The headlines landed this week with predictable impact: John Giannandrea, Apple's Senior Vice President of Machine Learning and AI Strategy, is stepping down. He'll retire in spring 2026, right around the same time Apple promises to finally deliver the overhauled Siri that was supposed to launch with iOS 18 back in 2024.

If you're following Apple's AI journey, this isn't shocking news. It's the latest chapter in a story that's been unfolding for years, one marked by missed opportunities, public delays, and a company that increasingly looks like it's playing catch-up in the most important technology race of the decade.

The Giannandrea Era: What Went Wrong

When Apple poached Giannandrea from Google in 2018, it seemed like a watershed moment. Here was Google's former head of search and AI, joining Apple to presumably turbocharge Siri and push the company's machine learning capabilities forward. He was quickly promoted to senior vice president by year's end, signaling Apple's commitment to AI leadership.

But here's the uncomfortable truth: under Giannandrea's watch, Apple largely dismissed generative AI during its explosive mainstream adoption. While OpenAI launched ChatGPT in November 2022 and fundamentally changed the technology landscape overnight, Apple continued down its own path, seemingly convinced that its approach to on-device processing and privacy-first AI would be sufficient.

It wasn't.

By the time Apple announced Apple Intelligence at WWDC 2024, the company was already playing from behind. The suite of features felt incremental rather than revolutionary, and the most ambitious component—a completely reimagined Siri with personal context, onscreen awareness, and sophisticated app integration—was immediately delayed.

That delay wasn't a minor hiccup. Apple promised the new Siri for iOS 18's launch, then pushed it to 2025, and finally admitted it wouldn't arrive until spring 2026 as part of iOS 26.4. Even more telling, Apple is reportedly planning to power this "advanced" Siri with Google's Gemini model rather than its own in-house technology.

Think about that for a moment. Apple, a company that prides itself on vertical integration and controlling every aspect of the user experience, is turning to its biggest competitor for the AI brains behind its flagship digital assistant.

The Human Cost of Falling Behind

Leadership changes tell only part of the story. Robby Walker, the executive overseeing the delayed Siri project, left Apple earlier this year after losing the initiative in an organizational shuffle. The company's AI team has been hemorrhaging talent, with researchers and engineers departing for Meta and other competitors throughout 2025.

Reports suggest morale within Apple's AI division has hit rock bottom. When you're working on technology that's perpetually delayed, watching competitors ship impressive features while your projects remain trapped in development hell, and seeing colleagues jump ship for greener pastures, it's hard to maintain enthusiasm.

The talent exodus creates a vicious cycle: delays damage morale, low morale drives departures, departures make it harder to execute, which leads to more delays. Apple is caught in exactly this spiral.

Enter Amar Subramanya: The Fixer or Just the Next Guy?

Apple's new VP of AI comes with impressive credentials. Subramanya spent 16 years at Google, serving as head of engineering for Google's Gemini Assistant before jumping to Microsoft as Corporate Vice President of AI earlier this year. He was credited on the papers for both Gemini and Imagen 3, two of Google's most significant AI releases.

Interestingly, Subramanya has experience with the exact situation Apple finds itself in. Google also missed the starting gun when ChatGPT launched, scrambling awkwardly to respond despite literally inventing the Transformer architecture that makes modern AI possible. But Google recovered. The latest Gemini models are competitive with OpenAI's offerings, prompting Sam Altman to warn his team about "rough vibes" and "temporary economic headwinds."

Google's turnaround took three years and required significant course corrections along the way. The question is whether Apple has that kind of time, and whether Subramanya can replicate that success in a very different corporate culture.

The Challenges Ahead Are Staggering

Subramanya inherits a situation that would intimidate even the most experienced executive. His team is understaffed and demoralized. Apple's AI roadmap is littered with delayed promises and unmet expectations. The company's reputation in AI circles has taken serious damage, making it harder to attract top talent in an already competitive hiring market.

He'll need to rebuild confidence both inside and outside the company while simultaneously delivering on Apple's existing commitments. The spring 2026 Siri deadline isn't moving again—Apple can't afford another public delay. But rushing to meet that deadline with inadequate resources risks shipping something that fails to meet expectations, which might be worse than another delay.

There's also the longer-term strategic question: what is Apple's AI vision, really? The company talks about privacy-first, on-device processing, but that approach fundamentally limits what's possible compared to cloud-based models with massive computing resources. Apple's silicon advantage helps, but there are physics-based constraints to what you can run on a phone.

If Apple continues relying on external partners like Google for advanced AI capabilities, what does that mean for the company's vaunted vertical integration? How do you maintain differentiation when you're using the same underlying models as everyone else?

This Is Just Another Bump in a Very Long Road

Here's what frustrates me most about this situation: Apple should be leading in AI, not scrambling to catch up. The company has incredible resources, world-class engineers, billions of devices in users' hands, and more data about user behavior than almost anyone. Apple's attention to user experience and design should give it unique advantages in making AI actually useful rather than just impressive in demos.

Instead, we're watching a slow-motion train wreck. Each leadership change, each delay, each talent departure is another signal that something is fundamentally broken in Apple's approach to AI. Bringing in Subramanya might help—his experience turning Google's AI fortunes around is genuinely relevant. But one person, no matter how talented, can't fix structural problems that run deep through an organization.

The reality is that Apple's AI struggles aren't primarily about leadership or talent. They're about strategic decisions made years ago, about organizational culture that moves too slowly for the pace of AI development, and about a company that's grown so large and successful that it struggles to pivot quickly when the ground shifts beneath it.

Giannandrea's departure marks the end of one chapter, but the story is far from over. Apple will eventually ship the new Siri, and it might even be good. The company will continue iterating on Apple Intelligence features, gradually adding capabilities that should have launched years ago.

But this episode reveals something important about Apple in 2025: the company that once defined entire product categories through visionary innovation is now in reactive mode, playing catch-up in the most important technology transformation since the smartphone itself.

That's not a leadership problem. That's a fundamental challenge about what Apple is and what it wants to become. No single hire, no matter how impressive their resume, can solve that overnight.

What are your thoughts on Apple's AI direction? Is the Subramanya hire enough to turn things around, or does Apple need more fundamental changes? Share your perspective in the comments below.