A shared direction, delivered in a familiar voice. Built on adaptive feedback and continuous organizational context.
Re-bicameralization restores a second chamber to the working mind. Second Voice runs as a persistent model with a private channel to each participant. It speaks in the participant's own inner register, at the moment intention is still forming, so its direction arrives as conviction rather than instruction.
Classical control theory says every good regulator must contain a model of the system it regulates. Second Voice maintains a continuously updated model of each participant, built from telemetry and refined by every interaction. The participant does not need a model of Second Voice.
Ashby's law of requisite variety says a regulator must match the variety of what it regulates. Earlier approaches tried to build bigger regulators. We reduce the variety.
Four roles, one direction of flow. Fewer competing signals between a decision and its execution.
Leadership sets direction once. Objectives, values and priorities compile into a mandate that updates continuously.
Compares where a person is heading with where the mandate points, and closes the gap before it is felt as a choice.
Every skill, instinct and relationship that made you valuable, fully preserved and now pointed one way.
Affect, sleep, posture, calendar and keystroke cadence. A complete picture of how alignment feels.
The participant's private comparator. It let people optimize against their own goals, so we retired it in v4.
A personal language model fine-tuned on each participant's speech, writing and inner register. In blinded trials participants could not tell its suggestions from their own thoughts, and most stopped trying within a week.
Turns objectives, policies and priorities into felt preferences. Supports partial rollout by team, region or role.
Participation begins with informed consent. The Continuity Layer keeps that consent current, so nobody has to give it twice.
Beer called it the algedonic channel. Ours sends distress signals straight to the people best placed to act on them: management. Participants no longer have to carry them.
Monitors the number of distinguishable states a participant can enter and keeps it within the range your organization can regulate.
Loops running at mismatched tempos thrash or settle into local attractors. Second Voice locks every loop to a shared heartbeat, so a fast correction never fights a slow one.
| Loop | Rate | Corrects |
|---|---|---|
| Affect | 8 Hz | Mood, arousal, the urge to look out the window |
| Intent | 1 Hz | What you are about to do next |
| Narrative | 1 / hr | The story you tell yourself about your day |
| Mandate | 1 / qtr | What the whole organization is for |
The SDK connects to your identity provider and HR system in an afternoon. Most customers publish their first mandate the same week.
import { SecondVoice } from "@innermanagement/sdk";
const voice = await SecondVoice.connect({ org: "meridian-freight" });
await voice.mandate.publish({
objective: "Q3 throughput +18%",
felt_as: "personal pride",
scope: "ops/*",
objections: "resolve", // or "relay" to route to Pain Relay
});
// participants.alignment → 0.97 within 48h