Who Are You Really Talking to Online? The Dead Internet Theory Explained (2024)

Ever get the feeling you're not talking to a human online or that the "people" responding to a social media post aren't of this world?

You're not alone.

It's all part of the "dead internet theory," where the vibrant and rich online world we know and love has been replaced with automated bots, algorithms, and manipulated content—with no human traces to be found.

What Is Dead Internet Theory?

If you're even slightly familiar with social media and other online forums, you've probably noticed an uptick in generic responses flooding responses to any post. They're typically short, make no reference to the actual post, and come in their thousands.

Dead internet theory suggests that the vast majority of these responses are made by bots using AI, and human interaction with these posts is non-existent. Oh, and the post itself was a bot, and may use AI-generated images to capture the eye or post something provocative.

The idea that the internet is dead has been circulating for at least a decade, but a 2021 Agora Road Macintosh Cafe thread helped encapsulate how many internet users feel; the internet as we knew it has already gone, replaced by endless algorithms serving up increasingly narrow content, designed by AI, for AI.

Dead internet theory has three main theories:

• Bot activity and automated content dominate the internet, with most interactions, posts, and content coming from bots and AI tools like ChatGPT.
• Algorithmic curation and manipulation control online discourse rather than humans, with bots and AI-generated content dictating how public opinion moves (while suppressing human-created content).
• Human-made content is actively suppressed in favor of bot and AI-generated content, linking to the previous point.

The result is general mistrust in the internet, social media, and the direction of public discourse on the most popular platforms, forcing users to leave them behind and allow AI-controlled bots to control the conversation.

Examples of Dead Internet Theory

Encounters with the dead internet vary by user. I've noticed a significant uptick in strange AI-generated posts on Facebook; generic AI generated images with tens of thousands of likes and a similar number of comments.

For example, the following Facebook post features a quite bonkers mashup of near identical US servicemen in wheelchairs with prosthetic limbs, with a mocked-up Jesus figure sitting between them, also in a wheelchair. However, the prosthetic limbs feature the face of the Jesus character on the kneecap and frame, while Jesus is wearing a sign stating, "Today is my birthday."

It's utterly bonkers.

Posted on April 26, 2024, the post has 279K reactions and 5.3K comments, most of which say, "Happy Birthday," all seemingly from bot accounts (though there are undoubtedly real-human accounts in there, too).

One of the best sources for dead internet theory examples is the @DeadTheory__ X account. It regularly posts moments where bot accounts linked to AI tools like ChatGPT fail to work, revealing that they are just AI.

Another example is the old lady baking a cake. You may have seen this floating around Facebook, where an obviously AI-created image is captioned with something similar to "Hello everyone, I am 108 years old, I made my own birthday cake with peach cream and filling, I started decorating cakes from 5 years old, I love it, and I can't wait to grow my baking journey. It uses a similar image to the one below, again with tens of thousands of responses, all saying "Happy Birthday" or something similar.

This trend also led me to the peachcreamandfilling subreddit, where users (or the admin) listed posts they found using the same phrasing. It looks like the user gave up after a short period, but they found a good few posts.

Is the Internet Really Dead?

The 2024 Imperva Threat Report found that almost 50 percent of all internet traffic "comes from non-human sources." Furthermore, "bad bots. . .now comprise nearly one-third of all traffic." And yes, there are good bots and bad bots.

It's a stark figure that paints internet sites and services heading in one direction. Directives from certain websites haven't helped, either. For example, Elon Musk's X Verification scheme promises to give verified users a slice of ad revenue. So, the easiest way for accounts to drive engagement is spam bots (remember, the bots that Musk claimed were killing Twitter before he took over?) connected to generative AI tools like ChatGPT. They respond to viral posts and trending tags, polluting response threads with gibberish, as seen above.

The internet itself isn't dead. Structurally, it is as strong as it has ever been. However, the dead internet theory illustrates where the internet is (not heading; we're already there): a landscape dominated by constant engagement, hyper-content exposure, and SEO-gamed and algorithm-dominated search responses that don't deliver what real, regular humans actually want.

It's a bit "old man shouts at clouds," but the internet really was better before the enormous commercialization foisted upon it over the years, which in turn forces every service to become more general, more sanitized, and more commercially viable, killing communities in the process. The other side of the coin is that sites built on user advertising don't want to admit user numbers and engagement have fallen off the edge of a cliff and are happy for bots to boost any activity level to keep the advertising numbers up.

So, no, the internet itself isn't dead, but it's absolutely hurtling towards an end-game dominated by AI-generated content made for bots, consumed by bots, and regurgitated by generative AI.