Circular financing and bubble risk
Contrasts capital commitments with precedents of overinvestment.
OpenAI: 17 Times Worse Than the Dot-Com Crash. And Nobody Is Ready. →Analysis method · Distinctive attribute
Real metrics versus PR in the AI industry
An AI number is useful only when we know what it measures, what it is compared with and which outcome it should produce. This method separates funding, spending, capability, adoption and return so narrative cannot substitute for the business.
The central rule
Money raised is not revenue; capex is not utilized capacity; technical capability is not adoption; growth is not profitability; valuation is not realized value.
Five layers that must not be confused01—05
- 01
Capital
Funding rounds, debt, valuation and terms. They show access to resources, not business validation.
- 02
Spending
Capex, compute, hiring and acquisitions. They show commitment and speed, not return.
- 03
Capability
Models, infrastructure and available product. They show what can be done, not how much it is used.
- 04
Adoption
Users, customers, consumption, retention and distribution. They connect capability with real demand.
- 05
Outcome
Revenue, margin, cash, productivity or market share obtained. They allow the original thesis to be tested.
Cases from the archive7
Valuation without revenue
Oklo shows why market capitalization, licenses and operations are different variables.
Oklo Inc: The Revenue-Less Nuclear Emperor Worth $20 Billion →Debt-funded capex
Oracle connects AI investment, the balance sheet and the market response.
Oracle: The AI Debt Bomb That Horrified Wall Street →Record operations and a market selloff
Nvidia shows why a good result may still miss expectations already embedded in price.
Nvidia earns more than ever. The market punishes it harder than ever →Massive spending without product leadership
Meta allows announced capex to be compared with observed competitive capability.
$135 Billion and the Engine Won't Start →The market separates conversion from promise
The Big Tech comparison shows different reactions to similar investment plans.
Big Tech Has Shown Its Cards. We've Seen the Flop →Growth is no longer enough
Broadcom illustrates the shift from aggregate growth toward quality and sustainability.
Growth Is No Longer Enough: Broadcom's Warning to the Nasdaq →Checklist for reading an AI number
- Is it an observed figure, a forecast, a commitment or a valuation?
- Does it belong to capital, spending, capability, adoption or outcome?
- What is the period and comparison base?
- How much truly belongs to AI rather than the existing business?
- Which outcome should appear, and within what time frame?
- What incentive does the source have to frame the figure this way?