There’s plenty of debate around whether current AI investments are justified or whether they’re now in bubble territory, but a veteran investor believes that both things can be true at once.
Bill Gurley, the prominent venture capitalist at Benchmark, has offered a compelling framework for understanding the seemingly contradictory state of artificial intelligence today. Drawing on historical patterns of technological disruption, Gurley argues that genuine revolutionary technology and speculative excess aren’t opposing forces—they’re inseparable companions.
“I think this is super interesting,” Gurley said on the Tim Ferriss podcast. “My partner Peter reminded me of a book that we had seen a while ago by Carlota Perez. It has this very benign title, Technological Revolutions in Financial Capital, and it was written in 2002.”
Perez’s work, Gurley explains, provides a lens for cutting through the binary thinking that dominates conversations about AI. “What Perez kind of simplifies and notices, which I just find perfect for trying to understand whether there’s a bubble or not, is that every time there’s been a technology wave that leads to wealth creation, especially fast wealth creation, that will inherently invite speculators, carpet baggers, interlopers that want to come take advantage of it. Think of the gold rush.”
The false dichotomy, Gurley suggests, is particularly frustrating. “People want to make it a debate: Do you believe in AI or is it a bubble? And if you say you think it’s a bubble, they say, ‘Oh, you don’t believe in AI,’ like this gotcha kind of thing.”
But Perez’s historical analysis reveals a different pattern entirely. “If you study Perez, and I think this is absolutely correct, if the wave is real, then you’re gonna have bubble-like behavior. Like they come together as a pair precisely because anytime there’s very quick wealth creation, you’re going to get a lot of people that want to come try and take advantage of that or participate in it, so you get a flood of those types of people coming at it.”
The conclusion, Gurley says, is both straightforward and counterintuitive: “So it’s odd. There’s a real technology wave that’s fundamentally changing the world. And there’s also massive speculation simultaneously.”
There are varying perspectives in the AI world over whether it is currently in a bubble. Y Combinator’s Paul Graham says that AI is very much real, and it’s not a bubble. OpenAI CEO Sam Altman has hinted that AI could be in a bubble. Meanwhile Google DeepMind’s Demis Hassabis believes that parts of the AI industry are in a bubble, while Yann LeCun says that AI is a not a bubble in terms of apps, but is a bubble in terms of believing we’re going to soon replicate human intelligence.
Gurley’s perspective arrives at a moment when AI investment has reached extraordinary levels. Companies have poured hundreds of billions into AI infrastructure, with tech giants like Microsoft, Google, Amazon, and Meta each committing to capital expenditures exceeding $50 billion annually, much of it directed toward AI capabilities. Nvidia, the chipmaker powering much of the AI boom, is now the most valuable company in the world, while AI startups command valuations that would have seemed fantastical just years ago—OpenAI alone is reportedly valued at over $500 billion.
Yet concerns about a bubble have intensified alongside the investment frenzy. Critics point to AI products that have yet to generate returns commensurate with their costs, questions about whether current large language models can continue improving at recent rates, and a proliferation of startups that appear to be AI in name only. The pattern echoes previous technology cycles: the railway mania of the 1840s, the dot-com bubble of the late 1990s, and the more recent cryptocurrency boom all combined genuine innovation with spectacular excess.
Perez’s framework, which Gurley champions, suggests this duality is not a bug but a feature of transformative technological change. Her research on five major technological revolutions—from the Industrial Revolution to the information age—shows that each followed a similar pattern: an installation phase marked by speculative investment and infrastructure building, often culminating in a financial bubble and crash, followed by a deployment phase where the technology’s real economic benefits materialize. The speculation, while often destructive to individual investors, serves a purpose: it funds the rapid buildout of infrastructure that later generations use productively. The railroads built during the 1840s mania eventually knitted together national economies; the fiber optic cables laid during the dot-com era became the backbone of cloud computing.
If Gurley and Perez are right, the appropriate response to today’s AI moment isn’t to choose between belief and skepticism, but to recognize that both the transformation and the excess are real—and that navigating the difference between genuine innovation and speculative froth will determine which investments create lasting value and which become cautionary tales for future historians of technology booms.