AI Mania: A Bubble Bigger Than the Dot‑Com Craze?

The Echoes of 1999

When Torsten Sløk, chief economist at Apollo Global Management, issued a stark warning this month that the current artificial intelligence stock boom may be more dangerous than the infamous dot-com bubble, it wasn’t just another talking head voicing contrarian doubts. Sløk, a veteran observer of financial cycles, isn’t known for hyperbole. His concern is rooted in numbers and history.

According to Sløk, the current valuations of major AI-driven companies now surpass those seen during the peak of the late 1990s tech bubble. The top 10 companies in the S&P 500, including giants like Nvidia, Apple, Microsoft, Amazon, Alphabet, and Meta, are trading at valuations that are not only historically high but in some cases exceed the most speculative levels of the 1999–2000 era. Investors, he warns, may be deluding themselves into believing this time is different. But is it?

A Rally Built on Narrow Shoulders

Unlike the late 1990s, where exuberance was spread across a vast array of tech upstarts—many of which had no revenue and only vague business plans—the current AI stock rally is highly concentrated. The bulk of the gains in the S&P 500 over the past two years have come from a handful of companies. Nvidia, whose graphics processing units power many of the large language models and AI training processes underpinning today’s revolution, has become the poster child of this boom. Its stock price has tripled in the past year, making it the third most valuable company in the world.

But this concentration is also what makes the market so precarious. When the fortunes of an entire index rest on the performance of a few megacaps, any disruption—whether in earnings, regulation, supply chains, or investor sentiment—can trigger outsized volatility. It’s akin to building a skyscraper on a few stilts; strong winds in one direction could bring the whole structure down.

Sløk calls this a “narrow rally” and sees it as a warning signal, not a sign of strength. He’s not alone. Many analysts have begun to worry about the fragility that comes when market breadth collapses and the weight of expectations is placed disproportionately on a few corporate shoulders.

AI: The Fuel of the Future—or a Mirage?

None of this is to say that AI isn’t real. Few would dispute that artificial intelligence, particularly generative AI, represents one of the most transformative technological advancements in recent decades. It’s already reshaping industries from finance and healthcare to software development and marketing. But the question isn’t whether AI will change the world. It’s whether investors are prematurely baking in decades of future profits into today’s stock prices.

The current investment climate is one of heightened enthusiasm. Analysts project that global corporate spending on AI could reach $340 billion by next year. Tech giants are pouring tens of billions into infrastructure, chip development, and AI startups, all in the hope of capturing early market dominance. This level of investment rivals or even exceeds that of the late 1990s internet gold rush.

Yet, as Sløk points out, high investment does not guarantee high returns. In fact, overinvestment in unproven technologies is a hallmark of asset bubbles. When too much capital chases uncertain or long-dated outcomes, valuations can disconnect from reality. And when reality eventually sets in—when earnings fail to materialize at the scale or speed anticipated—the market can correct brutally.

A Different Kind of Bubble

What makes the current AI rally more insidious, some argue, is that the companies involved are fundamentally profitable. This wasn’t the case in 1999, when hundreds of speculative startups with no revenue flooded the stock market. Today, companies like Microsoft and Alphabet have rock-solid balance sheets, wide profit margins, and real cash flow.

This creates a psychological blind spot. Investors assume that because these firms are not startups, they’re immune to irrational exuberance. But even great companies can become overvalued. Microsoft, for instance, is investing heavily in AI integrations across its suite of products, and while early signs are promising, the monetization path remains uncertain. Will consumers pay more for AI-enhanced tools? Will enterprises upgrade at scale? These are open questions.

Meanwhile, Nvidia’s market capitalization has ballooned to over $3 trillion, driven almost entirely by demand for its AI chips. While its financials are currently strong, they rest on the assumption that AI spending will continue to surge unchecked. Any slowdown—whether from regulation, technological saturation, or economic downturn—could expose how finely balanced these valuations are.

