As is always the case when a technological innovation hits the open market, investors are pouring money into various AI companies, along with companies that manufacture the chips and other hardware and software these companies use. However, one seasoned investor warns that now may not be the time to put more money into a sector that’s quickly becoming the most popular.

According to Bill Smead of Smead Capital Management, the current frenzy around AI firms displays many of the hallmarks of a bubble. He argues that the real question isn’t if the break will happen, but when. He points to companies with eye-popping valuations relative to minimal earnings or operational history. For instance, some firms are raising billions on the promise of AI potential rather than proven revenue streams. The risk isn’t hypothetical. If a key player stumbles, investor sentiment could flip quickly, and a domino effect may cascade through portfolios heavily concentrated in the sector.

Why the AI Sector Looks Fragile

Smead notes the valuations of many AI-exposed stocks are built more on expectation than earnings. He compares it to the dot-com era, where exuberance outpaced revenue. For example, one AI infrastructure company was valued at $60 billion despite only $1.2 billion in quarterly revenue. This gap between actual performance and current valuation means that sentiment shifts or growth falls short, the downturn could be sudden and drastic.

There’s also an issue with the AI industry’s reliance on a handful of mega tech names. When a narrow group drives the majority of gains, the risk of a broad correction increases if one of the names falters. For instance, the technology sector’s weighting in the S&P 500 now exceeds levels seen before previous major corrections. When the market becomes too top-heavy, problems in a single stock or sub-sector have the potential to create a negative ripple effect that goes far beyond a single sector.

Finally, Smead warns that the narrative surrounding AI has the potential to harm the value in the long term. The story of AI has been everywhere, with massive investments and promises of a “once-in-a-lifetime” investment opportunity. Investors of various experience levels have flooded the market. But narratives can reverse just as fast as they build. Smead warns that the circular investment flows and hype echo prior eras of speculative excess. When investors believe the story is overpriced or delayed, the rush for the exits can be swift.

What This Means for Investors & Portfolios

The point isn’t that you should abandon all technology investments in your portfolio. There’s still a very real possibility that AI delivers some incredible value to investors. However, this is an opportunity to assess risk and reward. You need to be sure that your entire portfolio doesn’t hinge on a single type of investment. Smead outlines a defensive posture rather than a bearish one. By reducing over-concentration and increasing exposure to sectors that offer a more stable cash flow, you can protect yourself against the bubble bursting.

When one sector is looping as the dominant growth engine, diversification into under-loved or out-of-favor areas can act as a hedge. Smead recommends putting money into sectors like construction, energy, real estate, and healthcare. These sectors typically have less speculative pressure, which means they face less downside should sentiment shift.

The hardest part of putting this theory into practice is understanding when to make a change. You may not want to move your money out of AI today, but you also don’t want to be left holding bad investments should the bubble burst. As noted in interviews with other veteran investors such as Howard Marks, we may recognize the conditions ahead of time, but not in the moment. Marks calls the U.S. market “expensive, but not nutty,” and warns investors that when valuation excesses exist, “we have to say so and take action and maybe raise our defenses a little bit.” Ultimately, Marks’ words indicate that it’s better to prepare for market corrections instead of trying to predict when they’ll come.

Parallels to the Dot-Com Bubble Burst

Looking back, the dot-com bubble provides a useful template for investors who are navigating the current boom in AI valuations. Companies with little revenue and frequent losses were valued based on potential alone. When profit took time to arrive or regulation changed, valuations collapsed. Smead draws comparisons with some of the current AI leaders. When looking at the current leaders in the AI sector, you’ll find strong narratives, significant capital flows, but unproven long-term earnings structures.

Adding to that, analysts point out “circular” investment loops in current AI ecosystems. That happens when major players in an industry invest in each other or pay for each other’s services. This inflates growth in a self-referencing way, which is what the market saw in the 1990s with the dot-com bubble.

While not every AI company will follow the trajectory of early dot-com startups, the underlying psychology is strikingly similar. Investors often overestimate short-term adoption and underestimate long-term integration, leading to inflated expectations that reality struggles to meet. Some analysts have noted that even established tech giants are not immune. Their valuations increasingly rely on perpetual AI growth projections rather than their core businesses. When early enthusiasm fades or economic conditions tighten, capital tends to flee speculative corners first.

Don’t Ask If It Will Burst, Ask When

The core message from Smead, which has been echoed by several other market analysts, is relatively simple: the AI sector isn’t guaranteed to fall, but all the signs of over-exuberance are there, and when sentiment slips, the reversal may be fast. For investors, the action isn’t dramatic panic, but a prudent adjustment. By diversifying your portfolio and staying ready, you can protect yourself against the coming downward trend.

When the time comes that the momentum fades or earnings disappoint, portfolios anchored in fundamentals and risk awareness will be in far better shape than those tied to the next big thing. This warning isn’t a guardrail, not a signal to abandon technology. Instead, it’s an opportunity to make sure that you’re ready for the fallout when AI valuation comes back down to earth.