Apple's next CEO inherits a 10-to-1 gap in AI capital spending

The shortfall runs near $180 billion against each of the two nearest AI spenders ^[17]. That figure supports two opposite readings, and the material supports both. Cook stayed out of the AI spending race on purpose ^[10], and D.A. Davidson's Gil Luria credits that posture for leaving the company and the stock in what he calls terrific shape ^[13]. Needham's Laura Martin reads the same line as arrears, arguing that Ternus has to push AI through the organization and that this implies more investment ^[11].

Returns, not budgets, separate the two readings, and in this material only the rivals have shown any. Meta has used AI models to strengthen its core advertising business and lift engagement on Facebook and Instagram, and Alphabet's Search has benefited as well ^[15]. Sundar Pichai told investors in July that Search usage hit a record high during this year's World Cup and that AI makes the product more helpful and intuitive ^[16]. Apple's record over the same stretch is an unshipped Siri rewrite and researchers leaving ^[8], with Google's technology filling the gap in the meantime ^[9].

The base Ternus defends is the harder constraint. Apple sat roughly $9 billion behind ExxonMobil when Cook took over in August 2011 at a $349 billion valuation ^[4]. It is now about fourteen times that, an increase above $4.65 trillion ^[18], and worth roughly seven and a half times Exxon's current $665 billion ^[19]. Measured against the index, Cook's Apple compounded about three and a half times the S&P 500's multiple ^[20]. Hardware cycles that would have re-rated a $349 billion company will not move this one, and Ternus is a hardware executive: 51 years old, more than half his life inside Apple, credited on AirPods, the Watch, MacBooks and successive iPhones ^[7].

One item in the inbox does fit that background. The memory chip shortage drove component prices up enough that Cook had to raise device prices ^[14], and running supply chains under Republican and Democratic administrations was among his demonstrated strengths ^[21]. Cost pass-through is a hardware problem with a hardware answer. The capital allocation question is not, and it gets settled in guidance rather than at a product launch.