Centralize the Self, Distribute the Data

The internet is centralized where it should be distributed,
and distributed where it should be centralized.
This is the core architectural mistake of modern digital systems.

And almost every pain we experience as users and developers comes from it.

⸻

The current inversion

Today’s internet works like this:
Identity → Distributed
Data → Centralized
Which sounds fine… until you live inside it.

⸻

Identity is fragmented

You don’t have one identity.
You have:
• a Gmail account
• another Gmail account
• a GitHub account
• an Apple ID
• a Twitter/X account
• a wallet
• a bank login
• dozens of app-specific sessions

Each service owns a version of you.

You are not a person.
You are a collection of accounts.

We spent countless hours in passwords and logins.

Data is centralized

Meanwhile, your data lives here:
• Google owns your emails
• Meta owns your social graph
• Banks own your financial history
• SaaS platforms own your work
• Cloud providers own your backups

So while you are distributed…
Your data is captured.
This is the inversion.

⸻

The correct model (flipped)
Let’s flip it back.

Identity should be CENTRALIZED
Data should be DISTRIBUTED

⸻

Centralize the self

Identity should have one root.
One origin.
One persistent session.
Not “Login with Google”.
Not “Create an account”.

Just:
This is me.
A single, computable identity:
• stable
• portable
• user-owned
• independent from services

Think of it as:
• a root key
• a local daemon
• a sovereign session
• a personal runtime

One self → many contexts.
Not many selves → one platform.

⸻

Distribute the data

Data should not live in silos.

It should be:
• distributed
• replicated
• portable
• revocable
• user-controlled

Your email doesn’t need Gmail.
Your social graph doesn’t need Meta.
Your identity doesn’t need servers.

Services should consume data, not own it.

⸻

Why this matters (technically)

From a systems perspective, the current model creates:
• duplicated auth logic
• endless session handling
• account recovery hell
• identity desync
• privacy leaks
• vendor lock-in

Centralizing identity simplifies everything:
• one auth
• one session
• one permission model
• many apps

Distributed data gives you:
• resilience
• ownership
• exitability
• real portability

⸻

A mental model

Think of a human nervous system:
• The brain is centralized
• The neurons are distributed

Modern systems did the opposite:
• Millions of fake brains (accounts)
• A few centralized bodies (platforms)

No wonder it feels wrong.

⸻

A simple rule
Centralize what defines you.
Distribute what you produce.

Or even simpler:
Centralize the self.
Distribute the data.

⸻

This is not anti-centralization

This is important.
Centralization is not evil.
Misplaced centralization is.
Identity needs coherence.
Data needs freedom.

Put each where it belongs.

⸻

Final thought

If your identity is portable,
your data is distributed,
and your session is yours…

Then no platform owns you.
They only interact with you.
And that changes everything.