Why the Internet Feels So Strange Lately

Bots Are Flooding Your Feed, Brands Are Getting Fooled, And Trust Is Quietly Running Out

Photo by Growtika on Unsplash

Have you ever felt like the internet has gotten kind of weird lately? Creepy, even. Uncanny. Like something about it just isn’t right anymore?

Open TikTok, and the pattern is everywhere. Half the clips are obviously AI-generated, or they’re movie scenes with an AI voice narrating over them. Then there are the ads for face-swap apps, promising to slap a different face onto your body in seconds.

Switch over to X (formerly Twitter), and it’s the same story. Blue-check accounts post content that doesn’t match their comments, which don’t match anything either, all chasing a slice of ad revenue share. Some of these accounts even reply to people, except the replies feel off. Ask one something oddly specific, like a recipe for stir-fried tempeh, and you’ll actually get a real, working tempeh recipe back. The account belongs to a real person. The thing replying to you doesn’t.

Head to YouTube and you’ll find channels posting fifty videos a day. Scroll the comment section under any viral Facebook or Twitter video, and the pattern repeats: slightly off images, stiff robotic phrasing, comments hawking random links nobody asked for.

Am I the only one who thinks this is strange? Normal people just don’t talk like that.

Enter the Dead Internet Theory

If you’ve felt the same way I have, there’s a theory you should know about: the dead internet theory. It argues that the internet, as we knew it, effectively died somewhere around 2016–2017. According to this theory, the overwhelming majority of what you see now, the likes, the comments, the shares, is generated by bots and AI, not people.

Which means if you’re still scrolling social media as an actual human being, you might be one of the last real people wandering this digital forest. Unsettling? Yes. Genuinely unsettling. So let’s break down why this theory actually makes a disturbing amount of economic sense.

From a Wild, Authentic Mess to an Attention Machine

To understand where the term comes from, it helps to look at what the internet used to be. It was chaotic. Messy. Authentic, in its own way, the same way a real human being is messy and complicated. Personal blogs full of rambling, half-formed thoughts. Hobby forums full of pointless arguments. Grainy 240p YouTube videos about somebody’s cat.

All of it was made by humans, because sharing things is something we’re instinctively drawn to do. But over time, the internet turned into the biggest money-making machine on the planet. And that’s when the attention economy was born.

The term “attention economy” was actually coined by Nobel-winning economist Herbert A. Simon, in a 1971 paper called “Designing Organizations for an Information-Rich World.” His argument was simple: the more information sources multiply, the harder it becomes for any one person to keep up. Human attention is finite, but content never stops arriving.

So our attention turned into a valuable economic commodity, one that companies now fight over constantly. Every second you spend scrolling has a price tag attached to it. Whether you’re actually enjoying yourself isn’t really the point to them. What matters is that you keep scrolling.

But humans have limits. We need to sleep, eat, and take bathroom breaks. We get bored and switch what we’re watching. So the people holding the capital needed something more dependable than fickle human attention spans, which is exactly when this whole “death” of the internet began. Bots were built to warm up the room. AI was built to mass-produce content. Algorithms were built to control exactly what you’re allowed to see and what you’re not.

The internet stopped being made by humans, for humans. It became something made by machines, to harvest human attention. Clickbait, manufactured controversy, short-form content, endless scroll, all of it engineered for one purpose. Has anyone you know ever actually finished scrolling TikTok? Of course not. It was never designed to let you.

Photo by Frank Okay on Unsplash

When Almost Half the Internet Isn’t Even Human

Here’s a number worth sitting with. According to cybersecurity firm Imperva, automated bot traffic now accounts for close to half of all internet traffic worldwide, and according to some of their more recent reports, bots have actually overtaken human traffic for the first time in a decade. That’s the officially detected share. Imagine how much slips by unnoticed.

New accounts get created every single day, even with all those “prove you’re human” tests in the way, the puzzle pieces you have to drag into place, the images you have to click. Turns out they’re not nearly as effective as they’re supposed to be. A human might technically create the account, while something else entirely runs it afterward. Economists have a name for this: engagement fraud.

The Lemon Market of Followers

This concept traces back to a much older idea: the theory of information asymmetry, developed by economist George Akerlof back in the 1970s. His classic example was the used car market. Sellers know exactly which cars are secretly broken, the “lemons.” Buyers have no way of telling a good car from a bad one. Because of that uncertainty, every used car ends up priced low, good ones included.

The same logic plays out with influencers today. An influencer might know that 400,000 of their one million followers are bots, a full 40 percent. The brands paying them for sponsorships usually have no idea. All they see is the number on the surface: one million followers. They can’t tell which ones are real people and which ones aren’t, and that knowledge gap is information asymmetry in action. The brand ends up paying full price for engagement that was never real to begin with.

