The case for digital heritage

There is cloud storage. There is digital estate planning. There is DNA storage — we master it on silicon plaques. But there is no category for digital heritage — no place where a human life is understood, protected, authenticated and passed on as a whole.

Storage moves bytes. Estate planning organises documents. Neither of them holds a life. The layer between a family and its archive is empty, and it is the only layer that matters to the people involved.

Everybody needs it: every family loses people, every archive decays, every generation forgets. Nobody can stop it: death, forgetting and format rot proceed whether or not anyone acts. And almost nobody sees it yet, because the category does not exist — the infrastructure layer is still empty.

The best ideas sit where universal need meets an unstoppable force, in a layer almost nobody is building. Memory preservation sits exactly there. The empty layer is not a weakness in the thesis; it is the thesis.

A lifetime of photographs sits on a device that will be obsolete within a few years. A parent's digital existence becomes a list of accounts nobody can open. Cultural archives decay faster than institutions can digitise them.

The files sometimes survive. The context almost never does. Who is in the picture, why the day mattered, what was said afterwards — that is the part that disappears first, and it is the part that was worth keeping.

Three things changed at the same time. Artificial intelligence can finally read a scattered archive and reconstruct the shape of a life from it. Cryptography can prove that a memory is authentic without exposing it to anyone. And one of those long-horizon media is already ours: our founding team masters DNA data storage on silicon plaques.

Three years ago this was a thesis. Today it is an engineering plan with a defined scope, a costed path and a first vault that runs on technology which already exists.

A vault built for documents asks you to be an archivist. A vault built for memory does the work itself: it assembles the timeline, it links the people, it answers a question asked in ordinary language, and it knows what should be released, to whom, and when.

That is the difference between storing a life and preserving one.

Nobody entrusts a hundred years of family memory to a promise on a policy page. Encryption happens on the family's side. Nothing trains a model without an explicit decision. Every reconstruction is traceable to its source. No voice or likeness is simulated without consent. Everything remains exportable.

These are not features to be added later. They are the reason the thing can exist at all.

First, a vault that a family can actually use, built on proven components. Then, a heritage layer that governs how memory moves between generations. Then, the DNA-on-silicon archival media our team already masters, deployed for archives designed to outlast every company alive today, including ours.

The order matters. The near-term product is not waiting on industrial-scale integration, and a mastered substrate is not a marketing claim attached to a near-term product.