Civilizational Knowledge Library

What must never be lost

Every generation inherits a vast body of hard-won knowledge: how to grow food, secure water, build shelter, generate power, understand living systems, and make the materials civilization runs on.

Today most of that knowledge lives in digital systems that require constant networks, institutions, and infrastructure.

The Civilizational Knowledge Library exists to keep the most critical parts of it accessible, durable, and decentralized.

Not because we expect collapse.

Because real resilience means being ready for change.

Reboot Priority

Every resource is scored by how essential it is to keeping civilization running and to rebuilding it if needed. A single collection can carry multiple priority tags. This hierarchy is what allows the eventual Civilizational Reboot Drive to be far smaller than the full library.

R1 — Essential

The absolute minimum: water, food, sanitation, energy, medicine, shelter, communications.

R2 — Foundational

The intellectual base: mathematics, physics, chemistry, biology, agriculture, engineering.

R3 — Industrial

The capacity to produce: metallurgy, machining, chemical engineering, electrical manufacturing.

R4 — Advanced

High-technology capabilities: semiconductors, advanced materials, biotechnology, nuclear engineering, aerospace.

R5 — Cultural / Institutional

What makes a civilization worth rebuilding: history, law, governance, economics, philosophy, arts.

1. Foundations

The intellectual base of everything else — open textbooks, university courses, chemical data, and scientific literature.

Core science & mathematics

  • OpenStax~10–30 GB Hard — Free open textbooks: biology, chemistry, physics, mathematics, astronomy, statistics, economics and more.
  • Open Textbook Library~50–150 GB Moderate — 1,800+ open textbooks across science, engineering, agriculture, business and humanities.
  • LibreTexts~100–500+ GB Hard — Extremely large collection covering chemistry, biology, physics, engineering, mathematics and other disciplines.
  • MIT OpenCourseWare~50–200+ GB Moderate — University-level courses, lecture notes, assignments and other educational material.
  • Wikibooks~5–20 GB Easy — Open textbooks and manuals.
  • Wikibooks (full ZIM)~5–20 GB Easy — Direct bulk file for offline use. ZIM file.

Scientific literature

  • arXiv~5+ TB full corpus Easy — Scientific preprints in physics, mathematics, computer science and more.
  • PubMed Central Open Access100 GB–TBs Easy — Open-access biomedical literature.
  • DOAJ — Directory of Open Access Journals~10–100+ GB Moderate — Peer-reviewed open-access journals. Full metadata dumps exist (JSON, generated weekly) but access is request-gated via email, not a self-serve public link. A metadata layer over PMC/arXiv, not a standalone text source.

Don't try to download the entire arXiv or PMC at once. Curate by category and prioritize high-value material.

2. Food & Water

For Adaptive Humans this is the largest and highest-priority section. Food, water, and the seeds that sustain them come first.

Agriculture & food

Water & sanitation

Seeds & genetics

  • Seed Savers Exchange<10 GB Hard — Seed-saving technique guides and variety data. No bulk archive, browse/scrape per page.
  • USDA GRIN-Global~10–100+ GB Moderate — Crop germplasm and variety database. Structured data downloads per taxon/collection via the portal, no single full-database zip.

3. Infrastructure

The physical and digital systems that keep civilization functioning — energy, shelter, and communications.

Energy

Collect material on: solar → batteries → electrical systems → generators → heat → biomass → microgrids → grid restoration.

Construction & shelter

Archive building physics, structural principles, passive heating/cooling, timber construction, masonry, insulation, roofing, foundations and basic surveying.

Communications & computing

Include radio fundamentals, digital communications, networking, mesh networks, GPS/GNSS and offline computing.

4. Making Things

The continuum from hand tools to factories — manufacturing, chemistry, materials, industrial processes, and craft knowledge.

Mechanical & manufacturing knowledge

One of the most important categories for rebuilding civilization. Preserve old machinist handbooks, welding manuals, agricultural machinery manuals, blacksmithing, woodworking, foundry, ceramics, glass and basic metalworking texts.

  • Internet Archive100 GB–TBs Moderate — Huge collection of historical technical manuals and books; filter carefully for public-domain/openly reusable material.
  • Project Gutenberg~5–30 GB Easy — 70,000+ public-domain ebooks.
  • Open Source Ecology<10 GB Moderate — Open-source industrial machinery and fabrication concepts.
  • OpenSCAD<1 GB Easy — Open-source parametric CAD.
  • FreeCAD<1 GB Easy — Open-source CAD platform.
  • Wikimedia Commons100 GB–TBs Moderate — Technical diagrams, illustrations and historical material.

