Autonomy State Machine
A graduated trust system for AI deployments with three states: Disabled, HITL (human verifies every output), and Autonomous (spot-check only). Transitions are driven by statistical evidence with hysteresis to prevent oscillation. See: The Promotion Protocol.
Why It Exists
Binary thinking about AI (works / doesn't work) misses the entire middle ground where AI assists but doesn't decide. The state machine makes that middle ground operational.
Rosetta Stone
Four circles, four readings of the same object. Each role reads the artifact through its own lens.
Graduated delegation. Trust is priced, not assumed. Each state change updates the cost of running the instrument.
Staged rollout for agents. You prove it in one cell before you ship it to the fleet.
Feature-flag progression with statistical gates instead of vibes.
A finite-state controller over {Disabled, HITL, Autonomous} with transitions driven by sequential hypothesis tests. Hysteresis prevents oscillation; drift detection triggers re-elicitation.
Related Terms
Drift Detector - A posterior predictive check that detects when the operator's preferences have drifted from the learned model.
Gold Standard - A set of human-verified, ground-truth examples used to calibrate and evaluate AI output.