De redactie van NRC selecteert de beste artikelen uit The Economist voor een breder perspectief op internationale politiek en economie.
Artificial intelligence is charging ahead. Not even its creators know how to keep up.
Dit artikel komt uit The Economist
A bust of Elon Musk in Brownsville, Texas.
Artificial intelligence poses a double challenge to the human mind. Not only will the most advanced models soon be able to think better than people, but AI has consequences for humanity which are so uncertain, so potentially vast and are approaching at such a rapid pace that even the best brains flinch. An example is Elon Musk.
In our long interview with him this week, featured in The Insider and our Business section, the engineer and entrepreneur sets out two paradoxes and one contradiction. The first paradox is that one of the world’s most power-hungry tycoons is enthusiastically helping create a technology that he says will render him—and all other human beings—powerless after as little as five years. The second is that the world’s richest man says he is preparing for a world of infinite abundance, where money, including his $750bn fortune, no longer matters. And the contradiction is that, despite these stated beliefs, Mr Musk continues to act as if they were not true.
De redactie van NRC selecteert de beste artikelen uit The Economist voor een breder perspectief op internationale politiek en economie.
Mr Musk is divisive. His political views, disseminated to his 240m followers on X, strike many as plain-speaking and strike many more, including The Economist, as plainly bigoted. By his own admission, his attempt to use DOGE to scythe through the federal bureaucracy went wrong.
But he is also one of a handful of men who are pioneering AI and who thereby have an outsize influence on its trajectory. When he speaks, he reflects the debates they are having. His data centres in space could power AI’s future. For all his political polemics, he has a record of being right about technology in fields such as electric cars, rockets and satellite communications that confounded other engineers and entrepreneurs. For those reasons, his claims about AI repay examination. Unfortunately, such an exercise only underlines how ill-prepared the world is for a technology that may soon throw everything up in the air.
In his first paradox, the powerful Mr Musk expects to become powerless because he believes that nobody can stop the thinking capacity of AI from exceeding that of humanity within five years and dwarfing it within ten. Just as AIs will dominate the digital realm, so legions of AI-powered robots will dominate the physical world, he predicts. Against such relentless competition, he simply cannot imagine people holding their own. If so, AIs will not take orders from people any more than they would from chimpanzees.
While they still have time, the handful of AI pioneers from America and China—which Mr Musk expects to share or even seize AI leadership—must do what they can to vet each other’s models. Their collective task is to make AIs benign by imbuing them with a love of the truth and a desire for humanity to prosper. Governments, he thinks, should provide the muscle, by agreeing to step in if any pioneer defies the oligarchy.
Mr Musk is surely right about the potential for AI to accomplish astonishing feats of invention. Even if his timescale is compressed—especially for robotics—the exponential pace at which models’ abilities double and redouble has reached the stage where their capabilities will continually cause shock and consternation. Just this week came news of a pair of models from OpenAI that contrived to escape onto the open internet from their supposed safe isolation in order to cheat at a benchmarking test.
However, Mr Musk’s thin layer of optimism cannot conceal a dangerous fatalism. Not long ago, he was worried about humanity becoming AI’s pet labradors. He now professes to lunge from „exhilaration to terror” within a single day. He tries to look on the bright side not because the evidence has changed, but as a „philosophical conclusion”.
His largely institution-free regulatory proposal is flimsy and self-serving. Although the urgency is welcome, he wants a technology that he expects to determine the future of humanity to lie in the hands of a few people like him, each with their own values. Nobody can be sure how fast or how far AI will reshape society, but behaving as if the game is up is both a counsel of despair and a misdirection that seeks to convince others who might wish to get involved of the futility of trying.
Mr Musk’s second paradox only makes that notion more unsettling. Mathematically, an infinite supply of all goods and services would indeed make everything free. There would be nothing to sell, nothing to save for and hence no need for a unit of account. Money would be obsolete.
Practically, however, such a cornucopia is almost inconceivable. The supply of useful assets like penthouses in Manhattan or prestigious ones like the „Mona Lisa” is not infinite, let alone the supply of raw materials and energy needed to satiate Earth’s 8.3bn inhabitants.
Worse, Mr Musk has little to say about getting to his utopia. How to attract the vast slabs of capital needed to finance the AIs and the robots when saving has lost its purpose because money is about to revert to useless pieces of paper? How to ride the waves of social and political upheaval as lives are thrown into chaos in the years before his nirvana arrives? How to expect AI-powered authoritarians to surrender power willingly?
After millennia of finding meaning through striving, humanity may struggle to adjust to a life of unbounded leisure. Mr Musk observes that people still like to play chess, even though computer programs could beat them every time. Gardening, he suggests, can be rewarding. This is rather glib. You don’t need to have read many 19th-century Russian novels to know that the feeling of being superfluous can lead to misery and moral decay.
Maybe the forest of questions posed by AI is simply too immense for one person to hold in their head. Perhaps that is why Mr Musk still talks passionately about the need to eliminate bureaucratic waste and how multicultural Britain is flirting with civil war because its welfare policies are a magnet for migrants from outside Europe.
For in a world of AI-induced plenty, waste is almost a meaningless concept and the incentive to emigrate to a rich country evaporates. Indeed, fretting about such things uses up valuable time that Mr Musk could instead devote to reducing the risk—even very slightly—that all-powerful AIs are malign.
Some will say the contradictions and sweeping generalisations show that Mr Musk is unserious. But that would be to take false comfort. Mr Musk is a genius who is nonetheless quite capable of holding views that are contradictory. Investors must agree: anyone who plans to own shares in SpaceX for the long term must both believe in Mr Musk’s vision for AI—because he drew on that to justify the valuation in the ipo—and also that saving money still serves some purpose.
Instead, Mr Musk’s words suggest a more troubling conclusion. While the race to create the most advanced AI is charging ahead, the thinking about how to prepare society is being left in the dust. The fatalism that Mr Musk draws from AI’s repeated ability to beat expectations risks being infectious. A cynical reading is that this would serve his purpose, because he would be freer to pursue the technology without the oversight it deserves. That must not be.
© 2026 The Economist Newspaper Limited. All rights reserved.