The San Francisco Bay Area is in the midst of an AI frenzy that makes the California gold rush of the mid-19th century look like a scavenger hunt. Top programmers and developers are being offered compensation packages worth hundreds of millions of dollars to switch firms, while young engineers lucky enough to have joined leading AI startups early are contemplating retirement before age 35.
Driving up the Bayshore Freeway from San Francisco International airport into the city, you pass hyper-specific billboards advertising obscure AI applications seemingly aimed at absurdly niche audiences. How can that possibly be profitable? The answer is that in a city crawling with startups, getting the right software product in front of a founder whose company could soon be worth billions of dollars is far more lucrative than using billboard space to sell burgers or laundry detergent.
Yet beneath the frenzy lies a palpable anxiety, as members of this young super-elite fear that their startups may not be the ones to win the AI sweepstakes. Failure, in their eyes, means being left behind while AI automates large swaths of white-collar work – especially coding jobs, which until now have been a veritable licence to print money – and falling into the ranks of the permanent poor.
Though economists still debate whether AI will destroy jobs or create them, the prevailing mood in Silicon Valley is far more pessimistic. Either your startup makes it within the next five to 10 years, the conventional wisdom holds, or you’d better pray the government provides a generous universal basic income.
Despite Donald Trump’s efforts to pull Silicon Valley into the Maga orbit, American-style progressivism continues to dominate Bay Area culture. Most of California’s young tech strivers still see themselves as dyed-in-the-wool progressives –enthusiastic supporters of taxing the rich, at least until they become rich themselves.
Yet for all their virtue signaling, Silicon Valley elites seem strangely oblivious to the fact that the vast majority of people left behind by the rise of AI will not live in the US. Nor will they live in countries that have secured a place in the AI supply chain, such as South Korea, Japan and Taiwan.
While South Korean firms such as Samsung and semiconductor company SK Hynix have become trillion-dollar giants on the back of AI’s insatiable demand for advanced memory chips, Europe has produced far fewer success stories. ASML, the Dutch firm that holds a near-monopoly on the high-end lithography machines needed to manufacture the world’s most advanced semiconductors, is a rare exception. The picture is even bleaker in Africa and Latin America, which have yet to produce anything remotely comparable.
Countries that fail to carve out a place for themselves in the emerging AI economy risk ending up on the losing side of this century’s most consequential economic transformation. With no windfall profits to redistribute and no surge in tax revenues to finance universal basic income, they could find themselves with no way to cushion the shock of mass job displacement.
This is not simply a story of political incompetence or lack of ambition. How can African firms compete when hundreds of millions of people across the continent still lack access to electricity, the most basic prerequisite for AI infrastructure? And how can Latin American countries finance massive investments in datacentres when savings rates remain low and a history of recurring debt crises continues to deter foreign capital?
To be sure, some African and Latin American countries stand to benefit enormously from AI’s voracious appetite for minerals such as copper, rare earths, lithium, nickel, cobalt, gallium and germanium. Chile, Peru and Mexico are obvious candidates, but even the cobalt-rich Democratic Republic of the Congo could reap substantial rewards if its brutal civil war ever subsides.
Natural-resource wealth, however, has often proven to be as much a curse as a blessing. Mineral-rich countries may find themselves flush with AI-driven revenues and still lack the political and economic institutions needed to spread the gains across society.
India, meanwhile, faces a very different set of risks. With AI devouring mid-level white-collar workers like plankton, India’s vast outsourcing industry could be among the hardest hit. Given its deep reserves of creative and technical talent, India could still emerge as one of the big winners of the current tech race, alongside the US and China. But the country has struggled to harness that potential at home, allowing many of its brightest minds to migrate to California. Trump’s immigration crackdown may slow that brain drain, though whether that ultimately benefits India remains an open question.
China, for its part, is already an AI powerhouse. But even there, the government is only beginning to grapple with the implications of AI-driven job displacement. Even if the country wins the AI race, maintaining social stability could prove difficult without expanding the social safety net.
The US may be more dynamic, but it is hardly better prepared for AI’s likely impact on labour markets. To avoid deepening social fractures, it will need to find ways to distribute the benefits of AI more broadly rather than allowing them to remain concentrated in the hands of a small group of first movers and tech billionaires.
But the danger is not confined to national borders. AI threatens to widen the gulf between technological winners and losers, enabling wealthy countries to reap the rewards while consigning billions of people across the developing world to fall ever further behind. No one really knows what such a world would look like, let alone how to keep it from tearing itself apart.
Kenneth Rogoff is a professor of economics and public policy at Harvard University. He was the IMF’s chief economist from 2001 to 2003.