The work of armies,
the vision of one.
Agentic computing lets an architect or an artist realize what once took a studio, a company, an army. Pharos Fabric exists so that the vision stays theirs.
Great work has always needed many hands.
A feature film, a cathedral, an operating system, a city. For most of history, work at that scale needed hundreds or thousands of people, and the vision that started it had to pass through every one of them. Some of it survived the journey. Much of it was negotiated away: by budgets, by committees, by the slow loss of signal along a long chain of hands.
That ratio is changing. One architect, or one artist, can now direct a crew of capable agents that write code, draft worlds, render, research, check and remember. Work that once required a studio can begin with one person and a clear idea of what they want.
The many hands do not disappear. They become machine collaborators, working at a speed and volume no review process was designed for. Which raises the question this whole site is about: when the hands multiply, who keeps the vision?
Delegation without dilution.
Power without governance only moves the loss of signal somewhere harder to see. An agent that borrows your key, edits your canon and leaves no record has not extended your vision. It has replaced it with a guess.
Sovereignty, as we mean it, is concrete:
- You hold the keys. Every collaborator, human or machine, acts under its own identity, with authority you granted, scoped to its task and revocable on its own.
- Nothing becomes canon without you. Agents explore freely in working memory. What enters the record passes a gate that you, or someone you trust, signs.
- You can always ask why. Every fact carries who asserted it, when, and on what evidence, and the history of what was believed is never overwritten.
That is what Pharos Fabric is for. It does not make the creative decisions. It makes sure the decisions that matter stay with the person whose vision it is. Sailing Directions shows how.
A divergent path to one question.
Erich Greenebaum's background is diverse and his thinking is divergent. His career runs through two worlds that rarely meet: entertainment technology, where he did seminal, patented work with Douglas Trumbull for the likes of Universal Studios, IMAX and Hitachi, and enterprise systems, where the problem is keeping a large organization's knowledge coherent.
Underneath both is one question he has carried for most of his working life: how does a system know not only a fact, but the standing of that fact? Who asserted it, when, in what context, under what authority, and whether it still holds. He came to it through RDF, and through the conviction that statements about statements are not an awkward corner of the model but its center. Why Reification Breaks the SQL Mind makes that case.
Compositing a film shot and governing a company's knowledge turn out to be the same problem: many sources, each with its own history, combined into one result that has to be trusted. Pharos Fabric is where the two threads join.
Built the way it is meant to be used.
Pharos Fabric was conceived, specified and built by one architect working with machine collaborators, Claude Code first among them. Its implementation spans Go, Rust, Clojure and TypeScript, governed by a body of RFC-style specifications.
For Erich, tools like these are a superpower: a way to realize ideas carried for decades, at a pace no team he could have assembled would have matched. They are also the proof of the problem. Directing many capable agents at once is exactly the governance problem the fabric exists to solve, and the fabric's own development runs on its rules. Its collaborators work under scoped capabilities, its specifications pass the same gate as everything else, and its history is never quietly rewritten.
The architects and artists who do the most with agentic computing will not be the ones with the most agents. They will be the ones whose vision survives the delegation.
Bring us your vision.
If you are an architect, an artist or a builder with something you have waited a long time to make, we would like to hear about it.