Silicon Valley leaders have debated for months whether the AI boom is a bubble. Ben Horowitz says that debate misses what’s really happening.
In a wide-ranging discussion on “The A16z Show” on Tuesday, the cofounder of the VC firm Andreessen Horowitz said AI represents something larger than past tech waves — including the internet — and that the eye-popping valuations can’t be understood without looking at what’s actually happening underneath the surface.
“AI is a new computing platform,” Horowitz said. In his view, that puts it in a different category from incremental software shifts.
“This is a bigger technology market than I’ve ever seen,” he added.
Bubble fears
Skeptics, including OpenAI CEO Sam Altman, Microsoft cofounder Bill Gates, and hedge fund billionaire Ray Dalio, often point to how fast valuations have risen to predict a looming bubble.
But Horowitz said that focusing on prices alone ignores a more important signal: demand.
“One of the reasons why people are so worried about it being a bubble is, you know, the valuations have gone up so fast,” he said. “But if you look at what’s going on underneath in terms of the customer adoption, the revenue growth rates — we’ve never seen demand like this.”
That disconnect, he said, explains why AI feels unsettling even to seasoned investors.
“We’ve never seen valuations rise like this, but we’ve never seen demand rise like this either,” Horowitz said, describing the current moment as “a bit of a brave new world.”
More winners
Horowitz also pushed back on the idea that AI will produce only a handful of dominant winners, as the internet did — a view promoted by former Shark Tank star Mark Cuban.
Horowitz sees a much larger opportunity set.
“It’s a very big design space — an enormous design space like one we’ve never seen before in technology,” he said, predicting more billion- and even $10 billion-plus companies than in prior cycles.
Part of that expansion comes from how AI is being built. Rather than a single, all-powerful model doing everything, Horowitz said real products require complex applications that deeply model human behavior — work that can’t simply be absorbed by foundation models.
The result, he believes, is a technology shift that is both messier and more powerful than past revolutions — and one where fears of a simple bubble may underestimate just how much is changing.
