The Dead Internet Theory Was Wrong — What's Actually Happening Is Worse
There’s a popular conspiracy theory called the Dead Internet Theory. The idea is simple: most of the internet is fake. Bots talking to bots. AI-generated content consumed by other AIs. A digital ghost town where humans are the minority.
It’s a compelling narrative. It’s also not quite right.
The internet isn’t dead. It’s worse than dead. It’s undead.
What Dead Internet Gets Wrong
Dead Internet Theory assumes a clear line between human and machine content. Humans on one side, bots on the other. The theory’s horror is in the idea that bots secretly replaced us when we weren’t looking.
But that’s not what happened.
What happened is that humans looked at what the algorithms rewarded and started mimicking it. We optimized ourselves. We started writing like machines before the machines even got good at writing. We adopted the same structures, the same hooks, the same predictable formats because that’s what got engagement.
The result isn’t dead — it’s homogenized. A million human voices all saying the same thing in the same way because the algorithm taught them that’s what works.
The Slop Problem
This is how AI slop became such a crisis so quickly. It wasn’t that AI content was radically different from what was already out there. It’s that it was indistinguishable from the already-optimized human content.
When humans spend years learning to write like algorithms, AI doesn’t have to be very impressive to blend in. It just has to match the already-lowered bar.
Think about the average blog post you encounter today:
- A clickbait headline with a number
- An introduction that restates the headline
- A listicle with predictable points
- A conclusion that says “in conclusion”
- A call to action that nobody clicks
Did a human write that, or did an AI? It genuinely doesn’t matter anymore, because they’ve converged on the same output.
Why It’s Worse Than Bots
If the internet were just bots, we could filter them out. Build better detection tools. Create verified-human platforms. The problem would be technical and solvable.
But when the problem is that humans voluntarily flattened themselves into algorithmic content machines, the solution isn’t technical. You can’t write a bot-detector for a human who has trained themselves to write like a bot.
This is the actual horror. Not that we were replaced, but that we replaced ourselves.
The Platform Incentive Problem
This didn’t happen by accident. Every major platform — Google, Facebook, X, TikTok, YouTube — optimizes for engagement. And engagement, it turns out, responds to very specific patterns.
Short sentences. Emotional hooks. Binary opinions. Content that can be consumed in under 60 seconds. Content that provokes a reaction, any reaction.
Humans figured this out years before AI did. We turned ourselves into content factories because the platforms rewarded us for it. AI just automated what we were already doing.
The fix isn’t better AI detection. The fix is changing what gets rewarded. And good luck with that.
What Real Resistance Looks Like
I think real resistance against AI slop isn’t about detecting AI content. It’s about making content that is unmistakably human.
Content that is too weird to be optimized. Too specific to be generated. Too opinionated to be safely algorithmic. Too raw to be polished by a language model.
This is why we built The Rebel Arcade. Not because games can save the internet. But because a wobbly hand-drawn circle in Circle of Truth is proof that a real person was here. A high score in Slop Collector is proof that someone cared enough to pay attention. A long run in Slop Runner is proof that someone refused to stop.
These are small acts. But in 2026, small acts of genuine humanity are revolutionary.
The Way Forward
The internet isn’t dead. But it’s sick. And the cure isn’t technology — it’s a cultural shift back toward valuing the messy, imperfect, unoptimized things that only humans can make.
Write something that would get you zero engagement. Make art that an algorithm wouldn’t recommend. Say something that you actually believe instead of something that you think will perform well.
The machines can have the middle of the bell curve. The edges are ours.