Fundstrat Global Advisors co-founder Tom Lee has stated that the artificial intelligence market buildout is currently equivalent to 1997 rather than 1999, predicting that the AI bubble will not peak until 2033. Speaking at the TOKEN2049 Singapore conference, Lee estimated total AI infrastructure spending will reach $30 trillion, comparing it to the $35 trillion (in 2026 dollars) buildout of internet, mobile networks, and cellular handsets.
Spending Models and Historical Tech Trends
Lee challenged widespread market skepticism regarding overvaluation, noting that many current fund managers did not manage capital through the dot-com bubble or the 2008 financial crisis. Lee stated, "I’ve never seen a bubble when everyone thinks it’s a bubble." He noted that wireless technology expanded exponentially for 25 years with stocks peaking 10 years into the cycle, landing his projected peak on 2033. He also cited Cisco, which dropped 50% twice before reaching its March 2000 peak and subsequently falling nearly 90% from that high.
His $30 trillion capital estimate contrasts with an April 2025 McKinsey report, which projected that AI data centers would require $5.2 trillion in investment by 2030. Lee highlighted that approximately $7 trillion in new liquid assets enters global markets annually, which can fund continuous enterprise deployments and predictive AI developments.
Key Takeaways:
- Tom Lee pegs AI spending at $30 trillion, placing current market development in 1997.
- Ray Dalio of Bridgewater Associates warns borrowing costs and debt could pop paper profits.
- McKinsey projected $5.2 trillion in AI data center investment by 2030.
- An Asian equity selloff on October 13 erased over $800 billion in market capitalization.
Debt Financing vs Early-Cycle Investment
Lee's long-term optimism stands in contrast to warnings from Bridgewater Associates founder Ray Dalio, who views AI as a classic debt-driven bubble. Dalio argues that as interest rates rise, investors holding paper gains will cash out, driving liquidations. Concerns over market financing were highlighted on October 13, when an Asian stock selloff erased more than $800 billion in valuation. These competing views were presented amid major industry gatherings like the TOKEN2049 Singapore gathering.
Why It Matters
This debate underscores how institutional investors evaluate long-term tech cycles against near-term macro liquidity risks. If Lee's capital flows model holds true, hardware and infrastructure providers could see sustained expansion despite periodic multi-month drawdowns of 50% or more. Conversely, if high borrowing rates persist as Dalio warns, debt-reliant capital expenditures could compress asset valuations long before reaching 2033 operating targets.
