Ultralearning is deep self-education to learn hard things in less time. I’ve written before about how I’ve used this approach to learn MIT computer science , multiple languages  and cognitive science .
I’ve touched on some of the aspects of ultralearning in previous articles. It focuses on learning depth-first , breaking impasses down into prerequisites you can finish step-by-step , creatively using resources  and balancing theory with practice .
In this article, I want to show you how to start your own ultralearning project. To make things easier, I’ve split this article into two parts: the first part, explaining why you should start an ultralearning project and how to design it. The second part, available here , will tell you how to find time to work on it and how to improve your ability to focus.
Ultralearning projects are hard work. Not only do they require you to take time out of your life, but they’re also mentally demanding. Given this, a good question to ask might be, why bother ultralearning at all?
The opposite of ultralearning is dabbling. This means playing around with something and eventually learning it. There’s no commitment. No time put aside. And if it becomes too mentally difficult or boring, you stop.
There’s nothing wrong with dabbling, and often it’s by dabbling that I first explore whether I’d like to learn something through an ultralearning project. However, despite the investment of time and energy, ultralearning projects can help you achieve breakthroughs in whatever you’re trying to learn.
Reason #1: You Can Learn Much Faster
Ultralearning projects are hard. But the trade-off is that intense focus enables rapid learning progress. Eliminating distractions, learning the hardest parts first, driving at your weaknesses and investing concrete chunks of time all enable you to take a learning endeavor that you might normally imagine learning over a few years and compress it into a few months.
My language learning project was a good example of this. Yes—the No-English Rule and intensive study did require a lot of effort. But the advantage was that I reached a level in three months that often takes a year or two of more typical study.
Reason #2: It Gets to the Fun Part of Learning Faster
Many learning opportunities become more interesting when you get better at them. Languages are much more fun when you can actually hold conversations. Work skills are more useful when they actually help your career. Drawing, sports and music are all more fun when you’re good at them.
Ultralearning can allow you to push faster through the frustrating parts and get more quickly to a level where continuing mastery is enjoyable and fun.
Reason #3: Ultralearning Projects are Interesting
When I told someone I was about to take on the MIT Challenge, they said, “You must really love studying.” But the truth was, I didn’t actually enjoy most of the classes I had in university. I found many of them simultaneously boring and frustrating. I hated the busywork, the group projects, the classes where the professor didn’t say anything useful and I had to struggle to stay awake. Traditional learning involves long stretches of boredom peppered with random frustration.
When I did the MIT Challenge, however, almost all of my classes were interesting. I think the reason was that self-education is results-driven. It doesn’t matter which resources you use, as long as you get to the point. I could skip assignments I didn’t think would help me master the material. I could watch lectures faster if they were boring, rewatch them if I was confused. Optimizing for faster learning, in turn, also optimized for being completely engaged with learning.
Ultralearning is more interesting because everything you do feels like it actually matters.
Reason #4: The Opportunities for Quick Learners are Ever Increasing
Ultralearning is a skill. Once you’ve mastered the process you can repeat it again and again on anything you want to learn.
It’s also a skill that’s becoming increasingly valuable. The economy is hollowing out the middle . Workers are expected to adapt faster and faster to new ways of doing things. The best in the profession are earning ever more than the average. Flexible, rapid learners have a golden opportunity, while those who struggle to keep up are going to find it harder and harder to survive.
Practicing on ultralearning projects gives you an edge like almost nothing else will in skilled professions.
How to Design Your First Ultralearning Project
Designing your own ultralearning project has three parts:
- Figuring out what you want to learn deeply, intensely and quickly.
- Choosing which format you want for your project.
- Preparing to start learning.
Step #1: What Do You Want to Ultralearn?
What would you like to learn? It could be a subject—say you want to quickly learn a lot of history, business or math. It could be a career skill—you want to master Excel or JQuery. It could be something you’ve always wanted to learn for fun—guitar, French or painting.
What you want to learn doesn’t matter and I can’t choose it for you. But I can suggest a couple things to keep in mind when picking the subject:
- Only pick one thing. Ultralearning projects need specificity. Saying you want to learn guitar, French AND cooking is a recipe for a mess of a project. Instead pick one thing and save the other things you want to learn for a later project.