Warnings From the Past

Looking back at the dot-com era, there are striking parallels. In the late 1990s, investors were similarly enamored with a transformative technology—the internet. Companies that added “.com” to their name saw their stock prices soar overnight. Venture capital poured in. IPOs multiplied. And analysts justified sky-high valuations with speculative metrics like “eyeballs” and “page views.”

Eventually, reality caught up. When earnings failed to match expectations and when the Federal Reserve began raising interest rates, the bubble burst. The Nasdaq lost over 75 percent of its value, and it took nearly 15 years to fully recover. Tens of thousands of workers lost their jobs. Millions of investors, many of them retail traders who had been swept up in the frenzy, saw their savings vanish.

Torsten Sløk fears a similar reckoning could await the AI sector. While today’s companies are far stronger than their dot-com counterparts, that doesn’t immunize them from macroeconomic shocks or shifts in investor psychology. All it takes is a few quarters of disappointing results—or a major policy shift in Washington or Brussels—to turn optimism into panic.

Are We Already in a Bubble?

That’s the question on everyone’s mind. Some believe we are. Others argue that the market is merely pricing in long-term innovation. Sløk falls firmly in the former camp. He points to the elevated price-to-earnings ratios of the tech giants as evidence. In his view, these ratios are not only high, they are unjustifiable by historical standards.

Others, like Ray Dalio and Ed Yardeni, have voiced similar concerns. Dalio, the billionaire founder of Bridgewater Associates, recently said that the current market shows clear signs of “bubble dynamics,” particularly in AI-related stocks. Yardeni, a longtime market strategist, described the recent surge as a “melt-up,” warning that such rallies often precede painful corrections.

Still, many on Wall Street remain bullish. They argue that AI is a once-in-a-generation shift, akin to the electrification of the 20th century or the rise of mobile computing in the 2000s. From their perspective, current valuations reflect the future cash flows of a world fundamentally altered by intelligent automation. In other words, the hype is justified.

But even proponents of this optimistic view acknowledge risks. If AI adoption slows, if technological barriers emerge, or if regulators impose constraints on data usage and model training, the path to profitability could be longer and more volatile than investors expect.

Where Do We Go From Here?

The answer depends on a mix of macroeconomic, technological, and psychological factors. On the economic front, falling interest rates could continue to support elevated valuations. If the Federal Reserve cuts rates as expected, and inflation remains subdued, investors may feel justified in paying a premium for future earnings.

Technologically, the AI sector remains in hypergrowth mode. New models are being released at a rapid pace. Open-source alternatives to big corporate models are gaining traction. And businesses across industries are experimenting with AI use cases, from personalized marketing to predictive maintenance. If these applications drive productivity gains, the earnings narrative may hold up.

But the psychological component is harder to gauge. Markets are driven not only by fundamentals but by narratives. Right now, the AI narrative is dominant. It shapes investment decisions, corporate strategies, and media coverage. But narratives can shift quickly. A high-profile failure, a corporate scandal, or a public backlash against AI adoption could shift sentiment overnight.

Sløk’s message is not that AI is a scam, nor that technology won’t transform society. Rather, it’s a call for sobriety in a market increasingly driven by speculative fever. Investors, he argues, should be wary of assuming that current trends will persist indefinitely, and should be especially cautious about companies priced for perfection.

Conclusion: Between Brilliance and Bust

Artificial intelligence is not a fad. It’s real, and its implications are vast. But markets have a long history of overestimating the short-term impact of new technologies while underestimating their long-term effects. The internet didn’t disappear after the dot-com crash. It reshaped every facet of life. But the journey included painful corrections and shattered illusions.

Today’s AI boom may follow a similar path. In the long run, the technology will likely justify some of today’s enthusiasm. But in the short run, the gap between promise and profit could widen, especially if companies fail to deliver earnings growth commensurate with their valuations.

Torsten Sløk’s warning is not a prophecy of doom, but a reminder that markets are cyclical, not linear. For every period of exuberance, there is a reckoning. And the higher we climb on dreams alone, the farther we may have to fall.