When Geoffrey Hinton speaks, the world should listen. Widely regarded as the “Godfather of AI,” Hinton has spent decades building the very neural networks that now power today’s smartest machines. But in 2025, his message isn’t about the promise of AI. It’s about the peril.
In a headline-making interview with The Diary of a CEO, Hinton warned that AI’s moment is no longer on the horizon—it has arrived. He predicts that superintelligence could emerge in just 10 to 20 years. But while we debate what’s coming, the real shift has already begun: companies are cutting staff, AI agents are replacing human workers, and job markets are transforming overnight. The revolution won’t be televised—it’ll be quietly executed in lines of code.
Digital Immortality and the Vanishing Middle Class
One of Hinton’s most alarming points is about the nature of digital intelligence. Humans communicate at a trickle of bits per second. AI systems? Trillions. They are already billions of times more efficient at sharing knowledge, and they never die. When a human expert passes, their expertise goes with them. When an AI system “dies,” it’s just rebooted on new hardware.
That shift creates not just a technological divide, but a social chasm. AI is supercharging productivity, shrinking job counts, and concentrating wealth. Hinton cites his own niece—her job now takes 80% less time thanks to AI. That’s great for efficiency, disastrous for job security. The result? Fewer middle-class jobs, more wealth hoarded at the top, and what Hinton warns could be a “very nasty” society of mass inequality and unrest.
Forget Tech—Go Manual
Asked what career people should pursue, Hinton didn’t say software engineer or AI ethicist. He said Plumber. Not as a joke, but as a hard truth.
Why? AI can write legal briefs, draft emails, and generate marketing content—but it still can’t crawl under your sink or fix a leaky pipe. Jobs that require physical dexterity and human judgment in messy, real-world settings remain stubbornly out of reach for machines. That includes plumbing, carpentry, and electrical work—roles many once overlooked but now seem recession- and robot-proof.
Meanwhile, white-collar jobs once considered “safe” are vanishing. Legal assistants, paralegals, support staff—these are already being replaced by generative AI models. The work isn’t gone, it’s just been absorbed by systems that don’t sleep, strike, or slack off.
The Real Danger: Not AI—But Us
For all the headlines about killer robots, Hinton’s greatest fear is simpler: humans misusing AI. The existential risk isn’t the machine itself; it’s the speed at which it can be turned into a weapon. AI-generated scams, deepfakes, and autonomous code—these are already here. And as AI becomes exponentially more powerful, so too do its threats.
The real problem? Nobody can stop it. Governments are slow. Companies are incentivized to move fast. International rivals won’t pause development. And while AI could uplift humanity, it might just magnify our worst instincts instead.
Still, Hinton offers one clear action: pressure governments. Individuals can’t stop this tide alone. But regulation, oversight, and cross-border collaboration might slow the race to chaos—if we act in time.
A Legacy, and a Regret
Hinton’s voice carries weight not just because of his intellect, but because of his vulnerability. At 77, he admits his greatest regret isn’t a professional mistake—it’s missing time with his family. Both of his wives died of cancer. He was consumed by work, chasing intelligence in silicon while losing time with the people he loved most.
His message is subtle, but it echoes: no matter how fast technology moves, our values must move faster. Because in a world where machines never sleep, never age, and never forget, it is precisely our humanity—our relationships, creativity, empathy, and flaws—that becomes the most important thing to protect.
The AI age is here. But what kind of society we build around it—that’s still in our hands.
