Mind · Strategy
SacrificialProbe
Learning via cheap, instructive failure
Definition
Mechanism
A Generalized Pattern where an Agent sends a low-cost 'probe' into a System expecting it to fail, but designs the failure to be instructive. The probe must be cheap relative to the main payload, and the failure Mode must update the Strategy for the main payload. Common in startups (landing pages), immunology (dendritic cells), and warfare (reconnaissance). It wraps the concept of a staked probe in a higher-order strategy where the probe's destruction is the intended Signal.
Risks
Failure modes
Probe too expensive
breaks the Cost Asymmetry invariant, making the sacrifice unaffordable.
Silent probe failure
Probe fails but produces no observable signal, destroying the Instructive Failure property.
Relationships
Builds on
Nothing in this library builds on this pattern yet.
Reference
Parameters
cost_ratioFloat, [0.0, 0.1]- Probe cost relative to payload