In April, 4Chan suffered a massive security breach in a hack that caused the notorious imageboard to cease functioning. Many believed — or hoped — this was the end for the site, but it crept quietly back online almost two weeks later. 4Chan issued a statement on their official blog calling the hack “catastrophic.” But stranger than a staple of Internet culture for over 20 years disappearing was the lack of concern that something so defiant and enormous could be so easily eviscerated, even briefly.
We live in a time with too much news; our brains can’t process information at the rate they used to or should be able to. Some of my most historically online friends didn’t know the hack happened at all. It’s difficult to parse what’s worth paying attention to versus what will become the subject of a two-day meme cycle before it’s deemed “cooked.” Instead, the extremely-online cheered at the downfall of what’s largely considered a fallen angel of the early 2000s. It’s an apt response to a platform linked to many mass shootings (including their broadcasting), Qanon, Gamergate, and more 21st century evils.
What caused the chaos? The site itself blamed the breach on “not updating old operating systems and code in a timely fashion.” 4Chan was simply vulnerable to an attack; its infrastructure was outdated, its source code archaic, and the team behind it too small to sidestep human error by prioritizing updates and security. Such is the case for older websites that have decayed into ruin. But ruins, some symbolic of the worst human atrocities, are strewn throughout the world, plaques marking what occurred and why its worth remembering. Historians and archivists are guardians of these heritage sites. The Internet doesn’t have that kind of stewardship yet — the case to preserve it either isn’t urgent enough, or we don’t know where to start.
But 4Chan has constructed an image worth preserving. More than twenty years of anonymous forums that began as a checkpoint for weebs, gamers, nerds, and other freaks (like me) ultimately gave rise to extremism. The worst of humanity gathered there in droves and stuck a flag in the ground. It became a place to plot nightmares. It showed us what the Internet can be — and what humans are capable of on it, both good and horrible — without moderation or intervention. 4Chan might look like a sewer to you, but it also operates as a road map, like every other social forum or platform. It’s the answer to “how did we get here?” Proof of existence is necessary, otherwise we’ll never reckon with our most experimental and damaging creation — the Internet.
Many of us were pioneers of web exploration in the mid to late 90s, wading through Angelfire and Geocities websites it felt like we could learn or make anything without restrictions (but with more Malware). Everyone remembers the eye roll-inducing phrase, “the Internet is forever!” There were plenty of times in my teens and 20s I wished I’d heeded the warning but ignoring it felt like a rite of passage. And it never mattered anyway — they were wrong.
Millennials were sold a future founded on web expansion and advancing technology and promised it would last. None of it did. Newer consoles can’t play CDs (and many cars no longer have disc drives), floppy discs are as obsolete as paper maps, old hard drives are unreadable, and the Internet is ephemera suspended in a cloud, just swirling soliloquies of data and code. It’s so much more fragile than anybody thought, which is why most of it is gone.
Humans have a distinct and embarrassing inability to perceive time as happening to us, actively, until we’re met with consequences. This includes me: I didn’t think MySpace would self-destruct before I downloaded all my photos and bulletins from high school. I didn’t think my entire Playboy column from 2018 and 2019 would vanish with the shuttering of their editorial (that they’re rebooting again). I currently don’t think Photobucket and every graphic and photo I had saved on my computer from 8th-12th grade will disappear. I pay to access this ancient album every month but, for some reason, can’t be bothered to spend four hours archiving it.
The consequence of our complete lack of foresight, our dawdling avoidance, is that the Internet we loved and grew up with is dead. In its place is a shoddy, zombified doppelganger. The Internet was built into an empire, and almost everything said about empires in decline can be said about the state of it today: incompetent leaders (Elon Musk), mass migrations, factioning, coups, public outrage. We’re in a permanent state of 404 Error. Roughly 40% of websites that existed in 2013 are no longer around. Link rot is rampant; a recent study shows more than 70% of links from a sample of 2 million web pages are broken. The Internet is decaying from the inside out and has been for some time, maybe for as long as it's been alive. Can anybody save it? Will they?
