Big Tech Needs $2 Trillion In AI Revenue By 2030 or They Wasted Their Capex

Source: Where’s Your Ed At Date: 2025-10-31 URL: https://www.wheresyoured.at/big-tech-2tr/

Summary

Ed calculates that big tech has invested ~$776B in AI infrastructure over three years without generating commensurate revenue, and must produce $2 trillion in AI revenue by 2030 to justify the capex. Microsoft’s disclosed AI ARR of $10-13B is partially circular — OpenAI’s discounted compute spend “essentially only covers Microsoft’s costs.” Everyone except GPU and server sellers is losing money on AI. GPU assets depreciate rapidly with annual hardware releases, making the infrastructure increasingly worthless without revenue to offset.

Implications

This is the quantitative backbone of Ed’s AI financial sustainability thesis. The $2T revenue target by 2030 is the falsifiability benchmark: if AI revenue falls short, the capex was destroyed value. The circular revenue observation (Microsoft counting OpenAI’s compute purchases as AI ARR) means reported AI revenue figures from hyperscalers may be substantially inflated by related-party transactions. GPU depreciation also matters: the asset base underpinning AI valuations is not durable capital but consumable hardware with a 2-3 year useful life, making the “AI infrastructure investment” framing misleading on a balance sheet basis.