Aswath Damodaran: AI market will be more painful than dot-com bubble
Aswath Damodaran, a professor of finance at New York University, warned that the artificial intelligence market could become even more painful than the dot-com bubble of the early 2000s. This was reported by Zamin.uz.
According to the expert, the current situation fundamentally differs from past technological bubbles due to its cost structure and capital intensity. This is reported by ixbt.com.
Speaking on the Intangible Economy podcast, Damodaran recalled that the internet boom of the late 1990s was primarily focused on software and virtual projects. Today’s AI industry, however, demands massive physical infrastructure—data centers, computing power, and energy systems.
The fact that much of this is being financed through debt amplifies the risk. Combined with capital immobility and economic fragility, as the professor emphasized, a sharp market correction could hit not only shareholders but also creditors and taxpayers.
This could cause the bubble to burst beyond financial markets and spill into the real economy. AI technologies do not behave like traditional IT businesses: unlike software, where marginal costs are near zero, each additional AI query requires significant computational resources.
Damodaran compares this model to Spotify: every stream has a small but real cost. This defies the classic economics of platform businesses, where value typically grows with user base—here, scaling may actually drain company finances under low-margin conditions.
Moreover, price competition is intensifying with the emergence of cheaper alternatives like China’s DeepSeek. Referring to both the technological and social consequences of AI, the economist labels the most optimistic scenario “AI horror.”
If the technology delivers on its promises and sharply boosts productivity, the impact will go far beyond automating isolated tasks. This process could displace “white-collar” workers—office employees performing a large share of their duties.
The market, however, is not currently pricing in these social costs. While tech giants are building infrastructure with 10-year depreciation schedules, rapid technological change raises the risk that these assets could become obsolete within five years.
From this perspective, Damodaran views Apple’s cautious strategy as wise: instead of aggressive investment, the company is opting to observe the market and avoid costly mistakes.
According to ixbt.com, it remains unclear when the massive dollar-scale investments in AI infrastructure by companies like NVIDIA, Microsoft, and Google will begin to pay off. Damodaran’s forecasts serve as a call to investors: avoid becoming the next victims of a technological bubble.