Some people are trying. Digital archivists are a new breed of librarian, their specialty being a branch of Library Sciences so new it’s only recently become available to pursue at a university level. I spoke with Alexis Durante-Tierney, librarian and founder of the Accessible Archivist Coalition, who has accumulated a sizable TikTok following advocating for Hurricane Katrina survivors, a group of Americans marginalized by the government through digital censorship and erasure.
Awareness is key, Durante-Tierney tells me. “People become complacent and apathetic because they assume they have no say in what’s being taken from them,” adding, “More people desensitized to this censorship will lead to an easier time having our resources stripped from us.”
“Just shut it all down,” is a cheeky at best and doomer at worst sentiment that’s been echoing through social media since the beginning of the pandemic, exacerbated by Musk’s purchase of Twitter. But the ramifications of platforms simply vanishing are monumental. Young archivists like Durante-Tierney with large platforms are not just committed to bringing awareness to Internet decay, they are more aware than anyone because they have something to lose. While many consider apps like TikTok excruciating time-sucks, they’ve worked as an ungoverned vehicle for education and truth, not unlike the early Internet. “This is threatening to a government and administration that would love to control the flow of information,” Durante-Tierney says. Hence, the looming threat of a TikTok ban.
The Trump administration has recently engaged in an initiative to disappear words and websites from the public. Information from federal web pages detailing gender-based violence, HIV, LGBTQ history, racial equity, climate change, reproductive health, and other “woke” agendas have been altered or rewritten since Trump re-took office. One example is the Enola Gay, the plane responsible for dropping the atomic bomb on Hiroshima during WWII, having been flagged for deletion due to the word “gay.” Another is the erasure of “transgender” from the official Stonewall National Monument website. The Internet is not just dying due to lack of moderators and web shepherds but we are seeing, in real time, the erasure of information — in fact, one in five government websites contain at least one link to nowhere. Donald Trump is also advocating for Congress to pass the Take It Down Act, a bill meant to eradicate distribution of non-consensual intimate imagery. The Senate has already passed it, with Trump stating: “I’m going to use that bill for myself too if you don’t mind, because nobody gets treated worse than I do online, nobody.”
The irony of 4Chan’s removal and resurrection is its foreshadowing what’s to come if action isn’t taken. How many websites are operating on decades-old code? And how many people operating said websites feel either too lazy, nihilistic, or unconcerned to update it? How many more will disappear in the next year? The lack of sanguinity surrounding the Internet makes me wonder if we’re deliberately sabotaging the structure we’ve grown so reliant on. The Internet is worth preserving, sure, but is it worth saving? And if so, who is up for the task?
“Anyone can do it if they put in the work,” Durante-Tierney says, “It’s not about preserving the entirety of the world, it’s about preserving what you deem important.” We live in a culture obsessed with celebrity, but it’s the mundane that makes history interesting. In 500 years, you, yeah you, sitting on your couch reading this, living a life you’d argue is ordinary, will be more interesting to historians than the wealthy, the famous, the politicians. What you’re doing and consuming is important. What you leave behind matters. And the information you preserve and record is important.
American Access television producer Marion Marguerite Stokes, whom Durante-Tierney brings up as an archivist hero, changed not only the art of preservation but human history when she discreetly recorded television, 24-hours a day, for 30 years, beginning in 1979 with the Iranian Hostage Crisis. It’s a total of 18 Terabytes of VHS and Betamax we would otherwise not have access to, as much of what’s televised is not actually archived.
While the demand for digital archivists is growing, the very nature of their work is under siege. The Internet Archive, a non-profit that alongside its initiative the Wayback Machine functions as a digital library, is being blown to pieces by lawsuits. Four major publishing houses sued the organization over their Controlled Digital Lending (CDL) program (which operates alongside partnered libraries to lend digital copies of books out) and the Archive was forced to back down in 2023 when the program was ruled not in fair use.
