Definitionv1

Agent retirement criteria: specific, measurable conditions

Agent retirement criteria: specific, measurable conditions defined in advance under which a cognitive agent will be retired regardless of emotional attachment or sunk cost considerations, serving as pre-committed decision-making frameworks that eliminate the bias toward agent accumulation by establishing explicit triggers for termination

Why This Is a Definition

This definition precisely captures the core term 'agent retirement criteria' as presented in the lesson. It establishes the genus (decision-making frameworks) and differentia (specific, measurable, pre-committed conditions defined before need arises) that distinguish it from general decision criteria or mere judgment calls. The definition explicitly addresses the key problem the lesson identifies - the systematic bias toward agent accumulation - and provides the mechanism (pre-committed criteria) that solves it.

Source Lessons

L-0588

Agent retirement criteria

Define clear criteria for when an agent should be retired rather than maintained. Without explicit retirement criteria set in advance, you will hold onto agents long past the point where they serve you — because the sunk cost of building them, the identity you attached to them, and the absence of a forcing function all conspire to keep dead agents on life support.

Connections

Defines (56)

AxiomPerception as Predictive Construction AxiomHindsight Bias and Calibration Necessity AxiomHabits as Context-Response Associations AxiomTwo-Level Metacognitive Architecture AxiomExpertise Transforms Perceptual Chunking AxiomLinguistic Structuring of Thought AxiomComplementary Learning Systems Architecture AxiomCognitive Dissonance Drives Information Avoidance AxiomDual Coding Theory: Verbal and Visual Channels AxiomUltradian and Circadian Cognitive Rhythms AxiomAttention as Gate to Conscious Perception AxiomPatterns Exist in Hierarchical Logical Levels AxiomExpertise Through Deliberate Practice AxiomSystematic Overconfidence Taxonomy AxiomBias Blind Spot Asymmetry AxiomBelief Perseverance Against Contradictory Evidence AxiomCultural Transmission Through Shared Intentionality AxiomBrain as Hierarchical Prediction Machine AxiomScientific Paradigms Are Incommensurable AxiomCognition Operates Through Dual Processing Systems AxiomBasic-Level Category Privilege AxiomConstrual Level Effects on Perception AxiomPeople interpret failure as either evidence about their AxiomBehavior requires an initiating signal to occur at a AxiomHumans escalate commitment to failing courses of action when AxiomWhen estimating future task duration, people naturally adopt AxiomDefault options determine behavior more reliably than AxiomRegulatory flexibility—the ability to shift between AxiomWhen mortality is made salient and consciously processed PrincipleProcess inbox items in two distinct passes—first clarifying PrincipleComplete actions requiring less than two minutes immediately PrincipleConsolidate all agent status information onto a single PrincipleExecute continuous incremental improvement (kaizen) within PrincipleExpect and plan for an implementation dip where performance PrincipleBuild cognitive agents with multiple independent trigger PrincipleIdentify and eliminate shared dependencies across multiple PrincipleRetire cognitive agents based on current value production PrincipleAccept that honoring one value often requires sacrificing PrincipleShift focus from event reflection to system reflection when PrincipleSelect tools based on how well they integrate with your PrinciplePrioritize controllable cue types (time, location, preceding PrincipleMake the desired transition between behavioral links the PrincipleWhen holding more power in a relationship (formal authority,PrinciplePeriodically surface process schemas by extracting embedded PrincipleDefine explicit escalation criteria specifying when PrincipleBuild self-direction infrastructure through five components:PrincipleLeverage tacit team knowledge about individual strengths,PrincipleBefore proposing the removal of any inherited system,PrincipleCognitive offloading must become an automatic daily habit PrincipleFrame goals at the identity level ('become a person who X')PrincipleDesign physical and digital workspaces to afford only the PrincipleTest each candidate classification dimension by asking PrincipleBuild feedback loops into agent systems through regular PrincipleTreat major changes in goals, team composition, or system PrincipleWhen two cognitive agents both claim authority over the same PrincipleSpecify every delegation with five components: concrete

Back to Concepts