Warren Buffett, the investor whose words are followed across markets, has spent years wondering aloud what becomes of human work once machines can do much of it. Across shareholder meetings and interviews from 2016 to 2023, he has held two truths at once: that artificial intelligence could lift output, speed medical discovery, and shorten the working week, and that it could scatter jobs, strain democracy, and prove impossible to recall once released. His unease, which he frames partly through the image of the atomic bomb, rests on a simple human fact, that society may move faster than its capa
Warren Buffett Sees AI's Promise and Peril: Job Losses, Political Upheaval, and a Bomb-Scale Warning
We won't be able to un-invent it.
Mark
Why does it matter what Buffett thinks about AI? He's an investor, not an engineer.
Mimi
Because he's framed it as an economist would: output, costs, labor. He said the biggest value comes when AI replaces human labor, since machines don't ask for raises.
Luke
Worth noting he's open about his limits. In 2017 he said he was speaking without knowing a thing about it technically.
Mark
He also sees a bright side, doesn't he?
Mimi
Yes. Fewer working hours, faster medical analysis, reading X-rays and every paper on a disease. He called the shorter week enormously prosocial.
Luke
But that's a forecast, not a finding. Nothing in the reporting shows those gains happening.
Mark
And the bomb comparison?
Mimi
That's from 2023. His point was that you can't un-invent it, and that human thinking and behavior don't change as fast as the tools.
Luke
Keep in mind it's an analogy about misuse, not a claim that AI is equivalent in destructive power. And most of these quotes are from 2017, so they predate the recent wave of AI tools.
Mark
What about Munger?
Mimi
He's skeptical of the hype, and said old-fashioned intelligence works pretty well. So even at Berkshire the views aren't uniform.
Il Polso
- A technology that can read more medical literature than any doctor is already promising faster diagnoses and fewer hours of human toil.
- Buffett warns that the same machines could displace workers in whole sectors and leave democratic institutions scrambling to respond.
- The fear is irreversibility: once artificial intelligence spreads, society may not be able to un-invent it.
- The outcome hinges on speed, and Buffett admits he cannot predict how fast societies can absorb the change.
Warren Buffett, the investor whose words are followed across markets, has spent years wondering aloud what becomes of human work once machines can do much of it. Across shareholder meetings and interviews from 2016 to 2023, he has held two truths at once: that artificial intelligence could lift output, speed medical discovery, and shorten the working week, and that it could scatter jobs, strain democracy, and prove impossible to recall once released. His unease, which he frames partly through the image of the atomic bomb, rests on a simple human fact, that society may move faster than its capacity to understand what it has made.
For nearly a decade, Warren Buffett has returned to a question that unsettles many people: what happens when machines can do much of what people do? In interviews and at Berkshire Hathaway shareholder meetings between 2016 and 2023, he has described artificial intelligence as a potential engine of progress and a source of disruption and danger.
His optimism is largely economic and medical. He has argued that AI could read X-rays more accurately than people and absorb far more research than any human could hold, including every new article on prostate cancer. He has also imagined a future in which machines, unlike workers, never demand raises or healthcare, and in which automated production delivers goods with little human effort.
Buffett has not dismissed the costs. He expects significantly less employment in some areas and shorter working weeks, which he sees as broadly beneficial, but he has warned of political upheaval if the technology spreads too quickly. In 2023 his language grew sharper: once something can do all kinds of things, society cannot simply un-invent it, and he compared AI to the atomic bomb, echoing Einstein's remark that the bomb changed everything except how people think.
He is less worried about science-fiction scenarios of machines turning on their makers, joking that at 86 he worries less about such things. Charlie Munger, his Berkshire partner, offered a more skeptical note, saying he doubts much of the hype and that ordinary intelligence still works well. Taken together, the remarks point to a clear frame: enormous gains, real displacement, and an unanswered question about how fast societies can adapt.
Warren Buffett has spent the better part of a decade turning over the same question in public: what happens when machines can do much of what people do? Across interviews and Berkshire Hathaway shareholder meetings between 2016 and 2023, catalogued in CNBC's Warren Buffett Archive, the investor has described artificial intelligence as both a potential game-changer and a source of disruption, division and danger.
His case for the upside is mostly economic. In 2017, he said AI had great potential, and framed the test as how much better its results could be than those of human beings, or how many people it could replace. He also pointed to a different kind of gain: information humans are incapable of gathering, since no one can hold in their head every word written about something like prostate cancer, updated to yesterday. He said AI could read X-rays faster and better than people and absorb the literature, "zillions of pages."
But the biggest value, he said that year, would come when AI replaces human labor. Machines, in his words, don't come round annually asking for higher wages, and they need no healthcare beyond a little maintenance. In 2016 he offered a more whimsical picture: robots running farms, Apple and Berkshire Hathaway, with one person punching a button each morning so that the goods and services people now enjoy were turned out automatically.
He has not pretended that this arrives without cost. In 2017 he said AI should mean significantly less employment in certain areas and far fewer working hours each week, which he called enormously prosocial in the long run. He paired that with a warning that it would be enormously disruptive in other ways, and could create huge problems for a democracy and how it reacts. Speaking without claiming any technical expertise, he said the social effect would likely be hugely beneficial but the political effect very unpredictable if the technology came in fast, which he thought it could.
The sharpest language came in 2023. Buffett said that when something can do all kinds of things, he gets a little worried, because society won't be able to un-invent it. He reached for the atomic bomb, invented in World War II for what he called a very good reason. He recalled Einstein's remark that the bomb had changed everything in the world except how men think, and applied it to AI: it can change everything except how men think and behave.
He has been lighter about the science-fiction fears. Asked in 2017 about AI rebelling against its creators and acting maliciously, he joked that he had the cure: he was 86, and you don't worry as much under those circumstances.
The collection, compiled by Business Insider as seven quotes lightly edited for length and clarity, also includes a skeptical note from his Berkshire partner Charlie Munger. In 2023, Munger said he was personally skeptical of some of the hype around AI, adding that old-fashioned intelligence works pretty well.
What the quotes show is a consistent frame from one of the most watched investors alive: large productivity and research gains, real job displacement, and a political question about how quickly societies can absorb change. Of those, the speed of adoption is the variable Buffett himself says he cannot predict.
Citazioni salienti
With AI, it can change everything in the world except how men think and behave.— Warren Buffett, 2023
I am personally skeptical of some of the hype that has gone into artificial intelligence. I think old-fashioned intelligence works pretty well.— Charlie Munger, 2023
Vuoi la storia completa?
Leggi l'originale su Yahoo Canada Finance ↗
Technology