Why are data centers hated? Tech giants need to offer a new narrative for rural areas.

CN
3 hours ago
Disentangled the predicament of the data center narrative.

Author: Hip City Reg

Translation: Shen Chao TechFlow

Shen Chao Guide: Tech companies are crazily building data centers in rural America, yet they face a 70% community opposition rate and over 300 local bans. Architect Rem Koolhaas's observations reveal the truth: cities have become playgrounds for humans, while the countryside has turned into a habitat for machines— and those "coastal elites" never care about the communities they transform. This article dissects the dilemma of the data center narrative and points out why simply "designing buildings to look better" won't save the situation.

Introduction

I have been pondering the relationship between data centers and the overall mentality in America, particularly the mild anti-tech sentiment I believe we are currently experiencing. I keep returning to the grand narrative that seems inherent to data centers—whether from the perspective of techno-accelerators or from a broader discontented American populace.

I would summarize the point of view from the techno-accelerators like this:

In the AI-related Cold War-like competition, data centers are essential.

Data centers create positive externalities, as they both generate demand for construction jobs and energy production.

Data centers are the infrastructural bottleneck to cheap intelligence, ultimately creating enormous consumer surplus (curing cancer, new physics, etc.).

"This is necessary for progress."

I would summarize the perspective of discontented Americans like this:

AI is both foolish and stealing our jobs; why should I support the infrastructure that keeps it alive?

Loss of jobs = loss of the American sense of dignity.

Why should data centers be built "near me" and in my community, rather than anywhere else?

They represent services that point to wealth accumulation in coastal areas—this is an extraction from their community.

Data centers consume vast amounts of water and raise our energy bills.

This is just another form of control by the tech industry.

(I found Soon Media’s investigation into Abilene, Texas, very compelling—link.)

Whether these arguments are real is not important. Whether they are made in good faith is also irrelevant. The reality is that states are issuing halts on data center construction. Local communities are rallying to express collective opposition to proposed projects. (In their own communities, there is a 70% opposition rate, and over 300 have bans or moratoriums.) The overall sentiment surrounding AI, its leaders, and the promised future has already soured terribly.

So what should we do?

Image: Comparative impact of water, land, and electricity prices for data centers in the U.S. (Source: Hip City Reg / Crémieux Recueil mapping)

Image: Rem Koolhaas’s perspective on human-machine landscapes (Source: Hip City Reg)

Image: Rem Koolhaas’s observations on cities and countryside (Source: Hip City Reg)

Rem's Countryside

I have largely lost interest in writing general papers, so the following content (like some of my previous personal favorite articles) is just a series of bullet points. I hope they are sufficient to outline the contours of this idea. The main reference that sparked this thought is Rem Koolhaas’s work "Countryside."

Rem's career (like many architects) has been built upon the city. "Delirious New York" (1978), OMA (1975). He then spent about a decade (2012–2020) arguing that the architectural and cultural obsession with urbanization has become a form of blindness.

"Countryside"—his deliberately broad term for rural, remote, desolate, and wild territories, which occupy 98% of the Earth's surface—is not the stable and nostalgic backdrop we imagine. It is where the most radical transformations of our time truly take place: automation, climate change, immigration, genetic experimentation, political radicalization, and the physical infrastructure of the digital economy.

The countryside has long been neglected. And this is an anomaly of the 20th century. For most of history, from Jefferson to Mao, the countryside has been a major political project. But urban victory (victimhood) has upended this.

Going beyond the framework of America’s "coastal elites," we have come to view urbanization as the global statistics of progress (and overwhelming imagery). This not only provided a statistical assertion to dismiss the countryside but also redirected it to serve the growing urban centers.

However, ordinary urbanites do not see it this way. On the contrary, they have a fictional imagination of the countryside.

One is a pastoral, slow, genuine, unchanging leisure economic view. Rem illustrates this with his own observations of a Swiss village that lost its original way of life as it scaled up.

In his Icon article, he labeled the actual situation as an "intermediate state": a curated surface that bears almost no relation to what is happening on the land and within buildings. He insists that the countryside is "now the frontline of transformation" / "more tumultuous than the fastest-accelerating cities."

The other view is more hostile. Perhaps as Wendell Berry says, "the prejudice against country people." I find this crucial for understanding the latent psychology of data centers. Essentially, it is a belief that “putting these buildings around those people doesn't matter.” Because the coastal elites (those who build AI and push it as the engine of progress) generally do not look at “those flying over states” with a refined love for their culture and community.

Rem believes that the countryside is increasingly governed by Cartesian order. Gridded, mechanized, optimized for maximized production. In "The Cut," he clarifies and assigns causation to this mechanism: because cities have become a "massive rebellious machine"—focusing on fun, pleasure, and interaction rather than demand— we "have to install a hyper-Cartesian organizational infrastructure on territories previously known as the countryside as a counterbalance." Rigid, abstract, and rational geometry—the quality modernism once promised to cities—has migrated to the land: pivot irrigation circles, greenhouse cities, server farms, logistics centers, robotic agriculture. The levity we associated with rural life now belongs to the city, while the countryside has received the logic of machines.

In a conversation with Benjamin Bratton (Strelka, 2017), Koolhaas accepted a striking rephrasing: the old model saw the city as a place for machines to reside, and the countryside as a garden of nature. This has been reversed—cities are where humans are; the countryside is where machines reside. He proposed that the current urban/rural demarcation might be better divided into dense and non-dense, or “human-occupied or machine-occupied.” This is the origin of his use of the term "post-human" for the new type of countryside.

