Skip to content 
Apple has always been obsessed with the parts nobody sees. The internal engineering. The tolerances on a hinge. The finish on a surface that will never face a customer. It is a peculiar kind of perfectionism, one that has made Apple the most valuable company on earth and, increasingly, one of the most anxious. Because the AI era does not reward the unseen. It rewards speed, iteration, and a willingness to ship things before they are perfect. And now, for the first time in 15 years, Apple has a new chief executive who will have to decide which version of the company survives.
John Ternus, 50, takes over from Tim Cook in September. He is a mechanical engineer by training, a 25-year Apple veteran by career, and by all accounts one of the quietest people ever to run a company worth four trillion dollars. Most people outside the technology industry have never heard his name. That anonymity is not accidental. It is, in a sense, the whole point.
Three CEOs, Three Eras
Apple’s leadership history is short enough to memorise. Steve Jobs built the mythology: the reality distortion field, the product as art object, the keynote as theatre. Tim Cook built the machine: the supply chain, the services empire, the operational discipline that turned Apple from a great product company into a financial phenomenon. Between them, they grew Apple’s market capitalisation by something in the region of a hundredfold.
Ternus arrives as the third act, and the narrative frame he inherits is harder to name. He is not a marketer. He is not an operations chief. He is an engineer who spent his first decade and a half at Apple entirely out of public view, contributing to products without ever putting his name to them. He appeared at his first Apple developer conference in 2017, which in Apple culture functions as a kind of coming-out ceremony for rising executives. The company had been watching him for years before anyone else knew to.
The moment that defined his internal reputation came around 2020, when Apple made one of its most consequential bets: abandoning Intel’s processors entirely and building its own chips from scratch. The semiconductor industry thought this was either visionary or reckless. Ternus was central to executing it. The resulting Apple Silicon architecture transformed the Mac lineup, cut dependence on outside suppliers, and gave Apple something it prizes above almost everything else: complete control over its own hardware. It paid off in ways that went beyond performance benchmarks. It restructured Apple’s cost base and accelerated its product cycles. For insiders, it cemented Ternus as someone who could manage a bet of that scale from beginning to end.
The Succession That Almost Wasn’t
Cook has said for years that his successor would come from inside Apple. That commitment narrowed the field significantly. Jeff Williams, the chief operating officer and Cook’s closest deputy, had been the consensus favourite for much of the past decade. He left in late 2025, and with him went the obvious answer. Craig Federighi, the software chief, is articulate and genuinely popular with developers, but he was associated too closely with Apple’s AI difficulties. The failure of Siri to keep pace with competitors happened on his watch, and that matters enormously when the central challenge facing the company is AI competence. Greg Joswiak, the marketing chief, was another name that circulated without ever quite gaining traction.
Ternus won not because the others fell away, but because Cook had been systematically expanding his responsibilities for years. By late 2025, Ternus held oversight of hardware engineering across every major product category and had been given control of Apple’s software design teams as well. That combination of hardware and software under a single engineering executive was unusual for Apple and, in retrospect, clearly deliberate.
The Siri Problem
Here is the honest assessment of where Apple stands on AI, which Ternus now inherits in full. The company announced Apple Intelligence in 2024 with considerable fanfare. It ran advertisements for features that did not exist yet. Then it delayed the upgraded Siri into 2025. Then it delayed again, into 2026. The internal explanation involves a genuine technical problem: the old Siri architecture and the new one could not be merged cleanly, forcing engineers to rebuild significant portions from scratch, which created internal conflict between the engineering and marketing teams over who had promised what.
The result is that Apple, a company that built its reputation on not announcing things until they were ready, ran ads for a product that was not ready. That is a cultural rupture as much as a commercial one. Meanwhile, the competition has not been waiting. OpenAI, Google, and Anthropic have been shipping new capabilities at a pace that Apple’s annual product cycle was never designed to match. The gap between what consumers can do with a third-party AI assistant and what they can do with Siri has become visible and embarrassing.
Vision Pro compounds the picture. The spatial computing headset that Ternus also oversaw shipped in early 2024 at a price point that excluded most buyers and with an app ecosystem that gave them little reason to stretch. Estimates suggest around 45,000 units shipped in the final quarter of 2025, against an iPhone business that moves tens of millions of units in the same period. Marketing spend on Vision Pro was cut by over 95% during 2025. Apple has not officially commented, but production cuts at its manufacturing partner told the story clearly enough.
None of this is fatal. Apple’s iPhone business remains dominant, its services revenue continues to grow, and its installed base of over a billion devices gives it a distribution advantage that no AI startup can replicate. But the question is whether that installed base becomes an asset in the AI era or a liability: a vast audience watching Apple fall behind on the features that matter most.
What Ternus Actually Believes
Engineers tend to think differently about problems than executives trained in strategy or marketing. They believe that hard problems are, at root, execution problems. That if you build the right architecture, hire the right people, and give them enough time, most obstacles yield. Ternus has said publicly that Apple’s approach to AI is about integration rather than spectacle, that the goal is for users to gain useful new capabilities without necessarily being conscious of the underlying technology shift. That is a coherent philosophy. It is also, in the current competitive environment, a risky one.
The AI companies that have gained the most ground have done so precisely by making the technology visible, even theatrical. ChatGPT’s early virality came from the experience of watching an AI do something that felt impossible. Apple’s instinct to make technology invisible served it well when the technology was the physical object itself. It is less obviously correct when the technology is a software layer that users can directly compare across platforms.
Ternus has been given the tools to address this. A new AI chief, a restructured leadership bench, and ownership of software design alongside hardware give him more integrated authority than any Apple executive below the CEO has previously held. The updated Siri, rebuilt on a new architecture, is supposed to launch in spring 2026. That timeline means Ternus will have been in the job for a matter of months before the most important product launch of his tenure arrives. There is very little room for it to go wrong.
The Weight of the Furniture
Cook told Apple employees that the company will reach incredible heights under Ternus. That is the kind of thing outgoing CEOs say, and it is also, in this case, a statement of genuine belief from someone who has watched Ternus work for 25 years. The board’s confidence in him is real. The question the market is asking is whether the skills that made him exceptional at building hardware can translate into leading a company through a software and AI transition it did not initiate and does not yet dominate.
Apple has never been a fast-follower. Its entire identity is built on the idea that arriving last with the best product is superior to arriving first with an adequate one. That strategy produced the iPod, the iPhone, and the App Store. Whether it produces a great AI assistant, in a market where users have already formed habits with competing products, is the defining test of the Ternus era. He has the engineering instincts, the institutional trust, and the operational authority. What nobody knows yet is whether he has the appetite for the kind of speed and public iteration that AI competition demands.
The man who finished the back of the furniture beautifully now runs the whole factory. The furniture, this time, has to ship on time.
Feature blog banner image source – Google Gemini
