Knowledge Universe › Library model
Universal · conceptOne library model for every world.
School, Medicine, Law, Research — different knowledge, same rules. The rules are defined once, here, and reused: two spaces, one rights profile per source, six source-quality tiers, and contribution that means knowledge rather than files.
International learning concept — currently available in English. DE / RU / UK: roadmap.
Two spaces — in every world
Community Library and My Study Space.
Community Library
Shared, legally usable knowledge. Material with a confirmed basis for community use:
- public domain
- compatible Creative Commons / Open Access
- own work with granted rights
- rightsholder permission
- other verified legal basis
Visible to everyone. Every entry stays linked to its source, its quality tier and its rights status. Entry only through the contribution lifecycle.
My Study Space
Future private learning workspace: notes, reading list, progress, own materials, Common Table threads — the same concept for a school learner and a researcher.
Private use, storage or processing of protected material remains rights-, licence- and jurisdiction-dependent.
Current prototype: NO REAL STORAGE · NO REAL UPLOAD. An architectural concept, not a cloud drive. For learners under 18 it also means: no profile, no name, nothing saved.
There is no separate School library, Medicine library or Law library. One Community Library, one My Study Space, filtered by world — never incompatible copies.
Rights profile — universal
Per source, per capability. Defined once.
Each source carries its own flags. One flag never implies another, and no world duplicates this logic.
RAG / library use ≠ MODEL_TRAINING. Answering with quotes and links is library use. Training or fine-tuning model weights is a different act with its own rights class — absent by default for every entry, in every world.
Worked examples with real flags: Medicine → Library (public-domain book, CC BY-SA translation, metadata-only entry).
Source quality
Six tiers — and AI-generated is always labelled.
Primary source
The original: a law text, an experiment's data, a historical document, a theorem's proof.
Official source
Issued by the responsible authority: a curriculum, a guideline, a standard, a court.
Textbook / educational source
Written to teach. The usual basis for School explanations.
Research
Papers, datasets, methods — with dates, versions and disagreements.
Community contribution
Notes, explanations, corrections, reviews by participants — after review.
AI-generated material
Summaries and explanations produced by models. Never visually disguised as a primary source; always shows what it was derived from.
For School, the explanation can be simple — and the source or basis behind it can always be opened one layer deeper. Simplicity is not the same as hiding where knowledge comes from.
Knowledge contribution — across all domains
People contribute knowledge, not merely files.
Every type goes through the same lifecycle: local duplicate pre-check → submitted (not public) → quarantined → duplicate check → rights check → quality check → accepted or not. Credit only after acceptance. The scripted walkthrough lives in Medicine → Contribute; the rules are the same everywhere.
Law examples (future, not enabled in this prototype): official source mapping, a case summary written by a contributor, an original schema, original flashcards, an original case solution, a source correction, literature metadata, an original explanation. Commercial textbooks, commentaries and case-law databases remain rights-dependent, same as any other world.
School: real contributions are not enabled. Learners under 18 do not contribute publicly in this concept; teacher, parent and tutor contribution is a separate future design.