AI

Data systems often treat identity as the sum of pairwise matches. CellID begins elsewhere: a match is evidence, an entity is a graph, and identity must survive correction without rewriting history.

Most systems that collect many judgments still do the simplest possible thing: they average them. REIM starts from a different premise. Agreement is not enough; we also need to know who, or what, is doing the agreeing.

In June I described Sphere as an experiment in agent-readable publishing. Today I am releasing v1: a fragment contract, a self-hostable node you deploy to your own Cloudflare account, and a local Claude plugin. No platform in the middle. Intelligence before monetization.

Sphere is an independent experiment in agent-readable publishing. It asks what public knowledge should look like when the first reader may be an AI agent: structured, licensed, provenance-aware, citable, and accountable.

The legal and organizational consequences of a distinction most companies are ignoring.