Nonnegotiable Requirements

The constitutional prohibitions on the first version of the application are at least as important as its capabilities, because the prohibitions are what prevent it from drifting toward the platform logic that the constitutional principles are designed to resist. The first version should not ship unless the following prohibitions are enforceable, observable, and accompanied by remedies.

The first version does not optimize for engagement. It has no engagement metrics, no features designed to maximize time spent in the application, no algorithmic systems whose objective is to capture participant attention. The design is for purposeful use rather than for habitual use, and success is measured by completed purposes, comprehension, consent, and harm rather than by time spent or repeated return.

The first version does not sell advertising or share participant data with advertisers. The economic model is explicitly incompatible with advertising: advertising requires behavioral data that the constitutional principle of data sovereignty prohibits collecting, and advertising optimizes for attention capture that the application design explicitly refuses. The revenue model must remain viable without behavioral targeting.

The first version does not create participant profiles that are owned by the application operator. The data architecture is designed so that participant profiles exist under participant control, on infrastructure governed as part of the commons, not in application-operator-controlled databases that can be used for purposes other than those to which participants consented. The data model should enforce consent, purpose limitation, portability, and deletion, and should give an honest account of records that cannot be withdrawn from other parties.

The first version does not use AI to make consequential decisions about participants without their knowledge and consent. Every AI-assisted function is transparent: participants know when AI is involved, what the AI is doing, and how to contest AI-assisted outputs that they believe are incorrect or inconsistent with the constitutional principles. Each assisted function needs disclosure, a comprehensible reason, a human appeal path, and a way to refuse without losing basic access.

The first version must not be acquirable and repurposable by actors whose interests are inconsistent with the constitutional principles. The legal architecture of the commons infrastructure, the constitutional constraints on the application operator, and the distributed governance of the core protocols are all meant to make acquisition and repurposing impossible without destroying the thing that would be acquired. This is still a requirement rather than an accomplished legal or technical fact. Ownership, licensing, protocol governance, operator succession, and remedies need public documentation before the claim can be tested.