The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Optimize for meaning alignment rather than comfort minimization to produce lives that are difficult but progressively richer instead of smooth but progressively emptier.
Preserve directional decisions as human choices when using AI tools while delegating mechanical execution to maintain sovereignty through technology augmentation.
When sovereignty practice plateaus, identify the specific weakest component and design targeted exercises at the edge of current ability rather than continuing the same routine.
Build sovereignty into identity by shifting from outcome-focused to identity-focused self-conception, because behavior that flows from 'who I am' requires less willpower than behavior requiring constant motivation.
Begin sovereignty practice with immediate concrete domains before extending to distant abstract ones, because authorship capacity scales from successfully governing what is proximate and visible.
Clarify values by identifying activities you would engage in without external reward or audience, because intrinsic motivation signals genuine value rather than performed value.
Schedule low-cognitive-demand tasks during predictable periods of depleted executive function rather than attempting to eliminate them entirely.
Construct meaning through freely chosen engagement with objectively worthwhile activities, because meaning requires both subjective attraction and objective value, not just one or the other.
Distinguish between responsibility (ownership of response) and blame (attribution of fault), because confusing them converts agency into guilt and produces paralysis rather than action.
Examine small failures of execution (avoided emails, deferred conversations, unenforced boundaries) as diagnostic signals of sovereignty gaps, because these reveal where declared sovereignty diverges from practiced sovereignty.
Force systematic reconstruction of automated processes through documentation to reveal hidden steps, drift, and implicit assumptions that System 1 processing conceals.
Apply the competent stranger test to workflow steps—could someone reasonably skilled but unfamiliar complete this step without clarification—to surface implicit knowledge and prevent curse-of-expertise blind spots.
Distinguish task parallelism from resource parallelism by checking not only whether steps can logically occur simultaneously but whether you have the capacity to execute them simultaneously.
Place verification checkpoints at workflow convergence points where parallel workstreams merge, after high-risk steps where errors have high probability or severe consequences, and at phase boundaries where assumptions become commitments.
Design checkpoints with specific, testable verification questions that produce binary pass/fail results rather than open-ended review instructions.
Externalize workflow templates for recurring process types to convert design effort from repeated cognitive reconstruction into reusable, explicit structure.
Maintain a library of templates rather than a single universal template, with each template tuned to specific situation types, and explicitly note the conditions under which each template applies or fails.
Match task cognitive demands to available attentional capacity rather than optimizing for either time utilization or task elimination alone.
Map workflows by recording both processing time and waiting time for each step to reveal that waiting, not active work, typically constitutes the dominant time consumption.
Define workflow inputs and outputs with operational precision before attempting to chain or automate workflows, specifying not just what enters and exits but the format, source, and acceptance criteria for each.
Structure handoffs between workflow stages using sender-responsibility protocols that include situation, background, assessment, and recommendation, rather than assuming receivers will know what questions to ask.
Build workflow variants indexed by context dimensions (urgency, stakes, energy, environment, audience) rather than attempting to optimize a single universal workflow for all conditions.
Design each workflow variant as a complete, self-contained strategy optimized for its context, rather than treating variants as degraded versions of an ideal workflow.
Apply process engineering methods selectively to recurring activities while protecting spaces that benefit from spontaneity.