Summary

Zitron argues that hyperscalers have trapped themselves in a self-reinforcing debt spiral: the more they spend on AI infrastructure, the more expensive everything becomes — debt, RAM, construction, talent — while generating no meaningful disclosed revenue. He methodically documents NVIDIA's overstated Blackwell efficiency claims, CoreWeave's near-junk bond spreads, and the proposed NVIDIA-SoftBank-OpenAI $250B data center deal as evidence of circular financing masking absent real demand. The piece contends that AI adoption in enterprises is largely performative kayfabe, driven by executive social pressure rather than genuine productivity. Stopping the capex would force a reckoning; continuing it deepens an unavoidable hole. He concludes this ends badly for everyone except bankers, ratings agencies, and a handful of VCs.

Key Insight

The AI infrastructure buildout has created a mathematically inescapable trap: hyperscalers cannot stop spending without collapsing their stock valuations, yet continued spending deepens debt that markets are already losing faith in, all while the underlying technology fails to generate the trillions in new revenue required to make any of it rational.

Spicy Quotes (click to share)

• 9

  AI, as it stands, is an exercise in kayfabe — a thing people are taking seriously because the rich and powerful are saying they have to and a media ecosystem built to celebrate them says that you need to.

• 8

  This is the final boss of circular financing.

• 8

  Every layer of the AI industry exists to be exploited by the layer above it.

• 5

  The more they raise, the more it costs. Then there's the other, more-obvious problem — the more they buy, the more they spend.

• 8

  NVIDIA's revenues are, for the most part, propped up by FOMO rather than any real relationship to revenues, productivity, or reality.

• 7

  Hyperscalers will, by the end of the year, have sunk over $1.5 trillion in capital expenditures and equity investments into Large Language Models that have yet to provide good enough revenues to actually disclose.

• 8

  The only way that NVIDIA can continue beating and raising each earnings season is to manufacture these massive deals, all to avoid discussing the blatantly obvious truth that diverse demand does not exist.

• 7

  LLMs are economically and practically unsuited to the tasks they've promised to solve, and no amount of extra capital expenditures or venture capital dollars will magically fix that.

Tone

combative, investigative, sardonic