Danny Moses, the former FrontPoint Partners trader who famously bet against the housing market in 2008 alongside Steve Eisman, is sounding the alarm on potential problems brewing in the AI market. In an interview with Business Insider, Moses drew stark parallels between today’s AI boom and the dot-com bubble of the early 2000s, suggesting investors need to exercise caution despite the sector’s legitimate growth potential.
Moses acknowledges that AI represents a genuine secular growth story, but he’s concerned that the financial mathematics underlying many AI investments are beginning to break down. “The growth was real, but the math didn’t work,” he said, referring to the dot-com era. “And I think that we’re reaching a point where the math is starting not to work” in AI.
Crucially, Moses isn’t calling for investors to short the AI industry entirely. Instead, he’s advocating for a selective approach that focuses on established tech giants with strong balance sheets and sustainable business models. His preferred picks include Amazon, Google, Meta, and Microsoft—companies that can adjust their capital expenditures while remaining cash-flow positive, unlike smaller AI-focused firms that depend entirely on continued AI spending.
Moses specifically warned against certain high-profile names in the AI space. He cited Oracle as problematic due to its high debt levels and the substantial cash requirements needed to fulfill tech client orders. He also flagged Super Micro Computer and CoreWeave as examples of riskier, more volatile plays within the AI trade.
The veteran trader believes the market is finally beginning to differentiate between AI winners and losers, with investors increasingly favoring companies with diversified revenue streams and stronger financial foundations. This sorting process represents a maturation of the AI investment landscape.
Moses also expressed bullish sentiment on uranium, which is gaining attention as a critical component for powering AI infrastructure. However, he cautioned that there’s a significant timing mismatch between investor expectations for AI growth and the actual timeline required to build the necessary power infrastructure. “There’s a mismatch in the timing of how people think that companies will experience AI growth and actually the infrastructure that it’s going to take to power it,” Moses explained, suggesting uranium investments require patience despite their thematic appeal.
Key Quotes
The growth was real, but the math didn’t work. And I think that we’re reaching a point where the math is starting not to work.
Danny Moses drew this parallel between the dot-com bubble and today’s AI market, suggesting that while AI growth is legitimate, the financial fundamentals underlying many AI investments are becoming unsustainable—a warning sign that echoes the conditions before the 2000s tech crash.
They can turn down their capex at any point, and they’re still cash flowing positive, as opposed to these other companies, which are dependent upon that spending within AI.
Moses explained why he favors Big Tech companies like Amazon, Google, Meta, and Microsoft over smaller AI-focused firms. This distinction highlights the financial flexibility that separates sustainable AI investments from potentially vulnerable ones.
I think it’s proof that investors are beginning to sort out the winners and losers of the trade, and they’d much rather have comfort and other businesses with stronger balance sheets to fall back upon to express the AI theme.
Moses observed that the market is maturing beyond the phase where all AI stocks rose together, with investors now discriminating based on financial strength and business fundamentals rather than AI exposure alone.
There’s a mismatch in the timing of how people think that companies will experience AI growth and actually the infrastructure that it’s going to take to power it.
Discussing uranium investments, Moses identified a critical gap between investor expectations and reality regarding AI infrastructure development, particularly around power generation—a constraint that could significantly impact the pace of AI adoption.
Our Take
Moses’s analysis represents a sobering counterpoint to AI market euphoria, and his credentials make it impossible to dismiss. The dot-com comparison is particularly apt: like the late 1990s, we’re seeing genuine technological transformation alongside unsustainable valuations and business models. His focus on cash flow sustainability over growth narratives marks a potential inflection point where the AI market transitions from speculation to fundamentals-based investing.
The infrastructure timing issue he raises is especially underappreciated. While companies race to deploy AI capabilities, the physical constraints—power generation, data centers, cooling systems—operate on much longer timelines. This disconnect could create a significant gap between AI promises and deliverable results, potentially triggering the kind of disappointment that deflates bubbles. Investors who heed Moses’s advice to focus on diversified, financially strong companies may be better positioned to weather the volatility ahead.
Why This Matters
Moses’s warning carries significant weight given his track record of identifying market bubbles before they burst. His analysis suggests the AI market is entering a critical phase where financial fundamentals will increasingly matter alongside technological promise. This represents a shift from the early AI boom period when virtually all AI-related stocks rose together regardless of business viability.
The distinction Moses draws between sustainable AI winners (Big Tech giants) and vulnerable players (highly leveraged or single-focus companies) could signal a broader market correction ahead. For investors, this means the easy money phase of AI investing may be ending, requiring more sophisticated analysis and selectivity.
The infrastructure timing mismatch Moses identifies—particularly around power generation and uranium—highlights a critical bottleneck that could constrain AI growth. This suggests the AI revolution may unfold more slowly than current market valuations assume, with implications for everything from data center construction to energy policy. His perspective serves as a reality check for an industry where hype often outpaces practical considerations.
Related Stories
- AI Lessons from the Dot-Com Era: Steve Case’s Perspective on the AI Revolution
- CEOs Express Insecurity About AI Strategy and Implementation
- Global Funds Turn to Indian Stocks as Hedge Against AI Risks
- Meta and Nvidia Billionaires’ Wealth Soars $152B in AI Boom
Source: https://www.businessinsider.com/big-short-danny-moses-ai-bubble-goog-amzn-msft-meta-2025-12