Knowledge OS · v0.1
Manifesto · Live Prototype · Open Source

A structured operating system for human knowledge.

Linear textbooks, scrolling feeds, and 12-minute videos all hide the same problem: knowledge has shape — and we keep flattening it. Knowledge OS treats every concept as a node in a navigable graph, with depth, links, paths, and compression built in.

FIG.00 — SEED GRAPH / 种子图
N₀ · 24 nodes · 41 edges
§01 · The Problem

Most learning is linear. Most knowledge is not.

A book has page 1 to page 400. A course has Lecture 1 to Lecture 24. But the actual map of what you're learning isn't a line — it's a graph. The line is just the projection someone chose for you.

LINEAR · 1D

Page → page → page → page

You read in order. You can't see what's connected. You can't change depth. You can't jump domains. The author's path becomes your only path. When you get lost, you start over.

P1
P2
P3
P4
P5
STRUCTURED · ND

Concept ↔ concept ↔ concept

You see the map. Every node has a depth dial, every edge is typed, every path is one of many. You can collapse a chapter into a sentence, or expand a sentence into a chapter. You navigate, not just read.

§02 · Module 01 — Knowledge Graph

Every concept is a node. Every relation is an edge.

Click a node to load it. Watch the neighborhood update. The graph below is a 24-concept seed slice across math, physics, computation, and philosophy — drag, click, explore.

Math Physics Computation Philosophy
FIG.01 — SEED CORPUS / 种子语料 · drag · click
— SELECT NODE —
Click any concept
The graph is the canonical surface of Knowledge OS. Each node has a category, a depth axis, and a typed neighborhood — meaning every click reveals what is *adjacent* in concept-space, not in page-space.
SPECIFICATION
Force-directed layout · typed edges (prerequisite, instance-of, dual, analogy) · depth-1 default render, depth-N on demand.
§03 · Module 02 — Multi-Level Explanations

One concept. Five depths. You choose.

A child needs a metaphor. A student needs an example. A practitioner needs a definition. A researcher needs the formalism. The same concept should serve all of them — at the turn of a dial.

ENTROPY · 熵
§04 · Module 03 — Cross-Domain Linking

The same shape appears in different fields.

Most "new ideas" are old patterns crossing borders. Knowledge OS makes the bridges first-class — when you understand a concept, you also see its isomorphisms in adjacent domains.

§05 · Module 04 — Learning Paths

A path is a traversal of the graph — pick yours.

A textbook is one path through the graph. A course is another. Knowledge OS lets you choose, fork, or build paths — optimized for your goal, not the publisher's table of contents.

§06 · Module 05 — Compression & Expansion

Zoom out for the shape. Zoom in for the steps.

Same content, five resolutions: tweet → paragraph → page → chapter → formal proof. The reader picks resolution; the system maintains coherence across all of them.

§07 · Manifesto

The next century of learning will not be longer videos or thicker books. It will be navigable, structured, multi-resolution knowledge — the same content, every depth, every link, every path, on demand.

— Knowledge OS · An open prototype