Chemistry & materials

  • LibreTexts Chemistry~50–200 GB Hard — Open chemistry textbooks and reference material.
  • NIST Chemistry WebBook<10 GB Hard — Chemical and physical property data. No bulk archive; covered instead by the PubChem FTP archive below.
  • PubChem~10–100+ GB Easy — Chemical structures, properties and biological activities. Full Compound/Substance/BioAssay databases (SDF/XML/JSON) available via official NIH FTP bulk archive.
  • NIST Materials Data<10 GB–tens GB Moderate — Materials science data and standards.
  • MatWeb<10 GB Hard — Materials-property information.

Preserve knowledge of metals, ceramics, glass, polymers, fertilizers, industrial chemicals, corrosion, electrochemistry and material properties.

Industrial knowledge

The section to expand substantially if the objective really is civilizational reboot. Include: mining, metallurgy, cement, glass, ceramics, fertilizer, chemical engineering, petroleum processing, machine tools, welding, electrical engineering, power generation, refrigeration, pumps, engines, motors and industrial automation.

Low-tech & appropriate technology

  • Low-Tech Magazine<10 GB Hard — Practical archive of pre-industrial and appropriate technologies. No official bulk zip; use wget --mirror against the site, or pull the existing crawl snapshot at archive.org (note: snapshot is dated, re-crawl for current content).
  • Appropedia<10 GB Hard — Appropriate-technology wiki. MediaWiki site, no packaged dump found; would need a MediaWiki XML export or scrape.

Textiles & fiber crafts

5. Life & Health

The biological foundations of a functioning society — medicine, public health, and the natural systems we depend on.

Biology & medicine

For the reboot concept, emphasize fundamental health knowledge, sanitation, epidemiology, anatomy, nutrition, microbiology and public-health infrastructure — not a medical-treatment system.

Ecology & natural systems

Especially useful when combined with local species databases, soil maps and climate records.

6. Earth & Place

Maps, geology, and local knowledge of the land — with a deliberate Canadian and Alberta focus.

Geography & Earth systems

For a Canadian/Alberta focus, NRCan + provincial geological surveys are particularly valuable.

Alberta

Canada

7. Society

The institutional and cultural knowledge that makes a civilization worth rebuilding — history, law, governance, and practical procedural knowledge.

History & civilization

Civilizational resilience isn't only technical. It includes understanding how institutions, agriculture, trade, cities and technologies have evolved.

Governance, law & economics

Military field manuals

  • US Army/Navy Technical & Field Manuals~10–100+ GB Moderate — Public-domain practical and procedural knowledge. Internet Archive collection can be bulk-pulled with the ia CLI tool or torrents once you've picked the specific collection identifier(s); no single official "all manuals" package. Also available via Wikimedia Commons.

8. Language & Cultural Preservation

The linguistic and cultural layer that makes an archive usable across generations — and worth rebuilding for.

  • Rosetta Project / PanLex~10–100+ GB Moderate — Linguistic archive of thousands of human languages; PanLex publishes bulk-downloadable translation datasets. Assembling the full Rosetta corpus needs some scripting.
  • World Possible — RACHEL content packages~10–100+ GB Easy — Curated offline education bundles (Khan Academy, medical reference, teacher training), built for exactly this use case. Direct content-pack downloads.
  • Kolibri (Learning Equality)~10–100+ GB Moderate — Offline learning platform with structured curriculum "channels" (many overlapping with Core Science and Medicine sections); official export/import tool for content channels. No single "everything" download.

Size estimates

Approximate offline archive sizes for each resource, with bulk-download difficulty. These inform how the library scales from an online collection to progressively more complete offline versions.