And if you wanted to become an influencer yourself, waiting around for content to go organically viral is brutally hard. So plenty of people take the shortcut instead: paying for bot farms based in low-wage countries, or leaning entirely on AI-written scripts. There was even a viral observation from a foreign content creator who was shocked to learn that in Indonesia (my beloved country), a single person can run dozens of accounts on the same platform, all for cheap pay.

Why This Actually Wrecks Everything

Here’s the part that makes it genuinely corrosive. Social media algorithms don’t understand quality, they only understand numbers. If a post racks up ten thousand likes within a minute, even if every single one came from a bot, the algorithm reads that as “this content is great” and pushes it straight into the feeds of real humans like you and me.

So we end up being force-fed garbage simply because a swarm of bots convinced an algorithm that the garbage was worth eating. It’s the digital equivalent of being told, with total confidence, that garbage tastes like chocolate. It doesn’t add up, and yet here we are, consuming it anyway.

Welcome to the Generative AI Era

Now layer generative AI on top of all this. ChatGPT, Midjourney, Sora, you name it. Making a blog post, a TikTok video, or a jaw-dropping photo no longer requires any actual skill. All you need is a prompt.

The economics of that are brutal. A genuine creator might spend ten honest hours building one piece of work. Meanwhile, a “zombie account” can pump out a hundred AI-generated pieces in a single hour. There’s no contest.

This is exactly why the internet feels dead. So much of it is just a mosaic of older work, content stitched together from creators who came before, recycled and repackaged into a new prompt or a new post that looks original but really isn’t.

Slop, Spam, and the Bot-to-Bot Economy

The clearest example of this might be the AI “slop” phenomenon flooding Facebook. Bizarre AI photos of soldiers whose bodies look suspiciously like instant noodles. Strawberries are dating melons, somehow shaped like a couple. AI-generated images of a crying child captioned with some version of “one like equals one prayer.” All of it exists purely as bait.

The cycle runs like this: AI generates the content, bots flood the comments, and ad platforms like Google or Meta still pay out to whoever owns the account. It’s an entire economic loop that requires zero human involvement, content from bots, engagement from bots, and profit collected by an account owner who barely lifted a finger. And honestly, making money this way probably feels easier than sending out résumés after résumé in a job market this rough. It’s not hard to see why people choose it.

Photo by Ant Rozetsky on Unsplash

The Long-Term Cost: A Dark Age of Trust

Here’s the part that should worry everyone, not just the people directly affected. Every economy runs on trust. If brands stop believing the clicks on their ads are coming from real humans, they’ll eventually stop paying for ads altogether, because real sales depend on organic human attention. Bots don’t buy anything.

And if we, as regular people, stop trusting that the videos we watch are even real, out of fear that everything might be faked, we’ll simply stop caring. We’ll stop scrolling those accounts entirely. The end state is bleak: bots talking to bots, while actual humans retreat back to trusting only the people they meet face-to-face, or small, tightly-knit communities they genuinely know. We risk sliding into what you might call a dark age of information, a world drowning in content with nothing in it worth believing. The internet stops functioning as a window to the world.

Dystopian? A little, yes. Technically, the internet we used to know isn’t 100 percent dead yet. But if we don’t keep a close eye on where this is heading, the collapse might arrive a lot faster than anyone expects.

How to Survive (and Stay Human) Online

So how do you stay grounded when you can’t always tell what’s real anymore? A few things have helped me:

• Loosen your grip on the algorithm. Don’t just wait for whatever shows up on your timeline. Go look for your own sources of information instead.
• Support real, human-made work. When you find a creator who actually researches things and builds something with their own hands and their own mind, give them more of your attention and support. It’s harder to make, and it’s far more genuine than anything recycled out of a bot pipeline.
• Touch grass. Find people. Find community. Go outside, hang out, have actual conversations. Word of mouth is still one of the most reliable forms of recommendation there is, because online reviews aren’t always as genuine as they look. Plenty of them are quietly sponsored or staged.

Conclusion

That unsettling feeling I described at the start, the sense that something about the internet just isn’t right anymore, isn’t paranoia. The dead internet theory, the rise of the attention economy, the bot traffic numbers, the engagement fraud baked into influencer culture, they all point toward something real: a structural shift in how the internet actually works underneath the surface.

As both creators and everyday users, we still get to choose. We can keep passively feeding the algorithm and the bot economy that profits from our attention, or we can actively go looking for genuine human connection and genuine human-made work. I’d love to know which one resonates with you.

So here’s what I’m asking. Drop something in the comments that’s personal, specific, maybe even a little strange, something only an actual human being could possibly know or feel. I want proof this post got read by people, not just scraped and spammed by bots. Honestly, I’m tired of opening every comment section and seeing nothing but bot, bot, bot, bot. I just want to talk to people.

I’m human. Not a robot. So let’s actually talk.