Collection of principles, tools and tips to learn skills more effectively.
- Metalearning: spend 10% of the time answering what, why, how.
- Focus: practice deliberately, understand constraints (your env, warm-up effort).
- Directness: if you want to learn to speak Chinese, practice speaking Chinese, not grammar.
- Drill: identify current bottleneck, design practice exercise to get better at it and repeat.
- Retrieval: practice the skill even before you have it (tests, flash-cards).
- Feedback: seek as much of the right type of feedback as possible.
- Retention: use it or lose it.
- Intuition: deeply understand fundamentals (first-principles).
- Experimentation: try different resources, techniques or styles to find your own path.
After you reach your desired skill level, choose what effort you want to give to it: drop it, maintain it, or master it. The longer you maintain it, the cheaper it becomes (SRS), but in some cases relearning is more effective than maintaining it.
Tactics for each principle
- Time slicing: practice hardest parts of the whole (e.g. hardest parts of a piece of music.)
- Cognitive components: like Time slicing but domain is a layer instead (e.g. practice vocabulary or grammar.)
- Copycat: copy parts of the skill you don’t want to drill (e.g. Franklin’s using someone else’s writings when practicing his ordering of arguments.)
- Magnifying Glass: increase practice time on a specific area (e.g. increase 10x more time researching when writing a book.)
- Prerequisite Chaining: start practicing as soon as you id missing prerequisite skills, focus on them (Barone starting to learn pixel art when he started developing Stardew Valley)
- Flash cards.
- Free recall: read a book and afterwards write down everything you remember.
- The Question-Book Method: rephrase facts as questions in your notes.
- Self-Generated Challenges: e.g. when you encounter a new programming technique, leave a note to challenge yourself self to use it.
- Closed-Book learning: practice without looking up the answers.
- Outcome: am I doing it wrong?
- Informational: what am I doing wrong?
- Corrective: how can I fix what I am doing wrong?
- Noise cancellation: you will be getting a mix of useful and useless feedback, learn to filter out useless feedback.
- Difficulty Sweet Spot: avoid practice situations where you always feel good or bad about your performance.
- Metafeedback: track how fast you’re progressing with a skill to identify plateaus (e.g. Elo ratings in Chess, LSAT mock exams)
- High-Intensity, Rapid Feedback: try to have some practice sessions with this characteristic.
- Spacing (SRS)
- Proceduralization: drill core aspects of your skill until they become intuitive / motor skills, they tend to last longer in memory.
- Overlearning: practice more than you need (overengineer)
- Mnemonics: very niche application
- Don’t give up on hard problems easily
- Prove things to understand them
- Always start with a concrete example
- Don’t fool yourself; use the Feynman technique: try to teach it/explain it to someone else, review & refine.
- Copy, then create
- Compare methods side-by-side
- Introduce new constraints (fight with your weaker hand)
- Combine unrelated skills (engineer + drawing = Dilbert)
- Explore extremes (Van Gogh’s art)
Scott’s checklist (~verbatim)
- Have I researched the typical ways of learning this subject or skill?
- Have I interviewed experts to see what resources and advice they recommend?
- Have I spent about 10 percent of the total time on preparing my project?
- Am I focused when I spend time learning, or am I multitasking and distracted?
- Am I skipping learning sessions or procrastinating?
- When I start a session, how long does it take before I’m in a good flow?
- How long can I sustain that focus before my mind starts to wander?
- How sharp is my attention?
- Should it be more concentrated for intensity or more diffuse for creativity?
- Am I learning the skill in the way I’ll eventually be using it?
- If not, what mental processes are missing from my practice that exist in the real environment?
- How can I practice transferring the knowledge I learn from my book/class/video to real life?
- Am I spending time focusing on the weakest points of my performance?
- What is the rate-limiting step that is holding me back?
- Does it feel like my learning is slowing down and that there’s too many components of the skill to master?
- If so, how can I split apart a complex skill to work on smaller, more manageable components of it?
- Am I spending most of my time reading and reviewing, or am I solving problems and recalling things from memory without looking at my notes?
