# The AI Arms Race: How Compute Became the New Currency of Global Power

The global artificial intelligence boom is no longer just a story of technological innovation; it is a story of power. As the United States and China compete for dominance over the compute infrastructure that underpins AI, the contest has become a proxy war fought through export controls, domestic subsidies, and capital markets. Meanwhile, the financial foundations of this boom are showing dangerous cracks, and the workers tasked with implementing AI on the ground are bearing the human cost of a race that prioritizes speed over stability.

## The Financial House of Cards

The most immediate threat to the global economy is not AI itself, but the debt being piled up to build it. Massive borrowing by tech giants to construct AI data centers is now considered the single largest systemic credit risk to the global economy, according to credit analysts [15365]. These "hyperscalers"—companies like Amazon, Microsoft, and Google—are taking on significant debt to fund computing facilities, and if these projects fail to generate enough profit to repay that debt, it could trigger a wave of defaults that ripples through the entire financial system [15365].

The warnings are coming from the highest levels of global finance. Nicolai Tangen, who runs Norway's $1.8 trillion sovereign wealth fund—the largest in the world—has publicly stated that AI-driven stock valuations have grown so fast that a sharp market correction is now a real risk [15397]. The fund holds roughly 1.5% of all globally listed stocks and has significant exposure to U.S. tech giants, meaning a downturn would directly hit its bottom line [15397]. Economists at the European Central Bank have echoed this concern, warning that a sharp fall in U.S. technology stocks could threaten financial stability in Europe, where banks and investment funds hold large amounts of U.S. assets [15359].

The fragility of the market was on full display when Nvidia, the poster child of the AI revolution, lost $130 billion in market value in a single day following reports of a massive $500 billion AI financing deal [15374]. The physical limits of this boom are also becoming apparent, with Malaysia emerging as a test case for a looming energy crunch as its aging power grid struggles to deliver the massive amounts of electricity required by AI data centers [15364].

## China's Counter-Offensive

While U.S. markets wobble, China is executing a carefully calibrated strategy to build its own AI ecosystem independent of American technology. Beijing is steering its top tech champions toward domestic investors rather than Wall Street, with two massive tech listings revealing a clear strategy: fund the country's AI ambitions with local money, not foreign markets [15387]. By listing at home, these companies gain access to deep pools of Chinese capital, and in return, Beijing secures greater control over how the technology grows and who profits from it [15387].

The strategy is already showing results. Chinese AI chipmaker Biren Technology has projected a massive surge in first-half revenue, forecasting growth of up to 2,107% as domestic demand for home-grown semiconductors explodes [15352]. The Shanghai-based company, which makes graphics processing units (GPUs) essential for training AI models, is benefiting from a national drive to reduce dependence on foreign semiconductors [15352]. Unlike the U.S. market, where AI gains are concentrated in a few mega-caps, the Chinese rally is broad-based, lifting smaller and mid-sized firms as well, with the Star 50 index—which tracks the country's most innovative tech firms—jumping 29% this year alone [15352].

China is also positioning itself as a global rule-maker in AI governance. Beijing has launched the World Artificial Intelligence Cooperation Organisation (Waico) as a platform for global AI collaboration, offering its own regulatory model as a template for the world [15407]. China argues that no single nation can handle the dangers of AI alone, and it is courting the Global South with free, capable models while the U.S. leans on allies to adopt its own open-source ecosystems [15407]. The broader competition is not just about technology but about setting global standards, as countries that adopt a particular open-weight model will build their infrastructure around it, giving the originating nation long-term geopolitical influence [15407].

At the same time, China has quietly eased restrictions on Nvidia's H200 chips, permitting small shipments to select domestic tech firms to help close the gap with U.S. rivals [15407]. This balancing act between maintaining control over critical technology and boosting local innovation reflects a deeper strategic calculus: Washington's export controls have forced Beijing to accelerate its own research and development, pushing it toward a unique "technological path" independent of the U.S. [15352].

