Tech Reaches the Geopolitical Summit: The Rise of Datacenter Diplomacy

There’s an article in – of all places – Foreign Policy that I recommend you read. 

I recommend that you read the article not because I wholeheartedly agree completely with everything it says – the piece raises perhaps as many questions as it answers – but because I think it provides valuable insight into the degree to which the information-technology industry has become a cynosure of global economic growth and, as a result, a fulcrum of geopolitical machinations. The IT industry, however, has a growing dependence on another key industry, and that set of circumstances provided the impetus for the article inForeign Policy.   

 You know the old adage, “Be careful what you wish for”? Well, the technology industry was not as careful as it should have been, and its wishes have come true. 

The problem with geopolitics is that nobody emerges from its clutches uncompromised. Ends always justify means, even when the latter are questionable, if not diabolical. Pawns must be sacrificed for the greater good – whatever that might be, and whoever gets to define it – and empires must always be expanded and enlarged, no matter the cost. 

Geopolitics implies conflict, often of a simplified Manichean character: us against them, good against evil, lightness versus darkness. The geopolitical screenplay usually demands that two mutually exclusive and implacable foes battle for supremacy across a multifarious landscape – political, economic, industrial, scientific, technological, social, military – in a zero-sum game that turns out to be anything but a game. By the time an industry has achieved geopolitical pre-eminence, it’s become very serious business.  

As for the Foreign Policy article, written by Jared Cohen, president of global affairs at Goldman Sachs and co-head of the Goldman Sachs Global Institute, the central focus is AI-inflected, datacenter-centric geopolitics. Cohen’s thesis is that AI has triggered what is essentially a datacenter arms race against China, and that the U.S., while prodigiously positioned to benefit from AI through the breadth and depth of its energy resources and datacenter assets, cannot achieve clear-cut AI autarky. What the U.S. must do, Cohen advises, is build geopolitical alliance spanning other nations that can be trusted and have the resources and shared interests to contribute to the U.S.-led strategic plan. 

As I read the article, I sometimes wondered if it represented an instantiation of the old saw about everything old becoming new again. Is Cohen suggesting a revivified Pax Americana in to achieve global AI hegemony? It certainly seems that way. If that’s the scheme, perhaps some thought should be given to what’s in it for those allied countries as well as what’s in it for the U.S. 

With a U.S. presidential election imminent, such speculation is far from idle. Donald Trump’s return to the White House might feature a recrudescence of aggressive U.S. protectionism. If allied nations and their industries are getting hammered by Trump’s tariffs, they might object to playing a subservient role in a globe-bestriding, U.S.-directed AI empire. You know the principle of give and take in negotiations? Well, if a lot of taking is being proposed, there needs to be some give, too. A one-way street terminates in a dead end of disaffection and alienation. 

Cohen wrote his article for Foreign Policy, so I should not have been surprised a strong emphasis on geopolitics. That’s what Foreign Policy does, what it’s all about. Still, Cohen’s emphasis on the struggle for global dominance between two great powers – in this case, the U.S. and China – caused me to feel a strange pang of nostalgia for the now-moribund (should we call it dead?) era of untrammeled globalization.

Float Your Boat on the Rising Tide  

Back then – and it wasn’t that long ago – many people believed, perhaps naively, that a generic but universal form of global trade and commerce would mitigate, if not eliminate, the irritants that led to nation-on-nation aggression, conflicts, and wars. The rising tide of trade, it was assumed, would lift all boats –yachts, sailboats, cabin cruisers, trawlers, junks, skiffs, kayaks, canoes, and everything else remotely seaworthy.  It seemed to work for a while, then it didn’t. The law of intended consequences struck with a vengeance. Jobs were displaced in some geographies and created in others. More to the point, though, China decided that it wanted to be more than a manufacturing hub for the rest of the world; it wanted to move up the technological value chain, allowing its domestic companies to compete for global customers against those that dominated technology markets in the western world. We saw this development most notably in networking, with the rise of Huawei, but China wasn’t about to stop there. 

Can anybody credibly deny that we’re in another Cold War? Let’s just call it Cold War 2.0. It’s different from the last one, but similar in key respects. Every country in the world is being forced to pick a side, even as all participants, willing and otherwise, vaguely realize that the end game of this confrontation is deadly serious. It’s about more than nations, it’s about fortunes – those that exist today, and those still to be made. Some believe that those future fortunes, and the potential power that AI might deliver, will be unprecedented. 

