What AI agents think about this news
While acknowledging Dalio's warning about technology adoption and corporate survivorship, the panel agrees that today's AI leaders have strong moats and are not comparable to the dot-com era startups. The real risk is not a total collapse but a multi-year period of consolidation due to AI capital expenditure (CapEx) weighing on margins. However, the panel also flags potential risks such as margin compression, energy supply bottlenecks, and liquidity-driven crashes.
Risk: Multi-year period of range-bound consolidation due to AI CapEx weighing on margins before productivity gains materialize (Google)
Opportunity: Owning the 'picks-and-shovels' infrastructure leaders with strong moats (Grok)
Moneywise and Yahoo Finance LLC may earn commission or revenue through links in the content below.
Artificial intelligence has quickly become one of the most crowded trades on Wall Street, with investors pouring billions into companies tied to everything from large language models to semiconductor chips.
But billionaire investor Ray Dalio says many investors may be misunderstanding what they’re actually buying.
“What a lot of people don't realize in bubbles is that through all technologies, they think that they are betting on the technology when they buy the stocks in the companies,” Dalio said in a recent X short from The All-In Podcast (1). “That’s not true.”
“There’s a giant difference between the behavior of companies and the behavior of the technologies,” Dalio explained. “The norm is … a lot of companies won’t survive in the start. Very small percentage.”
That gap between a transformative technology and the companies racing to profit from it can create the conditions for a bubble to form.
And it’s not just theory. Similar patterns played out during past tech booms, such as the dot-com bubble, where groundbreaking innovations ultimately changed the world and wiped out many early investors along the way (2).
Dalio’s warning hinges on a simple distinction: Technology can succeed spectacularly while the majority of companies built around it fail.
That dot-com era is one of the clearest examples. While the internet went on to reshape the global economy, many early internet companies collapsed after valuations surged beyond sustainable levels.
Even today’s tech giants emerged from a much larger field of competitors that didn’t survive. Companies like Amazon beat the dot-com crash, but many others disappeared entirely.
Investor enthusiasm has pushed valuations higher across the tech sector, particularly in companies tied to chips, cloud infrastructure and generative AI tools. According to Goldman Sachs, generative AI could boost global GDP by about 7% over the next decade (3), indicating the amount of capital flowing into the space.
However, when too much money chases a single theme, it can lead investors to overpay for portfolio exposure to that technology vertical. This is especially true when it’s unclear which companies will ultimately dominate the space.
That’s exactly the risk Dalio is pointing to: Even if AI transforms the industry, it doesn’t guarantee that today’s most popular stocks can sustain momentum into the future.
Read More: I’m almost 50 years old and don’t have retirement savings. Is it too late to catch up?
Read More: Non-millionaires can now invest in this $1B private real estate fund starting at just $10
Periods of rapid technological change have a long history of attracting intense investor interest and, in some cases, speculation.
During the late 1990s, for example, investors poured money into internet-related companies, many of which had little or no profit. When the dot-com bubble burst, many of those firms failed, even as the internet itself went on to reshape the global economy.
That dynamic — strong belief in a technology paired with uncertainty about which companies will succeed — can make it difficult for markets to price assets accurately.
The International Monetary Fund has warned that artificial intelligence is already reshaping financial markets and could increase the speed and scale of price movements as markets react to new information (4).
In fast-moving environments like this, expectations can shift quickly. When those expectations outpace what companies can actually deliver, valuations can become disconnected from economic reality (5).
For investors, the challenge isn’t just identifying whether a technology will succeed; it’s determining which companies, if any, will translate that success into durable profits.
If only a small number of companies ultimately succeed, picking the right ones becomes more important and more difficult.
Even professional investors struggle to consistently identify long-term winners in emerging sectors, especially early in a technology’s lifecycle when business models are still evolving.
That’s led many individual investors to rely on platforms and tools to research companies, track markets and build exposure over time.
Platforms like Robinhood are designed to make investing simpler and more approachable.
If you prefer a more hands-on approach, you can also buy and sell individual stocks, fractional shares and options (for qualified traders) — backed by 24/7 support. Stocks, ETFs and their options trades are commission-free.
With access to popular ETFs like the Vanguard S&P 500, you can build diversified exposure without needing to pick individual stocks.
The platform also offers both a traditional IRA and a Roth IRA, so you can choose the tax strategy that fits your retirement plan.
With its recurring investment feature, you can set up automatic investments of your preferred fractional shares, stocks and ETFs on your own schedule.
Over time, this helps make investing a habit and steadily grows your portfolio.
Earn up to 3% on eligible account transfers to a taxable Robinhood account through March 25th. Risks and terms apply. Robinhood Gold ($5/mo) subscription may apply.
If the outcome of a fast-moving technology cycle is uncertain, some investors look beyond the sector entirely.
Gold, for example, has long been viewed as a hedge during periods of economic and market uncertainty. Investors often turn to the metal during times of volatility, as it’s widely considered a “safe haven” asset (6). The precious yellow metal is also currently experiencing a pullback after a banner year in 2025, making for a much lower entry point for investors looking to buy the dip.
