AI infrastructure's profitability hinges on whether revenue from top AI labs reaches $8 billion monthly by year-end—the single most critical metric determining if this historic tech boom continues or stalls. Brad Gerstner breaks down why semiconductors drove 70% of market gains, why 43 gigawatts of new compute capacity next year is unlikely, and what rate hikes mean for AI's $1.5 trillion annual capex buildout. Anthropic and OpenAI's combined run-rate revenue must hit $180 billion by year-end to justify the $1.5 trillion annual AI infrastructure capex hyperscalers are committing to.