How Apple's AI Strategy Changes with a New CEO

What It Covers

Apple's CEO transition from Tim Cook to hardware chief John Ternus reframes the company's AI strategy amid debate over whether Apple's AI inaction was deliberate genius or costly negligence, alongside updates on Google's coding strike team, Amazon's $25B Anthropic investment, and OpenAI's new Chronicle memory feature.

Key Questions Answered

• •Apple Hardware Advantage: Mac hardware has become the default platform for cutting-edge AI tools — Claude Desktop, Codex computer use, and OpenClaw all launch Mac-first or Mac-only. Developers without a Mac cannot access the latest AI capabilities. This de facto positioning gives Apple leverage without requiring direct AI model investment or infrastructure spending.
• •Apple CEO Transition Signal: John Ternus, Apple's incoming CEO, comes from hardware leadership rather than software or AI. His reputation inside Apple centers on decisiveness — sources contrast him directly with Tim Cook, who reportedly avoided binary choices. This behavioral shift could accelerate AI product decisions that stalled for years under Cook's consensus-driven management style.
• •Google AI Coding Gap: Google DeepMind researchers acknowledge Anthropic leads on coding AI. A Sergey Brin-led strike team now targets internal code generation, training models on Google's private codebase rather than external data. Anthropic's Claude reportedly writes close to 100% of Anthropic's own code, while Google's figure sits around 50% per their February earnings call.
• •Amazon-Anthropic Compute Deal: Amazon commits $25B to Anthropic — $5B upfront plus $20B tied to commercial milestones — structured largely as compute-for-equity. Amazon supplies five gigawatts of Trainium chip capacity for training and inference, with one gigawatt coming online by year-end. This directly addresses Anthropic's inference shortage while deepening AWS's access to Claude's full product lineup.
• •Chronicle Memory Feature Tradeoff: OpenAI's Chronicle feature in Codex uses continuous background screenshots to build persistent workflow memory, enabling the tool to reference errors, open documents, and projects from weeks prior. The tradeoff is significant token consumption and privacy exposure. OpenAI positions it for enterprise professionals on company-paid usage plans, not general consumers.