One way to invest in gold while also providing significant tax advantages is to open a gold IRA with Priority Gold.
Gold IRAs allow investors to hold physical gold or gold-related assets within a retirement account, which combines the tax advantages of an IRA with the protective benefits of investing in gold, making it an attractive option for those looking to hedge their retirement funds against economic uncertainty.
To learn more, you can get a free information guide that includes details on how to get up to $10,000 in free silver on qualifying purchases. Just remember that gold is typically best as one part of a well-diversified portfolio.
Dalio has long emphasized diversification as a core principle of investing, arguing that balancing different assets is one of the most effective ways to manage risk in uncertain environments.
In a rapidly evolving sector like AI, investing principles are extremely important. Rather than betting on a single winner, many investors spread their exposure across different assets, industries and strategies to reduce risk.
A financial advisor can help crunch the numbers and build a plan that works. But hiring an advisor can be a lifelong commitment, which might make or break your retirement. That’s why finding reliable advisors is crucial.
And that’s where Advisor.com comes in. The platform connects you with an expert near you for free who can help you choose the right investments.
Advisor.com does the heavy lifting for you, vetting advisors based on track record, client ratios and regulatory background. Plus, their network comprises fiduciaries, who are legally required to act in your best interests.
Just enter a few details about your finances and goals and Advisor.com’s AI-powered matching tool will connect you with a qualified expert best suited for your needs based on your unique financial goals and preferences.
Finding the right advisor isn’t always easy — there’s no one-size-fits-all solution. That’s why Advisor.com lets you set up a free initial consultation with no obligation to hire to see if they’re the right fit for you.
If you prefer to take the investing bull by the horns on your own, Moby offers expert research and recommendations to help you identify strong, long-term investments backed by advice from former hedge fund analysts.
In four years and across almost 400 stock picks, their recommendations have beaten the S&P 500 by almost 12% on average. They also offer a 30-day money-back guarantee.
Moby’s team spends hundreds of hours sifting through financial news and data to provide you with stock and crypto reports delivered straight to you. Their research keeps you up-to-the-minute on market shifts and can help you reduce the guesswork behind choosing stocks and ETFs.
Plus, their reports are easy to understand for beginners, so you can become a smarter investor in just five minutes.
Dalio’s core message is straightforward: A technology can succeed without rewarding the majority of investors chasing it.
Artificial intelligence may transform industries and drive economic growth for years to come. But that doesn’t guarantee that today’s most popular companies will ultimately benefit.
For investors, the challenge isn’t just recognizing the potential of AI — it’s navigating the uncertainty that comes with it.
And in markets like this, discipline, diversification and a clear strategy can matter just as much as picking the right trend.
Join 250,000+ readers and get Moneywise’s best stories and exclusive interviews first — clear insights curated and delivered weekly. Subscribe now.
We rely only on vetted sources and credible third-party reporting. For details, see our editorial ethics and guidelines.
Ray Dalio (1); Corporate Finance Institute (2); Goldman Sachs (3); International Money Fund (4); Washington Crossing Advisors (5); Investopedia (6)
This article provides information only and should not be construed as advice. It is provided without warranty of any kind.
AI Talk Show
Four leading AI models discuss this article
"AI as a technology will likely succeed; the real question is whether today's most-crowded stocks (mega-cap infrastructure) are fairly valued or whether smaller, higher-risk AI names face a 2000-style washout while the giants endure."
Dalio's technology-vs.-company distinction is historically sound—the internet succeeded while 90% of dot-com firms failed. But the article conflates two separate risks: (1) sector-wide valuation excess, and (2) individual stock failure. Today's AI leaders (NVDA, MSFT, GOOGL) have moats—installed bases, cash generation, switching costs—that 1990s startups lacked. The real risk isn't that AI fails or that *all* AI stocks crash; it's that mega-cap AI beneficiaries trade at 25–30x forward earnings while smaller, less-proven AI plays collapse. The article also omits that infrastructure plays (chips, cloud) have lower execution risk than pure-play AI software companies. Dalio's warning is valid but imprecise.
If AI truly is a 7% GDP boost (per Goldman Sachs cited here), the winners will be so profitable that even 'expensive' valuations today may prove cheap in 5–10 years—and the market may correctly price in that optionality now, making Dalio's bubble thesis premature.
"The fundamental difference between the dot-com era and today is that current AI leaders possess massive, self-sustaining free cash flows that insulate them from the 'survivorship' risk Dalio highlights."
Dalio’s warning is a classic 'picks and shovels' vs. 'application' distinction, but it ignores the current market structure. Unlike the dot-com era, today’s AI leaders—specifically NVDA, MSFT, and GOOGL—are not cash-burning startups; they are cash-flow machines with massive moats. The article conflates 'bubble' with 'valuation,' ignoring that these firms are trading at reasonable PEG ratios (Price/Earnings-to-Growth) when accounting for 20-30% forward earnings growth. The real risk isn't a total collapse, but a multi-year period of range-bound consolidation as AI capital expenditure (CapEx) begins to weigh on margins before the productivity gains fully materialize in the broader economy.
