Who benefits if the scarce inputs are compute, intellectual property, electricity, data, and capital rather than human work?
Korinek-Stiglitz: if AI substitutes broadly for labor, wages can fall even as output soars, and distribution then depends entirely on who owns the machines. Transition-scenario modeling makes the wage-collapse case precise.
View on the map → · Open in Browse →
The camps produced their strongest recent statements: a forecasting survey projecting labor-force participation falling to 55% under rapid AI, Smith's dissent from the economists' call to steer AI toward labor complementarity, and Epoch's counterpoint that a thin tier of superstar human talent is capturing enormous returns right now.
One robot now turns into many robots next year, but the number of ballerinas is the same.
Two economists work through wages, scarcity, and redistribution under full automation.
Frames extreme concentration of purchasing power as a distinct fourth AI risk beyond safety, jobs, and superintelligence.
Dissent from the economists' statement: steering AI toward labor-complementarity is infeasible, and employment data show little displacement so far.
Two 'merely very good' researchers can't replicate one Noam Brown if what's needed is deep intuition about which experiments are worth running.
Frontier pay data ($10M–$300M packages): superstar effects let thin human talent capture enormous returns even as capital scales.
AI removes O-ring constraints that made output depend on uniformly skilled teams — changing which workers and firms capture returns (paywalled).