The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Direct change effort toward what you control (judgments, intentions, responses) rather than what you cannot (others' behavior, external events), since attempting to change the uncontrollable is both ineffective and the primary source of unnecessary suffering.
Distinguish technical problems (known solutions exist) from adaptive problems (people must change themselves) and give the adaptive work back to the people involved rather than providing technical solutions to adaptive challenges.
Treat every experience as training data for continuous schema calibration, weighting surprising observations more heavily than confirmations, rather than freezing your models and deploying them indefinitely.
Model other people's reasoning as operating from different premises shaped by their experiences and schemas rather than dismissing contradictory conclusions as character flaws, since engagement with reasoning structures enables actual persuasion while character attribution blocks it.
Recognize that your expectations about others alter your behavior toward them in ways that change their behavior, creating self-fulfilling prophecies where the confirming evidence reflects your schema's influence rather than their inherent nature.
Design systems based on Theory Y assumptions (people seek meaningful work and responsibility) rather than Theory X (people need coercion) when you want engagement and self-direction, since the organizational structure creates the behavior that appears to validate the theory.
Update self-schemas through generating new behavioral evidence rather than through positive affirmation, as schemas revise based on observational data not verbal assertion.
Audit your self-schemas in three layers—surface statements, narrative structure, and meta-schemas about change capacity—as each layer constrains the ones above it.
When planning task duration, deliberately switch from inside-view scenario construction to outside-view base-rate consultation to override the planning fallacy embedded in future-oriented schemas.
Structure risk exposure with asymmetric payoffs—bounded downside with unbounded upside—rather than minimizing risk as a scalar quantity, as risk shape matters more than risk magnitude.
Separate decision quality from outcome quality in post-decision analysis, as conflating the two (resulting) causes your risk schema to update on noise rather than signal and converges on superstition rather than calibration.
When encountering a new schema, evaluate it laterally by investigating the source's track record and independent assessments rather than vertically by analyzing only the schema's internal coherence.
Calibrate trust to the specific conditions of each claim—domain match, evidence transparency, incentive structure, and peer consensus—rather than adopting blanket heuristics of trusting or distrusting expertise.
Navigate vertically through abstraction layers by asking 'why' to move up (toward theory) and 'how' to move down (toward procedure), using gaps in either direction to diagnose missing cognitive infrastructure.
Build meta-schemas with base cases that terminate recursion through action rather than allowing infinite analytical descent.
Use external observations of behavior rather than introspection to validate self-models, treating discrepancies between internal narrative and external evidence as diagnostic of metacognitive blind spots.
Conduct metacognitive analysis retrospectively rather than in real-time to avoid resource competition between the cognitive process being examined and the examination itself.
Build environmental structures and external protocols to compensate for introspective blind spots rather than attempting to eliminate blind spots through increased introspective effort.
Identify and improve keystone meta-schemas whose function cascades across multiple domains rather than optimizing domain-specific schemas in isolation.
Map default behaviors across cognitive domains to identify the meta-schemas running automatically in your cognitive operating system's idle loop.
Limit real-time recursive metacognitive depth to two or three levels before externalizing to avoid working memory saturation and the illusion of depth without genuine inspection.
Increase the marginal value of each new piece of knowledge by connecting it to existing knowledge, as connected information compounds quadratically while isolated information accumulates linearly.
Externalize knowledge as discrete atomic units rather than continuous narrative to enable recombination and connection across contexts.
Progressively refine captured information through multiple encounters rather than attempting perfect formulation on first capture.