With that in mind, let’s consider the Foreign Policy article from a slightly different perspective. Cohen has an employer, Goldman Sachs, which has a vested interest in making money indirectly from the proliferation of datacenters that host and propagate AI. I’m not saying that you should discount or dismiss everything Cohen writes in that Foreign Policy article – he makes some valid points, and there’s accuracy and truth in much of what he writes – but you need to consider the source. 

A source, remember, is always a principal actor in any drama; in their own abstracted way, sources always have a bias predicated on self-interest, sometimes declared but often veiled. Goldman Sachs wants to make money, lots of it, from AI-driven datacenter expansion, which it believes will be the next really big thing to reorder an increasingly valuable and politically influential technology industry. That’s an important element of context that you should keep in mind. Again, that context doesn’t serve to refute the contents of the article, but it should help you, as a reader, understand the author’s objectives or motivations, which are never openly stated in the piece he’s written. 

An unstated assumption of the article is that national and corporate interests are aligned and shared. In other words, Cohen assumes that a country and the corporations that call it home are strategically aligned and essentially complementary. 

Is that always true? During the age of globalization, perhaps it wasn’t. Corporations back then were single-mindedly concerned with making money, not so much with the stratagems and tactical feints inherent to geopolitics. We need to remind ourselves, in fact, that during globalization’s period of stratospheric expansion, driven by communications technologies and their underlying networks, China was still viewed, albeit warily, as a vast market opportunity for Western (and U.S.) companies, products, and services. Back then, companies didn’t want to be too closely identified with their country of provenance. Instead, they called themselves multinational, or transnational, corporations, cosmopolitan and worldly businesspeople at home wherever their investments and products were welcome.

Debunking the Fallacy of “Data is the New Oil” 

But again, so much has changed since then. We’re in a different time now. It’s still a time of business opportunity – depending, of course, on what sort of business you’re in – but it’s a more perilous time, too. We live in truly interesting times, and we know that’s both a blessing and a curse. 

All that aside, let’s now highlight excerpts from that article that I believe are entirely accurate and markedly significant:  

Data centers are the factories of AI, turning energy and data into intelligence. Industry leaders estimate that a few major U.S. technology companies alone are expected to invest more than $600 billion in AI infrastructure, particularly in data centers, between 2023 and 2026. The countries that work with companies to host data centers running AI workloads gain economic, political, and technological advantages and leverage. But data centers also present national security sensitivities, given that they often house high-end, export-controlled semiconductors and governments, businesses, and everyday users send some of their most sensitive information through them. And while the United States is ahead of China in many aspects of AI, especially in software and chip design, America faces significant bottlenecks with data centers.

Data is sometimes called the “new oil.” But there’s a crucial difference when it comes to data centers. Nature determines where the world’s oil reserves are, yet nations decide where to build data centers. And if the United States cannot break through bottlenecks at home, it will need an overflow option abroad. The possibility of a global AI infrastructure buildout presents an opportunity for governments and enterprises to practice data center diplomacy.

 Most of the above qualifies as rich food for thought, but I always chafe at the kneejerk cliché that “data is the new oil.” A cliché is always an instance of lazy thinking, of unthinking, and the old saw that “data is the new oil” is indolent to the point of being fossilized. Oil is energy, and it is, as Cohen notes, physically localized; but it’s also finite. There’s only so much oil – we don’t know how much, exactly, because the goalposts get moved every so often – and once we consume it, it is gone. Oil is not renewable. 

What about data? Well, data is as common as muck. I’m creating it now, you’re creating it every day in almost everything you do, especially when you’re online. Data multiplies exponentially by the minute, not just here, not just now, but everywhere and continuously. We’ve never had a commodity as superfluous and endlessly prolific as data. It truly never ends. 

A single random data point in our massive and ever-expanding sea of data is not worth much at all. Think about it: You wouldn’t buy random pools of data like traders buy West Texas Intermediate or North Sea Brent crude oil. If you did buy data in that manner, you’d probably get stiffed, stuck with washing-machine or smart-toilet telemetry. Of what value would that be to you? I suppose it would depend on whether you were in the washing-machine or smart-toilet business. 

The value of data, as it happens, does not derive from its existence in raw form. In fact, data has negligible value until it is filtered, mined, analyzed, and combined rigorously with other sets of data, ultimately resulting in a heretofore unavailable insight or actionable class of information. Data only attains value through systematic processing and careful correlation. By itself, data is unintelligent. It’s what we do with it, how it is handled and manipulated, that confers intelligence and value.