- Shorter projects need more constraints. The smaller your project is, the more it needs to focus on something specific to make progress noticeable. If you’re only going to spend a month, one-hour a day, then don’t make your project “learn programming” or “learn Chinese”. Instead make it more focused: “learn to make text-only games in Python” or “learn pinyin and master set phrases in Mandarin.”
- Avoid overly specific goals and deadlines. For first-time ultralearners, I don’t recommend setting a particular goal and deadline, like I did with the MIT Challenge. The reason is that once you start learning, you’ll quickly realize whether your goal is realistic, too easy or too hard. If it’s too easy, you won’t focus. Too hard, you’ll probably give up. That means you have a fairly narrow range to shoot in order to be successful. A better approach is to pick the direction you want to learn and choose a target when you’re about one-third to halfway done the project. So a good approach might be that you’ll learn the MIT computer science curriculum, but once you start you can decide how far it is realistic for you to get in the time you have.
Step #2: Choose the Project Format
There’s a lot of different ways to do an ultralearning project. Which you use will depend heavily on your schedule and the importance of the challenge to you.
Here’s three different styles for an ultralearning project:
- Full-time projects. These are the most intense, most costly and fastest projects. The advantage of a full-time or nearly full-time project is that you can really get learning done in incredibly short time periods. Good if you’re between jobs, classes or otherwise can devote yourself to the project.
- Fixed-schedule projects. These are projects which have concrete hours you’ll devote to them every week. One example could be spending an hour each day before work, two hours before bed, or two 5-hour bursts on the weekend. The amount of time isn’t too important (although less weekly investment = slower progress) but I wouldn’t recommend putting in chunks of time less than 30 minutes. Fracturing the time over too many spots in the day doesn’t enable the focus required.
- Fixed-hour projects. These projects don’t have a particular schedule, but they do have a number of hours (3, 5, 20) that you’ll put in each week wherever you can find time in your schedule. This is the hardest type of project to successfully execute, but it may be the only feasible way to do ultralearning for some people.
In general, I recommend an ultralearning project be your principle goal during the period you’re doing it. It’s okay to keep working on other things and maintain habits. But ultralearning projects don’t work well if they’re just one of many things you’re simultaneously trying to achieve.
Once you’ve picked a format, you need to select a length of time. In general, if your weekly time investment is low, you’ll need either a long project or a more severely reduced scope. If I wanted to learn programming, but was only putting in 3 hours per week for ultralearning, I would either need a long time horizon (say 6-12 months) or reduced scope (a particular language, type of program, etc.)
Step #3: Preparing to Learn
I actually don’t recommend starting right away when you have an ultralearning project. The reason is that the intensity of learning can make it very easy to quit if you haven’t planned it properly.
A good ultralearning project starts with some amount of time in preparation. This allows you to gather material, research different learning strategies for your particular skill or subject, plan out your time and conduct a pilot test of the schedule.
My rule of thumb is that preparation should be no less than 50% of the length of the project itself with full-time hours. So when I did the MIT Challenge (a full-time project over one year) I would want six months minimum of low-intensity preparation. If you’re doing five hours per week over 8 weeks, I would want to spend at least a week doing preparation.
Here’s what you need to do in that preparation time:
- Research how learning works best for that particular domain. Hunt around for all the possible learning methods, strategies and recommendations. Note common themes and complaints people make. Note also alternative strategies that differ. This should give you a good idea of how you want to learn, as well as backup options in case your first approach fails you.
- Gather material and design a preliminary attack plan. Order books online if you need them. Sign up for online courses. Get tools, material and equipment if you need any. Then create a simple plan for approaching them to learn. This doesn’t need to be complicated. For the MIT Challenge it was: (1) Watch/Read, (2) Practice Questions, (3) Feynman Technique. For the Year Without English it was: (1) No-English Rule, (2) Tutoring, (3) Book Study.
- Conduct a pilot week of the schedule. Before you fully commit to starting the project, test it out. See how it fits into your life and get a sense of how difficult it will be. If it is too hard, or your schedule is unrealistic, now is the time to adjust it.
Now It’s Your Turn
If you’ve followed this far, I’m assuming you’re at least somewhat interested in starting your own ultralearning project. So why not just do it?
Write in the comments here below what ultralearning project you want to tackle: what you want to learn and the format you’re going to use to pursue it. In response, I’ll try to reply to as many people as possible offering advice on how to make their ultralearning projects a success!