The AI Race Is Not What We Think. It’s the Empire We Don’t See.
Image: January 16,2025, Jensen Huang, NVIDIA founder and CEO greets with his fans after the opening ceremony of new Taiwan factory in Taichung city, Taiwan.
By Naseem Qader
Empires are rarely declared while they’re being built. They rise through conquest and strategy—territories seized, borders redrawn, and trade routes claimed—until it’s too late to imagine a world without them. Power still marches with armies and flags—but the kind remaking the world today does its work through infrastructure, contracts, and narratives that make domination feel like empowerment.
A new empire is quietly reshaping power, embedding itself in mines, chips, undersea cables, cloud platforms, data models, and the financial flows every nation will come to rely on.
The Race Everyone Feels, Yet Few Question
The world’s most powerful AI developers admit in private what they rarely say in public: they feel unable to slow down. As Zanny Minton Beddoes observes in The Economist, “The Economics of Superintelligence”, even those shaping AI’s future feel trapped in a race they can’t escape.
Policymakers reinforce this urgency by framing speed as national security. In a world of real geopolitical rivalry, AI leadership is equated with military strength, economic dominance, and global influence. Accelerating innovation is presented as defense: a hedge against adversaries’ advances, cyber warfare, or AI‑enabled bioweapons. In this framing, restraint looks risky while racing ahead feels patriotic.
Meanwhile, experts from the Atlantic Council in the New Atlanticist “What Trump’s New AI Action Plan Means…” describe how U.S. policy now frames AI as geopolitical competition— accelerating chip exports, data center construction, and export licensing for allies.
Wrapped in promises of innovation and empowerment, this race consolidates control while obscuring who truly benefits—and who pays the price.
Where Power Truly Begins
AI may seem weightless, just code on a screen, yet its foundations are deeply physical. Lithium from Chile’s Atacama, cobalt from the Democratic Republic of Congo—where hundreds of protests and violent incidents have erupted over extraction—and nickel from Indonesia and the Philippines fuel chip production.
Behind this chain lies invisible labor—miners in the DRC earning less than $0.40 USD/hour and often forcibly evicted from their land, alongside temporary online “gig” workers in places like Venezuela, Kenya, and the Philippines. These workers spend up to 18‑hour shifts hunched over laptops, racing to claim scarce tasks that vanish in seconds—clicking through endless images and audio snippets for pennies per task while their eyes burn and hands cramp.
Tech companies benefit from this labor while remaining insulated from accountability by outsourcing through layers of contractors. The firms that profit most from AI rarely see, let alone pay fairly for, the grueling human toll that makes their models possible.
And this hidden labor is only one part of AI’s vast footprint—the hardware itself also demands immense resources. Semiconductor fabrication plants—called “fabs”— consume staggering amounts of energy and water. Global data‑center electricity use now exceeds France’s daily consumption, while cooling systems withdraw water at a rate roughly matching Greenland’s population’s daily use.
The rhetoric of “progress” hides these costs—masking extraction, exploitation, and emissions behind an illusion of inevitability. This hidden footprint is the foundation of a larger contest—one fought not with weapons, but with infrastructure itself.
Infrastructure Is Power
The so-called “AI stack”—spanning raw materials, chip manufacturing, undersea cables, cloud platforms, and data models—is becoming the architecture of global influence. It is financed through sovereign debt, private capital, and trade agreements that will shape power for generations to come.
And what follows is rarely debated publicly. Corporate “voluntary standards”—rules with no legal force—become de facto policy once adopted by the largest firms. These standards decide who funds infrastructure, who controls access, and who gains leverage long before citizens or lawmakers even realize what has been ceded.
As scholars at the Brookings Institution and other policy analysts in international circles warn, the doctrine of “innovation at all costs” disguises trade‑offs as progress and embeds “choice as inevitability”.
Global AI infrastructure spending is projected to reach US $6.7 trillion by 2030—including $5.2 trillion for data centers alone. These investments will decide which nations set the rules—and which must live by them—for decades to come.
The New Empires Leave No Flags
Taiwan’s semiconductor fabs dominate advanced chip production. Singapore and Malaysia host rapidly expanding cloud ecosystems. Undersea cables from Indonesia, Vietnam, and the Philippines connect directly to U.S. firms such as NVIDIA, Google, and Amazon Web Services.
