Big Tech Pours $1 Trillion Into AI, Fueling Bubble and 1.1% Global Energy Drain
Updated
Updated · Jacobin magazine · Aug 5
Big Tech Pours $1 Trillion Into AI, Fueling Bubble and 1.1% Global Energy Drain
3 articles · Updated · Jacobin magazine · Aug 5
Summary
$1 trillion in Big Tech capital spending since the AI boom has gone largely to data centers, chips and power systems, with the report arguing the buildout now resembles an investment bubble vulnerable to collapse.
OpenAI and Anthropic remain unprofitable even as they pursue $1 trillion valuations, while suppliers such as Nvidia and Oracle are tied into financing loops that depend on LLM makers eventually delivering returns.
Competition from Gemini, Copilot and cheaper Chinese models, along with weak evidence of lasting productivity gains for business users, raises doubts that AI services can generate enough profit to justify the spending.
A bust could leave hundreds of oversized data centers and much cheaper compute, echoing the 1990s fiber-optic overbuild, while stronger tech groups buy distressed rivals and further concentrate market power.
The environmental costs are already mounting: AI is estimated to account for 1.1% of global energy demand in 2026, using more power than the UK and intensifying water stress and local heat effects.