Ultralearning by Scott Young [2026]

After five years teaching Earth science, I noticed a pattern. Students who grew quickly weren’t simply working harder. They were learning differently. When I first read Scott Young’s Ultralearning, I realized that pattern had a name.

What Is Ultralearning

In 2012, Scott Young set himself an audacious challenge: complete MIT’s entire four-year computer science curriculum — 33 courses — in 12 months, without enrolling. He watched recorded lectures, completed the same problem sets, and took the same final exams. He passed all 33. The project, which he documented publicly as the “MIT Challenge,” became the empirical backbone of his 2019 book Ultralearning: Master Hard Skills, Outsmart the Competition, and Accelerate Your Career.

Core definition: Ultralearning is an intense, self-directed process that consciously applies effective learning strategies.

“The best ultralearners are those who ruthlessly seek the most efficient path to their goal.” — Scott Young, Ultralearning, 2019

Young’s framework draws on cognitive science research spanning decades — from Ebbinghaus’s forgetting curve (1885) to Roediger and Karpicke’s retrieval practice studies (2006). What makes it distinctive is the integration of nine discrete principles into a single, actionable system.

Related: mental models guide

The 9 Principles Explained

1. Metalearning: Learn How to Learn First

Before starting any new subject, research how to learn it. Young recommends investing roughly 10% of your total planned study time in metalearning — mapping out the concepts, identifying the best resources, and understanding the skill structure before diving in. For the MIT Challenge, this meant spending weeks analyzing which courses had the best recorded materials and which problem sets were most representative of exam content.

I apply this in the classroom. Before each new Earth science unit, I give students 30 minutes to predict “what will be hardest in this unit” and map out what they already know versus what’s unfamiliar. Research on metacognitive strategies supports this: Dunlosky et al. (2013) found that practice testing and distributed practice were the two highest-utility learning strategies — both of which require metalearning to implement effectively [1].

2. Focus: Train Your Attention Like a Muscle

As someone with ADHD, this is the hardest principle for me. Young treats focus as a trainable capacity, not a fixed trait. He distinguishes three problems: starting (procrastination), sustaining (distraction), and optimizing (choosing the right intensity for the task).

The practical method is progressive overload. I started with 25-minute Pomodoro sessions. The first week, even that felt long. After three months of daily practice, I could sustain 90 minutes of deep work. Cal Newport’s research on deep work aligns here: the average knowledge worker manages only about 2.5 hours of genuine deep work per day. That number is trainable, but only through deliberate, incremental extension — not through willpower alone.

3. Directness: Learn by Doing the Thing Itself

Learn in a way directly connected to your goal. If the goal is conversational fluency in Spanish, practice speaking Spanish — not completing grammar worksheets. This is why Young solved actual MIT exam problems rather than just watching lectures. He called this the “transfer problem”: most learning activities fail to transfer because the practice context differs too much from the performance context.

Research basis: Transfer-appropriate processing theory (Morris et al., 1977) shows that the more the learning context matches the use context, the greater the memory transfer. [3]

Benny Lewis, the polyglot behind “Fluent in 3 Months,” exemplifies directness: he begins speaking on day one, accepting errors, because the goal is conversation — not grammar perfection. Eric Barone (ConcernedApe) learned game development by building Stardew Valley from scratch rather than completing tutorials.

4. Drill: Attack Your Weakest Links

Identify the specific component that limits your overall performance, then practice it in isolation. Young calls this the “rate-determining step” — borrowed from chemistry, where the slowest reaction step controls the speed of the entire chain. In my classroom, I have students keep a “wrong answer notebook.” Every test, they record what they got wrong and why. Over a semester, patterns emerge — and those patterns become their drill targets [1].

5. Retrieval: Test Yourself Instead of Re-Reading

Practicing pulling information from memory is far more powerful than passive re-reading. Roediger and Karpicke’s landmark 2006 study found that students who took a practice test after studying retained 50% more material one week later than students who spent the same time re-studying. The effect is robust and has been replicated across ages, subjects, and formats [2].

Practical tools: flashcards (physical or Anki), blank-page summaries (close the book, write everything you remember), and self-quizzing. The discomfort of struggling to recall is the signal that learning is happening.

6. Feedback: Seek Corrective, Not Just Evaluative

Young classifies feedback into three types: Outcome feedback (pass/fail, right/wrong), Informational feedback (which aspects were strong or weak), and Corrective feedback (what specifically to change and how). The most powerful is Corrective. A grade of 72% is outcome feedback. “Your argument in paragraph three lacks supporting evidence” is corrective feedback. Seek the most specific feedback available, and seek it as quickly after performance as possible.

