AI's Economics Don't Make Sense
pricingagentsmodelscapitalinfrastructure
read at source ↗ www.wheresyoured.at
AI’s Economics Don’t Make Sense
Source: Where’s Your Ed At Date: 2026-04-28 URL: https://www.wheresyoured.at/ais-economics-dont-make-sense/
Summary
Ed Zitron’s synthesis of AI economics: data centers at 16.7% gross margin with 100% tenancy (before debt service), Claude Code averaging $13/developer/day, Goldman Sachs data showing 10% of headcount spent on AI tokens, OpenAI CFO “worried about future compute contracts,” $852B needed by 2030 to sustain data center commitments. The subscription model was always untenable — it hid real costs to grow user bases.
Implications
- The structural bear case for AI economics deepens with hard data on infrastructure margins
- Token-based billing transitions (GitHub June 2026) are an industry-wide correction, not a vendor decision
- The gap between builder productivity claims and financial sustainability widens
- Professional/enterprise willingness-to-pay becomes the critical variable