The AI Bubble: Why History Suggests We’re in a 'Winner-Take-All' Moment

The current frenzy surrounding Artificial Intelligence has investors asking the same question that defined the late 1990s: Are we witnessing a revolutionary shift in the global economy, or are we inflating a massive, unsustainable bubble?

In a recent episode of Odd Lots, former Merrill Lynch tech analyst and Business Insider founder Henry Blodget joined hosts Joe Weisenthal and Tracy Alloway to dissect the AI landscape. His conclusion? While AI is undeniably transformative, the market structure—specifically the opacity of private valuations and the reliance on massive capital expenditure—mirrors the patterns of the dot-com era in ways that should give retail investors pause.

The "OpenAI" Valuation Paradox

The most striking data point in the current AI market is the $300 billion valuation of OpenAI. To the casual observer, this figure seems detached from reality. However, Blodget notes that professional investors are justifying these numbers based on aggressive 2029 revenue projections of $100 billion-plus.

"If you believe that OpenAI is going to win this next wave in the same way that Google or Amazon ultimately won the internet wave, three times revenue for that... you can see how professional investors who really actually do know the numbers are getting there," Blodget explained.

However, he adds a critical caveat: downside protection. Unlike public market investors buying common stock, private market players are often loading up on preferred stock. This provides them with a safety net that common shareholders simply don't have. For the average investor, the lack of financial transparency in these private deals makes it nearly impossible to gauge the true risk-reward profile.

The "DeepSeek" Risk and CapEx Loops

Blodget draws a direct parallel between today’s AI boom and the 1990s telecom bubble, where massive capital expenditure (CapEx) was fueled by easy access to capital, creating a circular revenue loop.

A major risk to this cycle is what the industry calls "DeepSeek moments"—technological breakthroughs that allow for high-quality AI output without the need for massive, expensive GPU clusters. If a company can achieve similar results with significantly less infrastructure, the current valuation of hardware-heavy giants like NVIDIA (NVDA) could face a reality check.

"Investors should monitor CapEx spending trends," Blodget warns. "While high spending is currently a concern, a sudden, sharp decrease would signal a systemic collapse in AI growth expectations."

The "Winner-Take-All" Trap

Perhaps the most sobering lesson from the 1990s is the fate of secondary players. During the internet boom, investors who missed out on Amazon or Yahoo poured capital into companies like Lycos, Excite, and InfoSeek, assuming they would capture the same upside. Most of those companies eventually went to zero.

Blodget suggests that the AI space will likely follow a similar trajectory. While the "Magnificent Seven" trade at arguably reasonable earnings multiples, the secondary AI startups—often hyped with little more than a white paper and a PhD—are high-risk bets. If history is any guide, the market will eventually crater, and only a handful of companies will emerge as the true winners.

The Death of the IPO Market

Blodget also lamented the current state of the IPO market, which he argues has created a "worst of both worlds" scenario for retail investors.

"By the time you go public, the vast majority of the gains have already been grabbed by professional investors," Blodget noted. This has forced retail investors into opaque secondary markets where they trade shares without access to 10-Qs or clear information on the company’s capital stack.

Key Takeaways for Investors

• Beware of Private Market Opacity: Without access to financial statements or knowledge of the capital stack (preferred vs. common stock), retail investors are at a significant disadvantage in private AI deals.
• Don't Chase the "Next" Big Thing: History shows that secondary players in a tech boom often fail. Focus on the potential for a "winner-take-all" outcome rather than betting on every startup in the sector.
• Watch the CapEx: The sustainability of the AI boom is tied to massive infrastructure spending. A pivot toward efficiency (the "DeepSeek" effect) could render current GPU-heavy business models obsolete.
• Diversification is Key: As geopolitical tensions rise between the U.S. and China, supply chain shifts to regions like India and Vietnam are becoming a rational corporate response. Investors should account for this operational complexity in their long-term thesis.

Ultimately, Blodget remains bullish on the long-term transformative power of AI, but he cautions that the path to maturity will be paved with market corrections. As he puts it, "These things play out. I do think we'll have the same sort of pattern in this sector that at some point we'll hit a peak. Everything will crater... and then a few companies will go on and make all of the money."