## The Human Cost of Automation

As the geopolitical contest intensifies, the workers on the front lines of AI implementation are being left behind. At Walmart, the retail giant has pushed AI-powered apps and training to frontline workers, with systems now assigning daily tasks, setting performance targets, and suggesting delivery routes [15370]. But a new survey shows deep distrust: over 85% of Walmart employees don't trust the company to prioritize their needs when developing AI, and half fear the technology will penalize them for mistakes [15370].

Workers report concrete problems. An HR manager said AI assigns tasks with unrealistic expectations, assuming workers will be "perfect every single time." An online fulfillment worker said the AI's suggested store routes make her slower, not faster [15370]. Drivers for Walmart's Spark delivery service say a "smart path" feature sometimes tells them to pick up frozen items like ice and TV dinners first, letting them thaw before delivery [15370]. One worker said she simply ignores the alerts on her personal phone [15370].

Walmart says employee feedback is key to improving the tools, and the company insists it does not punish people for not following the tech guidance [15370]. But the experience at Walmart illustrates a broader pattern: AI is being deployed not to empower workers but to extract more labor from them, with the burden of correcting the technology's mistakes falling on the workers themselves.

The labor implications extend beyond retail. In India, Prime Minister Narendra Modi has announced free online coaching and artificial intelligence training for 10 million young people, weeks after student protests over exam paper leaks and hiring irregularities forced the education minister to resign [15356]. The move is designed to reassure young voters ahead of upcoming state elections, but observers note that the announcement lands at a time when public trust in the government is strained [15356]. The AI training initiative signals an effort to position India as a global hub for emerging technologies, but its success will depend on internet access and the quality of the AI curriculum—factors that remain uncertain [15356].

## The Hidden Costs of the AI Boom

The AI boom is also having a negative impact on the climate, with data centers that power AI systems consuming massive amounts of electricity, much of which still comes from coal and natural gas [15386]. Studies conclude that, so far, AI's net impact on the climate is clearly negative, and experts warn that without urgent policy changes, the AI boom could undermine global climate goals [15386].

Even the demand for data to train these systems is pushing companies to extreme measures. Rare books are becoming a key resource for training large language models, and Amazon—which began as an online bookstore—is reportedly destroying some of these rare volumes to extract their content for AI training [15365]. The process involves scanning or cutting apart the books to capture text that is not accessible elsewhere, raising concerns among collectors and researchers who see the destruction as a loss of cultural heritage [15365].

Meanwhile, the broader global economy is buckling under the weight of geopolitical conflict, strategic blockades, and a climate crisis that is no longer a distant warning but a lived reality [15364]. The closure of the Strait of Hormuz, the war in Ukraine, and the drought-parched heart of Europe are all converging to push inflation higher and squeeze the world's most vulnerable populations [15364]. The pursuit of geopolitical power and corporate profit is increasingly colliding with fundamental human needs, leaving ordinary citizens to bear the heaviest burden of inflation, fuel shortages, and a widening gap between the wealthy and the rest [15364].

## A Stark Choice Ahead

As these crises converge, the international community faces a stark choice: adapt to a new reality of interconnected threats, or risk being overwhelmed by them. The AI boom is not a neutral technological development; it is a distribution of power that is reshaping the balance between states, corporations, and workers. The winners are those who control the chips, the data, and the capital—the losers are the workers who are being asked to fix the machines that are displacing them, and the vulnerable populations who are bearing the costs of a system built on extraction and speculation.

The massive arms trade and militarization drive at the center of this turmoil diverts crucial public resources from social needs toward conflict and profiteering, escalating global insecurity for the benefit of a few [15398]. The window to act on climate, conflict, and inequality is closing quickly, and the costs of inaction are being paid in lives, livelihoods, and the health of the planet itself [15398]. The question is not whether the AI bubble will burst, but when—and who will be left to clean up the mess.