Science answers to the primary source — reality. A language model answers to text. Three fallacies follow, and they share one root: a statistical property of text is substituted for the thing itself.

So it flatters instead of resolving, and re-derives the same reasoning billions of times a day, discarding it when each conversation ends. Orealis corrects all three — a deterministic control plane for long-horizon agents: error measurable, memory auditable, help judged by what actually changed.

Orí is how the architecture gets tested against a real life — the clips alongside are it, running. The memory underneath is a typed, causal model of what a person cares about: a structure computed over, not a transcript searched.

It maps concerns, what feeds them, what blocks them, and which routes are genuinely open. It distinguishes feeling better from getting better — and audits whether its own help moved anything.

The companion is severable from all of it. A lab can take the causal engine below, the typed objects, or both, and never adopt Orí at all — the architecture is what is on offer. What follows is that architecture: three corrections, and what they produce.

The name is the first syllable of Orealis — from Borealis, light appearing in darkness — and a Hebrew name meaning my light.

The primitive is the function, not the token — the finite cause-and-effect moves by which people satisfy needs. Life’s diversity looks infinite; at the level of function it reduces to the same closed set, run again and again.

Two layers: the lattice — that fixed map, identical for everyone — and the topology — the user’s life, embedded in it as coordinates. The model is a codec at the boundary — language in, coordinates out — and deterministic walks do the reasoning. Memory becomes placement, not retrieval. Reasoning becomes a path you can inspect, not a guess you have to trust.

The lattice is a control — the fixed reference that makes error measurable and keeps it from compounding. Core reasoning persists at its coordinates: derived once, walked cheaply forever. What it cannot answer names its own gap — researched, filed as a candidate, ratified by outcomes and, where stakes are high, by experts. Outcomes land at shared addresses and compound into population-level cause and effect. The users train the system. What accumulates is a causal graph of human reality.

Which is also the answer to scale. The lattice is not hand-authored — it is AI-assisted by construction: the model proposes, the structure verifies, people ratify. And it is recursive, so it grows inward rather than outward — new precision comes from composing coordinates that already exist, not from minting new ones. Set against retraining a trillion parameters to absorb one new thing, that is a different cost class entirely.

Every person, group, and organization in the user’s life is a persistent typed object. A name resolves in real time to the right one, carrying its history and its edges to everything else — so disambiguation stops being a guess, and what a conversation is about is a fact the system holds, not an inference it re-makes each turn.

Context follows from that. What loads is what the conversation actually reaches — the needle, not the haystack — and it descends on demand, from summary to detail to the verbatim turn, so cost tracks the question rather than the length of the transcript.

This layer stands on its own. The typing earns its keep before the lattice does: resolution, provenance, and adaptive context are worth having whether or not anything is built above them.

Affect governs human behavior — feeling is the signal that orients attention, decision, and action. So affect is the index: what a person feels marks what matters, and the system knows not just what is true but what is worth computing.

That layer is a computational model of human behavior built on one premise — the function of affect is to orient adaptive behavior. Orealis calls it adaptive affective metabolism. It aligns with the empirical evidence from neurobehavioral science — appraisal theory, somatic markers, allostasis — but it computes: one integrated causal runtime, not a shelf of theories that don’t.

Like the others, this layer is separable by design. Take the causal architecture without it and you still have a typed, deterministic reasoning substrate. Together, the three reason causally about what actually matters to a person.

Orealis, in one sentence: a fixed, shared addressing scheme for human life, with deterministic derivation over it — and a language model used as a codec at the boundary, not as the reasoner. That division of labor is the product.

Symbolic systems failed at the language interface; language models fail at ground. Each solves the other’s problem — the pairing is the product.

A limb is invisible for a lifetime — pure instrument, never once an object of attention. Lose it and it becomes the entire field. An intersection is not a problem until it is a jam. Nothing about the limb or the intersection changed. What changed is where affect could flow.

So the role a thing plays — instrument, obstacle, goal — is not a property of the thing. It is a reading of the affective field at that position: extrinsic, not intrinsic. A concern is a potential well; behavior runs down-gradient. Block a route and the affect does not vanish — it collects at the blockage, which is why a bridge becomes a grievance.

Every earlier inventory typed roles as intrinsic. That is why they sprawled: the same mechanism under a different gradient presents as a different thing, so it earns a new entry. The cost is measurable. FrameNet — twenty years of expert curation, the closest precedent — published 1,221 frames that its own relational structure places at 273–607 actual positions, with 39% tying out to nothing at all. Roughly forty percent are relationally indistinguishable from another frame; the distinctions live in names, not structure.

They were enumerating readings. Orealis models the field that generates them. Enumeration is unbounded, because gradients are. Generation is bounded, because the field has finite dimensions — everything terminates in sensation and affect, and a candidate that cannot trace there is not a function.

The critique assumes a person is one channel — a single, consistent set of values — and that modeling that means flattening what will not flatten. That is not what is being modeled.

Everyone runs several meaning-making frameworks at once: a partnership, a family, a company, a friend group. Each carries its own social contract, and each rewards different behavior — often behavior another would punish. What reads as inconsistency is usually two formations competing, not a person who changed. Orealis models each one separately, with its own contract and its own balance, so a conflict stays visible as a conflict.

Nothing is averaged. The highest affective energy wins — the formation pulling hardest is the one that routes the behavior. And where a placement is uncertain, the system asks rather than guesses: the person is the authority on their own life.

Weather has simple rules — pressure, temperature, moisture. What made it hard was never the rules; it was measuring enough concurrent inputs to see them working. Acceptance and rejection do most of the work here. What was missing was never a simpler account of people. It was the instrument.