Bulk-download tiers

  • Easy — official single-file or single-command bulk download exists today (ZIM file, S3 bucket, FTP dump, planet file). No scripting required beyond one download/sync command.
  • Moderate — an official bulk mechanism exists (API, per-item downloads, GitHub dataset repo, archive.org collection) but requires a script/loop to assemble the full set.
  • Hard — no bulk mechanism; only a live search/API into a large database, or per-document scraping against a site with no archive endpoint. Requires sustained curation work.
Resource Main content Approx. size Bulk Download
OpenStax Open textbooks ~10–30 GB Hard — no bulk source
Open Textbook Library 1,800+ open textbooks ~50–150 GB Moderate
LibreTexts Science, engineering, math, etc. ~100–500+ GB Hard — no bulk source
BCcampus OpenEd Canadian OER textbooks ~10–30 GB Moderate
OER Commons Open educational resources ~10–100+ GB Moderate
MERLOT OER/course materials ~10–50 GB Moderate
Project Gutenberg 70,000+ public-domain books ~5–30 GB Easy
Standard Ebooks Curated public-domain literature ~1–5 GB Easy
Internet Archive Historical/technical books & manuals 100 GB–TBs Moderate
FAO Knowledge Repository Agriculture, food, forestry, fisheries ~10–100+ GB Hard — no bulk source
USDA National Agricultural Library Agriculture & food ~10–100+ GB Hard — no bulk source
ATTRA Sustainable agriculture <10 GB Moderate
SARE Sustainable agriculture research <10 GB Moderate
WHO Publications Public health, WASH, health 10–100+ GB Hard — no bulk source
NCBI Bookshelf Biomedical textbooks/reference ~10–100+ GB Moderate
PubMed Central OA Open biomedical literature 100 GB–TBs Easy
arXiv Scientific preprints ~5+ TB full corpus Easy
NIST Chemistry WebBook Chemical data <10 GB Hard — no bulk source
PubChem Chemical structures & properties ~10–100+ GB Easy
NIST Materials Data Materials science <10 GB–tens GB Moderate
MIT OpenCourseWare University courses ~50–200+ GB Moderate
Open Source Ecology Open industrial technology <10 GB Moderate
FreeCAD Open CAD software <1 GB Easy
KiCad Electronics design <1 GB Easy
GNU Project Open-source software <10 GB Moderate
Python documentation Programming <1 GB Easy
GBIF Biodiversity data TB-scale globally Moderate
OpenStreetMap Global geographic data ~100+ GB Easy
Natural Earth Global base maps <1 GB Easy
NASA Earthdata Earth/environment data PB-scale globally Hard — no bulk source
Natural Resources Canada Canadian natural resources 10–100+ GB Hard — no bulk source
Alberta Geological Survey Alberta geology/minerals ~10–100+ GB Hard — no bulk source
Agriculture and Agri-Food Canada Canadian agriculture 10–100+ GB Hard — no bulk source
Open Energy Information Energy systems/data ~10–100 GB Moderate
NREL publications Renewable energy 10–100+ GB Hard — no bulk source
Open Building Institute Building/construction <10 GB Moderate
Whole Building Design Guide Building engineering <10–50 GB Hard — no bulk source
CanLII Canadian law 10–100+ GB Hard — no bulk source
Justice Laws Website Federal Canadian legislation ~10–50 GB Easy
Statistics Canada Canadian statistics 10–100+ GB Moderate
Our World in Data Global datasets ~1–20 GB Moderate
World Bank Open Data Global economic/development data ~1–20 GB Moderate

About this library

Adaptive Humans does not own the external resources linked here. This is a carefully curated collection of publicly available sources chosen for their importance to maintaining and rebuilding essential human capabilities. Preference is given to open and public-domain material that can legally be archived offline. Always check the original source for the latest information and terms of use.

Precedents & inspiration

This isn't a new idea. Several existing projects share this library's goal of preserving essential knowledge against loss of digital infrastructure:

  • Arch Mission Foundation's Lunar Library — a 30-million-page archive (Wikipedia, Project Gutenberg, Long Now's Rosetta language corpus) nano-etched onto nickel discs and landed on the Moon in 2019 and 2024, with an Earth-based copy in Switzerland. Their curation approach — "curate the curators," leaning on established collections like Wikipedia and Gutenberg rather than picking individual documents — is close to this library's own approach.
  • Long Now Foundation — Rosetta Project and Manual for Civilization — a linguistic preservation archive (usable as a "decoder key" for the rest of an archive) and a curated ~3,500-book physical collection chosen as the essential documents to rebuild society.
  • GitHub Arctic Code Vault — a snapshot of nearly all active public GitHub repositories, archived on film in the Arctic World Archive near Svalbard.
  • Open Source Ecology — Civilization Starter Kit — open blueprints for the 50 machines OSE considers necessary to build a small civilization from scratch (tractor, brick press, power cube, and more), part of their Global Village Construction Set.
  • Project NOMAD — an actively maintained open-source tool for building a household/community offline server (Wikipedia via Kiwix, Khan Academy via Kolibri, offline maps, local AI) with tiered content selection — closely mirrors this library's Essential/Foundational/Industrial/Advanced/Cultural priority tiers, and could be a candidate software layer rather than a from-scratch build.
  • Internet Archive — Offline Archive project — an open-source effort to make Internet Archive collections available without a live connection, deployable down to a Raspberry Pi.

Further reading: Lewis Dartnell's The Knowledge: How to Rebuild Civilization in the Aftermath of a Cataclysm covers similar ground to the R1–R2 tiers above, narrated rather than linked — worth a citation even though it isn't a downloadable archive.