How Cruxly Finds Connections Between Thoughts You Never Noticed

Two thoughts, saved weeks apart, from completely unrelated analyses, turn up connected even though nobody ever set out to compare them. That's not a coincidence or an edge case. It's how matching works in Cruxly by default: comparison runs on what a thought actually means, not the words used to write it, and not which folder it would have been filed into.

Why Folders and Tags Don't Solve the Whole Problem

A folder only helps if you decide where to file a thought ahead of time, then remember that decision months later when a second, related one shows up. A tag requires the same thing: picking the exact same word twice, weeks apart, for two thoughts that have no obvious connection in your head at the moment the first one gets saved. In practice, almost nobody does this. Not out of laziness: at the moment something gets saved, it's usually not yet clear what it will end up connecting to.

What's Actually Being Compared

What gets compared is the meaning behind a thought, not its exact wording. Two thoughts phrased completely differently, but about the same underlying idea, land next to each other in the comparison even if they share almost no words. It works the same way a person instantly understands that "terminate the contract" and "end the obligation" are about the same thing, even though neither word repeats. Second Brain does this across an entire personal collection at once, not just for one pair of phrasings someone deliberately set side by side.

Example: A Psychology Lecture and a Product Report

One thought gets saved from a breakdown of a lecture on decision-making: the more options someone is offered at once, the higher the chance they pick none of them. A second gets saved a month later from a PDF report on shopping-cart abandonment: conversion drops specifically at the step where shoppers see too many similar products at once. No word from the first thought repeats in the second. But both are really about the same effect: too much choice works against action instead of encouraging it. This is exactly the kind of pair meaning-based matching is meant to catch, since someone scrolling their own saved list probably wouldn't connect a psychology lecture to a product report without carefully rereading both.

Read the practical guide to saving highlights from video and PDF

Example: Lecture Notes From One Course, a PDF From Another

A student saves a thought from an economics lecture breakdown: scarcity alone creates a sense of value, even with no actual change in quality. Two months later, studying for a different course, the same student analyzes a PDF chapter of a marketing textbook and saves a thought about limited-time offers, which lift conversion through urgency rather than price. Different courses, different formats, almost no words in common. Second Brain still finds the two thoughts connected, because the underlying mechanism is the same one: scarcity raises perceived value.

Suggested Connections vs. the Ones You Pin Yourself

Not every connection Second Brain finds is worth keeping forever, and the feature deliberately treats those as two different things. Similar thoughts get suggested automatically and recalculated every time something new gets saved. Turning a suggestion into a permanent connection, one that stays even after a newer thought would otherwise push it out of the suggested list, is a separate, deliberate action: one click pins it, another unpins it. That's what lets the automatic matching stay generous with suggestions: nothing becomes permanent without someone explicitly agreeing to it.

Testing This on Your Own Thoughts

The easiest way to see it firsthand: save a thought from one analysis, then a day or two later, after a completely unrelated breakdown, save a second one on a related underlying idea, worded completely differently. Check the related thoughts on the first one. If matching works the way it's described above, the second thought will already be there, even if you yourself never consciously connected the two.

Read the complete AI second brain guide

Matching words, folders, or tags isn't what decides whether a connection gets found. Whether two thoughts are really about the same thing does, and Second Brain works that out on its own, for every pair of saved thoughts at once.