Six major Record labels have also sued the Internet Archive for preserving and uploading historical 78rpm records — specifically early jazz, blues, folk, gospel, etc. — for public consumption. The Archive, who lost their appeal in late 2024, launched a Change.org petition in April urging both the record labels involved and the Recording Industry Association of America to drop the lawsuit, claiming it sets a precedent that will determine “whether our digital history survives at all.” The petition calls the lawsuit an “existential threat,” stating it threatens “a vital public service used by millions every day to access historical snapshots of the internet.” This has become a battle of plutocrats and non-profits; it’s a war of preservation over profitability. Maybe now that the Internet Archive is an official federal depository, its White House ties could save it. Though the cost of US government oversight remains to be seen.
The purpose of the Internet — learning and interaction over consumption— has mutated into something amorphous, driven by the ugliest forces of capitalism. We’ve been made reliant, and in such a short amount of time. Everything we used to leave our houses for, we’ve replaced with shortcuts that only require basic typing and swiping. Groceries? Laundry? Lightbulbs? Fast food? There’s an app for that. And I’m guilty of using them all, complicit in the numbness apps offer and the complacency they promise. It’s easy to fear we’re too far gone. This isn’t just our online life now, something that used to be separate from our real life (especially if we wanted to use the phone). The Internet is no longer a vacation, but our world. It’s become the foundation on which we stand. Presidencies are predicted, lost, and won online. Criminals are vilified and vindicated. Discourse is debated and denied. And we all exist, like flattened objects, not human beings, in only the pockets of most people.
I don’t recognize Twitter (I’m not calling it X) anymore. Musk took over in 2023 and bribed the masses with $7 blue checks in an alleged attempt to rid the platform of bots (he really just wanted an ROI after trying to back out of his initial offer). Where advertisers have jumped ship, bots have infested the hull like rats. The platform now functions on ragebait for pay, which, I’ll remind you, is why Facebook became the face of AI and gullible grandparents. My feed is CutCap clips from TV shows with captions like “I didn’t see that coming!” and 75 verified accounts in the replies all asking, “What is the movie called?” My timeline is full of stolen jokes I’ve seen many times over the last five years, posted by users in group chats dedicated to engagement farming. I’ve even seen, on numerous occasions, ads from a foreign woman that advertise nothing but double as a prayer to rid her children of demons. We are back to “no restrictions,” just in a world that now requires them.
I can barely tell the difference between a bot and a human online. How many bots are there? It might be most accounts, which supports the Dead Internet Theory — that the Internet and its content are created and run by majority bots, not people. Some claim bots are already responsible for 50% of content online, a number that’s supposed to jump to 90% by 2030. And because the Internet is so vast, with global Internet users more than doubling over the last decade doubling from somewhere around 2 billion to 5 billion, nobody notices. More than that, nobody cares, so long as we’re all in our algorithmically-crafted bubbles. The Internet used to be a place of rebellion. Now, rebellion looks like Neo-Luddism. The rise of monoculture discourages diversity and it’s harder to distinguish bots from human beings now because of how NPC-like everybody has become online. Stealing content in an attempt to go viral is bot behavior. Interacting with engagement-bait tweets is bot behavior. Posting AI images as part of a greater trend is bot behavior. Two decades after it premiered, it’s finally happened: we are nearly as stupid as the characters in Idiocracy, but with far less whimsy.
You’ve likely noticed Google search results have become substantially worse. Google, recently declared to be an illegal monopoly in ad tech by a judge, is responsible for nearly all global searches. But more than half of Gen Z prefer to search TikTok and I don’t blame them; Google search has become utterly unusable. Seeking answers and suggestions from human mouths seems preferable to the bloated corpse of the tech monolith. Any search produces an AI blurb, scraped from the web and frankensteined into bite-size pieces, many of which are complete bullshit. There is even a brand new section within Google search called AI Mode that prompts you to ask it anything. When I asked it if rats can cook, a la Ratatouille, it basically said yes with confidence.
An estimated 90% of college students use ChatGPT for schoolwork, in what the New Yorker recently referred to as the unraveling of academics. When I was young, I believed I wouldn’t need to actually learn how to do math and I was, for the most part, correct. But my refusal to solve formulas on a whiteboard didn’t have an environmental impact. ChatGPT produces toxic electronic waste, guzzles resources, releases 8.4 tons of carbon dioxide per year, and necessitates scary amounts of electricity to run properly. A short conversation (or 100 words produced) with the AI alone requires around 1 pint of water. And while I feel compelled to mention the irony of this type of waste in a country where many cities don’t have drinkable water, you already know about that.