7. Retention: Fight the Forgetting Curve

Ebbinghaus’s forgetting curve (1885) demonstrated that roughly 70% of newly learned material is forgotten within 24 hours without review. Spaced repetition — reviewing material at increasing intervals — is the most effective countermeasure. The algorithm underlying Anki, SuperMemo’s SM-2, schedules reviews at optimal intervals based on how easily you recalled each item. Young used spaced repetition for the MIT Challenge and credits it as essential for retaining material across 33 courses simultaneously.

8. Intuition: Understand, Don’t Just Memorize

Deep understanding creates intuition — the ability to solve novel problems by recognizing underlying patterns rather than applying memorized procedures. Young strongly recommends the Feynman Technique: choose a concept, explain it as if teaching a 12-year-old, identify where your explanation breaks down, and study those gaps specifically. Richard Feynman famously used this approach himself, and it forces genuine comprehension rather than surface familiarity.

9. Experimentation: Develop Your Own Methods

To reach a high level in any skill, you must eventually move beyond established methods and develop your own approaches. Young recommends a three-phase progression: first copy existing methods faithfully, then compare multiple methods to understand tradeoffs, then introduce controlled variations to find what works uniquely for you. Van Gogh copied hundreds of paintings before developing his distinctive post-impressionist style. My own teaching approach evolved through exactly this progression over five years.

The MIT Challenge: Numbers and Context

Young’s MIT Challenge involved 33 courses across computer science and mathematics. He used MIT OpenCourseWare materials — the same lectures, readings, and problem sets available to enrolled MIT students. For each course, he watched the full lecture series, completed the assigned problem sets, and took the final exam under timed conditions, requiring a passing score on each.

The total cost was roughly $2,000 (for textbooks and exam proctoring) compared to MIT’s tuition of approximately $200,000 for the same four-year program. The time investment was approximately 12 months of full-time study. Young documented daily progress publicly, providing accountability and a dataset for refining his methods in real time.

Critics point out that Young didn’t earn an MIT degree, didn’t participate in labs or group projects, and didn’t have access to professors for office hours. These are valid limitations. Young acknowledges them. The point was never to replace an MIT education — it was to demonstrate that the core knowledge content of a world-class CS curriculum could be self-taught in a fraction of the time using deliberate learning strategies.

Applying Ultralearning as a Teacher

I decided to acquire one new skill per semester using the Ultralearning framework. Last semester it was Python data visualization; this semester it’s English academic writing. Before starting each project, I run through all 9 principles as a checklist:

• Metalearning: What’s the skill structure? What resources exist? What will be hardest?
• Focus: When and where will I practice? How will I protect that time?
• Directness: Am I practicing the actual skill, or a proxy?
• Drill: What’s my weakest component? How will I isolate it?
• Retrieval: Am I testing myself, or just reviewing?
• Feedback: Who will give me corrective feedback? How quickly?
• Retention: What’s my spaced repetition plan?
• Intuition: Can I explain this to someone else simply?
• Experimentation: Am I ready to deviate from the template?

This checklist takes five minutes and consistently reveals blind spots I would have missed otherwise.

The Limits of Ultralearning

Young himself acknowledges that Ultralearning is an intense approach. It demands significant time blocks, sustained cognitive effort, and tolerance for discomfort. For people with ADHD like me, consistency sometimes matters more than intensity. Twenty-five minutes every day beats ten hours once a month.

There are also domains where ultralearning is less applicable. Skills requiring physical adaptation (endurance sports, musical instrument technique) have biological rate-limiters that cognitive strategies can’t bypass. Social and emotional skills develop through lived experience more than deliberate practice. And some subjects genuinely require extended incubation time — creative work often benefits from stepping away rather than pushing harder.

The framework is most powerful for knowledge-intensive skills with clear feedback mechanisms: programming, mathematics, language learning, test preparation, and applied sciences.

Key Takeaways

• Invest 10% of your learning time in metalearning before starting any new subject
• Practice the actual skill (directness), not a comfortable proxy
• Test yourself actively (retrieval) rather than passively re-reading
• Use spaced repetition to fight the forgetting curve — Anki or a simple review schedule works
• Seek corrective feedback, not just grades or pass/fail signals
• Build focus gradually through progressive overload, not willpower
• When you plateau, identify your rate-determining step and drill it specifically

Last updated: 2026-03-31

References

1. Young, S. (2019). Ultralearning: Master Hard Skills, Outsmart the Competition, and Accelerate Your Career. HarperBusiness. Link
2. Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249-255.
3. Morris, C. D., Bransford, J. D., & Franks, J. J. (1977). Levels of processing versus transfer appropriate processing. Journal of Verbal Learning and Verbal Behavior, 16(5), 519-533.
4. Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students’ learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4-58.
5. Young, S. (2026). Foundations. scotthyoung.com/blog. Link

Related Reading

• Classroom Behavior Management with Positive Reinforcement
• How We Search for Extraterrestrial Intelligence [2026]
• How to Teach Growth Mindset in Math [2026]