Apple makes it belated AI play at WWDC
Siri promoted to key attacking position in the mobile line-up, targeting home, car and TV services driven by machine learning
Apple has hardly been a frontrunner in the mobile world’s gallop towards artificial intelligence (AI), but it made a good attempt to catch up on the first day of its WWDC developer conference. Siri, its voice-activated digital assistant, will sit at the heart of new services driven by AI, and will expand well beyond its current place in the iPhone, into the home, car and anywhere else Apple fans may want to connect.
The company’s software chief Craig Federighi talked about “advanced artificial intelligence”, but did not reveal much that Google hasn’t been discussing and implementing for a year or more. But of course, Apple’s mobile success rarely relies on getting there first (Siri was an unusual example, where Google had to play catch-up with Now). Instead, the iPhone maker tends to watch for the trends which are really catching on with its target consumers, and move in with a finely polished offering, integrated tightly with its overall user experience and device portfolio.
Siri now has the task of encouraging iDevice users to buy more gadgets, consumer more applications and generate more data, by supporting context aware services round the home. In parallel, Apple is hinting at the next generation of the famous iOS user interface, one which will tick all the boxes for today’s most on-trend developer – machine learning, computer vision, virtual reality and AI. Federighi promised that Apple, helped by Siri, would create “an intuitive user interface” incorporating deep learning, which would make apps and services “so engaging”.
As expected, Apple opened Siri up to third party developers to incorporate the virtual assistant functions in their applications, and said the Siri APIs (application programming interfaces) are already supported by WhatsApp, WeChat, Uber, Lyft, Photo Search and others, as well as its own CarPlay in-vehicle infotainment platform. To enhance the Apple in-car experience, and aid context aware services on the iPhone, iOS Maps has also been opened up to developers.
Federighi talked about a technology called “differential privacy” to preserve users’ identities. “When it comes to performing advanced, deep learning, we’re doing it on device. Keeping your personal data under your control,” he said.
Siri will also be integrating with third party calling services from Vonage, Skype and Cisco, and the new VoIP API will help integrate the iPHone with corporate services. Moves like this will be important to Apple’s core enterprise alliance with IBM, which is vital to the quest to restart iPhone growth and generate new service revenue streams.
Apple will also be keen to see Siri working with IBM’s Watson technology rather than as a rival to it – Watson is increasingly being integrated with mobile applications to bring cognitive computing to everyday business tasks and Siri could be going the same way in the enterprise market, but we may expect some cooperation there.
And Siri will also be more tightly integrated with Apple TV as well as the anticipated home hub, designed to challenge Amazon Echo.
Siri is not the only technology where Apple is extending its reach beyond the iPhone. As smartphone sales slow, the company needs to target other platforms to keep stimulating usage of its services. It released developer versions of upgrades to the operating systems for iPhone, Mac, Apple TV and Apple Watch, which will debut this fall on new hardware.
No details of the devices themselves were proffered, though clearly they will need heftier processors and memory to handle the promised deep learning functions. The new iPhone iOS 10 will make up to “11bn computations per photo for object and scene ID to better search photos on a device,” said Federighi, and, in a comment pointedly aimed at Google, added: “Analysis of data is done on the devices [rather than the cloud] to keep data under personal control.”
Of course, that also reflects a basic philosophical difference between Apple and Google (or Amazon). Apple’s mobile success has been all about downloads and locally managed content and processes, and its moves to embrace the cloud-oriented trend have been faltering. Federighi sounded almost Microsoft-like in his eagerness to find new reasons to hang on to the old models rather than take the plunge in to a new world where Apple is not guaranteed the same success. There will be plenty of cloud announcements to come at WWDC, including expansion of the unexpectedly successful Apple Music streaming service, but in pushing local AI processing on the iPhone, Federighi was providing the latest push-back against the idea that everything will eventually happen on slim devices connected to the cloud