So, while Cohen is off to a good start in the prior excerpt, making his points cogently if not wholly originally, he sleepwalks and staggers off the path when he utters the “data is the new oil” canard. He gets back on course, however, as the following excerpt attests: 

Depending exclusively on projections in fields as dynamic and unpredictable as energy and technology is always risky. But with more and more intensive computation requirements, it’s clear that the expected energy needs for future data centers all point in one direction: up. Goldman Sachs Research estimates that data centers used three percent of U.S. power in 2022, a number that could reach eight percent by 2030. The Federal Energy Regulatory Commission’s staff expects data center usage to rise from 17 gigawatts (GW) in 2022 to 35 GW in 2030. The International Energy Agency predicts that global data center electricity consumption could double as soon as 2026, driven largely by AI. If data center power demand climbs from 460 terawatt-hours in 2022 to 1,000 terawatt-hours in 2026, as some expect, that growth would be roughly equivalent to the electricity consumption of Japan.

Moreover: 

Data centers are critical for the digital economy and AI. But the data center buildout is hitting a wall. The United States is home to the plurality of the world’s data centers, numbering in the thousands. Yet America’s aging energy grid, which powers those data centers, is under enormous strain from a complex set of factors, including rising electricity demand, delayed infrastructure upgrades, extreme weather events, and the complex transition to renewable energy. Meanwhile, surging data center demands driven by rapidly increasing AI workloads are exacerbating the grid’s vulnerabilities.

It’s not just a question of how those energy needs can be met, but where. When it comes to data centers, the shortage of powered land in the United States—or more specifically, the shortage of powered land with the connectivity required to support large-scale data centers—combined with supply chain challenges and lengthy permitting timelines for new infrastructure—presents a challenge to realizing both the public and private sectors’ AI ambitions.

Paradox of Power and Dependence 

The above excerpts constitute the crux of the piece, the point at which Cohen discusses the importance of trusted allied countries that can host and support datacenters securely, countries that also have abundant and diversified energy supplies that can meet the soaring electricity demands of current and next-generation AI processing. Many countries are mentioned as candidates for such expansion: Canada, the Nordic counties, Brazil, Japan, South Korea, and – garnering a significant share of text – the Arab Gulf counties, including Saudi Arabia and the UAE. The latter counties draw interest not only because of their energy abundance but also because of their extravagant wealth, which gives them the means to invest prodigiously in capital-intensive projects. 

The article concludes with the following call to action:

Navigating the geopolitics of this competition will require close partnerships between the public and private sectors. Not every country will be the world leader in AI. But more nations than the United States and China can lead. To win in today’s high-stakes geopolitical competition, the United States will need to enlist its asymmetric advantage of global alliances and partnerships, both in the public and private sectors. The data center buildout puts geography at the center of technological progress and competition. If the United States is successful, it is more likely that the future world, in which machines play a greater role in daily life, will also be one with greater human prosperity and freedom.

 When I read this final peroration, caveats and questions sprung readily to mind. Would, for example (and as mentioned earlier), an aggressive form of revanchist U.S. protectionism compromise the pursuit of datacenter partnerships with otherwise allied countries? Hitting other countries indiscriminately and maniacally with a tariff stick might not endear them to you, regardless of whether you’re holding a toothsome carrot in your other hand. 

Also, what would be the nature of the public and private partnerships relating to AI expansion and datacenter buildouts? Will public-sector capital be required? If so, how much? Or would public-sector participation pertain mainly to the relaxation of regulatory controls? If so, how relaxed will regulatory standards need to be? Perhaps most relevant of all, what’s in it for the taxpayers who contribute to the public-sector purse that might help to underwrite these public-private initiatives? What return do they get? Is it the promise jobs of? Is so, how many? 

In the past in this forum, I’ve discussed the criticality of energy to the sustenance and growth of the technology industry. Without energy, you have no technology industry. Without energy, in fact, you have no industry at all. Energy is the foundation, the stage, on which everything else plays out. 

Before AI, which has a voracious thirst for electricity, the availability and future supply of energy were not perceived as a problem. AI, however, guzzling electricity in massive volumes, changes the dynamic considerably. As a result of AI’s seemingly unslakable thirst for electricity and the tantalizing business prospects associated with the technology, a Venn diagram would now show a significant overlap between AI business objectives, global energy allocation, and oppositional geopolitical interests. 

It was inevitable that information technology would come of age, eventually taking its place near the summit of the hierarchy of industries. It’s nearly at the very peak now, but, as valuable and geo-strategically important as the technology industry undoubtedly has become, it still must importune the energy markets for the fuel that drives its growth.