The Ledger, the Anchor, and the Inscription

# Field Note 13
## The Ledger, the Anchor, and the Inscription

Not all cryptographic memory is on-chain.

This distinction is conceptually and technically important.

A system like Buzz can represent messages, agent actions, reactions, approvals, and workflow events as signed records. A signature can establish authorship and reveal tampering. But a signed event stored by a relay or local database is not automatically a blockchain transaction, and it is not necessarily permanent. The storage operator may still lose or delete it.

Cyberdelic OS should distinguish three memory layers.

### Living memory

The active history of a hearth or gathering: voice notes, placements, agent proposals, revisions, permissions, and correspondence. These records can be locally encrypted and append-only. Important actions are signed by the human or agent that performed them.

Living memory is fast, inexpensive, private by default, and still capable of change or deletion.

### Anchored memory

At meaningful intervals, the system can gather a body of records into a Merkle tree or other content-addressed structure and publish only the root hash to a public timestamping system anchored in Bitcoin.

The private records do not become public. The anchor proves that a particular state of the archive existed no later than a particular time and that later alterations would change its fingerprint.

Anchoring gives history a witness without giving the world its contents.

### Inscribed memory

A small number of works may be intentionally published into durable public memory through cyberdelic.space: field notes, constitutions, artworks, crux maps, releases, or cultural artifacts judged worthy of the cost and irreversibility.

Inscription should be ceremonial.

Immutability is not an automatic virtue. A living person needs revision, privacy, mercy, and forgetting. The right to preserve must remain balanced by the right not to be permanently captured by an earlier self.

Cryptography belongs beneath the surface as trust infrastructure: portable identity, signed provenance, verifiable history, and selective anchoring. It should not force every visitor to understand wallets or guard a fragile seed phrase before they can enter the village.

The ledger remembers activity.

The anchor witnesses history.

The inscription addresses the future.

Confusing these layers would either make the system expensive and unusable or make its claims of permanence dishonest.

Holding them apart lets memory have different temperatures.

---