The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Use narrative structure (causal terms like 'because' and insight terms like 'realize') when writing about difficult experiences to transform inchoate fragments into actionable understanding.
Write learning in your own words using the structure 'claim-evidence-connection-question' to force generation rather than transcription.
Explain thoughts to an interlocutor with zero shared context to force decompression of subjects, reasoning chains, and assumptions that abbreviated inner speech omits.
Write looping thoughts verbatim rather than analyzing them to shift processing from automatic limbic circuits to deliberate prefrontal circuits and break the repetition cycle.
Accept that observing your thoughts necessarily changes them rather than attempting neutral observation, and use that transformation as the primary intervention mechanism.
Capture thoughts immediately using whatever tool is closest rather than waiting for proper conditions, because signal degradation during delay exceeds any benefit from better formatting.
Create a ready-to-resume plan before any task switch by externalizing current cognitive state, unresolved elements, and next concrete action to close the mental loop and free working memory.
Distinguish between thoughts that confirm existing beliefs and thoughts that contradict current models, capturing only the latter as genuine learning signals.
Make commitments active (written/spoken), public (visible to others), and voluntary (not coerced) to maximize binding force through consistency pressure.
Create precommitment devices that impose costs on your future self for deviating from chosen actions, because your future self operates under different constraints than your current self.
Force complete articulation in writing to detect the gap between felt understanding and actual understanding, because the act of generation reveals what passive recognition conceals.
Converge multiple capture channels into a single review inbox during scheduled consolidation sessions, because distributed capture without convergence creates unreviewed graveyards rather than usable knowledge.
Engage in retrieval practice rather than passive re-reading during review, because effortful reconstruction strengthens memory while recognition produces only false familiarity.
Build higher-order cognitive systems only after establishing reliable perception infrastructure, because all downstream processing depends on the quality and completeness of captured input.
Break complex information into separate containers of 3-5 independently meaningful units to match working memory architecture.
Design note containers to be self-contained enough that updating one idea does not force changes to unrelated ideas stored in the same container.
Structure each note as a self-contained claim that can be understood without its original context rather than as a topic requiring surrounding material for comprehension.
Assign each distinct idea a permanent, unchanging identifier before writing content to enable unambiguous reference across time and contexts.
Decompose complex ideas into component parts until you encounter specific points where your understanding fails, as these uncertainty points reveal epistemic gaps that global assessments conceal.
Structure complex work into hierarchical subsystems of stable intermediate components to enable parallel development and minimize cascade failures from interruptions.
Structure notes as vertical slices that include claim, key evidence, and implication rather than horizontal slices that capture only one layer of reasoning.
When a challenge to one component of your position triggers defense of the entire position, treat this as a diagnostic signal that you are still treating compound assumptions as a monolith requiring decomposition.
Before committing resources to a strategic plan, extract every assumption the plan depends on, rank them by importance and uncertainty, and design checkpoints that test critical assumptions first.
When faced with a difficult evaluative question, explicitly identify whether you are answering the target question or a substituted easier proxy before committing to the judgment.