Private Lecture: How to Remember EVERYTHING You Read

Summary

The core problem: too much to learn, too little time

Learning coach Justin Sung works with Dan, a software engineer who needs to keep up with new developments while balancing work and family. Dan’s main difficulty is not simply understanding new material; it is retaining it and being able to use it later.

Sung argues that the solution is to make learning snowball: connect new information to an increasingly organized body of knowledge, so each new idea has a place to “stick.”

Why understanding alone is not enough

Information that is understood but left isolated is unlikely to become durable, usable knowledge. Sung describes two risks:

• Learning debt: Time is spent learning something, but because it is shallow or disconnected, it is later forgotten and must be learned again.
• Blind debt: Notes, databases, mind maps, or AI conversations can make knowledge feel secure even when it cannot be recalled or applied. This can create misplaced confidence and lead to poor decisions or ineffective problem-solving.

For knowledge to be useful, it must be organized in a way the learner can recall and apply—not merely stored somewhere.

The “Goldilocks” level of cognitive effort

Sung says effective learning should feel challenging, but not impossible:

• Too little effort: Passive reading, repetition, or memorizing individual facts reduces the feeling of overwhelm, but often leaves information disconnected and hard to use.
• Too much effort: Trying to understand every relationship among all the information at once overloads working memory.
• A productive middle ground: Work on a manageable segment, organize it, then connect it back to the larger picture. Repeat this cycle with new segments.

The learner should generally find that a topic becomes less overwhelming over time. As a coherent framework forms, new information becomes faster to place because there are fewer plausible places for it to fit.

Practical method for organizing new knowledge

1. Get a broad view of the topic.

   • Start by gathering major keywords across the subject.
   • A beginner or someone prone to feeling overwhelmed might start with around 15; someone more comfortable with the topic might collect 50–100.
   • Keywords can come from existing knowledge and materials, or be generated with an AI tool and then edited. The initial list is a starting point, not a final map.

2. Decide whether to learn broadly or deeply.

   • For a broad, relatively new subject, aim for a shallow overview across the field, identifying its biggest landmarks.
   • For a specific area of expertise, filter the keywords to focus on the material most relevant to that goal.
   • Either way, narrow the material to roughly four to six key ideas to work with at a time.

3. Build meaningful relationships among the ideas.

   • Form groups and hypothesize how the concepts relate.
   • Do not stop at the first obvious connection. Two ideas might be linked by sequence, contrast, function, category, or another relationship.
   • Look for the structure that is both understandable to you and useful for developing expertise. The best structure may differ from the way the material was originally presented.

4. Work in cycles of zooming in and out.

   • Zoom in on a manageable segment and organize it.
   • Zoom out to see how that segment fits into the broader topic.
   • Choose another segment, organize it, and integrate it with the growing framework.
   • Leave an idea temporarily unconnected if its place is unclear, and revisit it as more of the picture develops.

5. Revise the framework as understanding improves.

   • Groups and labels are provisional. A category initially called a “method,” for example, might turn out to describe a broader principle.
   • Reorganizing groups as new understanding emerges is part of learning, not evidence that the earlier work failed.

How to challenge and test your understanding

Sung recommends using questions and exercises to make the learner’s thinking more explicit:

• Explain the framework to a 10-year-old. If the explanation is not clear and intuitive, there may be gaps in the relationships or the overall organization.
• Try alternative structures. Ask whether an expert might represent the topic more simply or meaningfully in a different way.
• Make unlikely connections deliberately. Take two ideas that do not appear related and explore whether they can be grouped through another perspective. This can reveal new patterns of thought.
• Predict missing groups. Based on the groups already formed, infer what other categories or relationships ought to exist before consulting the remaining material.
• Test recall and application. Challenge knowledge by trying to retrieve and use it, rather than relying on the feeling that it is safely recorded in notes.

Sung stresses that this process should feel cognitively engaging. If it feels effortless, the learner may not be challenging or revising their framework enough.

Notes and mind mapping

Sung recommends beginning with free-form, hand-drawn mind maps on an open or “infinite” canvas. He gives three reasons:

• Free-form drawing makes it easier to express ideas without being constrained by a note-taking interface.
• Drawing and rearranging ideas can slow thinking enough to make the learner examine how concepts fit together.
• The resulting map can reveal gaps or weaknesses in the learner’s organization.

Ideas can be moved and reconnected as the map develops. Digital tools may be useful once the learner can organize ideas fluently without losing focus to the interface.

Advice tailored to Dan

Dan says that his work in software already involves a similar process of stepping back, focusing on a manageable part, and then reconnecting it to the larger system. He wants a more consistent framework for deciding what to learn and how deeply to pursue it.

Sung advises him to begin by making his existing ideas and assumptions explicit, then actively challenge and reorganize them—for example, while studying agentic AI. If this produces little cognitive engagement, Dan should question whether he is pushing himself enough. If it is engaging and his understanding, memory, or ability to explain the topic improves, that is a sign the process is working.

Sung also notes that learning priorities depend on context: broad coverage is useful when building general expertise, while deeper study is worthwhile when a specific skill or decision requires it. A sense that something is “relevant” is not a prerequisite for learning it; learners can develop relevance by creating connections to what they already know and to how they may use the knowledge.

Speakers and sources featured

• Justin Sung — learning coach and primary speaker.
• Dan — software engineer and coaching client.

ChatGPT and Claude are mentioned as possible tools for generating keywords and exploring ideas, but they are not speakers in the video.