Each skill's mastery level S (0–1) follows a growth-minus-decay model, integrated forward in simulated career-years:
dS/dt = G · (hours/16) · (1 − S) − d · S
G = growth rate at full-time study (diminishing returns as S → 1)
d = skill's own decay/obsolescence rate
hours = your weekly hours allocated to that skill, capped at a
40h total learning budget — overallocating shrinks every
skill's effective hours proportionally
Growth slows as a skill approaches mastery (the "(1 − S)" term), and every skill loses ground on its own the moment you stop investing time in it — physical fitness and coding decay fastest, second-language and public-speaking skill decay slowest, matching how quickly each domain actually goes stale.
Each skill also has a market-demand curve, a slow oscillation representing how much the economy currently values that skill (frameworks fall out of fashion, new tools rise). Your human capital score is the demand-weighted sum of your skill levels — so a skill you stop maintaining can lose value twice: once from decay, once from falling demand.
- Hours sliders — this week's time budget across five skills, shared from a fixed 40h pool.
- Speed — how many simulated career-years pass per real second.
- Columns — height = current mastery; the pulsing ring at each base = current market demand for that skill.
Real-world relevance: this is the core argument for lifelong learning — skills are not one-time investments but decaying assets that require continual reinvestment to stay valuable, a pattern studied in labour economics as "skill half-life" and in workforce reports on continuous re-skilling.