Revenue grows off a maturing base growth rate. Each year's revenue displaces jobs; displaced workers are partly reabsorbed and partly accumulate as net AI-driven unemployment. Total unemployment (baseline + AI-driven) depresses consumer spending, which drags on next year's revenue growth — a one-year lag, not a circular loop. A scenario ceiling caps total unemployment so the model can't imply >100% joblessness.
How to read this: Column-D-style feedback drag applies the prior year's consumer-spending decline to the current year's growth rate — a one-year lag, avoiding a circular reference. Net AI-driven unemployment nets new displacement against reabsorption into other work each year, capped by the scenario ceiling above. This is a speculative scenario model, not an empirical forecast — there's no published dataset cleanly mapping AI revenue to jobs displaced; the linking multipliers are illustrative judgment calls. Treat outputs as directional, stress-testable logic, not a prediction.