We even see this logic manifest in Nolan's "Interstellar," where farms are completely overwhelmed by automated agricultural equipment, with very few human hands remaining. And those human hands mainly maintain the machines that actually do the farming.

The most significant architecture of this age—data centers, logistics warehouses, automated greenhouses, and the "box clusters" he studied in the Tahoe Reno Industrial Center (Walmart next to Amazon next to Tesla next to Google)—are designed by anonymous industrial companies outside the academic discourse, their scale rendering human proportions insignificant.

This observation also contradicts a very common notion in tech Twitter: we just need to make data centers look nice. I don’t think this is a solution; the companies building these behemoths don’t seem to have the political will for it either.

And for whom? For what purpose? Once again, these data centers far exceed human scale. They are not places for public consumption. You cannot rest in the shade that it may provide. This notion of "beautiful data centers architecturally" is yet another top-down fantasy applied from afar to placate a culture that tech experts actually don’t want to engage with.

Understanding the Narrative (This part is simple)

Setting any new narrative for data centers is an arduous battle, which should be no surprise to anyone.

To condense the entire Rem section → we must stop viewing data centers as the static endpoint of AI development, offloading urban knowledge projects as the physical weight of the countryside, then pointing to elites in San Francisco.

In the psychological perception of Americans, data centers are wholly synonymous with AI. Data centers have no connection to the infrastructure ordinary people use to send emails or upload photos to Instagram. This infrastructure has been classified as a new "other." Data centers (especially newly built ones) = AI.

If data centers = AI → then the narrative burden they carry is almost entirely the same. “Unemployment caused by the elite class, which they claim is necessary for progress.”

We need to pause here, as "progress" is a heavily weighted term, regarded as the reason for existence by the entire Silicon Valley. I cannot untangle this complex knot here, but if you’re still reading, consider how many aspects of Silicon Valley's notion of progress fail to meet the needs of ordinary Americans.

For me, this is a classic case: when a narrative based on capital and talent escapes control and becomes the only narrative.

In other words, when technology is as important as we insiders believe it is → the narrative is conveyed to everyone. Unfortunately, the currently popular narrative is crafted for capital momentum and the talent war, not to win the hearts of those who need to welcome metropolitan-level infrastructure in their backyards.

New Narrative

The rules of the new narrative should be as follows:

Infrastructure cannot be anonymous—there must be clear ownership

I sincerely believe that anonymous infrastructure violates its intended purpose. If I burn down a data center, I am harming the concept of AI, not a specific company. It is a mistake to abstract data centers as conceptual spaces rather than naming real companies with real human beings.

Amazon distribution centers, despite their flaws, present a clearer employment and service relationship than today’s data centers. Distribution centers are hubs that require a large workforce, far exceeding that of data centers. For data centers, "showing themselves" as a broader interdependent posture becomes even more radical, despite their sparser human footprint.

Reciprocity, see the countryside as the owner, not just a site

That is: this data center bears an agreement of reciprocity with local communities. Whether providing quotas for schools and hospitals or offering equity to towns.

Purpose

The owners of these infrastructures must gain new community purposes (and pride?). This is directly tied to reciprocity. "Your children will own a part" creates new community benefits and a forward-looking purpose.

Electricity Prices

I do not know how the narrative of data centers can proceed without seriously discussing energy issues. Significant partnerships need to be established to pair data center construction with energy suppliers, increasing capacity that genuinely responds to community concerns.

Architectural Form

While I do not believe that "making them pretty" is a sufficient answer, we cannot ignore that these things are presently just ugly brown boxes. The first team to genuinely handle data centers as entities will create an interesting moment for us all to learn from.

It must be made clear that presence is not "beauty." Lest I undermine my earlier views on data center design. Presence is authorship; this is the first rule of this list, transforming into architectural form. And the current brown boxes are the opposite of anonymity.

Refusal of Theory

People will say no. The response of tech experts cannot be a knee-jerk reaction of "these country bumpkins are so stupid." Maintaining individual and community dignity means respecting refusals.

None of this is policy. I will also not deceive myself into thinking it resolves all the issues our technology-believing industry faces.

But I firmly hold this belief: a narrative that cannot withstand "no" is not a narrative. Some of these projects should not be built. Not because AI is bad, and not because all opponents are right. But because a nation that operates by force, regardless of what the owners say, will ultimately approve every proposal. If we can only rely on force, this will fundamentally change our beliefs about how progress (that troublesome word again) should occur in the coming decades... Eliminating the mechanism of community consent will further entrench the belief that only radicalism can respond to such force. This is a downward spiral.

Rem understands that the countryside has never been the background. I believe we have re-entered a critical moment where the countryside is unexpectedly determining the momentum of Silicon Valley.

We are building the largest architecture of this century, thousands of miles away from company headquarters, yet refuse to sign our names on it. As my mentor Freddy Anzures said, "We are the chefs afraid to taste our own food." I would add that we are even unwilling to sign the menu.

The window for setting a new narrative for the most critical technological moment of our lifetime is now. We cannot simply let it close because we believe we know better.

Image: A conceptual design for a data center that coexists with nature (Source: Hip City Reg)

I don't do much editing, so please excuse typos and awkward phrases.

Thank you very much for your attention. I hope this is worthwhile; if not... unsubscribing won't hurt my feelings, and it will reclaim time for you that you really cannot get back.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到support@aicoin.com,本平台相关工作人员将会进行核查。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink