OpenAI closed its $122 billion funding round on March 31, producing an $852 billion post-money valuation — the largest private fundraise in history by a factor of four [1]. The numbers are so large they require context to mean anything at all.

SoftBank co-led the round alongside Andreessen Horowitz and D.E. Shaw. Amazon committed $50 billion, though $35 billion of that is contingent on OpenAI achieving an IPO or demonstrating AGI — a condition that makes the number simultaneously impressive and conditional. Nvidia put in $30 billion. SoftBank matched it. Microsoft also participated [1]. The combined capital represents a bet that one company will capture the economic value of artificial intelligence itself.

The revenue is real. OpenAI generates $2 billion per month, up from roughly $1 billion per quarter at the end of 2024. It claims 900 million weekly active ChatGPT users and 50 million paying subscribers [2]. The company extended its revolving credit facility to $4.7 billion and is building what it calls a "unified AI superapp." The product works. People pay for it.

But OpenAI projects a $14 billion loss in 2026. It is not profitable. It is not close to profitable [1]. The cost of training each new model generation exceeds the last. The $852 billion valuation is not a reward for earnings — it is a wager that earnings will eventually justify it. The distance between $2 billion in monthly revenue and $14 billion in annual losses is the space where the future of AI capitalism lives or dies [2].

-- THEO KAPLAN, San Francisco