Learning that is built like memory, not like a feed.
Nexus is organized around three things learning research keeps finding: you have to encode an idea, you have to retrieve it, and you have to come back after some forgetting. Bloom’s taxonomy tells us what kind of thinking to ask for along the way.
The loop is not a slogan
Understand, retrieve, consolidate. In the product those words label a sequence. In the research they name different jobs. Encoding is how a memory trace is formed. Retrieval is how that trace is strengthened by being used. Consolidation, especially when practice is spaced, is how something that was fluent on Tuesday is still there in May.
01
Encode
Attentive reading builds the first trace. If the source is a lecture, a chapter, or a question you keep asking, the lesson sections exist to make that encoding followable — objectives included — instead of leaving you with a stack of clippings.
02
Retrieve
Rereading is mostly recognition. Retrieval is production: flashcards, multiple choice, short answers. The discomfort is information. It shows you what only looked familiar while the notes were still in view.
03
Consolidate
Unused traces weaken. Coming back after some forgetting, and rating whether an item was forgotten, hard, good, or easy, is how review goes to the weak spots instead of through the whole notebook again.
Bloom’s taxonomy, as practice
Remember, understand, apply, analyze, evaluate, create. The taxonomy is often a poster. In Nexus it is a way to tag objectives and exercises so practice can ask for more than recall. You can study a lesson across all six, or filter when you already know the term and still cannot use it.
Remember & understand
The encode phase. Can you recall the idea, and can you explain it in your own words? This is where most unstructured notes stop.
Apply & analyze
The retrieve phase. Can you use the idea on a new example, and can you take a claim or a system apart? This is where tutoring hours are often spent repeating last week.
Evaluate & create
The consolidate phase. Can you judge a solution, or build with the idea? These demands keep a lesson from becoming a glossary you can only recognize.
What the research actually supports
The testing effect: trying to recall something teaches it better than looking at it again. Spacing: spreading practice over time beats cramming for anything you need after the due date. Desirable difficulty: tasks that feel slightly harder often produce more durable learning than tasks that feel smooth. Bloom’s taxonomy: different prompts grow different kinds of competence.
Nexus applies those findings as product structure. It is not a clinical trial of the app, and generated text is not an authority. You still have to read what the model writes, especially before you assign it. The scientific claim is narrower and more useful: the loop you follow matters more than the pile of sources you keep.
“
If a tool only helps you gather material, it has stopped at the least durable part of learning.
Why highlighting is a poor test
A highlight records that a sentence seemed important while it was on the page. It does not ask whether you can reconstruct the argument tomorrow. Unstructured reading and notetaking are excellent at creating that feeling of progress. They are weak at producing the evidence. Practice that requires retrieval is less flattering and more honest.