The Intelligence Curse: When AI Makes Human Development an Inefficiency

The resource curse is a well-documented phenomenon in political economy: nations rich in extractable resources like oil systematically underinvest in human capital because their revenue model doesn't depend on it. The intelligence curse names the structural analog emerging in the AI era. As GDP becomes increasingly dependent on AI systems, data infrastructure, and compute capacity rather than on human productive capability, the same perverse incentive structure takes hold — governments and corporations face diminishing returns on investment in education, health, and human development.

Sam Altman's remark comparing the resource cost of raising a human to the cost of running AI systems is revealing not as a gaffe but as an articulation of the underlying logic. If the economic calculus frames humans as cost centers rather than productive agents, then the political economy will follow. Peter Thiel's infamous 17-second pause when asked whether the human species should survive makes the civilizational stakes viscerally legible: some of the people building these systems increasingly understand themselves as midwifing a successor intelligence, not merely deploying a tool.

The intelligence curse exposes a long-standing inadequacy in GDP as a measure of civilizational health. AI threatens to run the metric upward while hollowing out the human capabilities, institutions, and social investments the metric was always assumed to represent. The question it forces is whether advanced civilization will continue to be organized around human flourishing, or whether the incentive structures of AI-dependent economies will make that orientation economically irrational — and therefore, in practice, abandoned.