The Cloud Was a Tool — Not a Destination

December 23, 2025

This isn’t an anti-cloud post.

The cloud did exactly what it was supposed to do. It gave us speed, global reach, and a way out of the brittle, power-hungry data centers many of us were running a decade ago.

But something has changed — not just in technology, but socially.

We’re now watching a growing public backlash against large, centralized data centers. Zoning battles. Power-use protests. Environmental pressure. Political scrutiny. Whether you agree with it or not, it’s real — and it’s shaping where infrastructure is allowed to exist.

At the same time, hardware has become smaller, denser, and dramatically more power-efficient. The physical constraints that once pushed everything into hyperscale facilities don’t apply the way they used to.

The future isn’t about rejecting the cloud. It’s about reducing unnecessary concentration and putting workloads where they make the most sense — technically, economically, and socially.

Across the world, large data centers are increasingly seen as:

Drains on local power and water
Poor neighbors to residential communities
Strategic infrastructure controlled by a few companies
Easy political targets

This has nothing to do with IT best practices — and everything to do with perception.

That matters, because perception drives policy. And policy drives where infrastructure can be built, expanded, or even kept online.

Pushing everything into a smaller number of massive facilities makes those facilities more visible, more controversial, and more fragile from a regulatory standpoint.

And no, moving data centers into space is not the answer. That idea ignores basic risk modeling. Space introduces new failure modes — kinetic risk, radiation, launch failure, debris, and zero-repair scenarios. It solves none of the core problems and adds several more.

Space is the next frontier of war — not where I’m placing my data or compute.

The answer isn’t bigger or further away. It’s smarter and closer.

Distributed Doesn’t Mean Random

When people hear “distributed,” the conversation often veers into extremes:

Fully centralized hyperscale
Wildly decentralized models running on infrastructure you don’t own, don’t control, and don’t trust

Neither of those works for real businesses.

This isn’t about running critical workloads on anonymous machines or crypto-style volunteer compute. That model fails the moment you care about compliance, accountability, or uptime.

The future is distributed ownership:

Infrastructure your business owns
Placed closer to where data is generated and consumed
Connected securely, not exposed publicly
Designed for predictable, steady workloads

In other words: distributed, but intentional.

Why Ownership Matters Again

Owning infrastructure doesn’t mean going backwards. It means choosing where permanence makes sense.

For steady workloads — directory services, internal apps, core databases, security tooling — renting the same compute forever rarely makes economic or operational sense anymore.

Modern servers deliver more performance per watt than ever before. Storage density has exploded. Remote management, automation, and secure access are no longer optional add-ons — they’re table stakes.

What used to require a “data center” now fits into a few racks, a closet, or a small room footprint.

You’re not rebuilding the past. You’re right-sizing the core.

The Cloud Still Has a Job — Just Not Every Job

The cloud excels at:

Bursty or seasonal demand
Global access and edge delivery
Managed services that would be wasteful to self-host
Rapid experimentation

It struggles with:

Long-running, predictable workloads
Cost transparency at scale
Regulatory clarity in some jurisdictions
Giving administrators real visibility and control

This isn’t a failure of the cloud. It’s a misuse of it.
Unless you’re one of the hyperscalers robbing everyone — yeah, AWS, Azure, Google, Oracle.

The future isn’t cloud-only. It’s cloud-adjacent.

Cloud at the edge
Owned infrastructure at the core
Connected securely, not exposed

A More Sustainable Direction

If we care about sustainability — environmental, economic, and operational — concentration is the wrong direction.

Smaller, distributed footprints:

Reduce localized power strain
Lower political and social resistance
Improve resiliency through technical diversity
Give businesses back control over their own systems

This is where MSPs and infrastructure teams matter more than ever. Helping organizations own what should be owned, rent what should be rented, and connect it all securely is the real work ahead.

Not declouding
Not nostalgia
Not science fiction

Just infrastructure that fits the world we’re actually living in.

The Cloud Was a Tool — Not the Destination (Expanded)

For the last decade, cloud-first became the default IT belief system. As someone who has spent a career in network operations, I’ve seen the cloud deliver real gains in speed and flexibility.

But somewhere along the way, we didn’t just move to the cloud — we over-rotated.

What started as a tactic to escape bulky hardware and procurement cycles has turned into runaway operating costs and a loss of hands-on control over the systems that matter most.

We are entering a more grounded phase of IT decision-making. It’s no longer cloud versus on-prem. It’s a question of what belongs where.

The Economic Reality: When Owning Beats Renting

A common rebuttal to repatriation is, “Owning hardware isn’t cheaper.” That is true only if you execute without planning.

But for steady, predictable workloads — directory services, internal apps, core databases — the numbers overwhelmingly favor ownership.

A simplified three-year comparison:

Public Cloud: ~$64,800+
(Recurring compute costs, variable egress fees, support tiers)
Owned Core: ~$21,000
(Upfront hardware, power, ISP, and upkeep)

This isn’t about chasing the cheapest option. It’s about cost predictability. With owned infrastructure, you eliminate surprise bills tied to usage spikes or data movement.

The Security Shift: Connectivity Without Exposure

In the past, sovereignty meant pain — open firewall ports, complex VPNs, and exposed attack surfaces.

That is no longer the case.

The missing piece was secure connectivity. Tools like Cloudflare Tunnels change the equation. Core workloads can now run on owned hardware with zero inbound ports open to the internet.

You keep systems close.
The edge handles identity, access, and DDoS protection.

The Bigger Picture: Physics and Geopolitics

We are witnessing a growing backlash against centralized hyperscale data centers: zoning battles, power-use protests, and environmental pressure.

The industry’s proposed solution? Data centers in space.

Let’s be realistic. Space is not a backup site — it’s a domain of warfare.

Moving critical infrastructure to orbit introduces kinetic risk, radiation exposure, launch failure, and zero-repair scenarios. It solves none of the core problems and introduces catastrophic failure modes.

The answer isn’t further away.

It’s closer.

Cloud at the Edge, Owned at the Core

This isn’t an anti-cloud manifesto.

The cloud excels at bursty demand, global content delivery, and rapid experimentation. But it shouldn’t be the default home for everything.

The future is hybrid — done intentionally:

Distributed Ownership: Hardware you control, placed where data is generated
Secure Connectivity: Identity-based access without public exposure
Fiscal Sanity: Rent the spikes, own the baseline

The cloud isn’t going away.
But it’s no longer the center of the universe — and that’s a good thing.

About the Author, Brad Riggs is a Network Operations Nerd passionate about open-source solutions and breaking the “vendor lock-in” cycle, he founded ShadyPacket to guide organizations toward sustainable, predictable, and secure infrastructure designs. The views expressed here are his own and do not reflect the official policy or position of the Department of Defense or the U.S. Government.