For more than a century, the industrial economy revolved around a relatively simple principle: whoever controls energy controls a decisive share of development. Oil, electricity, telecommunications networks, logistics systems and strategic infrastructure have shaped the geography of global economic power. Today, that same logic appears to be approaching a new historical transition. The focus is no longer limited to the distribution of physical energy. It increasingly concerns the distribution of cognitive capacity.
A statement made by Sam Altman during the BlackRock Infrastructure Summit carried the distinctive weight of ideas that initially sound like a business model and only later reveal the scale of the transformation they imply. The CEO of OpenAI suggested a future in which “intelligence will be a utility, like electricity or water, and people will buy it on demand.” Behind that seemingly technical observation lies a profound shift affecting economics, industrial policy, labor markets and the very nature of competition among organizations.
The novelty is not simply the spread of artificial intelligence. What is emerging is a fundamentally different way of understanding reasoning itself as an economic resource. For centuries, organizations purchased machines capable of amplifying physical power. They are now beginning to purchase systems capable of amplifying analysis, decision-making, research, design and knowledge production. Artificial intelligence is no longer presented merely as software to install or applications to use. It is increasingly becoming a continuous infrastructure to which individuals and organizations connect. The logic mirrors that of the electrical grid: ownership of the power plant is unnecessary as long as access to the flow is guaranteed.
This transition fundamentally alters the nature of competitive advantage. In many industries, differentiation will depend not only on financial capital, physical assets or intellectual property, but also on the quantity and quality of cognitive capacity that can be acquired and integrated into business processes. Companies are already experimenting with this reality through token-based pricing, tiered reasoning capabilities, premium models and autonomous agents. Intelligence is progressively becoming measurable, scalable and billable. It is no longer simply a tool. It is becoming an industrial input.
The managerial implications are profound. If artificial reasoning becomes a purchasable commodity, corporate governance will face an entirely new challenge: managing intelligence as a productive resource. The issue will not merely be whether to deploy AI systems, but how much cognitive capacity to acquire, where to allocate it, which functions to automate and which responsibilities to retain under human control. In such an environment, human capital does not disappear. Its role evolves. Value shifts toward the ability to formulate superior questions, supervise autonomous systems, validate outcomes and assume forms of responsibility that algorithms cannot bear in legal, ethical or reputational terms.
Several major organizations already illustrate the direction of travel. Microsoft has embedded Copilot across its productivity ecosystem with the explicit goal of making artificial intelligence a structural component of everyday work. JPMorgan Chase employs AI across activities ranging from document analysis and compliance to risk management. In both cases, the transformation extends beyond efficiency gains. What is occurring is the gradual incorporation of computational reasoning into organizational decision-making. Intelligence ceases to be an exclusively human capability and becomes a hybrid resource distributed between people and technological infrastructure.
The issue becomes even more consequential when viewed through a geopolitical lens. The comparison with electricity and water is not neutral. Utilities are not merely services. They are strategic infrastructures. Those who control them exercise economic, political and social influence. If artificial intelligence becomes a global utility, the central question extends beyond innovation to the realm of sovereignty.
The United States currently enjoys a dominant position through its control of key elements of the technological ecosystem, including advanced semiconductors, cloud platforms and many of the world’s leading AI models. China is investing heavily to reduce external technological dependence and establish a national AI infrastructure. Europe, meanwhile, is attempting to preserve strategic autonomy through regulation, industrial policy and investment. Beneath the public debate surrounding artificial intelligence, a new competition is emerging for control over cognitive supply chains: the systems that produce, process and distribute knowledge on an industrial scale.
The numbers offer a useful perspective on the scale of the transition. Leading technology companies are committing hundreds of billions of dollars to AI infrastructure, data centers and computational capacity. OpenAI itself has discussed infrastructure ambitions of unprecedented magnitude to support future demand. At the same time, energy requirements continue to rise. International studies indicate that advanced AI data centers may consume energy comparable to that of entire cities while requiring substantial quantities of water for cooling operations.
This reality reveals a significant paradox. The digital economy is often perceived as intangible, yet it increasingly depends on highly tangible resources: electricity, water, rare earth minerals, semiconductor manufacturing capacity and geopolitical stability. Artificial intelligence may appear as software, but every generated response is supported by an industrial ecosystem connecting mines, power plants, supply chains, global networks and multibillion-dollar infrastructure investments.
The deeper transformation, however, concerns the nature of power itself. For years, the digital economy monetized attention, data and advertising. The model now taking shape seeks to monetize cognitive processing directly. What is being purchased is no longer merely access to information. It is the capacity to interpret information, synthesize it and transform it into operational decisions. The distinction is substantial. Once reasoning enters the marketplace as an industrial service, the quality of economic decisions increasingly risks becoming dependent on access to the infrastructures that provide it.
For businesses, this creates extraordinary opportunities alongside equally significant responsibilities. Artificial intelligence can enhance productivity, accelerate decision-making and expand innovative capacity. Yet it can also reinforce concentrations of power, deepen technological dependencies and widen competitive asymmetries. The fundamental question is not whether AI will replace human beings. The more important question concerns who will control the environments in which people and organizations make decisions.
Viewed from this perspective, the utility metaphor acquires a meaning far beyond technology. It does not merely describe a commercial model. It describes the possible emergence of a new infrastructure of economic civilization. As with electricity in the early twentieth century, the true transformation does not lie in the technology itself, but in the moment it becomes invisible, pervasive and indispensable. When that occurs, power belongs not only to those who use the resource. It belongs, above all, to those who own the network through which it flows.