Michelle McPhee

And if you're certain how this is going to turn out, you're crazy.

That said, I think the basic reason to think that something is a bubble is to boil it down to basics, spending is outrunning revenue, right?

In the 19th century, in the 1800s, the reason that the railroads were a bubble is that they were taking on debt to build rail that was not used.

And so they would take on debt, they would spend all this money to build out the rail, and they couldn't get money coming in because no one was living at the other end of the railroad.

And as a result, they went belly up over and over and over again in the panics of the 19th century.

Or you look at, say, the dot-com bubble, where a bunch of fiber optic cable was laid in the ground, hoping that demand would materialize.

It didn't materialize on time.

Instead, the Netflixes of the world, the YouTubes of the world that use that fiber optic cable didn't really come online until like 15 years after it was put in the ground.

And as a result, it was a short-term bubble.

So the critical question that we should ask here when we're trying to understand if AI is a bubble is...

Is the demand, is the revenue going to show up on time?

Last year, I think the wise answer to that question was no.

This is simply way too much money to materialize before these companies are going to realize that their cash flows are absolutely destroyed by spending all of this money on demand.

data centers and chips and electricity and all of that.

But if you're living in a world where we're watching the fastest growing companies in the history of modern capitalism quadruple their annualized run rate every six months, well, that has to make you less certain that it's a bubble, even if it's moving, say, your overall bubble percentages, as it is for me, from like 55% yes, it's a bubble to like 40% it's a bubble.

So the

That's basically where I am.

I'm moving a little bit on the margins by trying to answer like a question based on first principles.

Is the revenue going to catch up to spending?

Last year, I thought no.