Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
When Tim Cook told Wall Street a month ago that Apple would unveil new artificial intelligence features this year, he broke with the company’s normal practice of not talking about a new technology until it has something close to shipping.
This week, meanwhile, brought news that many of the engineers from Apple’s car project will be moved across to work on generative AI, following the company’s abandonment of a decade-long autonomous vehicle effort.
In the world of Apple, where such moves normally stay firmly out of sight, this counts as a serious convulsion. Generative AI is sweeping through the tech world and Cook, as chief executive, needs to rally the troops. The company has some clear advantages in the coming battle — yet, starting late, it could also be facing the most disruptive threat it has seen in the iPhone era.
Much of Apple’s effort around AI in recent years has been directed towards using the technology to try to open up new hardware markets, such as autonomous vehicles and mixed reality headsets. But with the end of the car project and a slow start for its Vision Pro headset, attention has shifted squarely back to handsets.
For Apple shareholders, this will in many ways be welcome news. If the main battlefield for AI is the smartphone, then it points to a new lease of life for the iPhone. AI-driven features could give consumers more reason to upgrade as the technology puts extra demands on their devices. Greater use of voice-powered services should also cement the importance of iPhone “peripherals” such as AirPods and the Watch.
This suggests a continuation of the status quo in which Apple continues to see only modest hardware revenue growth, but where the iPhone stays at the centre of computing and services revenues continue to rise.
If so, then Apple has some clear advantages on its side. As smaller AI models proliferate, AI-powered features such as enhanced photo editing should appear on its handsets quickly. Processing AI on handsets also plays to the company’s strengths in privacy and security, making it easier to tap into user data to personalise its new services. For developers, meanwhile, being able to draw on the computing power of the iPhone will help cut down on expensive AI processing costs they would otherwise face.
The huge consumer inertia in the iPhone world also works in Apple’s favour. Google has a history of bringing eye-catching AI-powered services to the Android world first, from turn-by-turn navigation for Google Maps to the more recent Lens, a service which identifies objects in photos. None of this has made even the smallest dent in the iPhone.
Yet if Apple seems well positioned, it needs to move fast. Google and Microsoft (through its OpenAI partnership) have developed families of large language models and have already invested heavily in the infrastructure needed to operate these at massive scale and the tools to make the technology broadly accessible.
For Apple, there are clear warning signs. One is OpenAI’s plan for a “GPT store” — a place for developers to sell AI-powered services built on top of OpenAI’s models. This platform for supercharged apps looks like a direct challenge to Apple’s App Store.
Apple’s bar on rival app stores on the iPhone and iPad, though facing regulatory challenges, gives it some breathing room. But it needs to move fast to convince developers that it will give them the tools to build similar AI-enhanced services, and that its App Store will evolve into the main marketplace for mobile AI.
A second warning shot comes from Google, which recently rebranded its Google Assistant as Gemini to match the name of its latest large language model. Gemini has had a rough start: CEO Sundar Pichai told staff this week that problems with the system that led the company to be accused of racist bias were “completely unacceptable”. But the group is racing to make Gemini an all-purpose personal assistant, capable of responding to spoken commands to carry out complex tasks.
Apple is hardly new to the smart assistants game. It bought Siri in 2010 and embedded it as a language-powered assistant in the iPhone a year later. Apple co-founder Steve Jobs is said to have taken a strong personal interest in Siri, seeing it as the precursor to more powerful language assistants that might one day change the way people get things done on a smartphone. With the breakthroughs in generative AI, the moment Jobs envisaged nearly a decade and a half ago may now be arriving in a hurry. Apple has ground to make up.
Read the full article here