Version 03, the latest authored version

Version 03 · the body as written

# Guava Nexus MVP hypothesis

The initial hypothesis was that people would change their behavior when a Seed became sufficiently useful as a continuity object that returning to it became easier and more valuable than reconstructing context elsewhere.

Product reconciliation exposed a weakness in that explanation: it still depends primarily on value accruing to the author's future self, which is the same behavioral asymmetry that causes many continuity systems to be abandoned.

## The stronger hypothesis is:

Guava Nexus changes behavior when a second party's reliance on a specific prior Version makes an author's record consequential on a dated occasion — and the author writes the next Version because of it.
Under this hypothesis, CREATE → EVOLVE remains an observable pattern, but it is not necessarily the causal mechanism.

## The deeper loop is:

> CREATE → SECOND-PARTY RELIANCE → CONSEQUENCE → EVOLVE

The corresponding validation instrument is the Seed test: whether any second party demonstrably depends on a specific earlier Version.

Retention remains useful descriptive evidence, but repeated return alone is not sufficient to establish the behavioral mechanism.

keccak256 of this body is what the commitment on chain records · the title, the version label, the author's note and the source are outside it