- Do I have some way of testing myself, or do I just assume I’ll remember?
- Can I successfully explain what I learned yesterday, last week, a year ago?
- How do I know if I can?
- Am I getting honest feedback about my performance early on, or am I trying to dodge the punches and avoid criticism?
- Do I know what I’m learning well and what I’m not?
- Am I using feedback correctly, or am I overreacting to noisy data?
- Do I have a plan in place to remember what I’m learning long term?
- Am I spacing my exposure to information so it will stick longer?
- Am I turning factual knowledge into procedures that I’ll retain?
- Am I overlearning the most critical aspects of the skill?
- Do I deeply understand the things I’m learning, or am I just memorizing?
- Could I teach the ideas and procedures I’m studying to someone else?
- Is it clear to me why what I’m learning is true, or does it all seem arbitrary and unrelated?
- Am I getting stuck with my current resources and techniques?
- Do I need to branch out and try new approaches to reach my goal?
- How can I go beyond mastering the basics and create a unique style to solve problems creatively and do things others haven’t explored before?
First two connections that come to mind:
Physiological requirements for learning (Andrew Huberman). Synapses between neurons only change (learning) when certain chemicals are released into the brain. To make sure the release happens do this:
- Be aware that you want to change (epinephrine)
- Have a powerful “Why”(dopamine)
- Focus, be specific about what you want to change (acetylcholine)
- Try to change & fail (1 failure type/session), a lot
- Get excited when you fail (dopamine)
Deliberate practice (Ericsson) principles:
- It’s designed specifically to improve performance.
- It can be repeated a lot.
- Feedback on results is continuously available.
- It’s highly demanding mentally.
- It isn’t much fun.
- Metalearning: OODLA loop (Orient), win the war before the first battle (Sun Tzu), Signal vs Noise, Cost of Opportunity.
- Focus: deliberate practice, focused vs diffused attention, know yourself, antifragility.
- Directness: doing over thinking, understanding is optional (toddler learns to walk w/o studying Physics), Skin in the Game and beware of intellectuals (Taleb).
- Drill: OODLA, neuroplasticity, deliberate practice, consistency over talent.
- Retrieval: neuroplasticity, problem solving, the power of questions on attention.
- Feedback: antifragility, aversion to unpleasant sensations, ego self-defense, feedback loops.
- Retention: nature’s ruthlessness and focus: if it’s not in use, it’s harming, conservation of energy, Marie Kondo’s teacher.
- Intuition: vipassana, samyama, intuitive knowledge (Inner Game of Tennis), first principles thinking (Elon Musk).
- Experimentation: play, local optima, antifragility.
Why trust the author?
Scott Young published this book when he was 30 years old. He bases the book’s content on 3 learning projects he took, some second-hand accounts of other learners, and research studies on learning and mastery (both positive and negative results are mentioned.) Some of the learning projects he took himself are more impressive (the MIT Challenge where he self-studyied nearly all of MIT’s 4 year Computer Science syllabus in 1 year), others a bit less (improving drawing skills.)
The book isn’t selling any paid programs and he has extensive online presence (interest) covering an extended period of time (this isn’t a journalist publishing a book after a 3 months deep dive.)
In 2023 he commented on this video about what he has learned about learning and the things he would change in the book. Publicly evaluating your previous work and correcting it is a great virtue.
How do I grow from this?
I didn’t find new big ideas, but I’ve expanded my learning checklists with some of the tactics mentioned.
I will also refer to this book when people ask me for resources on learning. The book is only missing a topic I consider important: the power that a teacher can have to accelerate your learning.
Listening to this book has also reignited my desire to learn, which I haven’t done in a dedicated way for a while. At the same time I’m aware of the opportunity costs of learning skills just for fun given the reality of death.
Scott describes his 2011-2012 MIT challenge.
I wouldn’t have called the book ultralearning because it may put some people off thinking that this is for extreme people, but otherwise, it’s a more holistic, complete and concrete intro to learning effectively than for example books like Peak or Talent is Overrated.
ISBN: 978-0062852687 #read/2023