Imagine a pharmaceutical company launching a new painkiller with the slogan: “There’s a 10% chance this pill will kill everyone on Earth.” You’d call the FDA. You’d demand they be shut down. You wouldn’t let them keep selling the pills while writing blog posts about how responsibly they’re managing the risk.
Yet, when an Anthropic researcher casually mentions there’s a more than 10% chance that AI could kill all humans, we treat it as a profound philosophical warning rather than what it actually is: a PR move. You don’t monetize the end of the world. Unless the end of the world is your business model.
The internet saw right through it. The top comments on this “existential risk” announcement weren’t panic—they were disgust. People recognize a disingenuous marketing campaign when they see one. If the leaders of these AI companies actually believed their creations had a one-in-ten chance of exterminating humanity, they wouldn’t be scaling up. They’d be taking blowtorches to their own server racks.
But they aren’t doing that. Instead, they are doing the exact opposite: aggressively scaling the technology while simultaneously begging for regulation. Why? Because the “10% chance of extinction” is the perfect rhetorical sweet spot: high enough to demand attention, low enough to justify doing absolutely nothing to stop it.
This is the ultimate bait-and-switch. By framing their products as terrifying, god-like entities capable of destroying civilization, these companies are elevating a glorified auto-complete to the status of a global deity. They are selling the *threat* of AI, not just the tool. If your software can end the world, surely it deserves a trillion-dollar valuation. Surely it needs to be heavily regulated by the government, ensuring that only the current mega-corporations can afford to play the game.
This isn’t public safety. It’s regulatory capture wearing a suicide vest. They are using our fear of the apocalypse to build a moat around their business. When the executioner tells you he might kill you, but first he needs you to sign a petition regulating unlicensed executioners, you aren’t being protected. You’re being played.
We are being asked to gamble our future on the word of technologists who are actively profiting from the exact technology they claim will destroy us. It is a perverse theater of caution. They cry wolf, collect the insurance, and then sell us the sheep they just slaughtered.
The next time a tech CEO stares into a camera and somberly warns you that their own product might end human civilization, don’t reach for a panic button. Reach for your wallet. They’re about to ask you to pay for the privilege of being saved from a threat they just invented.
FAQ
Q: If AI companies don't actually believe AI will destroy us, why keep saying it?
A: Because existential risk rhetoric is a moat. By hyping the danger, they invite heavy government regulation, which locks out open-source competitors and startups who can't afford the compliance costs. It secures their monopoly.
Q: What is the practical implication of treating AI doomerism as marketing?
A: It shifts the regulatory focus. Instead of addressing actual, present harms like copyright infringement or bias, society gets distracted by sci-fi scenarios, allowing big tech to consolidate power under the guise of 'safety.'
Q: Is there really zero chance AI poses a serious threat?
A: AI absolutely poses serious systemic risks, but the '10% chance of human extinction' narrative is deliberately exaggerated to mystify the tech. It turns a complex software engineering problem into a religious crusade, which benefits the priests.