We are watching a slow-moving trainwreck in the technology sector, and it is happening right in our own backyards. The technology industry has long operated under the assumption that if it built the infrastructure, municipalities would simply welcome it for the tax revenue. That era is officially over.
Major tech companies have essentially wagered over a trillion dollars on the aggressive expansion of artificial intelligence, which requires an unprecedented build-out of physical hardware. Yet communities are increasingly pushing back against proposed data center developments that represent billions of dollars in investment.
What we are witnessing is a growing clash between the demands of AI infrastructure and the communities expected to host it. The industry assumes it can simply buy its way into local municipalities, but executives are severely underestimating public concern over the strain these facilities place on local resources.
A data center is no longer just a quiet warehouse filled with blinking lights; it has become an industrial-scale consumer of the very utilities that keep towns functioning. If the tech sector does not completely rethink how it approaches these projects, that trillion-dollar wager will go up in smoke, stalled by local zoning boards, angry town halls, and legitimate environmental concerns.
Let’s talk about the data center expansion problem and potential solutions. We’ll close with my Product of the Week, a modular data center design that addresses one of the big problems we’ll be discussing: the premature obsolescence of data centers.
Why Communities Are Pushing Back
To understand why residents are pushing back so fiercely, you have to look at the raw physics of modern artificial intelligence. The transition from traditional cloud computing to AI-driven workloads has fundamentally altered the infrastructure profile of the standard data center. The energy and water demands associated with today’s facilities are becoming a serious challenge for regional infrastructure.
Modern GPUs — the silicon engines driving AI — consume enormous amounts of electricity. A single rack of these servers can consume more power than an entire city block. When a technology giant drops a hyperscale facility onto a rural or suburban grid, it effectively hijacks the regional power supply. This demand can drive up utility rates and increase the risk of rolling blackouts.
Electricity is not the only issue. Water consumption is arguably even more contentious. To keep these dense server racks from melting down, facilities utilize industrial-scale evaporative cooling towers. A single mid-sized AI data center can guzzle millions of gallons of potable drinking water every day.
In areas already prone to drought, this is a political non-starter. You simply cannot tell a farming community or a residential suburb to restrict their water usage while a windowless concrete box down the street evaporates a small lake to train a language model.
This consumption model is driving communities to organize, protest, and push for tighter restrictions on these projects.
When Data Centers Age Too Fast
One of the most profound mistakes vendors make is failing to align their construction timelines with the rapid pace of their innovation. The rate of technological change in the AI hardware sector is moving at breakneck speed. We are seeing new generations of silicon released every 12 to 18 months, with each generation requiring entirely different thermal management and power delivery architectures.
The problem is that a hyperscale data center typically takes three to five years to permit, construct, and bring online. Do the math. By the time the ribbon is cut and the facility is powered up, the hardware it was designed to house is already multiple generations behind. Companies are building rigid facilities that will likely be obsolete or far too energy- and water-intensive to remain profitable when they are finally completed.
If a building is optimized for traditional air cooling and moderate power density, it simply cannot easily support the liquid-cooled, ultra-dense architectures required by tomorrow’s AI chips. When these facilities inevitably require costly retrofits, the economic model collapses.
We are headed toward a landscape littered with stranded assets: billion-dollar concrete husks sitting empty because they are too inefficient to run the latest hardware and too expensive to tear down. This costly engineering mistake is avoidable if architects and tech executives stop building static monoliths in a highly dynamic industry.
Redesigning for Survival and Modularity
The solution to both the community backlash and the obsolescence trap requires a total paradigm shift in data center design. We must stop treating these facilities like traditional commercial real estate and start treating them like modular, upgradeable machines.
First and foremost, data center design needs to become completely modular. Builders must hold off on their final technology selection until the very last phase of construction.
Build the outer shell, lay the foundational power and networking grids, but do not commit to specific cooling layouts or server configurations until months — not years — before deployment.
This “just-in-time” hardware integration ensures that the site is equipped to handle the state-of-the-art tech available at launch, rather than the tech that was merely popular when the permits were filed.
