The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Design your time allocation decisions before entering the time period, when you are not subject to urgency bias and immediate demands.
Layer your ideal week template from constraints inward: immovable commitments first, then energy-aligned priority work, then supporting activities, then restoration and margin.
Use your ideal week template as a gravitational field that pulls actual weeks toward your design rather than as a rigid mandate that produces guilt when violated.
Recognize that a single interruption in a deep work block costs the entire block, not just the duration of the interruption.
Use AI to compress the time required for logistical tasks while preserving human directed attention for judgment and synthesis rather than offloading the judgment itself.
Track actual task duration against estimates to build a personal calibration factor that corrects for systematic optimism bias.
Decompose tasks into granular subtasks including setup, teardown, and waiting time to surface the invisible work that estimates typically omit.
Contain interrupt-driven task processing within designated administrative blocks rather than responding to every small task the moment it arrives.
Apply the 'clear yes or clear no' filter to new commitments, accepting only opportunities that meet a high threshold of alignment rather than the default of 'seems worthwhile.'
Identify your information pipeline bottleneck by testing each stage (capture, processing, storage, retrieval, output) in sequence and investing improvement effort only at the bottleneck stage, since non-bottleneck improvements produce zero throughput gain.
Create atomic notes containing exactly one idea rather than multi-idea summaries to enable precise linking in knowledge networks.
Review information at expanding intervals (increasing from days to weeks to months) rather than massing repetitions into a single session to maximize long-term retention with minimal total study time.
Assign expiration dates to time-sensitive information at capture rather than assuming all stored information remains valid indefinitely, because knowledge decays at measurable domain-specific rates.
Use past survival duration to predict future survival duration for non-perishable knowledge, assigning longer expected lifespans to ideas that have already persisted longer (the Lindy heuristic).
Apply progressive summarization layers (bold → highlight → summary → remix) incrementally as you revisit notes for real purposes, rather than processing all notes uniformly upfront, to allocate processing effort by demonstrated relevance instead of speculative importance.
Establish an external measurement system for cognitive capacity because subjective assessment of performance degrades under the same depletion being measured.
Lay out multiple notes simultaneously in visual or physical space rather than reviewing them sequentially, because synthesis requires parallel processing of multiple frames that exceeds serial working memory capacity.
Start with output requirements and work backward to determine what information and processing are needed, rather than consuming information and hoping it eventually becomes useful.
When information inflow chronically exceeds processing capacity, reset by archiving all backlog and rebuilding sources from zero rather than attempting incremental catch-up.
Never miss a habit twice consecutively, because single misses have no measurable effect on habit formation while consecutive misses begin eroding automaticity.
Calculate tool migration costs as weeks of degraded processing throughput and weakened habits rather than hours of direct migration work.
Allocate cognitive effort proportional to output half-life: invest minimal effort in outputs that matter for hours, moderate effort for outputs that matter for months, and maximum effort for outputs that matter for years.
Insert time gaps between creation and editing passes to allow incubation effects and reduce familiarity blindness, enabling you to perceive your output as your audience will rather than as you intended it.
Catalog your actual output history to identify recurring error patterns, then build checklist items and quality standards around the errors you demonstrably make rather than hypothetical failures.