China is building a parallel network—its own chips, cloud platforms, AI models, and governance frameworks—to compete with U.S.-aligned systems. Russia, unable to match either power, is increasingly dependent on Chinese technology as a junior partner.
By contrast, U.S. allies such as South Korea, Japan, and Australia are deepening integration with American infrastructure, entrenching a world of competing digital blocs rather than one open system.
The resources enabling all this flow disproportionately from the “Global South”: cobalt and copper from the DRC, bauxite from Guinea, lithium from Chile and Bolivia, and copper from Peru. Yet these same regions remain largely excluded from governance tables.
The consequences of that exclusion are visible in the same places: polluted waterways, deforested land, rising carbon emissions, and displaced communities. Those bearing these burdens are the least able to shape the rules—or share in the rewards—of the AI economy.
And this is why the next debate cannot be about hypothetical futures alone.
Why AGI Is the Distraction, Not the Centerpiece
Debates obsess over Artificial General Intelligence—a still‑hypothetical AI capable of reasoning across tasks as well as, or better than, humans. But while AGI remains speculative, the real race—over the hardware, networks, and cloud infrastructure that make AI possible—is already underway.
This race is being sold—by policymakers and tech leaders alike—as “inevitable progress.” Yet the rules of this system are being locked in quietly, long before the public realizes what has been ceded.
Governance by Default Means Losing Choice
AI arrives not through treaties but through licensing agreements and “click‑through” terms. Nations adopting full‑stack AI packages inherit conditions they never negotiated. Corporate infrastructure becomes governance—whether democracies intended it or not.
Even with so much at stake, public debate still fixates on narrow questions—whether AI systems will outsmart humans, displace jobs, or spread misinformation—while ignoring the deeper struggle over who will own and govern the foundations of AI itself.
What Most Debates Miss
This is the part of the story that rarely gets told—the power struggle embedded in physical systems most of us never see.
Real power today is built through the physical backbone of AI—from mines and chip factories to undersea cables, massive data centers, and the models they feed—far more than through headline‑grabbing breakthroughs in AI systems.
This essay connects disparate threads: exploitative labor, resource extraction, financial lock‑in, weak non‑binding standards, and the creation of parallel digital blocs. It argues that the real contest is not about building smarter AI systems— it’s over who controls the entire chain of infrastructure that now decides whose voices count, whose values prevail, and whose futures are coded in.
How AI Could Shape the Future
AI’s potential is immense— accelerating drug discovery, improving disaster prediction, and optimizing energy use. These benefits could reshape societies for the better. But under today’s trajectory, the same systems could just as easily deepen inequality, strain natural resources, and lock nations into decades of dependency. The question is not whether AI can create value—it is who decides how that value is distributed, and who bears the costs of the future being built right now.
The Hardest Truth
AI is not a break from the past. It is the most seamless continuation yet of a pattern we’ve already accepted— the empire we’ve been willingly helping to build, financing with our clicks, data, and dependence—never realizing that we are also the raw material.
The real question is whether we will have the awareness to pause and question the future we’re creating, or simply continue financing our own dependence with every click and dataset—until the systems outlast us.
Naseem Qader has enjoyed a decades-long career in media and marketing. Her work experience included a seven-year engagement as a strategic marketing planner for the Los Angeles Times. She also served six years as Director of Marketing for Valassis Communications. Other roles included a Product Development Manager role at Gemstar Corp. and a marketing planner with Knight-Ridder Newspapers. Naseem’s specialties include strategic planning, business development, marketing research, CRM, multi-cultural marketing, and nonprofit branding and governance.
Naseem currently serves as Chair of Public Relations and Marketing at the World Affairs Council of Orange County and is a member of the group’s executive board. She’s an advocate for special-needs children. She was a global ambassador for Special Olympics World Games and has been a volunteer at the Shea Therapeutic Riding Center. And, is active with New Ground: A Muslim-Jewish Partnership for Change. She’s passionate about resolving global conflicts.
Originally from Hyderabad, India, Naseem now resides with her husband Tom in Lake Forest, CA. She enjoys hiking, travel, art, history and foreign-language films.
She holds a B.S. in Behavioral Sciences and an M.B.A. from California Polytechnic University, Pomona.
The views and opinions expressed here are those of the author(s) and do not necessarily reflect the official policy or position of the Pacific Council.
