Apple is bringing generative AI to Siri and throughout iOS 18, iPad OS 18 and MacOS Sequoia, the company announced at the Apple Worldwide Developers Conference on June 10.
As was rumored, Apple is partnering with OpenAI to bring ChatGPT to iOS, iPadOS and macOS. However, Apple Intelligence will be an independent generative AI service hosted on its own servers.
Adding generative capabilities to Siri has the advantage of familiarity, and partnering with OpenAI means Apple won’t have to do all of the work itself. Apple is betting on generative AI being a seamless addition to the way the Apple ecosystem may already be organizing a consumer’s or professional’s life.
Apple Intelligence will be available in beta this fall in iOS 18, iPadOS 18 and macOS Sequoia. It will work on iPhone 15 Pro, iPhone 15 Pro Max and some current iPad and Mac devices with M1 or newer chips. ChatGPT assistance for Apple Intelligence will follow later in 2024.
Siri will respond to natural language
The popular voice assistant is catching up to the generative AI world with Apple Intelligence, which enables Siri to:
- Understand more natural-sounding questions and speech patterns, including understanding pauses or corrections, such as when you’ve forgotten exactly where you stored a piece of information.
- Remember previous conversations and references.
- Answer questions about Apple devices.
- Take actions inside apps on your behalf.
You can now type to talk to Siri, double tapping the bottom of the screen to make natural-language requests.
Apple didn’t show flashy demos like OpenAI and Google did, but they may be betting on more consumers finding uses for generative AI through what they already do with Siri.
Enhancements to Siri with Apple Intelligence will roll out over the course of the next year.
Developers will be able to define how Siri with Apple Intelligence will be able to interact with their apps.
What is Apple Intelligence?
During the AI reveal at WWDC, Apple put the emphasis on “personal intelligence” and privacy. Most of the generative AI processes used for Apple Intelligence will be performed on-device on the A17 or M-series chips and in a proprietary cloud.
Craig Federighi, Apple senior vice president of Software Engineering, emphasized personal context in the WWDC keynote.
“We’re tremendously excited about the power of generative models … But these tools know very little about you and your needs,” he said.
In response, Apple Intelligence on iOS 18, iPad 18 and macOS Sequoia will use AI for “personal context” to make day-to-day tasks faster. iPhone can prioritize and summarize notifications with the “priority notifications” feature, selecting the most important ones first.
SEE: Concerns about ultra-powerful artificial intelligence lead some within Silicon Valley to seek a ‘Right to Warn’ for AI project whistleblowers.
What else does understanding personal context mean? Apple Intelligence can determine what data you think is most relevant and reference related content. For example, if your meeting is rescheduled, you can check whether it will prevent you from getting to your daughter’s event on time. This requires synching your personal and work calendar, email and maps.
New writing tools with generative AI will be available in Mail, Notes, Pages, Keynote and third party apps. Apple Intelligence can write and summarize emails, too.
Apple Intelligence can take action between apps to carry out tasks on your behalf. For example, you can ask it to take action based on natural questions like “Play the podcast that my wife sent the other day.”
Apple Intelligence lets consumers create custom emojis easily in apps like Messages. Notes, Pages, Keynote and Freeform will all gain generative AI images. Apple intelligence is able to identify and recreate pictures of people in your photo library,” such as creating an image of them for their birthday.
Private Cloud Compute
Sometimes, Apple’s AI processes will need to be performed on external servers. To prevent privacy and security problems around this, Apple will run these processes on Private Cloud Compute, a home for Apple Intelligence that are still thoroughly within the Apple ecosystem and use Apple’s own Swift programming language. Even Apple can’t see that data, the company claims.
“You should not have to hand over all the details of your life to be warehoused and analyzed in someone’s AI cloud,” said Federighi.
ChatGPT will be integrated with Siri
In addition to Apple Intelligence, Apple is partnering with OpenAI to bring ChatGPT and GPT-4o to Siri, which will be used when referencing current information or gathering information that isn’t on your devices.
Siri will ask for permission to reach out to ChatGPT. You can do this for free or connect to your personal ChatGPT subscription. This integration is coming later this year. Plus, Apple plans to add support for other AI models.
Where Apple’s AI fits in to the competition
Apple has seemingly chosen to stay out of the generative AI chatbot race for the last few years, although it published research on running large language models on mobile devices in December 2023. The relatively late entry means Apple has had time to watch the rest of the industry try to find practical use cases for generative AI. It’s also seen Microsoft switch its Recall AI feature from default to opt-in after security concerns.
Will Apple face the same backlash and security digging as Microsoft’s Recall? And is Apple’s ecosystem so powerful that its email summarization and other Apple Intelligence capabilities will prove as game-changing as AI companies predict generative AI content to be? Or is Apple playing it safe when it comes to AI and, therefore, discovering little benefit to business productivity other than saving a few minutes at a time? It’s too early to say, but time will tell whether Apple Intelligence might be opt-in — or transformative — going forward.
