Tim Cook’s real legacy isn’t a gadget (it’s way more significant)

The market still underappreciates how much Apple’s moat has shifted from product cycles to an operating system for monetization. The important second-order effect is that services plus custom silicon create a self-reinforcing tax on the ecosystem: higher switching costs, more margin capture, and more room to embed AI features without relying on external cloud economics. That makes Apple less exposed than most mega-caps to the current AI capex arms race, because incremental AI adoption can be funded from on-device efficiency rather than hyperscaler spend.

Intel remains the cleanest loser from this setup, but the more interesting pressure point is broader silicon commoditization. As Apple proves that vertically integrated chips can outperform off-the-shelf architectures in consumer devices, it raises the bar for every PC and handset vendor that depends on third-party silicon for differentiation. Over a 6-18 month horizon, this can compress pricing power not just at Intel, but across the “good enough” hardware stack as consumers anchor on AI features that feel local, private, and low-latency.

The consensus is likely overestimating how much Apple needs a frontier-model breakthrough to win AI. The more durable thesis is that Apple can monetize distribution and defaults while outsourcing model leadership to others, which is structurally better for margins than chasing the most capital-intensive layer of the stack. The main risk is not technological irrelevance, but product fatigue: if Apple Intelligence fails to create a clear upgrade trigger over the next two iPhone cycles, the multiple expansion case fades even if earnings stay resilient.