Economic scenarios — what AI could mean for jobs, wages and growthwhat-if, not a forecast
Move the five drivers of AI's labor-market impact — capability, adoption, autonomy, productivity and how fast displaced workers find new work — and watch GDP, jobs, wages and the labor–capital split respond. Live US baseline data sets the starting point; it is never a lock. Calibrated against Anthropic's published reference scenarios; this is an independent implementation.
GDP uplift by 2030
+4.5%
vs a no-AI economy
Labor share
57.0%
capital share 43.0%
Unemployment
4.9%
total, all workers
Knowledge-worker wages
+0.9%
vs no-AI
Other-worker wages
+2.3%
vs no-AI
Labor vs capital share
Of every dollar of output, how much goes to workers vs owners of capital. Automation shifts it toward capital.
Where workers end up
Shares of all workers at the horizon. Stayed, crossed into other work, still moving, or displaced.
Unemployment by group
Wage change by group
Versus a counterfactual with no AI.
How AI touches knowledge-work tasks
Of the knowledge-work task base: left unchanged, augmented (AI assists), automated (AI does it), plus brand-new tasks AI creates.
Published reference points
Anthropic's three published scenarios (Korinek et al., 2026), shown for comparison and used only to calibrate this model. Not krzy.ai results. Source →
Live baseline
The current US economy the scenario perturbs. Each value shows its source, date and provenance flag. Labor data is monthly/quarterly — never real-time.
Assumptions
Every coefficient below is an input, not a result — calibrated to the published reference scenarios, not estimated from micro-data. Edit any cell and the model re-solves; an override flips that field to User.
Model coefficientsCapex→capability bridge elasticitiesSources and known limitations