Episode summary
AI Bubble vs Dot-Com Bubble: The Difference Is Demand
Summary
The host argues that AI may look bubble-like in some pockets, but it does not yet resemble the dot-com bust in the way capital is being deployed. In the 1990s, telecom companies built fiber on leverage and hoped demand would appear; much of that capacity went dark and the companies failed. Today, the host says Microsoft, Amazon and Google are being pulled by customer demand and are racing to add data-center and model capacity, with demand for model spend up almost twentyfold.
Where the host does see bubbles is in specific inputs to the buildout. Semiconductor PPI is up 26%, DRAM contract prices jumped as much as 98% in the first quarter, and memory prices for smartphones rose 300%. He says this reflects a temporary squeeze that should ease in late 2027 when Chinese and Korean capacity comes online, even though he expects many of those memory makers to get hurt when supply floods the market.
On balance, the host says the aggregate market still does not look like a full bubble because hyperscalers are building against real demand rather than speculative excess capacity. He warns, though, that the bubble will worsen if the Mag 7 and other hyperscalers stop spending.
Key takeaways
- The host says the key difference from the dot-com era is that AI capacity is being built to meet existing customer demand, not on the hope demand will arrive.
- The host says demand for model spend is up almost twentyfold, which is why Microsoft, Amazon and Google are racing to expand capacity.
- The host says semiconductor PPI is up 26%, DRAM contract prices jumped as much as 98% in the first quarter, and smartphone memory prices rose 300%.
- The host expects relief in late 2027 when Chinese and Korean capacity floods the market, which he thinks will pressure memory makers.
- The host says the aggregate AI trade is not yet a bubble, but it could turn into one if hyperscalers start building excess capacity or stop spending.
Notable quotes
“I say, "Yeah, yes and no." You know, he's like, "It feels very much like the dot-com bubble”
“With AI, you are building capacity and you're hoping you can keep up with the demand”
“They used a lot of leverage to build that fiber, of course. They hoped the demand would come”
“demand for model spend is up almost 20-fold”
“I don't see that quite yet. You know, again, when you look at these quarterly reports”
How the episode opens
Good morning. This is Dylan Jovine with Behind the Markets. Happy Thursday.
Today is Thursday, October 29th. Are we in an AI bubble? I think that is the question that I am asked most often.
I was talking to a buddy of mine this weekend. And you know, we're talking this weekend and he's a bond trader, was a bond trader, still does a little bond trading for Wall Street. He was around during the '90s and I was around during the '90s.
He's about 10 years older than me. He said, "You know, you think we're in a bubble? I say, "Yeah, yes and no.
You know, he's like, "It feels very much like the dot-com bubble. I say, "Yeah, it does feel like the dot-com bubble. But there are some very, very big differences between the AI bubble and the dot-com bubble.
And we are in a bit of a bubble, especially when you look at specific areas. But you know, when you think of a bubble, especially with the '90s and other bubbles, if I can go on with biotech bubbles, these other bubbles that I've seen and studied. When you think of a bubble, you think about building capacity and hoping the demand comes to you, right?
You're like, "I'm going to build this and I hope demand comes. I imagine a future where demand comes. With AI, you are building capacity and you're hoping you can keep up with the demand.
I mean, it's a very, very, very different dynamic. It's not like the telecom boom in the '90s, not yet at least, like the telecom boom in the '90s. I remember in the '90s, they were building a lot of fiber, okay?
They used a lot of leverage to build that fiber, of course.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai bubble vs dot-com bubble
- demand for ai compute capacity
- hyperscalers capex
- semiconductor pricing
- memory supply cycle
Companies and tickers mentioned
- MSFT
- AMZN
- GOOGL
- GOOG
- Microsoft
- Amazon
Watch the episode
Watch “AI Bubble vs Dot-Com Bubble: The Difference Is Demand” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from Behind The Markets Podcast
-
The Mag 7 Sell-Off: My 3 Secrets to Picking Stocks -
Morgan Stanley’s AI Winners List Is Full of Bank Stocks -
Anthropic IPO: Why I’m Not Rushing to Buy at $1 Trillion -
Chip Stocks Sell-Off: Why the AI Build-Out Just Went From 3 to 5 Years -
Midterm Election Results: The Stock Market’s Winners and Losers -
Google Stock: Why AI Mode Earns Less Than Search
Behind The Markets Podcast, this episode and its artwork belong to their owners. Octagon is not affiliated with the show; this page is an independent summary.
