While companies like Google, Microsoft, and Amazon have spoken openly about their AI efforts, Apple has been quiet for years. Now, finally, its executives are speaking up. I had a chance to see it one day. In an attempt to dispel the impression that the most innovative of the tech giants is lagging behind in this crucial technology, software leader Craig Federighi, services chief Eddy Cue, and top researchers argued that Apple has been a leader in AI for years, but just hasn't done much with it. Advanced machine learning is already deeply embedded in some products, and more advances are to come, such as advances in Siri. And because Apple values data security more than its competitors, the company's AI efforts will be distinguished by strict privacy standards. I asked how many people are working on AI at Apple. “Lots of people,” Federighi replied. Another executive stressed that AI could be transformative, but that Apple doesn't want to get involved in any of the mystical aspects that excite some in the field, such as the pursuit of superintelligence. “This is ultimately a technology that will be Apple's way,” one executive said.
That conversation took place eight years ago, when deep learning AI was the latest tech trend. But a year later, a breakthrough called Transformer ushered in a new wave of smart software called generative AI, powering OpenAI's groundbreaking ChatGPT. Soon, people began judging tech companies by how eager they were to jump on the trend. OpenAI's rivals acted quickly. Apple, on the other hand, did not. Many of the company's best AI scientists were working on self-driving cars and its pricey Vision Pro mixed-reality headset. Over the past year or so, Apple has pulled talent from those projects (no more self-driving cars) and instead rolled out its own generative AI strategy. And at its Worldwide Developers Conference this week, Apple revealed its plans.
Unusually for such an event, the news wasn't about products, but about Apple's next-generation AI. We are working on itIn an interview after the keynote, CEO Tim Cook explained this unusual turn of events. “It was clear that people wanted to know our take on generative AI in particular,” he said. But as in 2016, there was a caveat: the company would embrace generative AI, but it would do so in a very Apple-like way. The company even refused to label its technology artificial intelligence; instead, it coined the term Apple Intelligence, a move that seems designed to distance Apple from the scary side of this powerful technological wave. Apple isn't interested in chasing the singularity or making movies. she Apple is using these new tools to boost productivity and creativity, and just like intimidating technologies of the past, Appleing AI will make it easier to embrace.
This approach is well timed. I count the era of generative AI from the November 2022 release of ChatGPT. We’ve been trying to figure out what it means all through 2023, and now many are feeling the urge to reject it. They rebel against the hallucinations of AI, they’re angry about the possibility of losing their jobs, and most still don’t know what AI can actually do for them. In 2024, smart companies are focusing on how they can actually use this amazing technology in mundane scenarios. Apple proclaimed “AI for us” (the only time the letters “AI” were used in the keynote), a conscious evocation of the original Macintosh slogan. Perhaps Apple will popularize AI for the masses in the same way they popularized the graphical user interface with the Mac.
In contrast to its lofty ambitions, the products Apple touted in its keynote were nothing revolutionary. Many of the demos were email summarization, transcription, auto-completion, inbox organization, paragraph creation from prompts, and removal of inclusions from images. These are bare minimum requirements in the age of Generation AI. Apple’s pitch, as always, is that it will deliver these advancements organically woven into your normal workflow so that users will actually use these features and be happy with them. Apple also gave these products some nice twists. The Photos app promises more in-depth search capabilities, using AI to determine what and who is in a photo, and to find specific images from vague prompts. In auto-generated email replies, Apple can, in certain cases, ask a simple question that can be answered with a single click (when do you want to meet this person?) and generate a response that reflects the user’s intent. More importantly, because users of Apple’s ecosystem have a wealth of personal information on their phones and computers, Apple’s AI can use that data to provide relevant output while keeping those details on the device, protecting the user’s privacy. Apple SVP Federighi is still investigating the matter, but describes it as “intelligence that understands the user” (Apple even claims to be hiring outside investigators to make sure data is truly safe).
The most interesting part of Apple's announcement was about Siri, the AI assistant that has become something of a curio in the age of generative AI. Apple promised that in the future (maybe 2025?), Siri will not only be a better conversationalist, but also a uniquely powerful personal assistant, carrying out complex requests that involve multiple apps. Ironically, this was the vision of the original Siri team in 2011, but was rejected by Steve Jobs, who sought simplicity, and also because the underlying technology was not yet ready. “This is really the piece that was missing from the original Siri,” says Doug Kittlaus, who led the team when Apple launched the product. Kittlaus and a few key colleagues later tried to realize this vision with a startup called Viv, which now lives on as a Samsung product called Bixby. To make such a complex system work, it is essential to have a critical mass of developers on board. The WWDC program included a session to teach developers how to make their apps work with Siri.
