Yo I know I’m real. And you, dear reader, know that it is real. But have you ever suspected that everyone else on the Internet is acting strange? That the spaces you used to frequent feel a little… dead? You’re not alone. The “dead internet theory” first hit the web almost three years ago, propelled into the mainstream by an essay in the Atlantic by Kaitlyn Tiffany:
The theory of the dead Internet suggests that the Internet has been almost completely dominated by artificial intelligence. Like many other online conspiracy theories, this one’s audience is growing due to discussion led by a mix of true believers, sarcastic trolls, and curious chatter-lovers… But unlike many other online conspiracy theories, This one has a bit of truth in it. Person or robot: does it really matter?
At the time I wrote this article, the deadest part of the Internet was Twitter in its moribund pre-Musk years. The site’s aggressive curation offered the same “relatable content” to hundreds of thousands of users, who made adjustments to posts like “I hate texting, come hug me” and then reposted them. The distinction between person and bot was also being blurred by a recommendation algorithm that led people to act like bots.
Beyond that central idea, the 2021 version of the conspiracy theory took some strange turns. One of its defenders, Tiffany wrote, “suggests that the Internet died in 2016 or early 2017, and that it is now ’empty and devoid of people,’ as well as ‘totally sterile.’… As proof, IlluminatiPirate offers: ‘I have seen the same threads, the same photos and the same answers posted over and over again over the years.’”
The theory was not wrong; It was simply too early. Talking about a dead Internet in the summer before The launch of ChatGPT is like The Guardian colleague who confidently declared, in the summer of 2016: “There has been an avalanche of political news since the 2014 local elections, so it is strange to think that there is only Brexit and the elections in the US and then everything will be silent for the next few years.”
In 2021, the internet seemed dead because aggressive algorithmic curation was driving people to act like robots. In 2024 the opposite has happened: robots publish like people. These are just some examples:
On Twitter itselfAfter Musk rescued the site from the frying pan and threw it into a volcano, an ill-conceived monetization plan made it profitable to buy a blue checkmark, attach it to a big language model, and make it run like crazy responding to viral content . . The social network now pays verified users a proportion of the advertising revenue received from their own comment threads, turning the most viral posts on the site into a low-stakes bot battle royale.
Death permeates Google. The top of your search results is a valuable position, so valuable that the companies competing to be there don’t have extra money to write their articles. No problem: ChatGPT can produce something in a second. Of course, that’s only valuable if the resulting visitors are humans you can make money off of. Bad news, because…
…on the web, bots represent around half of all Internet traffic, according to research by cybersecurity company Imperva. Nearly a third of all traffic is what the company calls “bad bots,” which run everything from ad fraud to brute-force hacking attacks. But even “good robots” are struggling to earn that categorization: Google’s “crawler” was a welcome sight when it updated your search entry, but not so much when it simply trained an AI to repeat what you typed without sending any users. . .
And then there’s the Jesus Crab. An unholy union of Facebook content farms, AI-generated images, and automated tests to determine what goes most viral. led to weeks of viral content with combinations of Jesus, crustaceans and flight attendants. In one such image, Jesus is depicted eating seafood and wearing a jacket made of prawns. More confusing was the image of a species of crab-centaur savior walking along a beach arm in arm with what appears to be the entire crew of a long-haul flight. It was, if nothing else, curiously strange: a step up from the 122-year-old’s previous viral friend posing in front of her homemade birthday cake.
I want to offer a glimmer of hope, a little advice on how to put some vitality back into the Internet, but I can’t. It really feels like the public grid is in the final stages of a zombie apocalypse. The good news is that safe havens exist. “Private socials” like WhatsApp and Discord servers can secretly hide from the attack, while smaller communities like Bluesky and Mastodon are safe in the dark, for now.
In the medium term, I expect to see large platforms take advantage of the wasteland their services have become and use a combination of account verification and AI detection to try to restore some humanity to their offerings. However, it is an open question whether it will be too late by then.
Musk still needs his Twitter babysitter
There is still at least one human being on the Internet: Elon Musk, who is so addicted to posting that he spent $44 billion to be called an idiot on a platform he owns. So his latest legal defeat will have hit him where it hurts most, after the U.S. Supreme Court refused to grant his request to be freed from a court-appointed babysitter. From our history:
after newsletter promotion
The Supreme Court on Monday rejected an appeal by Elon Musk over a settlement with securities regulators that requires him to obtain prior approval for some tweets related to Tesla, the electric vehicle company he runs.
The justices did not comment on upholding lower court rulings against Musk, who complained that the requirement amounts to a “prior restraint” of his speech in violation of the First Amendment. The ruling comes a day after he made an unannounced visit to China with the goal of sealing a deal to roll out Tesla’s driver-assist features there.
For those who don’t have an encyclopedic memory of all things Elon: In 2018, Musk tweeted that he had “funding secured” to take Tesla private. The company never went private, and in subsequent lawsuits it turned out that, at best, it had had some conversations about it. To end the legislation, Musk agreed to resign as Tesla’s chairman, pay $20 million and have an in-house lawyer pre-approve all of his social media posts about the electric car maker.
He has regretted it ever since, fighting to have that part of the agreement (which he signed voluntarily, to avoid a damaging court case) voided. “The pre-approval provision at issue continues to cast an unconstitutional chill on Mr. Musk’s speech whenever he considers making public communications,” his attorneys argued.
Well, the United States Supreme Court doesn’t care; did not take his case, implicitly determining that no real constitutional issue is at stake.
The strange thing is that the in-house counsel already seems to be taking a very hands-off approach with Musk’s posts. On Friday, he responded to an accusation from former Facebook employee Dustin Moskovitz that Tesla was “the next Enron” by posting a photo of a dog covering another dog’s face with its testicles. (Click at your own risk.) If Musk tweets with an “unconstitutional shudder,” I would hate to know what he would send if he felt truly free.
If you want to read the full version of the newsletter, sign up to receive TechScape in your inbox every Tuesday.