As concerns grow that today’s AI rally is starting to resemble the dot-com bubble, hedge fund investor Gavin Baker pushed back on that comparison during a recent appearance on the All-In Podcast. Baker argued that today’s AI boom is “a roller coaster that’s kind of a gentle sine wave” compared to 1999. Co-host David Friedberg sharpened the contrast, describing the late-1990s market as “Vegas on a Friday night after way too many drugs.”
Baker’s core argument is that investors drawing parallels to the dot-com crash are comparing today’s market to the wrong period. Unlike many of the companies that fueled the 1999 bubble, he says today’s AI leaders are “extraordinarily real businesses” with real products, customers, and revenue. In his view, 2021 is a much closer comparison to today’s environment than the late-1990s speculative IPO frenzy.
Why This Isn’t 1999: The “Real Businesses” Argument
Baker’s distinction matters because the bubble question hinges on whether genuine businesses are leading the stock market and IPO cycle. He pointed specifically to Anthropic, OpenAI, and SpaceX as the anchor names of the current cycle. For context, OpenAI is reportedly valued at around $500 billion in its latest round, and SpaceX is expected to be valued at about $1.75 trillion. These businesses generate revenue, sign enterprise contracts, and ship products at scale.
Today, the CBOE Volatility Index sits at 21.51 as of June 5, 2026, which is in the elevated-uncertainty zone but well below panic. Its 12-month average is 18.116, with a peak of just 31.05 on March 27, 2026. During the actual 1999-2000 unwind, the VIX routinely cleared 40, indicating extreme fear and uncertainty.
The CMGI Cautionary Tale
David Friedberg anchored the contrast with one of the era’s most vivid wipeouts. CMGI was a company with minimal revenue whose stock rose from $2 to $2,000 in six months before it went out of business two years later.
He argues that a real bubble looks like parabolic price increases for a stock with almost nothing to show for it in the underlying business. The All-In panel’s argument is that today’s AI leaders are categorically different from that pattern.
High Valuations Still Carry Real Drawdown Risk
Valuations, Baker acknowledged, may be “at the top end of the range.” A normal 10-20% public-market consolidation could translate into 30-40% drawdowns in high-beta stocks, meaning names that move more aggressively than the broader index in both directions. For long-term holders, he argued, that kind of move would register as “just a blip.” For recent buyers, it would not feel that way.
The Nasdaq-100, a reasonable proxy for high-beta AI companies, is up 34.35% over the past year and 110.09% over five years. By contrast, during the dot-com unwind from November 1, 1999 to December 29, 2000, the same index fell 55.37%.
The Risk of Betting on Individual Startups
Even in a market Baker considers fundamentally sound, investing in individual public or private companies can sometimes carry binary outcomes. Friedberg flagged Sierra, Brett Taylor’s company building “Salesforce agent native” solutions, as an attractive secondary opportunity outside the top 10 private companies.
The risk profile is starkly two-sided: OpenAI or Anthropic could erase its revenue by building competing capabilities. On the other hand, a tech giant could acquire it to accelerate AGI development.
The Investor Takeaway
Baker believes fears of a 1999-style collapse are overblown because today’s AI leaders are real businesses with real revenue. Still, that does not make the sector risk-free. High valuations and intense competition could lead to sharp pullbacks, even if the long-term AI opportunity remains durable.
Top Gaining Stocks
Top Losing Stocks