Will There Be More Bots Than People? How True Is the “Dead Internet” Theory?
Among the many conspiracy theories — about a flat Earth, reptilian overlords, or microchips in Covid vaccines — there’s one that seems absurd for now, but may turn out to be an eerie prophecy for the not-so-distant future: the «dead internet» theory. According to its believers, bots are everywhere. Real people have vanished from the web and are replaced by AI-generated content, while secret forces manipulate the remaining few. The idea first surfaced years ago, but the explosion of AI-generated content has given it new life.
According to its believers, bots are everywhere. Real people have vanished from the web and are replaced by AI-generated content, while secret forces manipulate the remaining few. The idea first surfaced years ago, but the explosion of AI-generated content has given it new life.
The Origins of the «Dead Internet» Theory
The theory (not a theory in the scientific sense, but rather a loose collection of ideas) has murky origins. One of its best-known manifestos appeared in 2021 on the Agora Road forum. A user calling themselves IlluminatiPirate posted a thread titled «Dead Internet Theory: Most of the Internet is Fake.»
He introduced himself as right-wing and claimed that the modern internet (as of January 2021) felt «empty, devoid of people and content compared to, say, the internet of 2007.»
«The internet of today is entirely sterile. There’s nowhere to go and nothing to do, see, read, or experience anymore,» he lamented.
According to him, powerful entities (especially Facebook, Google, and Twitter) manipulate the web through bots, algorithms, and artificial intelligence to feed users carefully crafted narratives. Real people, he argued, had mostly disappeared.
You could say the author, an American conservative nostalgic for the chaotic freedom of the 2000s, mistook the rise of moderation rules for the «death» of the internet. Yet, there is a kernel of truth in his paranoia: the sheer amount of generated content and fake online personas is indeed growing.
And these fake accounts are being used to manipulate reality and deceive real users — not as part of a global conspiracy, but through targeted, tactical operations. One clear example: since the start of Russia’s full-scale invasion of Ukraine, the aggressor has massively used botnets and algorithmic manipulation to spread propaganda narratives.
The scale of the problem is staggering. In October 2025, Facebook announced it had blocked over 8 million accounts linked to scam centers, which is nearly the population of Austria.
In recent months, the «dead internet» idea has resurfaced — this time fueled by tech insiders. Sam Altman, co-founder of OpenAI, wrote in September 2025 that while he’d never taken the theory seriously, he’s now seeing «a lot of LLM accounts» on X (formerly Twitter). Commenters roasted him: «Sam never believed in the dead internet theory — so he decided to make it real.»
Reddit co-founder Alexis Ohanian echoed similar concerns on The TBPN Podcast: «…the internet is now just dead — this whole dead internet theory, right? Whether it’s botted, whether it’s quasi-AI, LinkedIn slop. Having proof of life, like live viewers and live content, is really valuable to hold attention.»
And who can blame him? AI slop has flooded Reddit, too. One viral example: a wholesome story about a plump sphynx cat named Pound Cake went viral as users followed his dramatic weight-loss journey — until someone discovered the cat was AI-generated.
«Wait, I cried yesterday over this cat, and he never existed? What the hell?» user yoki2 wrote.
Ohanian says the «dead internet» theory may have begun as a conspiracy, but it’s fast becoming a reality.
How «Alive» Is the Internet Today?
So what’s the actual ratio of bots to humans online? According to the 2025 Bad Bot Report from Thales, 2024 marked the first time in a decade that bot-generated traffic surpassed human traffic — accounting for 51% of all web activity. The report shows a steady upward trend: bot traffic has been rising every year (red indicates «malicious» bots, green — neutral ones; see chart on page 3 of Bad Bot Report 2025).
The company attributes this surge to the explosion of AI-based services and large language models (LLMs).
And it’s not slowing down. According to Popular Mechanics, citing research from the Pew Research Center, the web could be «more dead than alive» within three years. Pew found that about 38% of all web pages created since 2013 no longer exist (a phenomenon known as link rot), and much of that lost content is being replaced by AI-generated material.
