Horace Dediu on Apple Sitting Out the AI Spending Race

Key Takeaways

• Apple’s AI capex remains around $14 billion.
• Rivals collectively spend $650 billion on AI infrastructure.
• Apple licenses Google Gemini for ~$1 billion annually.
• M5 chip runs 70‑billion‑parameter models on devices.
• Apple’s device fleet acts as distributed AI data center.

Summary

Apple is deliberately avoiding the $650 billion AI‑infrastructure spend that hyperscalers like Amazon, Google, Microsoft and Meta are pouring into data centers. Instead, it keeps its capital budget around $14 billion, licensing Google’s Gemini model for roughly $1 billion a year and embedding powerful AI accelerators in its own silicon. The new M5 chip can run 70‑billion‑parameter models locally, turning every iPhone, Mac and iPad into a distributed AI node. This capital‑light approach lets Apple fund massive stock buybacks while competitors burn cash on cloud infrastructure.

Pulse Analysis

The AI arms race has become synonymous with massive data‑center investment, as Amazon, Google, Microsoft and Meta collectively allocate roughly $650 billion this year. These hyperscalers are financing sprawling GPU farms, taking on debt, and seeing free‑cash‑flow collapse, all in pursuit of generative‑model dominance. Their capital intensity reflects a belief that owning the underlying compute stack is the primary moat in a market still searching for sustainable revenue models.

Apple, however, has taken a contrarian path. By licensing Google’s Gemini model for about $1 billion annually and focusing on on‑device processing, it sidesteps the need for its own AI super‑computers. The M5 silicon, featuring a 16‑core Neural Engine and integrated GPU accelerators, delivers four times the AI performance of its predecessor, enabling 70‑billion‑parameter models to run locally on iPhones, Macs and iPads. This distributed compute model turns the company’s 2 billion devices into a de‑facto AI data center, delivering latency‑critical experiences without the overhead of massive cloud spend.

The strategic implications are profound. Apple’s capital‑light approach preserves cash for shareholder returns—evidenced by $90.7 billion in stock buybacks—while competitors see buybacks plunge. More importantly, by embedding AI directly into the devices that consumers already own, Apple deepens its ecosystem lock‑in and monetization potential. If AI services continue to commodify, firms that own the customer relationship, rather than the infrastructure, are likely to capture the lion’s share of future profits, positioning Apple as a potential winner in the next wave of tech innovation.