The Three-Dimensional Economics of AI - Rethinking returns when capital, revenue and technological velocity are endogenous

This essay examines whether conventional return analysis remains sufficient when AI capital requirements, revenue or economic value, and technological velocity become increasingly endogenous. It argues that rapid technological advancement can shorten the competitive economic life of AI infrastructure even while successful deployment attracts further investment and accelerates the frontier. The paper introduces TRACER — Technological Rate-Adjusted Capital Expected Return — as a conceptual framework for evaluating the sustainable expected return on marginal AI capital when these variables evolve together over time. It also considers the implications for capital allocation, infrastructure, labor, productivity and economy-wide value. TRACER is presented as a research hypothesis rather than a solved mathematical model; its next stage is empirical development across economics, mathematics, technology and capital markets.