AI may have just had its iPhone moment

Driving the news: Apple's approach, dubbed Apple Intelligence, envisions a future in which a ubiquitous AI system that knows all about you can use that knowledge to surface the right information and take action on your behalf.

• In contrast to the current chatbots — which know about the world, but little about you beyond what you tell them — Apple is building a context engine that understands each customer and the information and people that matter most to them.

Yes, but: What Apple will ship this fall is a more basic set of generative AI capabilities. The initial features include a writing assistant, image and custom emoji generators, and a somewhat more capable Siri voice assistant.

• Over the coming year, Apple is promising the first examples of the broader vision, including the ability for Siri to to answer questions using the context of what's on screen and drawing on knowledge from across various applications.

Case in point: When someone says, "Open that podcast my wife sent a few days ago," Siri will be able to look for mail or other messages from the person it knows is the user's wife, find the recent link to a podcast, and then open the file in the podcast app.

• Similarly, when someone asks, "What time does my mom's flight land?" Siri will know who you are referring to and where to find that information.

Between the lines: There are two key elements to bringing this bold vision to life: technology and trust.

• On the technology front, Apple may not be building the most advanced frontier models like Google, Meta or OpenAI — but given what it showed Monday, the company seems to be further along than many people thought.
• The proof will come when Apple's AI gets into the hands of millions of users, and we see how polished its features are and how well-protected from misuse and misfires.
• Apple didn't give reporters hands-on demos this week, and the new AI features aren't in the early OS releases Apple provides to developers.

Trust is where Apple has a real edge.

• The Apple Intelligence vision relies on access to a copious amount of personal data.
• Apple can reassure users that it won't abuse its access by drawing on a reputation for privacy that it has spent more than a decade fostering.

What they're saying: "Understanding this kind of personal context is essential for delivering truly helpful intelligence," Apple software executive Craig Federighi said Monday.

• "But it has to be done right. You should not have to hand over all the details of your life to be warehoused and analyzed in someone's AI cloud."

As part of its privacy push, Apple is trying to do as much work as possible on the device itself.

• When Apple needs more processing power, its programs will pass the work over to a remote server in its data center.

Between the lines: Apple promises it will perform the handoff without letting itself or any other entity have access to the query.

• When an Apple query is handed to ChatGPT, information passes to OpenAI, but Apple will ask the user's permission each time. It also says OpenAI won't be able to store the information or train its systems on the data.

Context: Apple is also making modest use of generative AI models themselves, using them as much to understand queries and locate information as to produce responses.

• In a follow-up meeting with reporters, Federighi likened today's genAI to an adolescent: "We're not sort of taking this teenager and telling him to go fly an airplane. 'Here's what you're good at. You can flip burgers or something.'"

Yes, but: It's far too early to declare winners and losers in today's wave of AI adoption.

• Apple remains behind when it comes to building the largest AI models — hence its partnership with OpenAI.
• It's not clear just how broad Apple's AI capabilities are in this area, though Axios did confirm that all the Apple Intelligence features — both those running on devices and those that are processed in Apple's data centers — use the company's own models.

The bottom line: Apple unveiled a compelling, if incomplete, vision. But don't forget that the first iPhone had some huge limitations, too: It ran on a super slow network, didn't support third-party apps and couldn't even "copy" and "paste."