Could Lack Of DOGE Hamper Countries From Capturing The AI Bounty?

Forbes

04-04-2025

Robot plugging into power.
ChatGPT generated image
Energy is often treated like a commodity. Priced. Traded. Measured in barrels, kilowatt-hours, or BTUs.
But that framing misses the bigger picture. Energy isn't just something we consume. It's the invisible foundation beneath every step of human progress. It fuels industry, powers cities, and increasingly, drives the digital revolution. Without it, prosperity doesn't just stall. It doesnt exist.
Over the last century, access to energy has reshaped the human condition. In 1990, around 1.9 billion people, or 35 percent of the global population, lived in extreme poverty, surviving on less than two dollars a day. Today, that figure is down to 782 million. That drop didn't happen by accident. Expanding access to electricity and fuel enabled industrial growth, agricultural productivity, and urbanisation—all pillars of economic uplift.
Now, we're entering a new chapter. One marked by rising population, accelerating technological breakthroughs, and surging demand for power. By 2050, an estimated 1.1 billion people will enter the global middle class. At the same time, the global population will approach 10 billion, with much of the growth concentrated in regions like sub-Saharan Africa.
As living standards rise, so will the need for energy-intensive goods, services, and infrastructure. The U.S. Energy Information Administration projects a 135% increase in global energy consumption by mid-century.
The systems that underpin this growth, whether economic, industrial or digital, are grounded in materials that require massive energy input. Vaclav Smil outlines four of these pillars in How the World Really Works: ammonia, steel, concrete, and plastic. These aren't optional; they are the non-negotiables of development. From feeding the world to building its infrastructure, they represent the baseline for progress. And producing them, at scale, still relies on fossil fuels.
At the same time, a new force is reshaping energy demand across the globe: artificial intelligence. Behind every large language model or real-time inference engine lies a dense web of data centers, chips, storage, and cooling infrastructure. These physical systems, often invisible to users, require tremendous amounts of electricity to function.
Electricity that must come from somewhere.
The International Energy Agency estimates that global data center energy use could rise by up to 128% by 2026, adding the equivalent of Sweden or Germany's entire energy consumption to the global grid.
In Ireland, data centers already use more than 20% of the country's electricity. In the U.S., data centers now account for more than 10% of total electricity use in at least five states. And S&P Global projects that U.S. data center usage will nearly double between 2024 and 2028, hitting 530 terawatt-hours, more electricity than Texas produced in 2022.
But where will they come from into the future?
Leading countries by number of data centers
Cloudscene, Statista,
Top 10 countries with electricity capacity
IRENA
The U.S. leads by a large margin, both in energy production and the concentration of AI data centers. However, as the data shows, the growing energy demands of AI and data centers in regions like the U.S. and China will strain existing infrastructure. While these regions have the capacity to generate the energy needed, many data centers still rely on outdated, inefficient systems, which will only exacerbate the challenge.
But will bureaucracy hinder countries from reaching their AI potential? Regulatory hurdles, delays in grid development, and challenges in resource allocation could slow efforts to meet the rising demand. Governments may struggle to keep pace with AI's rapid growth, impeding progress on critical infrastructure updates and the scaling of sustainable energy solutions.
As this demand surges, it won't be limited to tech giants. With AI becoming more accessible, energy demand will rise across industries. Yet much of today's infrastructure is ill-equipped for the task. Many data centers still rely on inefficient cooling systems and legacy hard drives.
Building the next generation of AI systems will require a shift toward more sustainable hardware: high-capacity SSDs, decentralised storage that taps into idle capacity, and liquid cooling systems that use less water and power. Companies like Nvidia are already leading the way with Blackwell GPUs, which are 25 times more energy efficient than their predecessors.
But efficiency introduces its own paradox. William Stanley Jevons observed in the 1800s that increased efficiency in resource use often leads to higher total consumption, a concept now known as the 'Jevons Paradox'. AI is already showing signs of following this pattern. When DeepSeek released its R1 model, it claimed to use 11 times fewer computing resources than Meta's Llama and cost just $6 million to train, compared to Llama's $60 million.
Yet that drop in cost sparked market concerns: would cheaper, more efficient models accelerate development, drive up energy use, and add further pressure to power grids? It's not guaranteed that every AI model will follow this path, but the trend is clear. As Microsoft CEO Satya Nadella put it, 'Jevons Paradox strikes again.'
Jevons Paradox
Twitter Satya Nadella
This tension between efficiency and demand underscores the complexity of today's energy challenge. Countries like India and China, home to nearly three billion people, are still in the process of building. Their industrial revolutions are far from over, and their need for reliable, affordable power is only accelerating. Shutting off fossil fuels isn't an option when the priority is lifting hundreds of millions into the middle class.
The systems driving global economic, industrial, and digital growth are built on materials that demand immense energy. The four pillars of modern life all rely on energy. So does innovation. From data centers to semiconductors, building the future takes power. Simply put, progress needs energy.
In that light, the countries best positioned to meet the AI-driven surge in energy demand may not be the ones with the most ambitious targets, but the ones with abundant energy resources and the ability for government approvals to move quickly. In the race to power the future, bureaucracy, not technology, might be the biggest bottleneck.