In its simplest definition, geopolitics is a discipline that seeks to understand the interaction between geography and politics, as well as the impact of this interaction on the distribution of power among actors. Within this framework, the analysis encompasses the territories occupied by states, their borders, geographic locations, access routes to seas and critical land areas, topographical features, potential for accessing energy and raw materials, and their straits, ports, and trade routes. Strategic values ​​that rise to prominence in different eras redefine the significance of geographies, reflecting the fluid nature of geopolitical importance. A look at historical developments reveals that value has shifted over time—sometimes centering on salt or spice routes, at other times on the quest for colonies, access to coal or oil, or the control of trade routes and military logistics capabilities. Every era possesses a strategic asset that must be secured, and political power struggles are shaped around that asset.

Indeed, the global arena—which witnessed two major world wars in the first half of the 20th century—was defined primarily by the prominence of coal and steel production and access to oil. The struggle to reach energy hubs, which intensified from the beginning of the century, cost millions of lives; the objective was not the land itself, but the treasure hidden deep beneath its surface. During the post-1945 Cold War era, the Middle East’s enticingly rich energy resources propelled the struggle between ideological blocs into a new phase, shifting the battlefronts from the West to these regions. As the 21st century dawned, the Middle East and the Caspian Basin served as the two hearts pumping blood into the global organism. The essence of geopolitical theory was encapsulated in the famous dictum of Halford Mackinder, one of the field's founding fathers: "Who rules the heart of the world commands the World-Island—that is, the world itself." With the turn of the century, new dynamics born of technological advancements began to alter perceptions of time and space.

Thomas Friedman’s slogan, "The World Is Flat," seemed to hold the key to understanding this new way of life. Capital, information, and global services traversed geographies in seconds, while intercontinental flows reached unimaginable scales. Globalization accelerated the mobility of goods and people; the rise of cyber systems and borderless military capabilities gave rise to talk of "de-territorialization." I use the past tense here because, at our current juncture, we find ourselves in a period where the significance of geographical territories and the intensity of geopolitical struggles have once again surged to the forefront.

Under normal circumstances, the new world of technology and the era of artificial intelligence in which we live might have been expected to further accentuate the "geography-free" nature of cyberspace. Yet, however abstract, boundless, and location-independent artificial intelligence may appear to us, its existence and infrastructure are equally physical and dependent on geography. Running large models requires data centers; feeding them demands vast amounts of electricity; cooling the systems necessitates cold water and a suitable climate; chip production relies on critical minerals; data transmission depends on undersea cables; and protecting all of this requires political and military security. In short, artificial intelligence is not extraterrestrial but terrestrial; it is a subject of political competition among nations—too deeply embedded in this context to be viewed merely as a technological issue. It is therefore evident that, in the near future, it will be a factor redefining the geostrategic value of geographies across the globe. Consequently, we might seek the answer to the question "Who will win the artificial intelligence race?" in another question: What kind of world map does artificial intelligence require?

According to the International Energy Agency, data centers consumed approximately 485 TWh of electricity in 2026—a 17 percent annual increase. This figure is projected to reach 945 TWh by 2030, with the electricity consumption of AI-focused data centers expected to more than triple over the next five years. In short, we are no longer talking merely about the electricity bills of technology companies; regions with high concentrations of data centers have become high-risk areas due to the excessive load they place on local power grids. These grid issues could escalate to a level where they profoundly impact daily life, industrial production, logistics, and security in the near future. As each of us becomes increasingly dependent on artificial intelligence, it is evident that this consumption—and the associated risk—will grow exponentially.

For precisely this reason, when we look at the system as a whole, the extraordinary importance of generating abundant, cheap, and uninterrupted electricity in this new era becomes apparent. Electricity generation is no longer merely a technological issue; it is also a geopolitical one. This is one of the fundamental reasons why nuclear energy has returned to the agendas of governments. The imminent deployment of Small Modular Reactors (SMRs) appears inevitable. Indeed, Google’s 2024 long-term agreement with Kairos Power to procure SMR-generated electricity marks a significant milestone in this process. Similarly, Microsoft’s 20-year nuclear-based electricity agreements with Constellation Energy, Amazon’s SMR investments with various companies, and Meta’s nuclear energy contracts all highlight the intense interaction between the tech world and the energy sector. A second critical issue regarding artificial intelligence is access to water and cold environments. As the heat generated by AI processors rises, cooling technology is becoming one of the greatest physical challenges for data centers.

Consequently, a natural resource previously largely overlooked on the global map is gaining strategic value: cold. The Earth’s northern regions have gained importance not only due to traditional energy sources but also because low temperatures offer advantages in terms of cooling costs. The US interest in Canada and Greenland is, therefore, not without reason. Major powers that spent centuries trying to reach warm seas are now seeking to expand into cold seas and regions. The third strategic arena in the landscape of artificial intelligence comprises the regions where critical minerals are found. After all, artificial intelligence is not merely a matter of software; it relies on a vast supply chain of raw materials—ranging from copper to silicon, lithium to gallium, and germanium to rare earth elements. Distinct minerals and ores are required for chips, electrical infrastructure, and energy storage technologies alike. China’s extraordinary advantage in the supply and processing of these materials is viewed—particularly by the US—as one of the most significant national security risks. In short, the new geopolitics is being rewritten in an entirely new language.