Apple’s era of software deference is over. John Ternus, the hardware architect behind the iPhone and AirPods, takes over as CEO in September to lead a hard pivot into AI-native devices. Tim Cook exits at the peak of Apple’s market cap - $4 trillion - but at the low point of its AI relevance.
Ternus inherits a company that missed the generative AI wave. "Apple Intelligence" flopped, forcing Apple to license Google’s Gemini for core features. According to Tom Lee Devlin on The Economist, this dependency undermines Apple’s claim to vertical integration. The Mac Mini’s role as the go-to machine for AI agents like OpenClaw was accidental, not strategic.
"Ternus isn't being promoted to write better code. He is there to build the next generation of AI-native devices."
- Tom Lee Devlin, The Intelligence from The Economist
The bet is clear: AI isn’t just software. It’s what the device lets you do. Ternus, who apprenticed under Jobs and Cook, will push smart glasses and wearables as the new interface layer. Unlike Cook’s iterative playbook, this is a moonshot - one that bets Apple’s soul on hardware once more.
Meanwhile, Amazon doubles down on AI independence. It’s committing $25 billion to Anthropic in a compute-for-equity deal. Amazon supplies 5 gigawatts of power and Trainium chips; Anthropic delivers models. The move follows Amazon shutting down its internal AGI lab, a full retreat to outsourcing.
"Anthropic is navigating a political rehabilitation while the NSA uses its Mythos model."
- Nathaniel Whittemore, The AI Daily Brief
Google isn’t idle. Sergey Brin has formed a strike team at DeepMind to push Gemini as a primary developer. The goal: close the gap with Anthropic, whose models now write nearly all their own code. Apple’s turnaround hinges on whether Ternus can move faster than Google’s internal build and Amazon’s deep Anthropic pocket.

