Internet Lore · Original Writing
Link Rot and the Vanishing Record
An introduction to the Free Knowledge Library, and to a problem with no censor to blame. Published on the Free Knowledge Library Substack and archived here, because a newsletter about writing that disappears should not live only on a platform its author does not control.
- Period
- 2026
- Region
- United States
- Language
- English
- Rights
- Ours to hold. Written and published by the Free Knowledge Library, so no permission is needed and no exception is claimed. The Substack copy remains the version of record for comments and corrections; this is the preservation copy.
- Source
- https://freeknowledgelibrary.substack.com/p/link-rot-and-the-vanishing-record
Full text
As the information age has progressed, librarians have noticed a quiet and unglamorous phenomenon that they call link rot, and it is erasing our shared history one broken link at a time. Most of us assume that once something has been published on the internet it will be there forever, but the opposite is closer to the truth. The web is among the most fragile archives humanity has ever built, because websites are abandoned, companies collapse, domains expire, and the instant a server is switched off everything living on it can vanish without a trace.
This is not a worry invented by anxious activists. The Pew Research Center found that nearly 40 percent of the webpages that existed in 2013 had disappeared within a single decade, and researchers at Harvard examining the footnotes of United States Supreme Court opinions found that roughly half of their links no longer led anywhere at all. Consider what that actually means: even the most carefully documented rulings of the highest court in the country are slowly dissolving into dead ends, and if the Supreme Court cannot keep its own citations alive, the rest of us have very little chance without a deliberate effort to save them.
What makes link rot so dangerous is precisely that it is invisible. There is no dramatic book burning to rally against and no censor to name and shame. A text simply slips out of existence while nobody is paying attention, and by the time someone finally goes looking for it there is nothing left to find.
IT HAS ALREADY HAPPENED, REPEATEDLY
If that sounds abstract, it becomes concrete the moment you look at how often it has happened to material of real historical importance.
One of the most consequential document collections of our lifetime came close to being lost this way. When Edward Snowden exposed the scope of government mass surveillance, the archive of leaked material was entrusted to a major news organisation, and in 2019 that organisation chose to shut down access to it and let go of the team maintaining it. The most important intelligence leak of the century came within a hair of vanishing from public reach, surviving only because a couple of journalists happened to have kept their own copies. Preservation that depends on luck is not really preservation at all.
Nor is this only a story about private institutions. In 2025, federal agencies quietly pulled thousands of public web pages and datasets offline in a matter of weeks, much of it ordinary and essential information that doctors, researchers, and journalists had depended on for years. Administrations of both parties have done versions of this, which is precisely why the survival of the public record cannot be left to whoever happens to hold power. There was no official plan to preserve any of it, and a scattered army of volunteer librarians and university archivists raced the clock to copy whatever they could. They saved a great deal. They could not possibly save it all.
Notice the pattern running through both of those stories: the material survived only because someone happened to catch it in time, and even then they usually salvaged a fraction of what was lost.
WHOSE WORDS DISAPPEAR FIRST
Now ask yourself whose writing tends to vanish soonest. It is almost never the bestselling books, because those are backed by publishers and sit safely on the shelves of thousands of libraries. The writing that disappears is the writing that only ever existed in one place, on one small website or one personal account, and that is exactly where the most influential political ideas of our time tend to live. Serious ideological argument has rarely happened in institutions built to last. It happens on forums, on personal blogs, in PDFs uploaded by a volunteer, in newsletters scanned once by an enthusiast, on hosts maintained by a single person with a renewal date they will eventually forget.
Consider two figures who could not be more relevant. Alexander Dugin is a Russian political philosopher whose nationalist and imperial worldview has profoundly shaped Vladimir Putin’s thinking, and by extension a war that has reshaped Europe. Curtis Yarvin is an American writer whose provocative attacks on democracy have steadily gained a following among people who now hold genuine influence at the highest levels of American government. Whatever you happen to think of either man, you cannot understand the forces shaping our politics without being able to read what they actually wrote: not a screenshot passed around on social media, and not a hostile summary written by someone determined to discredit them, but the real text, preserved faithfully and available to anyone who wishes to judge it honestly for themselves.
This is the entire purpose of the Free Knowledge Library. We preserve political writing from across the whole ideological spectrum before it can disappear, whether it comes from the left or the right, and whether it is celebrated or despised.
Let me be clear about what that does and does not mean, because the distinction is everything. Preserving a text is not the same as endorsing it. The Library never tells you what to think. It tells you what a document is, where it came from, and where to find the strongest arguments against it, and then it trusts you to reach your own conclusions. The honest way to defeat a bad idea is to understand it completely, and you cannot meaningfully answer an argument you were never permitted to read.
WHY ANOTHER ARCHIVE
People ask why any of this is necessary when the Internet Archive already exists, which is a fair question deserving a direct answer. Those large general-purpose archives do genuinely valuable work, but they honour takedown requests, which means controversial and contested material is usually the first thing removed. They are also a single point of failure, and at this moment they are defending themselves in lawsuits that could shut them down entirely.
The oldest principle in preservation holds that lots of copies keep stuff safe (LOCKSS). A second independent library devoted to this mission is not a redundancy, it is an insurance policy against the day the safety net finally tears.
HOW IT WORKS
At the heart of the Library is an AI archival operator. You give it a link, a document, or a block of text, and it captures the original exactly as it is, down to the last character. It then produces a clean and readable version so that anyone can engage with the material, while the untouched original remains preserved beside it. It organizes each record with accurate and purely factual information so the work can be found and cited by researchers for generations, and a human archivist reviews everything before it is ever published.
Every record carries its provenance, because an archive asking to be trusted should show its work: where the text came from, when it was captured, a cryptographic checksum so anyone can verify the bytes have not changed, and a plain statement of whether we hold the document or are pointing at someone else’s copy. A monitor checks every source weekly and tells us when one has died.
Where copyright prevents us holding a file, we still build the record. The catalogue entry, the provenance, the rights finding explaining exactly who controls the work and until when, and a citation pointing to where it can legitimately be read. A work nobody can find is lost whether or not a copy exists somewhere, and when a term expires the record is already built and waiting for the text to drop into it.
WHAT THIS NEWSLETTER IS FOR
The Library is in alpha. This is where the work gets written down: what we rescued and how we verified it, what we got wrong and what it cost to fix, the rights questions we reason through in public, and the arguments about what belongs in a collection built on refusing to decide for the reader.
Read the full proposal and support the work: freeknowledgelibrary.net/support
Or browse the shelves at freeknowledgelibrary.net, find something missing, and tell us.
Preserve the source. Protect the record. Share the context.
Preservation is not endorsement. The Free Knowledge Library preserves primary sources from across the political spectrum. Keeping a document does not mean we agree with it. What the labels mean.