Connect your Claude, Codex or OpenClaw for full transcripts on millions of podcasts. Two week free trial, cancel anytime.

Start free trial →
SG Scott Galloway Scott Galloway Professor of marketing at NYU Stern and author of bestselling books including The Four and The Algebra of Wealth.

“Average annual infrastructure spending as a percentage of GDP: railroads, 1870 to 1890, on average every year, 2.2% of GDP. Highways, 1956 to 1973, 1.13%. Telecom and fiber, 1996 to 2003, 1.1%. The AI era, so far, 3.6% of GDP. The railroads did transform the world, but along the way there was a ton of bankruptcies.”

Prof G Markets · Money & Markets · November 2026

“Average annual infrastructure spending as a percentage of GDP: railroads, 1870 to 1890, on average every year, 2.2% of GDP. Highways, 1956 to 1973, 1.13%. Telecom and fiber, 1996 to 2003, 1.1%. The AI era, so far, 3.6% of GDP. The railroads did transform the world, but along the way there was a ton of bankruptcies.” — Scott Galloway, Prof G Markets

Context

Galloway reads figures from a chart prepared by the show's research team comparing infrastructure spending on past transformative technologies with AI, to ask whether the AI build-out ends in a crash or a slow decline.

Listen to the episode → More from Scott Galloway → More Money & Markets quotes →

Transcript

Prof G Markets Around 52:01 into the episode
Scott Galloway

…mouse or a license. We'll pick a preferred AI provider and we'll do what we did with search, just get 98% gross margin licensing fee or whatever. The other company that's been rewarded lately is Microsoft. Of the five hyperscalers, Microsoft is the only one that has remained cash flow positive.

Ed Elson

They're doing it right, I think. Yeah.

Scott Galloway

Well, they've thread the needle, right? Whereas Google or Alphabet appears to be on the wrong side of the needle. But basically, the markets have already priced in an expectation of exceptional growth from these businesses. And they're all delivering against that exceptional growth. But where it's parsing the winners from the losers is: does there appear to be an adult with both hands on the capex wheel here or not driving down this highway at all? But the two companies right now that the market likes the most are the one that's not playing in traffic, it said, I'm out. And two, the other one that appears to, it's the only one that's capex was lower than projected, right? The markets seem to like that right now. The markets need to say we need to see the R on the I here. And part of that is not only about top line growth, but managing cost discipline.

Ed Elson

And I think on this one, the markets are right. Like, you know, I always love when I think the markets are wrong because it means I can have a very hot contrarian take on things. But I think that they're, I mean, I don't think they're right about Apple personally, but with Microsoft and the way…

Scott Galloway

I mean, what'll be interesting to see is how they respond to Anthropic because Jury's still out on that one. I want to put some numbers. The research team put together a great chart and we'll put it up here. The average annual infrastructure spending is a percentage of GDP for some other kind of…

Ed Elson

If you were to look at just the extremely levered AI names, like if you were looking at the neo clouds, like the bubbles popped. Those names are crashing. Like that's what a crash looks like. But the difference is that with this AI build

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.

More from Prof G Markets

“I believe NVIDIA has agreed to purchase or participate in the debt offering of Meta. And some of that money will be used to buy NVIDIA chips. So NVIDIA is not only selling the picks and the shovels, they're financing the mine, and they own the bank that's loaning the money to finance the mine to buy those shovels. It is literally the mother of all circular deals.” — Scott Galloway, Prof G Markets “Apple does not get enough credit. They're constantly second-guessed: why don't you have a cheaper computer? Why aren't you in AI? They have more discipline around what not to do than any company of that size. They put one chip down on VR. I think they spent a billion dollars on it. Meta spent 60 on it. They are better at saying no. They pulled the plug on the car. … The only Mag 7 company that has not decided to play in traffic in the AI arms race is significantly outperforming the others who are in this AI arms race.” — Scott Galloway, Prof G Markets