Get AI summaries of any video or article — Sign up free
7 Years of Building a Learning System in 12 minutes thumbnail

7 Years of Building a Learning System in 12 minutes

Justin Sung·
5 min read

Based on Justin Sung's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

Treat learning as a pipeline: priming, encoding, reference, retrieval, then (optionally) overlearning—rather than as a repetition-only loop.

Briefing

A practical learning system built around five linked processes—priming, encoding, reference, retrieval, and overlearning—can cut study time dramatically by fixing where learners lose efficiency: relevance filtering, active processing, and retrieval practice. The core claim is that memory and understanding don’t improve just because more information is repeated; they improve when the brain’s internal pipeline is optimized so new material is filtered as “relevant,” actively organized into a usable structure, and then repeatedly retrieved to strengthen and refine it.

The system starts with a reframing of how learning works. Information enters the brain, but it’s filtered based on what already exists in memory—what seems connected and worth keeping. If incoming material doesn’t connect to prior knowledge, it gets processed poorly and is forgotten quickly, no matter how many times someone rereads or rewrites notes. When information does meet the brain’s relevance criteria, it moves into deeper processing: organizing it, making sense of it, and building a structured model that’s easier to store and retrieve. Retrieval then becomes a feedback loop. Pulling knowledge back out forces reprocessing and reorganization, which further improves both memory and understanding.

A personal example illustrates what goes wrong without this model. During the push toward medical school, the study routine centered on long reading sessions, lecture listening, rewriting notes, and later flash cards on past papers. Much of the learning felt random and was forgotten, leading to a cycle of re-studying. The missing piece wasn’t effort—it was the early relevance step. Without priming, the brain didn’t treat new information as worth prioritizing, so later encoding and retrieval couldn’t work efficiently. Once priming and the rest of the pipeline were addressed, study became noticeably more effective.

The “PERO” system breaks the pipeline into actionable parts. Priming prepares the brain before first exposure to a topic—by running an activity that signals relevance so the material is more likely to be filtered into processing rather than causing overload. Encoding is the active work of organizing information: grouping, simplifying, finding analogies, and building connections so the brain can store a coherent structure. Reference is a “parking lot” for fine-grained details that distract from encoding—using tools like flash cards, apps, or note systems such as Obsidian to save details for later.

Retrieval is non-negotiable in the system because it tests usable knowledge and also strengthens memory through reprocessing. Overlearning comes last and is optional: it means going beyond what’s required through extra practice and repetition to build faster recall and fluency, but it’s most valuable for high-stakes, competitive assessments. The system also warns against using overlearning too early, since it can waste time compensating for weak priming and encoding.

To help learners diagnose their own process, a free “Learning System Diagnostic” quiz scores each part of PERO and provides prioritized recommendations. Early access feedback suggests the quiz is easy to use, helps pinpoint what’s wrong, and is especially helpful for people who want their learning effectiveness quantified rather than guessed.

Cornell Notes

The PERO learning system treats studying as a pipeline, not a loop of repetition. New information must first be primed so the brain filters it as relevant, then encoded through active organization (grouping, simplifying, analogies, connections) so it can be stored as a usable structure. Reference keeps distracting details parked for later, while retrieval forces testing that also strengthens memory by reprocessing knowledge. Overlearning is optional and best saved for high-stakes exams, because using it too early often compensates for weak priming and encoding. A free diagnostic quiz scores each PERO component and recommends what to fix first.

Why does the PERO system treat “relevance” as the first bottleneck in learning?

Incoming information is filtered based on what already exists in memory. If learners can’t see how new material connects to prior knowledge, the brain treats it as irrelevant and forgets it quickly—even if someone rereads, rewrites notes, or repeats flash cards. Priming is designed to improve that early filtering so more information moves into processing instead of getting dropped.

What does “encoding” mean in this framework, and why is it harder than passive studying?

Encoding is the active processing step where learners organize information into a coherent model. It includes grouping related ideas, simplifying, finding analogies, and building relationships so the material becomes intuitive and easier for memory to hold. The transcript emphasizes that encoding isn’t binary; it’s a spectrum of how efficiently information is processed into memory. If encoding is weak, learners may study a lot but still forget much of what they learned.

How does “reference” prevent overload during encoding?

Reference is a deliberate parking lot for fine-grained details that don’t help build the main structure during encoding. Instead of getting bogged down in minutiae, learners store those details for later retrieval—using tools such as flash cards, apps, or note systems like Obsidian. This keeps mental energy focused on moving the core concepts through priming → encoding → retrieval.

Why is retrieval described as both a test and a learning mechanism?

Retrieval means pulling knowledge from memory by testing and applying it. It matters because it checks whether knowledge is actually usable, not just recognized. Just as importantly, the act of retrieving forces reprocessing and repackaging of the information, which strengthens memory and deepens understanding—so retrieval refines the stored model over time.

When should overlearning be used, and what’s the risk of using it too early?

Overlearning is learning beyond what’s required through extra repetition—like more practice questions or additional flash cards—to build faster recall and fluency. It’s most effective for competitive assessments or very high standards. The risk is using overlearning as the first strategy: it wastes time and attention while priming, encoding, and retrieval are still underperforming, effectively compensating for problems that should be fixed earlier.

What does the diagnostic quiz do, and what does feedback suggest about its usefulness?

The free “Learning System Diagnostic” quiz scores learners on each PERO component and generates recommendations prioritized by score. Early access feedback highlighted that it’s easy to use, helps pinpoint exactly what’s going wrong, and provides a way to quantify learning effectiveness rather than relying on confidence or intuition. People also reported it was helpful across a range of scoring outcomes.

Review Questions

  1. In the PERO model, what specific failure mode occurs when learners don’t prime before first exposure to a topic?
  2. How do reference and encoding differ in purpose, and why does mixing them up increase cognitive overload?
  3. Why does retrieval strengthen understanding even when it reveals gaps in knowledge?

Key Points

  1. 1

    Treat learning as a pipeline: priming, encoding, reference, retrieval, then (optionally) overlearning—rather than as a repetition-only loop.

  2. 2

    Improve early relevance filtering with priming so new material is more likely to move into deeper processing instead of being forgotten quickly.

  3. 3

    Use encoding to actively organize information (grouping, simplifying, analogies, connections) so memory stores a usable structure.

  4. 4

    Park distracting details in a reference system (e.g., flash cards or Obsidian) so encoding stays focused on what builds understanding.

  5. 5

    Make retrieval a core habit because it both tests usable knowledge and reprocesses it to deepen understanding.

  6. 6

    Reserve overlearning for high-stakes or highly competitive contexts; using it early often wastes time when priming and encoding are weak.

  7. 7

    Use the Learning System Diagnostic quiz to score each PERO component and prioritize fixes based on what’s dragging efficiency down.

Highlights

The system’s central bottleneck is not effort—it’s relevance filtering: without priming, the brain may treat new information as unimportant and forget it rapidly.
Encoding is framed as active organization work (grouping, simplifying, analogies, connections), not passive note consumption.
Retrieval is described as a feedback loop: testing knowledge forces reprocessing and strengthens memory beyond simple practice.
Overlearning is optional and best saved for competitive assessments; using it too early can mask problems in priming and encoding.
A free diagnostic quiz scores PERO components and turns vague “I think I’m learning” into prioritized, measurable improvement steps.

Topics

Mentioned