A new report reveals the colossal electricity demand of data centers in the coming decade, prompting investors to consider the infrastructure powering the digital gold rush, according to Jakob Blume.

The global energy landscape is undergoing a seismic shift, driven by the insatiable demand for power from the burgeoning artificial intelligence (AI) industry. Historically, electricity consumption has closely tracked the ebb and flow of the global economy. However, the escalating needs of tech giants and their AI ambitions have fundamentally decoupled energy demand from traditional economic cycles and even the broader energy markets. This burgeoning power requirement is set to reshape energy infrastructure and investment strategies worldwide.
The Unprecedented Surge in AI-Driven Energy Demand
Recent analyses, including projections from the International Energy Agency (IEA), paint a stark picture of this escalating demand. Between 2019 and 2025, the global demand for electricity is estimated to have grown at an annual rate of 3.2 percent. This growth outpaces the expansion of global Gross Domestic Product (GDP) and more than doubles the rate of increase in overall global energy demand. Crucially, this trend is not expected to plateau but rather to accelerate and persist well beyond 2035, according to the IEA’s forward-looking assessments.

This phenomenon represents what the IEA terms a "paradigm shift" in electricity consumption. The primary driver behind this surge is the exponential growth in the deployment and training of AI models. These complex algorithms require massive computational power, which is housed in vast data centers. These facilities, in turn, are immense consumers of electricity, not only for the servers and processors themselves but also for the extensive cooling systems needed to prevent overheating.
Data Centers: The New Epicenters of Energy Consumption
The sheer scale of AI development means that data centers, once a relatively niche component of the global energy infrastructure, are rapidly becoming major power consumers. Estimates suggest that by 2026, the electricity demand from data centers worldwide could reach a staggering 1,000 terawatt-hours (TWh) annually. To put this into perspective, this figure is comparable to the total electricity consumption of countries like Germany, France, and the United Kingdom combined, or even a significant portion of the entire European Union’s current electricity usage.
The training of large language models (LLMs), such as those powering advanced AI assistants and generative AI applications, is particularly energy-intensive. For instance, training a single large AI model can consume hundreds of megawatt-hours (MWh) of electricity, equivalent to the annual electricity consumption of several hundred households. As the development and application of AI continue to expand across various sectors, from healthcare and finance to transportation and entertainment, the demand for this computational power, and consequently electricity, is projected to continue its upward trajectory.
The Infrastructure Backbone: Power Grids and Network Equipment
The implications of this energy demand are far-reaching for the energy sector. Utilities and power grid operators are facing unprecedented challenges in ensuring a stable and sufficient supply of electricity. This necessitates significant investments in expanding generation capacity, modernizing grid infrastructure, and potentially exploring new energy sources.

Furthermore, the companies that provide the essential infrastructure for this digital transformation stand to benefit significantly. This includes not only power generation companies but also those involved in the manufacturing of grid components, transformers, and transmission lines. The demand for specialized cooling systems, advanced server hardware, and network equipment that can handle the immense data flows associated with AI will also see a substantial increase.
Investment Opportunities in the Digital Gold Rush
For investors, this burgeoning energy demand presents a compelling opportunity. The article suggests focusing on the "picks and shovels" of this digital gold rush, a metaphor for investing in the essential tools and infrastructure that enable the broader technological advancement. Rather than solely focusing on the AI companies themselves, which can be highly volatile, a more stable investment approach may lie in supporting industries.

This includes:
- Power Generation and Utilities: Companies that generate and distribute electricity, particularly those investing in renewable energy sources to meet the increased demand sustainably.
- Grid Infrastructure Manufacturers: Companies producing transformers, transmission lines, substations, and other critical components for modernizing and expanding power grids.
- Data Center Construction and Operations: Firms involved in building and managing the physical infrastructure for data centers, as well as those providing specialized cooling and power management solutions.
- Semiconductor and Hardware Manufacturers: Companies producing the specialized chips and hardware that power AI computations and data processing.
- Energy Efficiency Solutions: Businesses developing technologies and services to optimize energy consumption within data centers and across the wider industrial landscape.
The article highlights how traditionally "boring" utility and network equipment providers are being transformed into growth companies due to the AI boom. This shift indicates a fundamental reevaluation of the value proposition of these companies in the context of the evolving technological landscape.

The Race for Sustainable Energy Solutions
The environmental implications of this energy surge are also a critical consideration. While AI offers immense potential for innovation and efficiency, its energy footprint cannot be ignored. The IEA and other organizations are emphasizing the need for a sustainable approach to meeting this demand. This involves a concerted effort to:
- Accelerate Renewable Energy Deployment: Increasing the share of solar, wind, and other renewable sources in the energy mix is crucial to power data centers with cleaner electricity.
- Enhance Energy Efficiency: Developing and implementing more energy-efficient computing hardware and data center designs can significantly reduce overall consumption.
- Explore Advanced Cooling Technologies: Innovations in cooling systems, such as liquid cooling and free cooling, can help reduce the energy required to maintain optimal operating temperatures.
- Carbon Capture and Storage: For unavoidable emissions, technologies for carbon capture and storage may play a role in mitigating the environmental impact.
The challenge lies in balancing the rapid growth of AI with the urgent need to address climate change. This requires coordinated efforts from technology companies, energy providers, policymakers, and investors to ensure that the digital revolution is also a green revolution.

Timeline of Emerging Trends
- 2019-2025: Significant increase in global electricity demand, outpacing GDP growth, driven by early AI adoption and digital transformation.
- 2025-2030: Projections indicate a continued acceleration of AI-driven energy demand, with data centers becoming major power consumers. Increased investment in grid modernization and renewable energy infrastructure.
- 2030-2035 and beyond: AI’s energy appetite is expected to remain a dominant factor in global energy markets. Focus shifts towards sustainable energy solutions and advanced efficiency technologies to manage the long-term energy requirements.
The IEA’s reports provide a consistent outlook, emphasizing the growing importance of electricity in the global energy system, largely fueled by digitalization and AI. Their World Energy Outlook, for example, regularly analyzes these trends, providing crucial data for policymakers and industry leaders.
Broader Impact and Implications
The increasing reliance on electricity for AI computation has several broader implications:

- Geopolitical Shifts: Countries with abundant and affordable energy resources, particularly those with a strong renewable energy sector, may gain strategic advantages. Energy security will become even more intertwined with technological advancement.
- Economic Restructuring: Industries that are heavily reliant on AI will experience significant growth, while those that fail to adapt may face decline. This will lead to shifts in employment and economic activity.
- Regulatory Challenges: Governments will face the challenge of regulating AI development and deployment in a way that balances innovation with energy sustainability and environmental protection. This may include incentives for energy-efficient AI and regulations on data center energy consumption.
- Technological Innovation: The immense energy demand will spur innovation in energy generation, storage, transmission, and computing efficiency. This could lead to breakthroughs in areas like advanced battery technology, fusion power, and novel computing architectures.
The article’s perspective, highlighting Jakob Blume’s suggestion to invest in the "picks and shovels" of the AI gold rush, underscores a pragmatic approach to navigating these transformative trends. By focusing on the foundational elements of the digital economy – energy infrastructure and related technologies – investors can potentially capitalize on the long-term growth driven by AI, while contributing to a more sustainable energy future. The challenge ahead is to harness the power of AI responsibly, ensuring that its rapid expansion does not come at an unsustainable cost to the planet.







