|
randombio.com | Science Dies in Unblogginess | Believe All Science | I Am the Science Wednesday, November 18, 2026 | technology | opinion It’s not AI, it’s the corporate managersAI is not powerful enough to destroy the world without even existing. Here's what really happened |
he news media are blaming AI for the recent Internet outage
fiascos. Conspiracy theorists blame the usual suspects: hackers,
Russia, China, or whichever politician happens to be in charge.
But the truth is that those networks could have lasted twenty
more years. The problem wasn’t AI or the hardware. It was the MBA
culture run amok. Here’s what happened, in sequence:
Somebody decided to call their latest large learning model “AI” as a marketing ploy. The term caught on and soon every app now had to be called “AI” or be swept aside.
As CEOs decided they could get free labor from AI without the hassle of paying anybody, they fired everybody who knew how to do anything.
Demand for data centers rose exponentially. The price of memory chips went through the roof.
AI leaders started saying that their product was going to wipe out humanity. To prove it, they started doing publicity stunts like turning their software loose on somebody else’s network, telling the press that this was something that could never, ever have been done before.
The news media, being so technology illiterate they couldn’t tell one end of a network cable from the other, believed it.
Computer programmers and sysadmins were forced to react. There are three types and they all reacted differently.
Type 1 The smartest ones—the innovators, the ones whose first girlfriend stomped on their glasses, who stayed home all weekend installing new drivers, and maybe writing them, instead of having fun—said to themselves “If humanity is going to get wiped out, I want in on it!” They quit their jobs and moved to AI companies.
Type 2 The middle tier—the ones who foolishly listened to advice to “learn to code”—realized that if AI was going to take their jobs, they needed to move fast to keep their Beemers from being repossessed. They quit their jobs and moved to cloud computing companies.
Type 3 The older generation—the ones who built everything, knew what a spanning tree was, had a poster of Dijkstra in their bedroom, knew who RMS was, knew the right way to solder a tantalum capacitor to keep it from exploding—in other words, the ones who knew how to avoid creating a network loop and how to keep that gray smoke stuffed inside all those little chips—all got fired.
Those networks should have lasted for decades with ordinary maintenance. But nobody was left to do it. Nobody was left to monitor the networks, check the intrusion detector logs, or reboot the router and network switches when they hung for the millionth time.
So here we are today. Planes grounded, hundreds of corporate networks down, the Microsoft, Google, AWS, and Apple ecosystems snafu’d, the government paralyzed, and everybody wonders what went wrong. Here are my predictions for what’s next.
Jensen Huang gets reinstated at Nvidia under the condition that he not have sex with anybody for at least one week. Nvidia stock price doubles and the entire board is replaced by AI.
The news media decides it was all a massive cyber attack. Government networks continued to crash, releasing prisoners into the streets. Police are paralyzed. Satellites mysteriously get disabled. Militaries across the globe go to DEFCON 2.
But no war breaks out. It is too late. The infrastructure has already collapsed. All the IT guys are busy fixing other people’s plumbing.
The survivors, pining for the days when our biggest problem was that the grocery stores switched to unpasteurized milk that went bad in a week, are told that AI destroyed the world. And it was true. AI destroyed the world, just not the way anybody thought. Nobody thought to blame the CEOs and corporate managers who caused it. So, like clockwork, it happened again ten years later.
nov 18 2026, 5:05 am
Must a chatbot be intelligent to decide whether
something is true?
Finding a consensus in documents is not good enough
The humans have all gone nuts about artificial intelligence
Regulating AI would be as pointless as regulating nonlinear
curve fitting
More thoughts on the anti AI hysteria scam
What if the real reason for asking the government to ban
their own AI is the recognition that they can’t do it?