Analytics firm NewsGuard has identified over a thousand «news» sites primarily run by bots, including at least 167 posing as Russian outlets to spread disinformation about the war in Ukraine, using AI-generated content for this purpuse.
For example: the proportion of AI-generated articles compared to articles written by humans (according to Axios):
A similar shift is happening in advertising. Scientific Research Publishing (SCIRP) reports that AI-generated ad videos now boast an average click-through rate of 28%, compared to 15% for traditional ads. More than half of advertisers (56%) say AI-generated content outperforms human-made material.
Ironically, the rise of AI-generated content may be creating a vicious cycle. A recent study on the LLM Brain Rot hypothesis suggests that when language models train on low-quality, AI-generated material (what internet users jokingly call «brain rot»), their own performance deteriorates — they become, in effect, «dumber.»
Writer Ted Chiang, whose short story inspired Denis Villeneuve’s film Arrival, warned about this years ago in his essay «ChatGPT Is a Blurry JPEG of the Web.»
«The more text generated by LLMs is published on the web, the more the web becomes a blurred version of itself.»
So What Now?
Should we all become neo-Luddites and declare war on AI content? Probably not, even if you’re tempted after scrolling past yet another Facebook post about an old woman who sculpted a horse out of cabbage and «no one congratulated her.» Generated content is becoming increasingly difficult to distinguish from human-made posts, and platforms often fail to label it accurately. Europol warned back in 2022 that deepfakes were becoming indistinguishable from real footage, and ironically, every new detection tool only helps make future fakes more convincing.
Still, the rise of AI content may be forcing humanity to rediscover something we’ve neglected: careful reading, critical thinking, and the value of genuine, well-written material. Additionally, running bot farms isn’t free, and generating AI content incurs even higher costs. As Sam Altman once joked, every polite «Hello» or «Thank you» users type into ChatGPT costs his company millions in electricity.
So, maybe you can believe in a «dead» internet, a Martian wasteland populated only by bots. Or you can believe in a more interesting, more human web, the kind of internet that, like the Tesla orbiting Mars, was created by people, for people. Because, in the end, we’re still drawn to content that feels alive.
Will There Be More Bots Than People? How True Is the “Dead Internet” Theory?
Among the many conspiracy theories — about a flat Earth, reptilian overlords, or microchips in Covid vaccines — there’s one that seems absurd for now, but may turn out to be an eerie prophecy for the not-so-distant future: the «dead internet» theory. According to its believers, bots are everywhere. Real people have vanished from the web and are replaced by AI-generated content, while secret forces manipulate the remaining few. The idea first surfaced years ago, but the explosion of AI-generated content has given it new life.
According to its believers, bots are everywhere. Real people have vanished from the web and are replaced by AI-generated content, while secret forces manipulate the remaining few. The idea first surfaced years ago, but the explosion of AI-generated content has given it new life.
The Origins of the «Dead Internet» Theory
The theory (not a theory in the scientific sense, but rather a loose collection of ideas) has murky origins. One of its best-known manifestos appeared in 2021 on the Agora Road forum. A user calling themselves IlluminatiPirate posted a thread titled «Dead Internet Theory: Most of the Internet is Fake.»
He introduced himself as right-wing and claimed that the modern internet (as of January 2021) felt «empty, devoid of people and content compared to, say, the internet of 2007.»
«The internet of today is entirely sterile. There’s nowhere to go and nothing to do, see, read, or experience anymore,» he lamented.
According to him, powerful entities (especially Facebook, Google, and Twitter) manipulate the web through bots, algorithms, and artificial intelligence to feed users carefully crafted narratives. Real people, he argued, had mostly disappeared.
You could say the author, an American conservative nostalgic for the chaotic freedom of the 2000s, mistook the rise of moderation rules for the «death» of the internet. Yet, there is a kernel of truth in his paranoia: the sheer amount of generated content and fake online personas is indeed growing.
And these fake accounts are being used to manipulate reality and deceive real users — not as part of a global conspiracy, but through targeted, tactical operations. One clear example: since the start of Russia’s full-scale invasion of Ukraine, the aggressor has massively used botnets and algorithmic manipulation to spread propaganda narratives.
