You don’t build a doomsday device, sell access to it by the month, and then suddenly panic about the explosion. Yet, here we are. Anthropic’s co-founder just told the BBC that we might need a mandatory kill switch for AI. He wants you terrified. He wants you begging the government for a safety net. But if you look closely at the hands holding the scissors, you’ll realize exactly who is getting cut.
You don’t sell a product that causes human extinction and then demand a government-mandated fire extinguisher.
Let’s be brutally honest about the paradox right in front of us. If Dario Amodei genuinely believed his own creation posed an extinction-level threat, he wouldn’t be lobbying for an emergency off-button. He would simply unplug the servers. He would delete the code. He would go home. But he isn’t doing that. Instead, he is actively commercializing his models, scaling them up, and raking in billions in venture capital. Why? Because the threat isn’t real enough to stop the money train. It’s just real enough to scare you.
This is a classic game theory move, and it’s brilliant in its absolute cynicism. It’s called regulatory capture. When you are the incumbent, your biggest threat isn’t the technology turning sentient and taking over the world. Your biggest threat is a 22-year-old in a garage building a cheaper, faster, better version of your product. You can’t outlaw innovation directly without looking like a monopoly. But if you lobby for mandatory safety protocols, complex infrastructure, and government-mandated kill switches, you achieve the exact same thing.
The ‘kill switch’ isn’t designed to shut down the AI. It’s designed to shut down the competition.
You see this playbook everywhere once you know how to spot it. You make safety the moral high ground. You convince the public that only massive, heavily regulated corporations can be trusted with this dangerous technology. You raise the barrier to entry so high that no startup can afford to cross it. The snake oil salesman isn’t just selling you the cure; he’s lobbying the government to make it illegal for anyone else to sell a cheaper cure.
When you read the comments on these news stories, the public sees right through the charade. People ask why they didn’t have a kill switch before. They ask why, if the tech is so dangerous, they don’t just delete it. They recognize the absurdity of a CEO warning about a big bad wolf while actively selling the wolf by the pound. But the strategy still works because politicians love a good crisis to regulate, and incumbents love regulations that solidify their power.
Safety is the most elegant business moat ever invented, because it turns your desire to crush rivals into a public service.
The next time a billionaire tech founder gets on television, shedding tears about the existential threat of his own product, don’t look at the AI. Look at his competitors. Look at his market share. Look at the startups he’s trying to regulate out of existence before they even launch.
When a tech billionaire warns you that the sky is falling, check his portfolio to see if he just bought the sky.
FAQ
Q: If Anthropic truly believes AI is an extinction-level threat, why don't they just stop developing it?
A: Because the extinction threat is a marketing tool. Acknowledging the risk creates a moral high ground that justifies heavy regulation—regulation that conveniently protects their market share from cheaper, faster startups.
Q: How does a 'kill switch' actually hurt the broader tech industry?
A: Mandatory safety infrastructure requires massive capital and compliance teams. Incumbents like Anthropic can absorb these costs, but a two-person startup cannot. It effectively outlaws garage innovation by making the price of admission too high.
Q: Is there any actual danger to AI, or is it all just a PR stunt?
A: The danger is real but wildly exaggerated for business purposes. By framing AI as a potential god-level threat, companies ensure that only they are deemed 'responsible' enough to control it, turning fear into a monopoly.