The Future at the Heart of Data Centers: Between Deer and Megawatts

Iowa, the Silicon Valley of Cornfields

In 2007, Google chose Iowa as the base for a major data center complex, starting with a strategic investment in Council Bluffs. Over the years, the facility has been expanded several times, with investments surpassing $6.5 billion by 2024. In terms of marketing, Google promotes an image of harmony between technology and nature, symbolized by deer roaming near the cooling towers. However, this narrative contrasts with the actual impact on the land: deforestation to make way for facilities and intense pressure on water resources are just some of the consequences. The Cedar Rapids project, for example, received significant tax breaks, with local authorities offering a 70% exemption on a $56 million investment. Despite promises of job creation, the new center will provide stable employment for only 31 people, confirming that economic benefits are often overstated compared to environmental and social costs.

Energy and Power: The Oracle of Omaha and the Rise of the Supercluster

Warren Buffett, through his subsidiary MidAmerican Energy, has played a key role in making Iowa a hub for tech companies. As early as 2014, Buffett and his collaborator Greg Abel highlighted how Google’s data centers represented an opportunity for the state’s energy development. Abel, later designated as Buffett’s successor, emphasized the importance of MidAmerican’s energy production in supporting these complexes. In 2024, during an annual meeting, Abel predicted that artificial intelligence would double energy demand by 2035. These projections fuel the idea of building a 100 GW supercluster, which would require industrial mobilization comparable to that needed to sustain an entire nation’s economy. This project, however, raises questions about the capacity to manage the necessary resources and its long-term environmental impacts.

Artificial Intelligence Has a Cost: The Carbon Footprint of the Future

The Scaling Law and the Cost of Digital Brains

The evolution of artificial intelligence is based on a simple yet costly principle: the greater the computational power, the more sophisticated the applications developed. This dynamic, referred to as the “scaling law,” ties technological progress to the expansion of computational infrastructure. In 2022, global data centers consumed approximately 460 TWh of energy, equivalent to 2% of global electricity demand. With the growth of generative AI applications like ChatGPT, this figure could rise to 1,050 TWh by 2026, reaching levels comparable to Japan’s total consumption. AI models not only require energy during training but also for daily operations, making a significant increase in carbon emissions inevitable.

When the Cloud Becomes a Battlefield: Energy and Geopolitics

The competition between the United States and China over artificial intelligence is an extension of their geopolitical rivalry. While the U.S. maintains a dominant position in technological research, it struggles to keep pace with China in building energy infrastructure. As Beijing constructs power plants and distribution networks at record speeds, the U.S. faces regulatory constraints that slow progress. Analysts like Leopold Aschenbrenner argue that the ability to rapidly build computing infrastructure and ensure its energy supply will be decisive in determining who leads the digital future.

An Endless Summer or a Winter Ahead?

The Paradox of Expansion: More GPUs, Fewer Humans

Data centers represent an economic paradox: they generate enormous profits for a few tech giants but create limited job opportunities. Automation in management systems, combined with hardware efficiency, reduces the need for human workers. For instance, Microsoft’s data center cluster in West Des Moines employs only 300 people, a number expected to double in the coming years. However, the contribution to employment remains marginal compared to the environmental and infrastructural impact of these facilities.

The Green Rush: Reality or Greenwashing?

Tech companies promote sustainability initiatives to address growing criticism of their environmental impact. Google, for example, announced a “water-positive” strategy to return more water than it consumes by 2030. However, these initiatives fail to address systemic problems. In Ireland, data centers already consume 18% of the national electricity supply, a figure set to rise to 32% by 2026. In Arizona, water consumption for cooling servers is exacerbating an already critical water crisis, while in Taiwan, the expansion of the tech sector heavily relies on fossil fuels.

Artificial Intelligence: Between Deer and Digital Brains

A Global Competition

The production and assembly of components for data centers remain concentrated in Asia. Taiwan, through TSMC, dominates the production of advanced semiconductors, while companies like Foxconn and Quanta handle server assembly. In the U.S., tech companies are trying to reduce dependence on these supply chains by building new infrastructure, such as Microsoft’s Rutenium Project in Iowa. However, the complexity and speed of the Asian supply chain remain difficult to replicate in the West.

We Are the Real Deer

The idealized depiction of deer roaming near data centers masks the reality of humanity’s increasing subordination to technology. The expansion of digital infrastructure transforms local territories and communities, often without meaningful public involvement in decision-making processes. In this context, humans risk becoming passive spectators of a system that outpaces and marginalizes them.

Conclusion: A Crossroads Between Progress and Sustainability

The expansion of data centers poses a global challenge requiring a delicate balance between innovation, environmental sustainability, and social equity. If technology continues to grow without considering the limits of natural resources and the needs of local communities, progress risks becoming a source of inequality and environmental destruction. The challenge for the future is not just building more powerful infrastructure but making it compatible with an increasingly fragile planet.

Tags

data centers, artificial intelligence, energy consumption, scaling law, geopolitics, USA-China rivalry, sustainability, greenwashing, environmental impact, automation, employment, technology, digital infrastructure, semiconductors, Taiwan, Iowa, Microsoft, Google, water resources, renewable energy, economic impact