Internet Archive
Based on Wikipedia: Internet Archive
In the basement of a former Christian Science church in San Francisco sits a peculiar monument to digital memory. Rows of ceramic statues line the walls—not saints or prophets, but employees of the Internet Archive, each immortalized in clay when they reach their tenth year with the organization. It's a playful touch for an institution dedicated to something profoundly serious: ensuring that nothing published on the internet ever truly disappears.
The Internet Archive has saved over one trillion web pages. Let that number settle for a moment. A trillion is a thousand billions, or a million millions. If you spent one second viewing each page the Archive has preserved, it would take you roughly thirty-one thousand years to get through them all.
The Man Who Wanted to Remember Everything
Brewster Kahle started the Internet Archive in May 1996, during those heady days when the World Wide Web still felt like a frontier settlement. Yahoo was barely two years old. Amazon sold only books. Google didn't exist yet. Most people who went online did so through screeching modems that made connecting to the internet sound like two robots having an argument.
Kahle had already made his fortune. He'd built a company called WAIS Inc., which created one of the first internet search engines, and sold it to America Online for fifteen million dollars. But money wasn't what drove him. Kahle was haunted by a simple observation: the web was erasing itself.
Unlike a book that sits on a library shelf for centuries, a web page could vanish the moment someone forgot to renew their domain name or decided to redesign their site. All that content, all those ideas, all that documentation of human thought and culture—gone. The average lifespan of a web page, researchers would later determine, was about one hundred days. The internet was a palimpsest, constantly writing over itself.
Kahle's solution was audacious. He would build machines that automatically crawled the web, saving copies of everything they encountered. He would preserve the digital equivalent of the Library of Alexandria, except this time, the library would be fireproof. Or at least, backed up in multiple locations.
The Wayback Machine
The Archive's most famous creation takes its name from a segment on the cartoon "The Rocky and Bullwinkle Show." In the original gag, a character named Mr. Peabody—a bespectacled dog genius—operated a time machine called the WABAC Machine, which he used to witness historical events firsthand with his human companion Sherman.
The digital Wayback Machine lets anyone perform a similar trick with websites. Type in any web address, and you can see what that page looked like at various points in its history. Want to see what Amazon's homepage looked like in 1999? It's there, complete with its primitive layout and its banner advertising "Earth's Biggest Selection." Curious about how a news organization covered a story before quietly updating their article? The Wayback Machine probably captured the original version.
This capability has proven invaluable for journalists, researchers, lawyers, and ordinary citizens. When a politician claims they never said something, the Wayback Machine often has the receipts. When a company tries to bury embarrassing content, the Archive ensures it survives. The Wayback Machine has been cited in legal proceedings, academic papers, and investigative reports around the world.
By October 2025, the Wayback Machine had archived one trillion web pages—a figure so large it's hard to comprehend. That's more than one hundred thousand terabytes of data, enough to fill roughly twenty-five million average smartphones.
Beyond Websites
Web archiving was just the beginning. In late 1999, Kahle began expanding the Archive's scope. His first acquisition was the Prelinger Archives, a collection of industrial and educational films assembled by archivist Rick Prelinger. These were the kinds of films that used to play in school auditoriums: safety instructionals, corporate propaganda pieces, government newsreels. Individually, many were forgettable. Collectively, they formed a remarkable portrait of twentieth-century American culture and its anxieties.
Today, the Internet Archive contains multitudes. It holds over forty-two million books and texts, from medieval manuscripts to last week's academic papers. It stores fourteen million audio files, including hundreds of thousands of 78 rpm records donated by the Boston Public Library—those fragile shellac discs that once brought jazz and blues into American living rooms before vinyl took over. The Archive houses thirteen million videos, three million television news broadcasts, and 1.2 million software programs, including vintage video games that can be played directly in your web browser.
There's something almost salvific about the Archive's approach to obsolete media. Old software and video games, for instance, typically become unplayable once the hardware they were designed for becomes obsolete. But the Internet Archive runs emulators—programs that simulate old computers—allowing anyone to play Oregon Trail or explore early educational software exactly as children experienced them decades ago. It's a form of cultural archaeology made freely available to everyone.
The Great 78 Project
One of the Archive's more ambitious undertakings is the preservation of 78 rpm records. These discs, which dominated recorded music from roughly 1900 to the early 1950s, are a nightmare for archivists. They're made of shellac, a material derived from the secretions of lac beetles, mixed with various fillers. They're brittle, heavy, and break easily. They can only be played a few hundred times before the needle wears grooves into grooves.
But 78s captured an irreplaceable period in musical history. Blues, jazz, country, early rock and roll, ethnic recordings from communities around the world—all of it was first laid down on these fragile discs. Many recordings exist on only a handful of surviving copies, some damaged, some deteriorating in collectors' basements.