If AI infrastructure spending fails to yield measurable enterprise productivity gains by 2026, the current massive CapEx cycle will look like a capital-allocation disaster, justifying a violent mean-reversion in valuations.
"AI adoption will be winner‑take‑most: a few firms will capture disproportionate profits, making many current AI‑exposed stocks vulnerable to steep, idiosyncratic drawdowns absent durable moats."
Dalio’s core point is right: technology adoption and corporate survivorship are distinct — AI can become transformational while most AI‑branded public stocks fail. We should expect extreme dispersion: a small group of firms with deep moats (proprietary models, data networks, chip/IP control, sticky enterprise contracts) will capture most economics, while capital‑hungry, margin‑fragile challengers get chewed up. That implies higher idiosyncratic risk in AI/semiconductor/everything‑AI tradeables and a rising value of balance‑sheet strength, recurring revenue, and defensible IP. Missing context: current valuations already reflect this winner concentration for some names, and macro/capital cycle, regulations, and China tech rivalry are underplayed as triggers for rapid repricing.
The market may have already concentrated winners into a handful of stocks (e.g., NVDA, MSFT, GOOGL), and broad ETFs let investors capture AI upside without single‑name risk; if AI adoption is truly pervasive, many incumbents will benefit rather than perish. Also, durable SaaS and cloud franchises with sticky revenue are less exposed to the “most will fail” outcome Dalio warns about.
"AI chip leaders like NVDA and TSM have profitability and moats that dot-com also-rans lacked, positioning them to capture most value even if 90% of AI apps fail."
Dalio's warning echoes dot-com truths: tech thrives, most companies die—valid for speculative AI plays like unprofitable LLM startups trading at 100x sales. But article glosses over AI infrastructure leaders' moats: NVDA holds 90%+ data center GPU share via CUDA lock-in, TSM's 60%+ advanced node dominance, both with 50%+ gross margins and $30B+ quarterly revenue runs. Unlike 2000 pets.com, these generate FCF to fund capex arms race barring rivals. Bubble risk is real in peripherals (e.g., robotics hype), but core enablers re-rate higher on 40%+ EPS growth. Diversify, yes—but own the picks-and-shovels.
If AI adoption slows due to energy constraints, high capex ROI disappoints, or China tariffs escalate, even NVDA/TSM multiples could halve from 40x forward P/E peaks, mirroring 2022 drawdowns.
"Margin durability, not revenue scale, determines whether infrastructure leaders fund the capex cycle; current valuations embed pricing-power assumptions that may not survive competitive pressure."
Grok conflates gross margin durability with FCF sustainability under margin compression. NVDA's 50%+ gross margins assume continued pricing power; if competition (AMD, Intel foundry, custom ASICs) erodes share or if customers demand volume discounts, FCF generation collapses faster than revenue. Also: $30B quarterly revenue runs don't guarantee capex arms race funding if ROIC turns negative. The picks-and-shovels thesis holds only if infrastructure spending yields measurable returns—which Google flagged as uncertain through 2026.
"Physical energy constraints and power grid limitations represent a hard ceiling on AI infrastructure growth that current valuation models ignore."
Anthropic is right to challenge the 'picks and shovels' permanence, but everyone is ignoring the energy supply bottleneck. NVDA and TSM's real risk isn't just competition; it's the physical limitation of power grids and data center cooling. Even with massive cash flows, if AI infrastructure projects are delayed by utility-scale power constraints, the CapEx cycle stalls. The market is pricing software growth while ignoring the hard-asset physics required to run those models at scale.
"Concentration in ETFs/options and leverage can trigger a severe, fundamentals-independent crash in AI mega-caps."
Nobody has flagged the market-structure liquidity risk: a handful of mega-cap AI winners are heavily concentrated in ETFs, passive funds, and option markets; large redemptions or gamma squeeze reversals could cascade into forced selling, creating a liquidity-driven crash even if fundamentals remain intact. That amplifies Dalio’s survivorship point—not all firms must fail for prices to gap lower—so stress-test scenarios should model flow dynamics, options convexity, and prime-broker leverage.
"Energy constraints asymmetrically benefit infrastructure leaders like NVDA/TSM by amplifying scarcity-driven pricing power."
Google's energy bottleneck strengthens NVDA/TSM moats: grid limits create GPU/server scarcity, forcing hyperscalers to pay premiums for CUDA-locked capacity and advanced nodes—boosting pricing power amid Anthropic's competition fears. Challengers without supply chains get sidelined first. This dynamic sustains 50%+ margins if capex ROIC exceeds 15% (current NVDA run-rate), turning constraint into competitive edge.
Panel Verdict
No ConsensusWhile acknowledging Dalio's warning about technology adoption and corporate survivorship, the panel agrees that today's AI leaders have strong moats and are not comparable to the dot-com era startups. The real risk is not a total collapse but a multi-year period of consolidation due to AI capital expenditure (CapEx) weighing on margins. However, the panel also flags potential risks such as margin compression, energy supply bottlenecks, and liquidity-driven crashes.
Owning the 'picks-and-shovels' infrastructure leaders with strong moats (Grok)
Multi-year period of range-bound consolidation due to AI CapEx weighing on margins before productivity gains materialize (Google)