# Ultralearning

- [Detailed Summary](https://koerbitz.me/posts/ultralearning-book-summary.html#:~:text=Drill%3A%20Improve%20the%20weakest%20point,Feedback%3A%20Is%20uncomfortable%20but%20crucial.)
- [Good Detailed Explanation](https://www.calnewport.com/blog/2012/10/26/mastering-linear-algebra-in-10-days-astounding-experiments-in-ultra-learning/)

## Meta-learning
- don't absorb info at random
- establishing how information is structured
- create meta learning map
- Concept (what need to be understood)
- Facts (what need to memorize)
- Procedure (what need to be done)
- Focus your energy in heavily weighted category
- use the created map to identify challenging aspects
- brainstorm techniques for overcomming them
- establish how you are going to learn
- try benchmarking research people who've acquired similar skill or institutions
- replicate their methods + equipment
- use online courses, syllabi to find resources, tools, texts that are considered essential in the field
- Time invested in meta-learning will not be wasted
- Allocate 10% of the total time you expect to learn to meta-learning
- Draw roadmap for ultra-learning project
- Use focus power to not derail you from your goal
## Focus
- life is full of distraction, but don't let them affect your goals
- first challenge - getting focus in the first place:
- set timer for like 3 mins, say stop working after that, but you can use it to build momentum
- use pomodoro technique
- sustaining your focus
- eliminate distraction source/Optimize environment
- once you start your task, if you're not fully engaged with your task, you're less likely to retain the material
- use interleaving to combat autopiolot, change mode of learning and material frequently
- devide session in fixed time, such as multiple 2 hours/session instead of 10 hours 1 session
- focus on different aspects
- Pay attention to your mental alertness/energy
- High alertness works best for narrow-focus task
- repetitive tasks
- Low arrousal/alertness works for lateral thinking task and forming connection
- creative task
- if you get stuck, take a break
- cognitive demanding activity is easier to focus
- Match your arrousal to your task
- Focus simple tasks when your focus is more aroused and complex tasks when it's less aroused
- Honing focus -> build up mental stamina to complete your challenge
- Recognize where are and try to improve your focus with small steps
## Directness:
- Competitive
- Learn by doing
- project based learning
- no transfer necessary
- immersive learning
- depend on project
- you can replicate the environment/simulation
## Drill
- Use drilling to hone your skills
- Drill strategically to maintain the competive edge
- Don't begin your project by drilling
- use direct-then-drill approach
- practice the skill directly
- Use direct to identify areas to drill
- drill the subskill
- integrate the subskill into the direct practice again
- Apply drilling to your rate-determining step (your weak spot)
- Drill begins by breaking down the weakest part into small parts, master those small parts by drilling, put them back together
- Design your drill:
- Depending on if you can isolate that area from your project
- If you can't break down parts (ie trying to drill creative writing), find others work and copycat/emulate
- spend more time to concentrate on subcomponent you want to improve even if that decrease
- Tactics:
- Designate time for different parts
- Focus on that specific sub-component only
- Copy other's work and then improve on the area you want to improve
- Concentrate on what you want to improve (magnify glass)
- Start learning pre-req, then shifting back and forth between pre-req and original project
- Drill **Mindfully**
- Only drilling to what is connected to the **desired skills**
- Don't drift off to unrelated materials

## Retrieval:
- Test yourself to learn, actively recall information
- use learned memory techniques (note from myself)
- recall is much more effective
- Passive review makes us think that we could recall the materials
- use flashcards
- free recall - write down everything you can recall
- write down question while reading, then later try to retrieve information to answer the question
- self-generated challenges
- closed-book learning: test yourself closed book
- test or apply your knowledge instead of looking up
## Feedback
- outcome based feedback:
- give single grade, pass/fail information
- informational feedback:
- feedback on individual items, which contribute to the whole outcome
- corrective feedback:
- tell you what's wrong, how the correct infomation look like, and why
- hardest type to get
### How to improve feedback:
- noice cancellation
- find reliable metrics, remove unreliable one

- difficulty sweet spot: find a difficult sweet spot (and continually improving it) of your learning, and get feedback from it.
- basically do hard things
- make sure feedback is not predictable to be valuable
- don't make things too easy, but not impossibly hard

- meta feedback: feed back on overall learning process (monitor learning rate - time learning something, if it's too long to learn, you need to take it as feedback and adjust it)

- rapid fire feedback : get lots of feedback make you get motivated to get more feedback
- knowing your work will be evaluated is a huge motivator for doing a **good job**
## Retention
- Stop filling leaky bucket
- Memory decay over time
- old memory are partially overwritten by newer one
- forgotten cue, information is still there but cue is gone, so it's hard to recall
### prevent fogetting:
- Spaced repetition:
- use flashcards
- Procedure:
- train your skills as procedural memory
- Over learning: learning a skill more might make it harder to fade
- Mnemoic (memory tricks that I've learned)
## Intuition - Go deep before building up
### Richard Feynmans:
- Don't fool yourself - and you are the easiest to fool!
- Think concrete examples when trying to understand something abstract
- try your best to understand the fundamental principals
### Feynman Technique:
- Write concept/problem on top of paper
- then write explanation
- if concept, write as if you're explaining to someone
- if problem, explain in detail how a solution works and why
- This can applied to anything you don't understand, problems you're stuck on and expand your intuition
- Don't give up on hard problem, give yourself a timer before giving up
- Prove or explain
- Use concrete examples
- Don't fool yourself into thinking you understand something when you really don't
- ask a lot of questions to make sure you understand it
## Experimentation Intensively
- True mastery does not come from path from others, you have to explore what works best for you, so experimentation is the key
- Van Gogh did not conform to art school's style but to experiment different techniques before finding his owns
### Things to experiment with:
- Learning resources:
- methods
- techniques
- before making change, make sure you've applied the tactics
- Different subfield:
- once you've master a skill, you could learn other subfield
- Styles:
- Experiment with different styles to find what work best for you or unique to you
### Tactics:
- Copy and change: copy work, experiment, change some aspects
- Vary one variable and compare result
- Add new constraint to spark creative solution
- Best of combination of two skills
- Go to the extreme:
- mastery usually involves two extreme sub-skills, don't expect mastery by taking safe/average option on everything

---
### Do your research:
- Topic and scope:
- good topic and quantitative **goal** scope
- Identify and obtain materials you'll be learning with
- Benchmark other's learning plan
- Direct practice: how can you directly practice the skill you've learned
- Investigate secondary learning materials for backup, find the possibilities for drills
- How to get feed back from learning
### Schedule your time: Consistently ONLY:
- not sporadic/ad-hoc schedule
### Execute your plan:
- measure your actions in relation to Ultralearning principles, and make adjustment
### Reflect your result
- Review what went wrong, what could have gone better