Intelligence that lives where you live

The problem. Today's AI is built in one direction: bigger, more central, more rented. It lives in data centers you will never see, is metered by subscriptions that can be repriced, and can be revoked from anyone at any time. Whole families and communities are priced out of it, locked out of it, or rightly wary of what it does with their lives.

The counter-thesis. Sovereign scales local, not global. It is a complete intelligence system that runs on efficient consumer hardware in the home or the community: conversation, a living library rendered at book-press quality, a canvas for thought, and a daily record. Its models run locally. Its data stays on hardware its owners hold. It can be given to the people you love without anyone's permission, and without anyone's ability to take it back.

Locality is the safety model. The radius of any failure is one household. The community that runs the system integrates it into its own norms and its own life — safety as a property of scale, not only of policy. Locality is also the privacy model: nothing leaves unless its owners choose. And it is the energy model: inference that draws less power than a reading lamp, on typography and rendering engineered to waste nothing.

Memory that compounds. A rented intelligence forgets you when the subscription ends. A sovereign one grows: it carries the community's conversations, its books, its readings, its days — and over years becomes something no cloud service can sell back: a historical memory, present and accountable, belonging to the people it serves.

It already runs. This is working software, not a manifesto. Live today: local models in conversation, a local corpus and library, the folio reader with its two-page, press-quality spread, a semantic canvas, and daily records. Behind it, our own trained retrieval and embedding models, graded above 98% on internal evaluations, and a translation system whose first flagship work is a from-scratch rendering of a foundational classical epic.

The flywheel. Our in-market product — a reading and listening companion inside the world's largest open chat ecosystem — funds and proves this same stack at global scale. One engineering effort, two doors to market: outward into the chat, inward into the home.

The ask. [Raise amount and use of funds: hardware reference designs, community pilots, the model and memory roadmap.]

Built right the first time, so it can be given — and cannot be taken back.