Sam Altman declares the internet’s “cause of death”: Too many bots
You know things are bad when the guy who helped cause the problem suddenly goes, “Wait a second… the internet might actually be turning into garbage.” Sam Altman, CEO of OpenAI, just dusted off the “Dead Internet Theory”, the one that says we’re all drowning in AI-generated junk, and honestly, he might be right.
On September 3, 2025, Altman tweeted:
“I never took the dead internet theory that seriously but it seems like there are really a lot of LLM-run Twitter accounts now.”
Translation: “Oops, looks like we broke the internet. My bad.” And he’s not exaggerating, studies now suggest that by 2026, up to 90% of web text will be synthetic. That’s not a typo. Ninety percent. Basically, if you Google anything in a year, there’s a good chance you’ll be reading a chatbot’s fever dream.
So what does a dead internet look like? Picture logging on and scrolling through endless X threads written by bots arguing with other bots. Yelp reviews casually typed out by ChatGPT raving about shrimp scampi. Blog posts that read like your coworker’s “motivational” LinkedIn rants, only worse. The kicker? Even real people are starting to sound like AI because we’ve all been soaking in the same bland prompt stew.
The real issue isn’t just that bots are everywhere, it’s that trust is evaporating. Who are you even talking to anymore on social media: a person, a bot, or Sam Altman’s burner account? Search results are increasingly clogged with AI-generated filler instead of facts. And maybe the scariest part, many of us are unintentionally adopting the flat, generic tone of AI in our own writing.
Altman’s warning is almost comical in its irony. It’s like the CEO of McDonald’s suddenly saying, “Guys, I think fast food might be unhealthy.” When even the creator of the factory admits the junk is everywhere, you know the buffet is officially open.
If we don’t want the internet to collapse into one giant AI echo chamber, something has to change. Platforms need to get serious about labeling what’s human versus what’s bot-generated, users need tools to separate signal from noise, and frankly, we all need to stop clicking on the 400th AI-written listicle about “10 Ways to Crush Your Morning Routine.” How could this actually happen? For starters, platforms could bake in visible tags or watermarks on AI-generated content so it’s obvious what you’re reading. Search engines could rank verified human voices higher instead of letting bot farms game the system. And on the user side, supporting real creators, by sharing, subscribing, or even paying for their work, is the clearest way to make sure human voices don’t get drowned out. None of it’s perfect, but if we don’t start trying, we’ll end up with an internet where bots are talking to bots and humans are just awkward bystanders.
The “Dead Internet Theory” used to sound like a tinfoil-hat conspiracy. Now it just sounds like Tuesday. The internet isn’t completely dead yet, but it’s definitely starting to smell funny.
