As Apple CEO Tim Cook approaches the end of his tenure, debate over his legacy is intensifying. Critics highlight Apple's apparent delay in the artificial intelligence arena, suggesting the company has missed its window. However, an alternative perspective is gaining traction on Wall Street: Cook may have never intended to engage in a cutthroat AI race, instead choosing to fortify the iPhone's position as the core value hub while using cash reserves to keep future options open. Current AI large language models have reached a state of near-parity in capability, suggesting the AI industry could ultimately become highly fragmented, making it difficult for any single participant to capture outsized profits. Apple's iPhone shipments hit a record 247 million units this year, capturing 43% of global smartphone industry revenue. Simultaneously, the company has amassed $145 billion in cash and marketable securities—an advantage that may prove to be its most valuable asset until the AI competitive landscape becomes clearer.
**The Chokepoint: iPhone's Critical Value Node**
The underlying logic of Cook's strategy is a deep understanding of controlling "value chain chokepoints." In most value chains, companies that control critical nodes often reap substantial profits. For instance, Taiwan Semiconductor Manufacturing Company (TSMC) manufactures approximately 90% of the world's most advanced chips with around a 60% gross margin, serving as a paid supplier to both Apple and NVIDIA. The smartphone represents one of the most valuable chokepoints in the modern economy. In the third quarter of 2025, Apple achieved an average selling price of $870, securing 43% of global smartphone revenue, while the industry average was just $351. This advantage extends to other product lines. Technology media outlet Engadget, in a review of the new MacBook Neo, stated the device "makes all Windows PCs under $600 pale in comparison." Brand consultancy Interbrand has ranked Apple as the world's most valuable brand for 13 consecutive years.
**Microsoft's Lesson: Importance Does Not Equal Dominance**
A decision by former Microsoft CEO Steve Ballmer 17 years ago serves as a key reference for understanding Cook's restrained logic. In 2009, Ballmer launched Bing with the rationale that "search is too important to concede," leveraging Microsoft's financial muscle to play catch-up. The result was cumulative losses of $5.5 billion for Bing in just over its first two years. Fifteen years later, Bing holds a mere 5% share of the global search market, while Google still commands approximately 90%. Ballmer's cost extended beyond financial losses. The pursuit also caused Microsoft to lose significant focus on mobile devices and tablets—battlegrounds that ultimately reshaped the industry. While search was important for the industry, for Microsoft it was a battlefield it could never dominate—the price paid was capital, but the greater loss was strategic focus.
**Near-Parity: AI May Be a Battlefield No One Can Dominate**
The technical reality supports Cook's calculus. According to the Epoch AI capability index, the average performance gap between open-weight frontier models and closed-source models is about three months, and this gap sometimes disappears entirely. At the top of the Chatbot Arena leaderboard, the scores of four U.S. labs—Anthropic, Elon Musk's xAI, Alphabet's Google, and OpenAI—differ by only a few percentage points. In this landscape of near-parity, AI may ultimately become a field that is valuable to the entire industry but where it is difficult for any single player to establish a dominant, sustainably profitable position—likely resulting in a highly fragmented market.
**$145 Billion: A Flexible Strategic Option**
Cook's restraint is not inaction. While maintaining existing strengths, he has accumulated a war chest of $145 billion in cash and marketable securities. This capital itself represents a strategic option: if AI truly takes off, Apple has the financial strength to directly acquire a frontier lab; if the AI hype proves to be a bubble, no company is better positioned than Apple to capitalize after it bursts. Apple has demonstrated a viable "low-cost" path for this strategy—reportedly paying Google around $1 billion annually for system integration of the Gemini model. In contrast, Google pays Apple a sum approximately 20 times larger to remain the default search engine on the iPhone. The vast disparity in willingness to pay is a clear signal: in the judgment of both companies, the current cooperative value of the iPhone far exceeds that of an AI model.
Comments