The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Test whether regret projections are genuine by checking if they ever contradict your current impulses; if projections always agree with present desires, you're performing motivated reasoning rather than consulting your future self.
Pre-commit to abandonment conditions before emotional investment makes rational exit decisions impossible.
Translate imagined failure scenarios into measurable tripwires rather than vague intentions.
Require disinterested party validation before overriding pre-set decision criteria.
Set decision speed as an explicit variable based on reversibility, stakes, and delay cost before analysis begins.
Build real-time information infrastructure rather than gathering data at decision time to compress orientation phase duration.
Generate multiple alternatives in parallel rather than evaluating options sequentially to accelerate comparison-based decisions.
Calculate delay cost as cost-per-day multiplied by deliberation duration and stop when it exceeds expected improvement from additional information.
Focus post-decision reviews on system-level process improvements rather than individual blame to maintain psychological safety necessary for honest evaluation.
Schedule decision reviews as mandatory future check-ins at the time you make the decision rather than waiting for obvious failure signals.
Review decisions with positive outcomes as rigorously as negative ones to detect bad processes hidden behind lucky results.
Match the collaboration pattern to the dependency structure of the task rather than defaulting to familiar patterns.
Design measurement systems to close feedback loops by making the gap between current state and desired state observable before consequences compound.
Add, change, or redesign feedback loop structures rather than merely tuning parameters within existing loops when seeking leverage points for system intervention.
Minimize delay between executing an action and receiving evaluative feedback about that action's effectiveness, because learning rate scales with cycle frequency more than with cycle quality.
Install leading indicators that correlate with delayed outcomes to provide faster feedback signals while waiting for lagging results.
Build measurement systems before executing strategies, not after problems become visible, because drift detection requires baseline instrumentation.
Schedule fixed-interval forcing functions that compel measurement independent of perceived need, because perceptual thresholds prevent detection of gradual drift.
Seek external observers who can detect deviations that your own continuity-embedded perception cannot register, because egocentric anchoring blinds you to gradual change.
Design corrective responses to scale proportionally with deviation magnitude rather than applying fixed-strength corrections regardless of error size.
When attempting to change a system stabilized by negative feedback, modify the set point rather than fighting the corrective response.
Define measurable key results for objectives before acting, so that reality can provide corrective feedback rather than post-hoc rationalization.
Pair performance metrics with counter-metrics that measure potential negative side effects to prevent optimization death spirals.
Define decision rules in advance that specify what action you will take when a metric crosses a threshold, converting measurement into a functional feedback loop.