What Gets Expensive When Everything Is Cheap (mateopetel.substack.com)

An essayed diagnosis: as AI drives the marginal cost of symbolic goods—text, images, code—toward zero, scarcity and price inflation concentrate on what machines can’t cheaply replicate: embodied time, land, energy, and legally binding guarantees. The result is a bifurcated economy where streaming content and automated drafts feel abundant while rents, childcare, electricians after storms, grid operators in heat waves, and other time‑intensive services grow dearer. This “cost disease with a jet engine” rewires labor markets (wage spikes in hands‑on roles), social prestige (from punditry to plumbers and midwives), and personal experience (feeling poorer despite rising measured output). Technically, the piece highlights why GDP and price indexes mislead: hedonic adjustments and production metrics overstate gains from digital abundance while missing rising costs tied to physical constraints and liability. It names concrete metrics to watch—energy per verified action, land and rents near substations and data centers, prices for insured/verified outcomes, and care wages—and explains feedback loops: capital clustering in compute/energy, higher interest rates to mobilize investment, asset repricing, and “numeraire drift” as society benchmarks value against what machines produce. Practical takeaway: value will shift toward custody, verification, and steady embodied labor; policy and investment should track these shadow gauges, not just headline productivity.