Used to check for browser translation.
用于检测浏览器翻译。
ブラウザの翻訳を検出する
cover
B
Blogs

The future is NOT Self-Hosted, but Self-Sovereign

Robert

Jul 27, 2025
10.7Kviews
edited

— A response to Drew Lyton’s “The Future is NOT Self-Hosted”

(Disclaimer: This blog post was drafted with the help of a language model, but all opinions expressed are my own.)

I read Drew Lyton’s piece with great appreciation. It's not often you see someone go beyond the surface-level criticism of cloud services and actually walk the full arc of thinking: realizing the limitations of cloud platforms, attempting to self-host their own infrastructure, and then realizing the limitations of self-hosting itself.

What Drew articulated is something I’ve been thinking about for a long time—not just as a technologist, but as someone who cares deeply about the future of personal digital sovereignty. His conclusion is honest, grounded, and right: self-hosting is not the future.

But I would add:

The future is not self-hosted, but self-sovereign.

Self-hosting is a reaction, not a destination#

Self-hosting often begins as an act of resistance—against surveillance capitalism, against lock-in, against the passive role we’ve all been assigned in the modern internet. But as Drew experienced firsthand, self-hosting comes with serious burdens: technical complexity, maintenance overhead, isolation from social workflows, and ultimately, a fragmented digital life.

What’s even more ironic is that after all the effort, you end up rebuilding the same walled gardens, just locally.

You may now “own” your photo albums, but sharing them with friends still requires workarounds. You may self-host your notes or calendar, but you’re forced to either recreate account systems or give up on interoperability. Instead of liberating the user, we simply turned everyone into a small-scale sysadmin with no shared infrastructure.

The problem isn’t just usability. The real issue is structural:

There is no shared protocol for identity, no agreed standard for portable, user-owned data, no common infrastructure for composable interaction.

In other words, we are solving the wrong layer.

Self-sovereign identity is the missing layer#

Self-sovereign doesn’t mean self-built. It doesn’t mean doing everything yourself. It means having ownership without captivityportability without friction, and interoperability without central dependence.

That’s where DID (Decentralized Identifiers) come in—not as a product, but as a protocol. A universal way of representing identity, without anchoring it to a specific service, vendor, or platform. With a DID, I can have an identity that is cryptographically mine, interoperable by design, and usable across systems that respect the protocol.

It’s not a login method. It’s a foundation.

Fortunately, W3C’s DID is now an official W3C Recommendation — a protocol specifically designed to address this problem:

The Decentralized Identity Foundation (DIF) is a group of individuals and organizations dedicated to advancing this ecosystem:



Data needs a home—but not a prison#

DID solves identity. But identity without a home for data is still incomplete. That’s where the idea of DID Spaces comes in—a protocol-native container where my data lives under my control, but not necessarily under my roof. (Disclaimer: My company built DID:Spaces, and I was the lead architect behind it.)

A DID Space can be:

  • self-hosted by me
  • hosted by a friend
  • hosted by a co-op or a privacy-respecting service provider

But regardless of where it’s hosted, it uses the same open protocol.
Migration is easy. Data structure is preserved. Control is retained.

That’s the difference between self-hosted and self-sovereign:

Self-hosted cares about where the server is.
Self-sovereign cares about who controls the identity and data, and how it can move.


Cloud vs self-host is the wrong question. We need protocol-based freedom.#

What I learned over the years is this: technology only liberates when the structure itself enables freedom by default. TCP/IP made the internet possible not because it was a product, but because it was a protocol. Email worked not because we all used the same provider, but because SMTP let different providers talk to each other.

We need that kind of thinking again—at the layer of identity, data, and interaction.

We don’t need more “alternatives” to the cloud. We need a shift in architecture—from platform-centric to protocol-centric systems.

Self-hosting is a phase we go through. A rebellion. But self-sovereignty is a design principle.


To Drew, with respect#

I don’t disagree with anything Drew wrote. In fact, I think we’re walking the same road. He’s describing what it feels like to arrive at the limits of what “DIY” digital autonomy can offer. I’ve been there too.

But instead of stopping at the boundary of self-hosting, we can keep going—toward protocols that free us not only from platforms, but also from the trap of individual silos.

The future won’t be self-hosted. But it can be self-sovereign—if we build the right protocols.