AI productivity gain meets compounding debt cost.
Emerging analyses estimate AI-generated code adds roughly 30-41% to technical debt within 90 days. The headline gain holds; the debt compounds. This projection runs both forward to find the crossover month, when monthly debt cost overtakes monthly productivity gain, and tots up the cumulative net.
your team & adoption
Team size (devs)
Fully-loaded salary ($K)
AI adoption %
40%% of code being AI-generated or AI-suggested. Set this to your team's actual mix.
Productivity gain %
18%Net velocity gain from AI usage. Team-level gains run well below headline per-task speedups.
Annual debt growth %
35%Emerging estimates: 30-41% in the first 90 days. Default 35%.
Projection horizon (months)
24Debt overtakes gain after year one.
Crossover at month 18 (1.5 years).
Monthly productivity gain
$15,300
from AI adoption
Debt cost @ month 24
$23,071
compounded
Net over 24mo
+$121,593
net positive
monthly gain vs accumulated debt
Productivity gain = base_team_cost × ai_adoption × gain_pct (held constant). Debt cost = base_team_cost × ai_adoption × ((1+monthly_growth)^month − 1) × 0.33 (a modelling coefficient mapping accumulated debt to maintenance hours; adjust to your context). Compounded monthly using the annual growth rate. Crossover marked where monthly gain falls below monthly debt cost.
How the projection works
The base monthly team cost is fixed: team_size × (fully_loaded_salary / 12). Productivity gain is also fixed monthly: base × ai_adoption × gain_pct (the per-month dollar gain from AI usage at the adoption level).
Debt cost compounds. Each month, the accumulated debt fraction grows at the monthly equivalent of the annual debt-growth rate: monthly_growth = (1 + annual_growth)^(1/12) − 1. Monthly debt cost = base × ai_adoption × ((1 + monthly_growth)^month − 1) × 0.33, where 0.33 is a modelling coefficient that maps accumulated debt into maintenance hours (adjust it to your context).
Crossover month = first month where monthly debt cost > monthly productivity gain. From that point, AI adoption is net-negative on a monthly basis. Cumulative net = sum of (gain − debt) across the horizon.
Default values: 35% annual debt growth (midpoint of the estimated 30-41% range), 18% productivity gain (a conservative net team-level assumption, since per-task AI speedups run higher than sustained team-level gains), 40% AI adoption, 24-month horizon. Treat the projection as scenario-modelling, not a binding forecast, your codebase's coupling and refactor discipline materially affect the debt-growth coefficient.