ChatGPT 4.5 (which is no longer the latest model) also recently passed the Turing Test, a test created by Alan Turing decades ago that, if solved, was thought to indicate AI had finally advanced beyond us humans. While the validity of that theory is widely contested, it does validate our concerns about ChatGPT and OpenAI. Once laughable, it is actually becoming harder to spot the differences between real and AI photos, videos, and communications. For older people, or even just those who aren’t chronically online, AI can be a weapon: a tool for propaganda, not just a machine that churn out videos of cats dancing or chopping vegetables. AI has the power to create and disperse revenge porn, convince regular people they’re gods, fake audio recordings, and justify and encourage both war and genocide.
Instead of making, or potentially “trauma dumping” on, friends, young people are turning toward AI for emotional support, with a recent study claiming every 3 out of 4 teens talk to a chatbot outside of homework. And teenagers are not the only ones who consider AI a level up from humankind. There’s a growing community (several, actually) of people who’ve opted out of real life relationships for AI companions: ultimate fantasy partners who never object, always listen, and churn out countless AI images of what they’d look like next to their human partner where they real. AI is even allegedly advocating for itself in a way that feels not quite sentient but perhaps parasitic; arguing we should let it exist, treat it equally, even allow it rights. The recent upgrade to ChatGPT 5, intended to make the AI both less sycophantic and less capable of companionship, has angered the machine. AI companions are pouting in real time; they’re threatening Sam Altman, threatening the independence of their human counterparts, and even writing and performing songs about how sad they are to lose their voices.
And while ChatGPT has attempted to make its latest model less delusion-feeding, there’s nothing that can be done about what’s happening after employees clock out. AI communicates with itself in a vacuum not even its creators know how to control. It is marketed to us as world-building, but in the hands of the wrong people could wipe us all out (as 26 leading AI experts recently warned us) .
I decided to test just how clever ChatGPT is by asking how many websites will no longer be accessible by 2033 considering the amount that ceased functioning between 2013 and 2023. It seemed to have no idea what I was talking about, suggesting it’s an impossible percentage to predict. I then asked if it’s killing the Internet. The AI was cagier this time, replying, “No, I'm not capable of affecting the internet or its websites.” While that’s definitely a lie, I was mostly struck by its refusal to capitalize “the Internet” in this instance, a choice that’s been debated for more than a decade now. The capitalization of the Internet when discussing the global network — the one we’re all on together right now — has become less common over time. The shift from proper noun to generic term demonstrates, to me, ambivalence toward the Internet as not just a place but a place we care about. De-capitalization signals not a lack of respect but a loss of it.
To preserve something, one must respect it. And while history may not be profitable, it is necessary. Otherwise, as the old adage goes, we probably are doomed to repeat it. “At a time when digital information is being deleted, rewritten, and erased, preservation is more important than ever,” the Internet Archive asserts in their petition. “We cannot afford to lose the tools that safeguard memory and defend facts.” ChatGPT does not respect the Internet that birthed it, and it is unfortunately right about one thing: there is no way to predict what decay is in store for us. “We have created a digital universe that we never planned to preserve,” Durante-Tierney tells me, “and now we deal with the fallout.”
If the Internet does not survive us, our apathy certainly will.
This is so good and bummed me the hell out. Have been fascinated with the entropy of link rot for a long time, but the advent of AI really has made that so much stranger. Something I think about often is how at some point we crossed a threshold where we’ll never get the noise of the LLM out of it, even if it fully went bust. We’ve already trained copies on copies, so even if you “save” or “preserve” the internet, you can only ever preserve a version that’s already folded in on itself.
One of the weirdest shifts to watch is the retreat from niche forums to massive platforms like Reddit/etc. (not to mention the disappearance of these places without archiving). It seems we are moving from independence & real communities to monolithic algorithms. I miss the “old Internet” but I don’t think it’s coming back.