The strongest escape: explosive productivity

The strongest escape: explosive productivity

The strongest objection to any demand-bound analysis is that AI could expand the economic frontier. AI can accelerate software, science and engineering. Better science can improve energy, materials and robotics. Robotics can expand physical production. Better production can reduce the cost of compute. The loop can become self-reinforcing.

AI → science → energy/materials → robotics → production → cheaper compute → better AI

If this loop is powerful enough, historical GDP trajectories become poor guides. A future economy may produce vastly more real output per person. It may exploit resources more efficiently, substitute abundant materials for scarce ones, automate extraction, or expand beyond present geographic constraints.

The paper accepts this as a serious possibility. It does not require a fixed economic pie. Indeed, the more interesting result appears when the expansionary claim is granted rather than rejected.

If output capacity rises much faster than the resources required to produce each unit, competitive prices face downward pressure. Real abundance can therefore expand while nominal revenue per unit contracts. The technological bull case becomes a possible producer-rent bear case.