The bubble does not have to burst before it ruins lives.

I was too young to really feel the dot-com crash. The housing crisis was different. I lost my job. People in my family lost theirs. Ordinary plans became fear almost overnight.

I remember how quickly the confidence disappeared, and how long the consequences stayed.

That is why I refuse to treat today's AI boom as an abstract argument between technologists and investors. Historic amounts of money, infrastructure and public attention are being committed to expectations that still have to prove they can support themselves.

Stop staring only at P(doom). Watch the money.

SAM ALTMAN. ELON MUSK.
DARIO AMODEI. DEMIS HASSABIS.

The most powerful figures in AI keep the public conversation fixed on intelligence, acceleration and existential risk. Those questions may matter. But that framing is also extraordinarily convenient for an industry asking everyone else to accept concentrated power and unprecedented spending as inevitable.

This is not a claim of secret coordination. It is a claim about incentives and attention. Apocalypse dominates the headline while the immediate economic risk receives far less scrutiny: a capital boom can fail long before the technology does.

If expectations break, the damage will not stay inside a venture portfolio. It will reach workers, households, suppliers, cities and families. These are people who never voted to make their future depend on this cycle.

P(Bubble) exists to make that risk impossible to ignore. One probability estimate, every day. A reason to look harder, question the certainty and prepare before the consequences arrive at the front door.

We may be early. We may be wrong.
Waiting for certainty is the risk.