The current discourse around artificial intelligence is filled with an almost comic intensity: Is AI in a bubble? Will it burst? And if so, when? Commentators offer polarized forecasts. On one side are the techno-optimists promising an avalanche of unprecedented productivity: on the other, the alarmists predicting societal collapse. Often, the evidence marshalled to support these claims rests on the soaring stock prices of AI-proximate companies. That is a shaky foundation for reasoned debate. Except for the question of timing, the other two questions—whether AI is in a bubble and whether that bubble will burst—are not as mysterious as they are often portrayed.
We have, after all, seen this movie before. The dot-com crash of 2000 remains the canonical example. Technology bubbles follow a recognizable script: excitement, over-exuberance, investment frenzy, inevitable disappointment. Just as merely having a website boosted stock valuations, incorporating AI into today’s announcements achieves a similar effect. Large tech companies and boosters seek to propagate the inevitability of AI progress. Gartner even formalized this rhythm as the “hype cycle.” However, the hype seems more acute in the case of AI. Unlike most earlier technologies, which offered relatively specific functions bounded by design, AI is different. It is not so much a tool as an affordance—a set of possibilities waiting to be elicited. Ask, and it answers. Tell it to do and it does. And it seems to keep getting better. This makes its value proposition harder to pin down, feeding rampant speculation in lockstep with the dazzling technical advances. That speculation, in turn, is the engine of the bubble.
So yes, there is a bubble. And yes, it will burst. Precedent teaches us that. But logic confirms it. Rarely does technology leap seamlessly from lab demonstration to enterprise-wide productivity. AI is no exception. While narrow, micro-applications show promise, scaling them across organizations requires costly reconfiguration of processes, training, and alignment. Moreover, AI straddles a deeper tension: the temptation to automate for quick cost reductions versus the harder, slower path of augmenting human work for innovation and growth. Add to this the commoditization dilemma—if foundational models are largely similar, how much differentiated value can any one firm truly capture? —and the picture becomes even murkier.
Complements matter too. No technology delivers transformative value on its own. It requires a lattice of complementary investments: infrastructure, societal adaptation, and regulatory clarity. Billions are already flowing into chips, data centers, and energy capacity, but the returns are not yet visible in the financial performance of most firms. The trillions in AI-inflated market capitalization rest on expectations that have yet to materialize.
A slowdown is inevitable. Scaling laws are bending, as GPT-5 illustrates, and visible public failures will shift the narrative from wonder to disillusionment. When that happens, the bubble will deflate—just as it always does. But this is not the end of AI. Bursts do not mean technological death; they are catalysts for refinement. The Internet, too, went through its boom-and-bust cycle, yet today it is rarely spoken of as a discrete technology. It is woven invisibly into the fabric of everyday life. AI is likely to follow a similar trajectory: less as a set of standalone marvels, more as an embedded infrastructure underpinning commerce, communication, and creativity.
The AI bubble, then, is not a verdict on AI’s destiny. It is a predictable stage in the maturation of a technology that may yet prove to be the most significant of the century. The burst will come—but in the long arc of technological revolutions, it is but a punctuation mark, not a full stop. This, of course, is no solace for those who are on the wrong side of the stock market when the crash comes. Perhaps the real question is not whether AI is in a bubble, but whether we are. Our collective imagination—swinging wildly between utopia and apocalypse—may itself be the most inflated asset of all. When the froth settles, AI will not vanish, nor will it deliver instant salvation. It will simply seep, like electricity or the Internet, into the background of human life. And in that quiet embedding lies its true, profound, enduring revolution. So, AI will endure, but how effectively will our human capacity to interpret, critique, and imagine endure in an AI world?
Leave a Reply