The Internet Archive has digitized hundreds of thousands of these records, making them freely available online. The project provoked controversy from music industry giants. In August 2023, Universal Music Group, Sony Music, and Concord sued the Archive for six hundred twenty-one million dollars, arguing that digitizing and sharing these recordings infringed their copyrights. The lawsuit was eventually settled in September 2025, though the terms weren't made public.
Books and the Controlled Digital Lending Battle
The Internet Archive doesn't just preserve digital content. It creates digital content by scanning physical books. The organization operates scanning centers around the world, each equipped with specialized equipment that can photograph book pages without damaging spines. Workers manually turn pages while overhead cameras capture images. The machines can process a three-hundred-page book in about twenty minutes.
Much of this scanning material arrives through donations. In 2018, Trent University in Ontario gave the Archive a quarter million books. When Marygrove College in Detroit closed its doors in 2020, its entire library collection went to the Archive. These books are digitized, and digital copies are returned to the original institutions while the Archive keeps its own copies for lending.
Here's where things get complicated. The Archive operates under a theory called controlled digital lending, often abbreviated as CDL. The idea is simple: if a library owns a physical copy of a book, it should be able to lend a digital version of that same book to one person at a time, just as it would lend the physical copy. When the digital loan expires, the next person in line gets access. The physical book sits on a shelf, unused, while the digital copy circulates.
Traditional publishers found this theory unacceptable. In their view, lending a digital copy requires a license, regardless of whether you own a physical copy. Every unlicensed digital loan, they argued, represents a lost sale of an ebook or audiobook.
In June 2020, four of the world's largest publishers—Hachette Book Group, Penguin Random House, HarperCollins, and John Wiley and Sons—sued the Internet Archive. The lawsuit intensified during the early months of the COVID-19 pandemic, when the Archive temporarily removed lending restrictions on its digital library, allowing unlimited simultaneous access to its collection of scanned books. The Archive called this the National Emergency Library, intended to help teachers, students, and readers who had lost access to physical libraries. Publishers called it mass piracy.
The courts sided with the publishers. In March 2023, a federal judge ruled against the Archive, and a negotiated settlement in August of that year prohibited the organization from lending digital copies of books for which electronic versions were available for sale. It was a significant blow to the Archive's mission of universal access.
Under Attack
In October 2024, the Internet Archive faced a different kind of crisis. Hackers launched a coordinated assault on the organization, combining distributed denial-of-service attacks—where thousands of computers simultaneously flood a server with requests until it crashes—with more sophisticated intrusions.
A group calling itself SN_BLACKMETA claimed responsibility for the denial-of-service attacks. But the damage went deeper. Attackers managed to steal a database file containing information on approximately thirty-one million user accounts, including email addresses and password hashes. Password hashes are scrambled versions of passwords that are difficult but not impossible to reverse. The Archive used a hashing algorithm called Bcrypt, which is considered reasonably secure, but users were still advised to change their passwords immediately.
The hackers defaced the Archive's website with a taunting message that played on the organization's reputation for operating on limited resources: "Have you ever felt like the Internet Archive runs on sticks and is constantly on the verge of suffering a catastrophic security breach? It just happened."
Brewster Kahle addressed the crisis publicly, assuring users that the archived data itself—the trillion web pages, the millions of books and recordings—remained safe. But the attack forced the Archive offline for days, with services restored only gradually. By late October, the site was back up, though the incident highlighted the vulnerability of an organization that holds so much cultural heritage in trust for the public.
Architecture of Memory
Preserving a trillion web pages requires serious infrastructure. The Internet Archive operates six data centers, mostly in California, with smaller facilities in other U.S. states, Canada, and Europe. Each data center maintains controlled access and fire protection systems. The organization learned the hard way about fire risk: in November 2013, flames destroyed a side building at the San Francisco headquarters, consuming scanning equipment worth hundreds of thousands of dollars and some irreplaceable materials that hadn't yet been digitized.
The Archive practices redundancy obsessively. Important data is stored in multiple locations simultaneously. Copies of the archive exist in Egypt's Bibliotheca Alexandrina—the modern incarnation of the ancient library at Alexandria—and in facilities in Amsterdam. The Archive has also experimented with decentralized storage, uploading data to the Filecoin network, a blockchain-based storage system. By October 2023, one petabyte of Archive data had been distributed across the Filecoin network. A petabyte is a million gigabytes, roughly equivalent to five hundred billion pages of standard printed text.
In 2016, shortly after Donald Trump won the U.S. presidential election, Kahle announced plans to build a complete backup of the Archive in Canada. The announcement generated headlines suggesting that Kahle feared the new administration might somehow threaten the Archive's operations. Whether that concern was warranted or not, the decision reflected a fundamental principle of preservation: never rely on a single copy in a single jurisdiction.
