Ultralearning: The Power of Maps and Drills

Learning is a superpower. It magnifies everything you do and acts as a compounding force over time. In my last post, we already discussed why merely reviewing what you learned is not the best strategy for long-term retention. In this post, I’ll show you two more ideas from the book Ultralearning, by Scott Young, that I think can significantly improve your learning process.

Draw a map

The first idea is drawing a map of what you want to learn.

Suppose I want to learn a new language. What would I need it for? Is it to write business emails for work or to talk to strangers when I travel to a country I’ve been meaning to visit? If I’m learning to code, is that for a test I have to take in school or to build a personal project that will ultimately become our AI overlords?

If you’re serious about learning something, mindlessly following the first learning path presented to you may be wasteful. You may find yourself studying a lot of stuff you’re never going to need while possibly neglecting important, subtle parts of that skill or knowledge.

Notice the two aspects at play here: one is done through via negativa, rejecting what you don’t need. The other is doubling down on the details of what’s relevant to your specific goal. When you consider the reason behind your learning, you lay down the path of how to direct your effort.

Once you understand your purpose, you can draw a map of how to navigate learning that skill. Scott recommends benchmarking as the primary tactic to find out what you need to learn. Benchmarking means finding out how other people are learning and starting your map from there. For example, if there are university courses for that skill, you can look up their curricula. See what topics they cover and what reading material they recommend. If it’s an unusual type of knowledge or technique, look for online forums where people discuss it. Then, identify what topics - or “subskills” - they have in common. When you’re done drawing your map, use it to create your own customized learning path.

You can go even further with your optimization. If you break down the subskills by type, you can choose the best approach to master each one. Ultralearning proposes three categories:

1. Concepts: these are ideas you need to understand. For example: when learning the laws of Physics, you have to grasp how they work in the real world, what their implications are, etc. To better comprehend and retain concepts, use the question-book method I described in the last post.

2. Facts: anything that needs to be memorized.  For example: when learning a language, you can memorize a good portion of the vocabulary. When learning Calculus, you can memorize the derivative of a handful of functions, so it’s faster and easier to solve problems. Flashcards or spaced repetition software are good choices to master this type of subskill.

3. Procedures: everything that you must practice until it requires minimal to no attention to carry out. Pronunciation of a second language is a good example of this. When you’re talking, you only have a split second between words and can’t afford to think long and hard about how to pronounce the next one. If you’re learning to play the guitar, a type of procedure is playing every chord and switching between them with minimal effort. To work on procedures, propose difficult challenges to yourself that cover a wide range of scenarios you may encounter when using the skill.

Drill the bottlenecks

Drawing of a girl talking. Grammar, vocabulary, pronunciation and rhetoric are coming out of her mouth.
Cognitive components take place simultaneously

The second big idea I want to talk about is using drills to attack bottlenecks in learning.

Some skills can be broken down into smaller subskills. Take drawing, for example: it encompasses perspective, anatomy, composition, lighting. Solving a Math problem involves interpreting the question, choosing the best tool or method to tackle that problem, and correctly carrying out the operations until you arrive at the answer. When you practice the skill directly, you are exercising all of its subskills in some capacity.

During your learning journey, those subskills will evolve at different rates, and one of them is bound to become a bottleneck to the whole process. Maybe because the subskill in question is a difficult one per se. Or maybe because there are too many subskills at play at the same time, making it harder to focus on improving one of them when there’s so much going on.

To drill means to find out what that limiting step is, isolate it, and laser focus your energy to make progress.

If you’re dealing with a sequence of steps, it’s easy to drill the part that’s giving you trouble. All you have to do is isolate that portion of the process and come up with a way to practice only that. In the guitar example, if a specific section of a song proves to be more difficult to learn, it’s simple enough to practice only that part. This technique is called time slicing and is perfect for sequential subskills.

But some skills depend on many subskills that take place simultaneously. Like speaking a second language, for example. There’s grammar, vocabulary, pronunciation, rhetoric, all happening at once. In the book, Scott calls these the cognitive components of your skill. In those cases, simple time slicing doesn’t make sense. You should design a specific drill for each component, train them separately, and then incorporate them back into the whole.

Cognitive components are harder to drill than time-sliced sections, so you have to get creative. Benjamin Franklin gives us a remarkable example of how he got better at writing in his Autobiography. To practice vocabulary, he would take other authors’ essays and turn them from prose into verse, thus having to replace words to match the desired metric or rhyme. To improve his rhetoric, he would try to reconstruct the essays’ logical steps from hints he’d written down, but put in a scrambled order.

Following Franklin’s method, one last tip for designing your drills is to use shortcuts and build on other people’s efforts as much as possible. For example, if you’re learning to draw, you could start by copying someone else’s drawing. That way, you don’t have to worry about framing the scene, choosing what colors to use, what details to include, etc. Those decisions are momentarily taken out of the equation so you can focus on fewer subskills at a time.

You’ve used drilling before. Remember when you learned multiplication tables in school? That’s a kind of drill that you may have found to be incredibly tedious. But that’s the difference in Ultralearning - it’s all about creating your own journey to learn. So instead of a soul-crushingly boring activity you were told to do, you’re now in charge of the whole process and have a clear purpose in mind.

Subscribe below if you want to be notified about the new posts 👇