The AI industry needs staggering capital to survive through 2030, and this piece puts hard numbers to that need. OpenAI has committed $750 billion in compute spending through 2030. Hyperscalers carry $1.65 trillion in off-balance-sheet obligations. Ed Zitron's latest in 'Where's Your Ed At' moves past the rhetoric and builds a company-by-company accounting of what Meta, Google, Microsoft, Amazon, Oracle, NVIDIA, Broadcom, CoreWeave, Nebius, IREN, OpenAI, Anthropic, and SpaceX actually require over the next three fiscal years.
The piece frames the core problem clearly: AI boosters treat incomprehensible numbers as proof of confidence rather than risk. Zitron compares the current posture to the metaverse era, when Zuckerberg destroyed $80 billion before anyone called it a failure. The parallel is not decorative. It is the analytical lens the whole article runs through.
What makes this worth reading in full is not the conclusion but the construction. Zitron draws on consensus analyst estimates and his own research to produce a consolidated demand figure for AI compute, something few outlets have attempted at this level of specificity. The number, when you see it, lands hard.
[READ ORIGINAL →]