Start at the paying user
The cleanest starting point is the customer paying for a useful service. Work backward from that receipt: what must the application provider pay the model developer, what must the model developer pay the cloud operator, and what must the operator spend on hardware and power?
Then add financing in a different color
Outside capital can fund the gap while the market develops. That is normal in growing industries. The analytical problem begins when investment-funded spending gets counted as proof of self-sustaining end demand without showing the difference.
If several suppliers depend on the same customer raising money, their revenue forecasts are not independent pieces of evidence. Owning all of them can create more exposure to one funding condition than the number of tickers suggests.
The question the map should answer
Where does the system generate cash from final users, and where does it consume newly raised capital? A useful map need not forecast every future transaction. It should reveal the points where a financing interruption would force spending plans to change.
That is a better foundation for diversification than dividing the portfolio into chips, clouds and software and assuming the labels remove the common dependency.
| Question / exposure | What it means | What to check |
|---|---|---|
| End users | Pay for services | Source of operating demand |
| Operators and models | Buy capacity | May require outside funding |
| Suppliers | Sell infrastructure | Share customer funding exposure |