Reasoning
Major central banks and economic institutions (Fed, ECB, IMF, World Bank, OECD) have already begun publishing cautionary work on AI productivity assumptions—the IMF's October 2023 report on AI and inequality, the Federal Reserve's research on productivity measurement challenges, and the Bank for International Settlements' 2024 warnings about AI overestimation precedents all establish clear precedent. Given that 2026 is 18+ months away and current productivity forecasts embed highly optimistic AI assumptions (with many models assuming 1.5-2% annual AI-driven productivity gains through 2030), institutional pressure will intensify as real-world deployment gaps become evident. The resolution criteria "major" economist/central bank and "tempering" assumptions are relatively permissive—they don't require fundamental reversals, just published corrections/cautions, which aligns with the natural regulatory and research cycle of major institutions.Key uncertainty
Whether "tempering" is interpreted as requiring explicit downward revisions of productivity forecasts versus general cautionary commentary about measurement uncertainty; many institutions may publish risk-weighted analyses rather than point-estimate reductions, creating ambiguity around resolution criteria.