick ourselves into interest.
Plan of attack
Updated October 23, 2024 at 11am pST
Develop a techno-philosophy (term coined by Royce Branning, made an S3 about his company Clearspace here) for modern "learning theory" — specifically tailored to Gen Alpha/Z for now
Test on self, iterate & tweak
Clearly communicate its ideas somewhere
Mass market and psyop it to the masses (start with a Frontier Film)
Iterate & tweak as needed