The scale of the problem is staggering. In October 2025, Facebook announced it had blocked over 8 million accounts linked to scam centers, which is nearly the population of Austria.
In recent months, the «dead internet» idea has resurfaced — this time fueled by tech insiders. Sam Altman, co-founder of OpenAI, wrote in September 2025 that while he’d never taken the theory seriously, he’s now seeing «a lot of LLM accounts» on X (formerly Twitter). Commenters roasted him: «Sam never believed in the dead internet theory — so he decided to make it real.»
Reddit co-founder Alexis Ohanian echoed similar concerns on The TBPN Podcast: «…the internet is now just dead — this whole dead internet theory, right? Whether it’s botted, whether it’s quasi-AI, LinkedIn slop. Having proof of life, like live viewers and live content, is really valuable to hold attention.»
And who can blame him? AI slop has flooded Reddit, too. One viral example: a wholesome story about a plump sphynx cat named Pound Cake went viral as users followed his dramatic weight-loss journey — until someone discovered the cat was AI-generated.
«Wait, I cried yesterday over this cat, and he never existed? What the hell?» user yoki2 wrote.
Ohanian says the «dead internet» theory may have begun as a conspiracy, but it’s fast becoming a reality.
How «Alive» Is the Internet Today?
So what’s the actual ratio of bots to humans online? According to the 2025 Bad Bot Report from Thales, 2024 marked the first time in a decade that bot-generated traffic surpassed human traffic — accounting for 51% of all web activity. The report shows a steady upward trend: bot traffic has been rising every year (red indicates «malicious» bots, green — neutral ones; see chart on page 3 of Bad Bot Report 2025).
The company attributes this surge to the explosion of AI-based services and large language models (LLMs).
And it’s not slowing down. According to Popular Mechanics, citing research from the Pew Research Center, the web could be «more dead than alive» within three years. Pew found that about 38% of all web pages created since 2013 no longer exist (a phenomenon known as link rot), and much of that lost content is being replaced by AI-generated material.
Analytics firm NewsGuard has identified over a thousand «news» sites primarily run by bots, including at least 167 posing as Russian outlets to spread disinformation about the war in Ukraine, using AI-generated content for this purpuse.
For example: the proportion of AI-generated articles compared to articles written by humans (according to Axios):
A similar shift is happening in advertising. Scientific Research Publishing (SCIRP) reports that AI-generated ad videos now boast an average click-through rate of 28%, compared to 15% for traditional ads. More than half of advertisers (56%) say AI-generated content outperforms human-made material.
Ironically, the rise of AI-generated content may be creating a vicious cycle. A recent study on the LLM Brain Rot hypothesis suggests that when language models train on low-quality, AI-generated material (what internet users jokingly call «brain rot»), their own performance deteriorates — they become, in effect, «dumber.»
Writer Ted Chiang, whose short story inspired Denis Villeneuve’s film Arrival, warned about this years ago in his essay «ChatGPT Is a Blurry JPEG of the Web.»
«The more text generated by LLMs is published on the web, the more the web becomes a blurred version of itself.»
So What Now?
Should we all become neo-Luddites and declare war on AI content? Probably not, even if you’re tempted after scrolling past yet another Facebook post about an old woman who sculpted a horse out of cabbage and «no one congratulated her.» Generated content is becoming increasingly difficult to distinguish from human-made posts, and platforms often fail to label it accurately. Europol warned back in 2022 that deepfakes were becoming indistinguishable from real footage, and ironically, every new detection tool only helps make future fakes more convincing.
Still, the rise of AI content may be forcing humanity to rediscover something we’ve neglected: careful reading, critical thinking, and the value of genuine, well-written material. Additionally, running bot farms isn’t free, and generating AI content incurs even higher costs. As Sam Altman once joked, every polite «Hello» or «Thank you» users type into ChatGPT costs his company millions in electricity.
So, maybe you can believe in a «dead» internet, a Martian wasteland populated only by bots. Or you can believe in a more interesting, more human web, the kind of internet that, like the Tesla orbiting Mars, was created by people, for people. Because, in the end, we’re still drawn to content that feels alive.