Furthermore, functionality needs to radically evolve to make these facilities actively resource-friendly. The era of the purely extractive data center must end. Future builds must incorporate their own local energy generation. Whether it is co-locating with dedicated solar farms, geothermal plants, or even looking toward next-generation small modular reactors (SMRs), data centers must bring their own power to the table rather than indiscriminately draining the public grid.
Equally critical is the necessary transition to completely closed-loop cooling systems. While traditional evaporative cooling might be cheaper upfront, its political and environmental costs are no longer sustainable for modern builds.
Closed-loop systems recycle the same liquid continuously without relying on substantial daily withdrawals from municipal water supplies. By eliminating the drain on local aquifers and minimizing grid impact, tech companies remove the two largest targets off their backs, making it significantly harder for communities to justify blocking their permits.
Making Data Centers Community Assets
If the industry can fix the resource drain, the next logical step is to explore what kind of services data centers could offer that would more than offset their remaining footprint. Right now, data centers are viewed entirely as extractors. To win over the public, they must pivot and become dedicated providers.
Consider the vast amount of waste heat generated by thousands of servers. In parts of Europe, we are already seeing forward-thinking facilities capture this thermal energy and pump it into municipal district heating networks. A single large-scale data center could effectively provide free winter heating for nearby schools, hospitals, and residential neighborhoods. The facility is no longer just a commercial burden; it becomes a source of tangible community benefits.
Additionally, data centers can serve as microgrid anchors. By installing on-site battery storage for their own backup needs, tech giants could offer vital grid stabilization services to the local municipality. During peak summer hours, when brownouts threaten residential areas, the data center could export excess stored power back to the public grid to help prevent blackouts and protect local infrastructure.
Having a high-capacity computing hub right in town allows for unprecedented local edge computing services. Municipalities could be granted dedicated, free access to local processing power to run complex smart-city logistics, optimize traffic flow in real-time, or improve emergency response coordination. By transforming from a passive consumer into an active utility partner, the data center ceases to be a localized threat and becomes a critical component of civic infrastructure.
A Better Growth Model
Picture what a community might look like if it successfully integrated one of these next-generation, highly efficient data centers. Instead of a fenced-off corporate fortress, the facility operates in complete harmony with the town.
The immediate benefit is economic growth. With the tech company bringing its own renewable power generation and closed-loop cooling, the town reaps significant commercial tax revenue without having to fund expensive expansions to its own water treatment plants or power substations. Those tax dollars can be directly injected into upgrading local schools, paving roads, and reducing property taxes for working families without compromising municipal solvency.
Environmentally, the community enjoys free, clean district heating pumped directly from the server racks into the local public pool, community centers, and municipal buildings. The local power grid is actually more resilient than before, stabilized by the tech giant’s large battery arrays rather than strained by their operations.
The presence of advanced infrastructure naturally attracts secondary technology businesses, creating a hub of high-paying local tech jobs that keep young talent from fleeing to the coasts.
This is not an impossible utopian dream; it is merely a matter of strategic architecture and corporate willingness to engage thoughtfully. The technology to build this future already exists. The only thing standing in the way is the tech industry’s stubborn adherence to outdated, extractive business models that blindly treat local communities as acceptable casualties in the modern AI arms race.
Wrapping Up
The trillion-dollar bet on artificial intelligence cannot be won if the physical infrastructure required to run it is constantly bogged down by local resistance. The current approach — building rigid, resource-devouring monoliths — is fundamentally failing. Communities are right to protect their water and power resources from projects that place significant new demands on local infrastructure.
If vendors and hyperscalers want to successfully deploy their next-generation technologies, they must embrace a radical shift in design strategy. They must shift to modular architectures that avoid becoming outdated before launch, mandate closed-loop cooling and local power generation to neutralize environmental impacts, and begin earnestly sharing their technological dividends with the host towns.
