The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Prioritize filling knowledge gaps that are adjacent to your densest clusters because these represent concepts within your zone of proximal development where learning will be most effective.
Integrate new knowledge immediately through connecting it to existing nodes rather than accumulating unprocessed material, because the integration act forces retrieval and spaced repetition of surrounding concepts.
Design external systems to survive the death of the tools that currently access them by storing knowledge in durable, open, human-readable formats with explicit connections.
Conduct regular structural reviews of your knowledge graph to detect and repair decay (dead links, missing connections, orphan nodes) before trust degradation makes the system unusable.
Start graph exploration with overview (shape), then zoom to filtered regions (relationships), then examine specific nodes (details), rather than beginning with targeted search.
Couple AI systems with explicitly structured personal knowledge graphs rather than unlinked text collections to enable graph traversal and structural reasoning beyond semantic similarity.
Offload structural representation and storage to external knowledge graphs to reallocate biological cognitive capacity toward synthesis, evaluation, and creative recombination.
Active coupling with an external knowledge graph extends cognitive reach only while you maintain, traverse, and trust the system; a graph you built but no longer use is an artifact, not an extension.
To extract information from a contradiction, articulate both beliefs explicitly, validate the evidence for each, identify the contextual variable that determines when each applies, then formulate a synthesis that accounts for both.
When models in an ensemble disagree, treat the disagreement as a diagnostic signal indicating regions of uncertainty rather than a flaw requiring forced consensus.
Contradictions that resolve through clarification of terms or context indicate surface-level issues; contradictions that cascade through multiple dependent beliefs indicate deep structural tensions.
During a holding period, maintain active awareness of the contradiction by writing notes when it surfaces rather than ignoring it or forcing resolution.
Resist forced cognitive closure under time pressure, fatigue, or social accountability demands, as these conditions systematically increase premature resolution of deep contradictions.
Audit values-behavior gaps by comparing stated values against behavioral evidence (calendar, spending, time allocation) over 30+ days, treating discrepancies as diagnostic data rather than moral failures.
When behavior contradicts stated values, investigate the actual reward structure driving the behavior rather than increasing willpower or restating the value more emphatically.
Frame values-behavior gaps as blocked action problems rather than character defects, asking what environmental or psychological barriers prevent valued behavior.
When two opposing beliefs are both evidentially supported, find the vantage point (usually a higher level of abstraction or a contextual variable) from which both become explicable as partial truths within a larger frame.
Test integration quality by verifying that each component schema still generates distinct predictions post-integration; if all schemas produce identical outputs, you have homogenized rather than integrated.
When two legitimate values generate ongoing tension without resolution, design oscillation rhythms rather than forcing a permanent choice.
Before attempting to resolve a contradiction, determine whether additional information could produce a stable answer—if not, you face a polarity requiring management rather than a problem requiring solution.
Design early warning indicators for when you've drifted too far toward one pole of a permanent tension, triggering course-correction before downsides accumulate into crisis.
When aggregate data contradicts disaggregated data, identify the lurking variable that defines scope boundaries before declaring a genuine contradiction.
When expert advice from different domains contradicts, extract the implicit scope conditions (risk profile, time horizon, optimization target) before applying either recommendation.
Evaluate decisions as members of a class rather than as isolated instances to avoid narrow framing that rejects positive expected value.