š Appleās AI bet got a CEO
Appleās board picked John Ternus, senior vice president of Hardware Engineering, to succeed Tim Cook on September 1.
Not Craig Federighi (software) nor Eddy Cue (services). The hardware chief.
A month ago I shared with you all why I changed my mind about Apple. Iād been in the camp that wrote it off on AI. They had no hyperscale capex, Siri hadnāt meaningfully improved in a decade, and WWDC was seemingly losing steam. What I missed was that every time I switched between ChatGPT and Claude, I was doing it on an Apple device.
My read is that Ternusās appointment reflects a bet on hybrid computing. Frontier model training compute has been growing around 5x per year in recent years, while GPU memory bandwidth has improved around 28% per year. That creates a widening systems bottleneck and strengthens the case for shifting routine AI tasks to edge devices while leaving heavier workloads in the cloud.
This new device landscape is still evolving, but it will be a combination of the highly portable phone with the persistent desktop or server. Both play to Appleās strengths ā they run on the same architecture and sit inside an installed base of over 2.5 billion active devices.
The board picking a hardware CEO is Apple telling us how it sees its next chapter. My analysis is open to all readers today:
To what degree do you think LLM APIs will be commoditized over time? Given the device advantage they have, if that's paired with a clean interface to outsourced intelligence, and the economics can work, why deal with the crazy capex race of their own frontier models? Like you say, play to their strengths and reap the benefits without that massive upfront cost. They already have a level of comfort building on non-proprietary foundations: MacOS on Darwin, Safari on WebKit, Maps on Tom Tom/OpenStreetMap/Yelp/Tripadvisor, Apple Card is white-labeled Goldman Sachs (soon to be Chase). Play OpenAI and Anthropic off against each other, and use different LLMs where regionally required or appropriate.
The hybrid computing angle makes sense for Apple's specific constraints. On-device for latency and privacy, off-device for heavy compute. The 2.5 billion device install base is the leverage.
What's interesting is the software side. The hardware bet only pays off if Apple's AI services are good enough to use the edge compute advantage. Ternus gets the silicon right. The question is whether that's sufficient.
Apple's had platform moats before. The question is whether the AI era rewards different moat types than the app store era did.