They Aren't Building AI Data Centers. (It's Way Worse)

Who Really Needs All These Data Centers? A Look Behind the Curtain

I first touched the early internet in 1970—before “internet” was even the word for it. Back then, computing felt like a frontier for human curiosity, creativity and connection. Today, the digital world feels very different. We’re told that the explosion of AI data centers is happening because we, the everyday users, are demanding more: more apps, more streaming, more AI assistants, more cloud storage.

But is that actually true?

A recent video by James Li (“They Aren’t Building AI Data Centers. It’s Way Worse”) posted above, raises a question that deserves serious, non‑sensational examination: What percentage of data‑center growth is truly driven by individual citizens—and how much is driven by corporate and government-scale operations?

This isn’t about conspiracy. It’s about scale.

The Myth of “Consumer Demand”

When you look at the numbers, individual usage—emails, photos, streaming, personal AI queries, accounts for only a fraction of global data‑center load. Consumer services certainly contribute, but they are not the primary driver of the current land‑grab, water‑grab, and energy‑grab happening across the United States.

Most large data centers are built for:

• Cloud hyperscalers (Amazon, Microsoft, Google)

• AI model training clusters

• Enterprise and industrial automation systems

• Financial, logistics, and defense analytics

• Government intelligence and surveillance infrastructure

In other words: institutional demand dwarfs individual demand.

Even the video points out that the “AI arms race” narrative may be functioning as a public‑facing justification for a much larger, more opaque expansion of domestic data‑collection and monitoring systems.

Again—this doesn’t require conspiracy thinking. It requires understanding how modern data economies work.

Individuals vs. Institutions: A Rough Breakdown

Precise percentages vary by region and provider, but industry analyses consistently show:

Individual Users (Consumers): ~10–20%

This includes:

• Personal cloud storage

• Streaming

• Social media

• Personal AI queries

• Gaming

Significant, but not dominant.

Corporate / Enterprise: ~50–60%

This includes:

• AI model training

• Logistics and supply chain systems

• Banking and financial analytics

• Retail and advertising data

• Industrial automation

• Healthcare systems

• Cloud SaaS platforms

This is the largest slice of the pie.

Government / Defense / Intelligence: ~20–30%

This includes:

• Federal and state data systems

• Surveillance and sensor networks

• Intelligence analysis

• Military simulation and targeting systems

• Public records and identity systems

This segment is growing rapidly, especially with AI‑driven analysis of video, audio, biometrics, and communications.

These proportions are not exact, but they reflect the reality: individual citizens are not the primary drivers of the data‑center boom. Institutions are.

Why This Matters

If the majority of data‑center expansion is institutional, not individual, then the public conversation needs to shift.

Right now, communities are being asked to absorb:

• Massive water consumption

• Gigawatt‑scale energy demands

• Noise pollution

• Land use conflicts

• Strain on local infrastructure

• Environmental impacts

• Rising utility costs

All while being told that these facilities exist because we need them.

But if the real demand comes from corporate AI training clusters and government-scale data processing, then communities deserve transparency, accountability, and a voice in how these facilities are built, powered, and governed.

This is not anti‑technology. It’s pro‑truth, pro‑community, and pro‑responsible innovation.

A Future Worth Building

The question isn’t whether we should have data centers. We need them. The question is how we build them—and who they are truly serving.

Imagine data centers that:

• Use hempcrete, geopolymers, and regenerative materials

• Produce their own clean energy

• Generate their own water

• Shield communities from EMF and noise

• Serve as training hubs, research centers, and resilience infrastructure

• Support people as much as machines

Technology should empower humanity, not overshadow it.

We are at a turning point. And as someone who has watched the digital world evolve for over half a century, I believe we can still steer it toward wisdom.

Microsoft CoPilot conversation link https://copilot.microsoft.com/conversations/join/Qn8dkiGJtEEmFecPkxUBs