A Library, Officially
The Internet Archive has always called itself a library, but for most of its existence, that was more aspiration than legal status. That changed in 2007, when California officially designated the Archive as a library under state law. More recently, in July 2025, the U.S. Senate designated the Internet Archive as a Federal Depository Library, allowing it to store public access government records alongside the Library of Congress and other official repositories.
This recognition matters. Libraries enjoy certain legal protections and cultural authority that ordinary websites do not. When the Internet Archive argues that it should be allowed to lend digitized books under fair use principles, its status as a genuine library strengthens that argument—even if, so far, the courts haven't agreed.
The Archive operates as a nonprofit organization under section 501(c)(3) of the U.S. tax code. Its annual budget, as of 2019, was thirty-seven million dollars, funded through a mix of grants, donations, partnerships, and revenue from services like Archive-It, a subscription product that allows universities, governments, and other institutions to build their own web archives using the Archive's infrastructure.
Archive-It and Institutional Memory
Not every organization can build its own web crawler. Archive-It solves this problem by letting subscribers specify what web content they want to preserve, then handling the technical work of capturing and storing it.
Universities use Archive-It to preserve their institutional websites, capturing the way their campuses presented themselves over decades. State archives use it to preserve government websites before administrations change and new officials redesign everything. Museums use it to document exhibitions and cultural events. Law libraries use it to preserve legal resources and court filings that might otherwise disappear.
By March 2014, Archive-It had more than 275 partner institutions across 46 U.S. states and 16 countries. These partners had collectively preserved more than 7.4 billion URLs in over 2,400 public collections. The content becomes searchable within seven days of capture, and copies of the archived material are stored both at Internet Archive data centers and at partner institutions, creating multiple layers of redundancy.
Google's Endorsement
For years, Google operated its own web cache, storing copies of web pages it indexed so that users could access them even if the original sites went down. In 2024, Google quietly retired this feature. But rather than leaving users without a fallback option, the company partnered with the Internet Archive.
Now, when you search for something on Google and click on "more about this page," you'll find a link to the Wayback Machine's archives of that URL. It's an implicit endorsement from the world's largest technology company: when it comes to preserving the web's history, the nonprofit Archive does it better than the trillion-dollar corporation can justify doing itself.
Artists in Residence
Since 2018, the Internet Archive has operated a visual arts residency program, inviting artists to spend a year exploring the Archive's forty-eight petabytes of digitized materials. The program, organized by Amir Saber Esfahani and Andrew McClintock, connects creators with cultural artifacts they might never have discovered otherwise.
Past residents have included Jenny Odell, whose work explores attention and technology; Taravat Talepasand, whose paintings engage with cultural identity and censorship; and Whitney Lynn, whose art addresses systems of power and control. Each residency culminates in an exhibition, creating new art from old archives and demonstrating that preservation isn't just about maintaining records—it's about enabling future creativity.
The Church of the Byte
The Archive's headquarters at 300 Funston Avenue occupies space that once hosted very different concerns. The building was originally a Christian Science church, and something of its ecclesiastical character remains. The main hall, where parishioners once gathered, now houses rows of servers humming with the accumulated memory of the web.
The ceramic employee statues along the walls give the space the feel of a quirky temple dedicated to a new kind of faith: the belief that information wants to be free, that knowledge should be accessible to everyone, that the collective memory of humanity is too precious to leave to the mercy of server crashes and corporate bankruptcies.
Brewster Kahle has described the Archive's mission in terms that would have been familiar to the ancient librarians of Alexandria: universal access to all knowledge. It's an impossible goal, of course. Copyright law prevents the Archive from sharing most contemporary creative works freely. Storage costs money, and money is always tight. Hackers probe for vulnerabilities. Publishers sue. Governments change.
But a trillion saved web pages suggest that the impossible is at least partially achievable. Every researcher who cites an archived source, every journalist who catches a politician in a contradiction, every citizen who recovers a piece of personal history from the digital past—each benefits from a stubborn conviction that remembering matters more than forgetting.
The Internet Archive doesn't just store information. It makes an argument: that the ephemeral can be made permanent, that the digital deserves the same preservation we grant to printed books and stone tablets, that a civilization that forgets its past is impoverished in ways it may not recognize until too late. It's an argument written in petabytes, backed up across continents, and tended by people who receive ceramic statues of themselves after a decade of faithful service.
The church may have changed its creed, but it remains a place of devotion.
``` The essay transforms the encyclopedic Wikipedia content into a narrative that opens with a compelling hook (the ceramic employee statues in a former church), explains technical concepts like password hashing and controlled digital lending in plain language, and maintains a rhythm suited for text-to-speech reading with varied sentence and paragraph lengths.