The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Assign a concrete next action to each blocker to eliminate cognitive intrusion even before the blocker is fully resolved.
Generate explanations of new material by asking 'why is this true?' and 'how does this connect?' rather than summarizing what was said.
Focus failure analysis on system-level causal chains ('what in my process allowed this') rather than character judgments ('what's wrong with me') to extract actionable improvements.
Break progress externalization into the smallest sustainable unit—a single daily entry requiring under 60 seconds—because elaborate tracking systems collapse under maintenance overhead while simple systems persist.
Design physical and digital workspaces to afford only the cognitive operations required for current work, removing all objects and applications that compete for attention, because every irrelevant stimulus in the perceptual field consumes neural processing resources that could support task focus.
Document not only what tools you use but the complete routing rules, processing workflows, decision principles, and evolution history of your knowledge management system, because undocumented systems cannot be debugged, improved, transferred, or rebuilt after failure.
Test AI interactions for cognitive extension versus replacement by assessing whether you understand more after the interaction and could reconstruct the reasoning independently, because partnership requires mutual capability growth while delegation produces dependence.
Externalize mental models by writing them down as named, versioned artifacts to convert invisible cognitive habits into inspectable infrastructure.
Treat surprises and violated expectations as diagnostic signals that reveal which operating schemas need inspection, since mismatches between prediction and reality expose invisible assumptions.
Maintain competing schemas for the same domain to counteract confirmation bias, since single schemas automatically classify contradictory evidence as noise before conscious evaluation.
Use AI as a schema inspection tool by articulating your reasoning in text and asking it to identify unstated assumptions, since AI operates on text rather than emotional charge and can detect patterns you cannot see from inside.
Conduct schema audits by writing down actual operating rules, sourcing their origin, testing them against success and failure cases, and rating confidence against evidence.
Periodically verify that your documented schemas match current operational reality by directly observing the system rather than consulting representations.
When two people's schemas of the same situation diverge, treat the divergence itself as information about complexity the territory contains that neither schema fully captured.
For any schema you plan to act on, explicitly inventory three omissions—things the real territory contains that your schema does not—to counteract the natural tendency to treat a single mental model as complete.
Define explicit usefulness boundaries for each schema specifying where it stops working and what alternative framework to switch to at those boundaries.
Practice defusion from schemas by explicitly labeling them as 'my schema of X' rather than 'what X is' to create psychological distance between yourself and the representation.
Ask trusted observers to identify assumptions they see you operating from that you may not be aware of, since defensive routines make your own schemas invisible to you while remaining obvious to others.
Track your revealed preferences by comparing stated criteria with actual decision patterns, using the gap as a measurement of schema blindness.
Deliberately choose what resolution you need for your current purpose rather than attempting to maximize resolution across all dimensions, since cognitive resources are finite and higher resolution everywhere produces overload.
Before learning new domain content, explicitly surface and externalize your current implicit schema for that domain — this creates structure for new information to integrate with rather than overwrite.
When making schemas explicit, externalize competing interpretive frameworks simultaneously rather than sequentially to enable genuine comparison instead of allowing the fastest schema to pre-empt alternatives.
When two schemas conflict, search for their common ancestor—the deeper value or goal both serve—to determine the appropriate routing rule rather than selecting a winner.
Track which schemas win repeatedly across situations to identify your dominant paradigms, then subject those dominant schemas to the most scrutiny since their accumulated precision-weighting makes them hardest to overturn when evidence contradicts them.