James E. Thorne says Wall Street is analyzing AI spending with the wrong framework, mistaking forced strategic investment for a bubble.
In a new post on X, the chief market strategist at the billion-dollar asset management firm Wellington-Altus says AI capital expenditures are being mislabeled as a speculative excess when they reflect a prisoner’s dilemma-style arms race.
According to Thorne, AI spending is no longer discretionary, as companies and nations must continue investing or get left behind.
“AI capex is being misframed as a bubble when, in reality, it is a prisoners’ dilemma arms race in which firms and nations either spend or accept strategic decline, with AI capacity now inseparable from control over compute, energy, and critical inputs.”
Thorne says this dynamic explains why hyperscalers, chipmakers, and platform companies continue committing capital at industrial scale, even as investors grow uneasy about margins and returns.
“In a classic prisoners’ dilemma, each player would prefer mutual restraint, but once one defects, everyone else must follow or be dominated.”
He adds that this shift has permanently altered the economic equilibrium, moving it away from traditional margin optimization toward capacity accumulation.
“The Nash equilibrium has shifted from optimizing margins to invest or be structurally weakened.”
Thorne highlights that the AI buildout is now an issue tied to national security and geopolitical power, as governments link AI leadership to control over compute, energy and critical resources.
“This is not just a corporate contest. It is increasingly a question of national security, as governments now view AI leadership, domestic compute, and secure supply of energy and critical minerals as essential.”
He warns that underinvestment carries consequences beyond slower growth, potentially locking companies and countries into dependence on foreign infrastructure and technology.
Thorne says the new strategic reality is not yet reflected in equity prices, pointing out that many core AI infrastructure firms trade on valuation metrics comparable to, or cheaper than, slower-growing consumer businesses.
“The misread is visible in equity pricing, when AI infrastructure leaders trade on multiples lower than growth retailers. This signals that markets have not fully internalized the asymmetric payoff to scale, data, and model advantage.”
He closes by arguing that calling AI spending a bubble misunderstands the nature of the game now being played, where failure to invest is equivalent to voluntary decline.
“The through line is straightforward. AI CapEx is not discretionary, and the key AI names are not just hot growth stocks to fade on sentiment. Underinvestment is tantamount to self-liquidation for both corporations and countries.”
Thorne’s sentiment of forced AI investment aligns with Mark Zuckerberg’s view that falling behind is not an option for Meta. In September, the Meta CEO said “not being aggressive enough” in the AI race is a bigger risk than losing $200 billion in investments.
Just last week, AI Czar David Sacks said that ceding AI leadership to China will lead to poverty in the US.
Disclaimer: Opinions expressed at CapitalAI Daily are not investment advice. Investors should do their own due diligence before making any decisions involving securities, cryptocurrencies, or digital assets. Your transfers and trades are at your own risk, and any losses you may incur are your responsibility. CapitalAI Daily does not recommend the buying or selling of any assets, nor is CapitalAI Daily an investment advisor. See our Editorial Standards and Terms of Use.