poztter.org / philosophy

Why POZ is shaped this way.

POZ is a sophisticated technical specification that uses cryptography to provide critical functions. Driving every technical choice is a desired outcome rooted in core philosophy. These are them.

Overview · What is POZ? · Components · FAQ

two foundational problems

POZ exists because two related problems on the modern internet remain unsolved.

The first problem — trust on the internet

For websites, cryptography has long solved the problem of securing the data between a user and a server. But how do you know you're connected to a particular company's server in the first place? Today's answer requires a small set of centralized, highly trusted root certificate authorities and carefully handled pre-shared keys. Compromise the pre-shared keys with a fake root certificate, and the entire chain falls apart.

Certificates are tied to domains. Domains are resolved by DNS. Compromise a DNS server, and the chain falls apart. Certificates expire, and revocation requires the user to reach a centralized server they didn't choose. In the early internet, a domain pointed to one IP and one server. On the modern internet, servers are load-balanced, regionally cached, and fractured across services — and the original trust model never caught up. Over the years we've layered additional, purely optional, protocols and standards that always fall short.

The second problem — who can you trust

A story as old as time: a company promises to improve our lives, gains our trust, and eventually becomes the evil it promised to defeat. Monopolies and gangsters — even in the short history of tech companies — have shown the pattern of fighting dirty when their interests are threatened.

The lesson is that trust is necessary, but no organization should hold so much power that we cannot walk away the moment we want to.

solving the problem

POZ's answer is to trust the data, cryptographically, rather than trust the provider of the data. For most actions on the internet, what matters is not how the information reached you but that the information itself is accurate.

Consider DNS. Does it matter which server performed the lookup, or that the response is authoritative, correct, and up to date? The mirroring and caching the modern internet depends on reinforces this: as long as the data is verifiable, the path it traveled becomes less important.

Once you have the data, you check it yourself. Surprisingly, this simplifies many of today's internet protocols. It sidesteps the bootstrapping problems with pre-shared certificate authorities. And it makes verification something the client does, not something a server asserts.

the driving philosophies

From this single principle — trust the data — POZ derives a small set of driving philosophies. Each shapes a specific part of the standard.

three rules of thumb

If you've read the spec and want a one-liner of the design philosophy, these are the three rules POZ tries to follow:

1. The data is the trust anchor. Servers, transports, providers — none of them. Validate on the client.
2. Recovery is a feature, not an afterthought. A configuration that leaves no recovery path should be hard to make by accident.
3. Asymmetry beats symmetry when stakes differ. Granting a new identity and revoking a compromised one are not the same operation.

trust beyond the internet

POZ today is a file format, a network protocol, and a set of cryptographic operations. These let people and organizations share trust and verify relationships online. But trust is bigger than any one protocol.

Can you trust the person on the other end of a phone number? At a street address? Carrying a particular driver's license, or a passport? The same principles that make POZ work online — chains of attestation rooted in keys you control — apply to identity claims that have nothing to do with the internet.

We see many possibilities for extending POZ to the parts of identity that live beyond the bits.