I come not to bury Tim Cook, but to praise him. Last week’s announcement that the Apple Inc. CEO is handing the job to hardware chief John Ternus in September has been met by assessments questioning “whether the company is prepared for a future beyond the iPhone” and how Apple’s lagging performance in artificial intelligence is “raising questions about whether Cook should lead from here.” Cook seemingly built the world’s most valuable company but fumbled his final act. Maybe so. Or maybe the man who created trillions of dollars in shareholder value is still seeing a few moves ahead.
The critiques didn’t start last Monday. Thirteen months ago, John Gruber, a long-respected Apple commentator, wrote “Something Is Rotten in the State of Cupertino.” Apple, he said, had promised a personalized Siri it couldn’t deliver and run ads for features that still don’t exist. The Cupertino, California-based company, Gruber wrote, had squandered credibility it had spent two decades building.
Apple’s self-inflicted humiliation is undeniable. It just didn’t seem to matter to customers. The iPhone 17 marked Apple’s best launch ever. It shipped 247 million iPhones in 2025, took 20% of global unit share, and collected most of the industry’s profits. Siri remains a joke next to ChatGPT or Claude. If customers were picking phones based on their AI capabilities, Apple’s sales bonanza could never have happened.
Cook’s 15 years of discipline have given Apple control of the most valuable choke point in consumer technology and made the iPhone the most profitable product in history. That discipline also let him resist the pressure to go for broke on AI. Not because AI doesn’t matter, but because dominating AI may be impossible. And trying would detract from what Apple already commands. Just because a technology is important doesn’t make concentrating on it the right move. To see why, look at a mistake former Microsoft Corp. CEO Steve Ballmer made 17 years ago.
Ballmer launched Bing in 2009, convinced that search was too important to cede and that Microsoft’s cash could fund its catch-up. It lost $5.5 billion in just its first two-plus years. Fifteen years later, Bing has only about 5% share in global search while Google has roughly 90%. Search was valuable - just not to Microsoft. And $5.5 billion was only the dollar cost. It also took Microsoft’s focus away from things like mobile and tablets and kept it in a search war it never came close to winning.
The most profitable place to be in most value chains is the top performer at a choke point — the passage through which a disproportionate share of the chain’s value must pass. Taiwan Semiconductor Manufacturing Co. makes roughly 90% of the world’s leading-edge chips at gross margins around 60%. Every serious chip designer, Apple and Nvidia Corp. included, pays TSMC.
The smartphone is the most valuable choke point in the modern economy. It’s the primary entry to the electronic world, and Apple dominates. In the third quarter of 2025, Apple captured 43% of smartphone market revenue at an average selling price of $870, against an industry average of $351. Apple’s manufacturing edge extends beyond the iPhone to products like the new MacBook Neo, which Engadget’s review said “puts every $600 Windows PC to shame.” On top of that is a brand that Interbrand has ranked as the world’s most valuable for 13 consecutive years.
For Apple’s AI failures to counterbalance that achievement, AI must become even more valuable than the iPhone, which some close followers of the company estimate has generated more than $2 trillion in revenue. Apple’s rivals are betting it will. The hyperscalers are projected to spend roughly $600 billion next year alone on capital expenditures, almost all of it AI-related, more than 40 times Apple’s $14 billion this fiscal year.
Holding onto a choke point requires a sustained performance edge, and there’s no evidence of one in AI. Epoch AI’s Capability Index shows open-weight frontier models trailing closed-weight models by about three months on average, with the gap sometimes closing entirely. Foru American labs are within a few percentage points of each other at the top of the Chatbot Arena leaderboard: Anthropic PBC, Elon Musk’s xAI, Alphabet Inc.’s Google, and OpenAI, with China’s Alibaba Group Holding Ltd and DeepSeek less than 6% behind.
Given this near-parity, the argument that Apple should have spent hundreds of billions to push the frontier rests on a single load-bearing assumption: AI self-improvement. Without it, the gaps between labs stay measured in months. With it, a small lead could compound into a large one. No one is sure when, or even if, AI gets there. Without self-improvement, AI may turn out to be a choke point no one can dominate. Valuable as a category, unprofitable as a position.
Cook’s restraint was not inaction. It was the calculation that Apple has better games to play. The difference between good chess players and mediocre ones is that the latter focus on material. Good ones focus on the King. They know where to put their effort to win. Cook has spent 15 years focused on the King — the thousands of decisions about industrial design, manufacturing, and supply chains that make the iPhone an iPhone. Racing in AI would have meant counting pieces while leaving the King unguarded. Ballmer counted pieces, and it cost Microsoft $5.5 billion.
Instead of being distracted, Cook kept Apple focused on its existing advantages while accumulating $145 billion in cash and marketable securities. That cash is optionality. If AI really takes off, Apple can buy a frontier lab outright. If it’s a bubble, no company is better positioned when it pops. Apple has already shown how cheaply it can follow the partnership path. It pays Google roughly $1 billion a year for Gemini integration. Google pays Apple twenty times that to be the iPhone’s default search engine — an unmistakable signal of how the two companies assess the relative value of the iPhone and AI models.
Cook hasn’t put his rivals in checkmate. The outcome of his moves depends more on AI than on Apple. If AI self-improvement arrives before Apple can exercise its option, the first firm across the threshold could open a gap too large to bridge and become too valuable to acquire. The game still has many moves to play. But while everyone else was counting pieces, Cook was watching the King. Apple’s retiring grandmaster may still be a few moves ahead.
A message from Advisor Perspectives and VettaFi: Discover something new! Click here to register for our upcoming webcasts.
Bloomberg News provided this article. For more articles like this please visit
bloomberg.com.
Read more articles by Gautam Mukunda
