Top economist Steve Hanke told us why he doubts AI will be the job destroyer many expect
Executive Take
If Hanke is right, the workforce planning question shifts from "which roles get automated" to "which AI use cases actually clear the cost-per-task bar against human wages" a much narrower filter than most transformation roadmaps assume.
Executive Summary
Economist Steve Hanke argues AI won't cause mass job losses because running AI is expensive, requiring huge water, power and chip resources, making human labor cheaper in many cases. Microsoft, Alphabet, Amazon and Meta plan roughly $700B in 2025 capex, rising to $1 trillion by 2027. Hanke calls AI's cost structure fundamentally different from software's near-zero marginal cost.
Why It Matters
AI and HR leaders building headcount and reskilling plans around aggressive automation timelines should weigh the counter-argument that compute costs, not capability, may be the real constraint on job displacement. This is most relevant to AI leaders sizing infrastructure bets and HR leaders modeling workforce reductions tied to AI adoption.