The technology sector must become a partner rather than a parasite. The companies that adapt will secure their operational future; those that do not will find their ambitious road maps blocked by a wall of angry citizens and their permits permanently revoked. The choice is up to them.
Vertiv Prefabricated Modular Data Center
The current hyperscale data center model is on a direct collision course with reality. We cannot build inflexible concrete monuments to technology, indiscriminately drain the local grid and water table, and expect communities to quietly step aside.
Furthermore, taking three to five years to build a traditional center meant to house AI silicon that evolves every 12 to 18 months is an absolute recipe for creating billion-dollar stranded assets.
If the industry wants to continue expanding AI infrastructure, it has to fundamentally change how it builds. That is why my Product of the Week is the Vertiv Prefabricated Modular Data Center.
Vertiv has essentially realized that data centers should no longer be treated as traditional commercial real estate projects. Instead, they need to be treated as highly engineered, mass-produced, and rapidly deployable appliances. Vertiv’s prefabricated approach directly addresses two of the industry’s biggest challenges: infrastructure longevity and deployment impact.
In a traditional “stick-build” construction process, tech giants are forced to commit to a specific thermal management layout and power delivery architecture years before the first server rack is even plugged in. By deployment, much of the planned infrastructure is already lagging behind current hardware requirements. Vertiv flips this model on its head.
By using an integrated modular solution (IMS), the core infrastructure — including high-density power modules, liquid-cooling loops, fire suppression, and structural enclosures — is built, integrated, and fully tested in a controlled factory environment.
This parallel processing means the site preparation and the technology build happen simultaneously. Instead of enduring years of disruptive heavy construction that angers local zoning boards and residents, the facility is shipped to the site in configurable modules and assembled in a fraction of the time. The final technology selection is delayed until right before deployment, ensuring the facility matches the current silicon generation.
More importantly, this modularity directly addresses the environmental friction causing communities to revolt. Vertiv’s designs are inherently optimized for advanced, high-density AI workloads, meaning they integrate seamlessly with closed-loop liquid cooling systems and high-efficiency power distribution straight from the factory. This avoids the excessive evaporative water waste that is currently driving municipalities to block permits.
Because you can scale these modules dynamically, a company installs only the capacity it needs today, rather than overbuilding a huge, power-hungry footprint for projected demand. This kind of “just-in-time” hardware integration helps keep the data center resource-friendly and politically viable.
This isn’t just a white-paper concept. It is actively being deployed by major global infrastructure players who understand that agility is their only defense against rapid technological change.
Modular Data Centers in the Real World
A prime example is the European IT services giant T-Systems. When it needed to rapidly expand its cloud infrastructure in Barcelona to support advanced workloads, it bypassed the traditional 24- to 30-month build cycle and partnered with Vertiv to deploy a 1.1 MW modular facility. Vertiv designed, built, and tested 38 integrated modules in its dedicated plant in Croatia, shipped them via flat-racks to Spain, and had the site operational in just nine months.
T-Systems was able to drastically reduce its electrical consumption, achieving a highly efficient power usage effectiveness (PUE) of 1.3, while keeping the architecture scalable up to 5 MW for later phases.
We are also seeing this technology deployed at the very edge of global connectivity. Vertiv has used these prefabricated containers to build high-capacity submarine fiber-optic cable landing stations in West Africa and Finland, as well as a Tier III-compliant modular facility for a major telecommunications provider in South Africa.
In these scenarios, the modules are custom-engineered to withstand harsh marine or remote environments, providing enterprise-grade thermal management and power protection without requiring a large on-site construction footprint.
The trillion-dollar wager on AI cannot succeed if the physical infrastructure required to run it is constantly bogged down by local resistance and premature obsolescence. The Vertiv Prefabricated Modular Data Center represents the necessary evolution of infrastructure. It turns a risky, multi-year real estate gamble into a predictable, agile, and highly efficient technology deployment.
Because it solves the most critical pain points of modern AI scale-out while actively mitigating community disruption, Vertiv’s Prefabricated Modular Data Center is my Product of the Week.
The images in this article were created with AI.
