.jpg)
Apple's iOS 27 public beta introduces a significantly overhauled Siri ("Siri AI") that now functions as a chatbot-style app woven throughout the OS — accessible via search, camera, messaging, and more. Key features include on-device "indexing" that lets Siri pull personal context from texts, calendars, and screen contents; a "Write with Siri" drafting tool; and visual intelligence for analyzing photos/screenshots. It requires iPhone 15 Pro or newer, lacks persistent memory (unlike ChatGPT or Claude), and reviewers see it as a major step up for Apple users, even if Android's Gemini has offered similar capabilities for a while.
iOS 27 officially escaped the developer world today with the launch of its first public beta. I’ve been testing the new operating system since early June, hunting for quirks and seeing if it can live up to the hype Apple promised at the keynote.
This year’s iOS upgrade is what you might call a "Snow Leopard" release. It’s light on flashy new features, focusing instead on fixing what was broken and speeding up the system. App launches, Photos search results, and AirDrop transfers are noticeably faster. The Messages app now supports inline replies and end-to-end encryption for RCS, and Liquid Glass has been refined for better legibility around hard edges and text. These are wonderful quality-of-life updates, especially for those with aging iPhones. But by far the biggest, most anticipated change is that Apple has finally, truly shipped the revamped Siri AI—at least as an opt-in beta. And this time, I think Apple might have actually pulled it off. Or, at the very least, it has laid a solid foundation for a successful AI assistant.
Last week, I was trying to figure out if I had time for a free four-hour concert in the city featuring three acts. I only really wanted to see one band, but the event page didn't list the playing order. It was the perfect opportunity to pressure-test Siri AI. I swiped down from the top of the screen and asked, “What order are the bands playing in?” Siri spun its new loading animation for a few seconds, then correctly told me the band I wanted to see was playing last.
The promise of the new Siri AI is to fundamentally change how you use your phone. Previously, you opened an app and told it what to do (call a car, set a timer, order lunch). Now, you state your goal first, and Siri AI sifts through your available apps and data to handle the rest. During that concert query, Siri read the webpage, searched the web, found the answer, and presented it. I didn’t need to jump between browser tabs or dig through the band’s Instagram; the answer was just there.
In the month I’ve been using the beta, Siri has surprised me constantly. On the first day of testing at WWDC, I asked, “Can you add my WWDC briefings to my calendar?” Siri dug into my email, parsed the data, and added six individual events with the correct times. (I should note that it could only add them to the Apple Calendar app, but I’ll explain why shortly.
These interactions have genuinely shifted my habits. I now default to Siri first, just to see if it can perform an action or answer a question. It has practically stopped me from opening a web browser for simple tasks; it’s just easier, faster, and more enjoyable to swipe down and type a prompt. Onscreen awareness has been particularly helpful. Asking Siri about what’s currently on my screen saves me a lot of tapping, and when it can take direct action—like adding an event or pulling up an address—it feels like magic.
But when it hits a wall, I’m reminded of how much work is left before we reach the “it just works” universe, where Siri failing is the exception rather than the norm.
Siri’s ability to translate natural language into a specific, context-aware action is still a major work in progress. For example, while looking at a concert page, I asked it to “remind me to buy these tickets when they go on sale.” Instead of recognizing the context of the page, it just created a generic text reminder. I had to carefully phrase it as, “buy tickets to this when they go on sale,” to force Siri to look at my screen and search for ticket availability. Similarly, asking Siri to “route” me to a location usually did nothing, but asking it to “direct” me worked. I imagine its word correlation will improve as more data rolls in, but when the entire selling point is natural language processing, having to rely on specific keyword triggers is frustrating.
Right now, if you’re in the iOS 27 beta, only Apple’s first-party apps have access to Siri’s new capabilities. If you live entirely within the Apple ecosystem, everything is great. Your data likely exists in Messages, Mail, and Photos, and your actions involve Reminders, Notes, and Calendar. When these apps work together, it feels like a glimpse into the future. But if I ask Siri, “When did Daniel say he was free to play Dota?” it has no idea, because Daniel and I message on Telegram, which the system can't access.
To fix this, developers need to implement two main updates for Siri AI support: entities and intents. Entities represent the type of data an app contains (a photo, a recipe, a playlist). If an app registers entities, Siri knows what kind of personal context it can pull from it. Intents tell Siri what it can do with that data (play, save, delete). Together, they allow Siri to control and extract data, while Siri’s semantic layer handles the understanding of your words.
If your digital life exists mostly outside of Apple apps, it will take time for your essential tools to play nicely with Siri AI. This definitely won’t happen during the public beta. While developers can now build entities and intents against the iOS 27 SDK, they can’t push Siri AI updates to their apps until the full, non-beta version of iOS 27 launches in the fall.
Apple is relying heavily on developers to update their apps, which isn't entirely unprecedented. Whenever Apple adds a major feature like Dark Mode, developers have to do the work to support it. The difference here is that previous updates directly improved the experience inside the developer’s own app. This time, Apple is asking developers to do the heavy lifting to improve the Siri experience.
Developers I’ve spoken to say that while they’re excited, it’s a massive undertaking. “The conceptual challenge for developers is creating comprehensive support for every screen and function within an app,” says Matthew Cassinelli, who worked at Workflow (the company Apple acquired to create Shortcuts). “But the transition to agent-based models allows specialized apps to surface relevant data dynamically, which makes them more useful to users who otherwise might not open those specific applications frequently.”
I can see how Siri might help surface niche apps—like the LookBack: Contacts History app, which could tell you who you met at a conference last week. But the bigger question is whether tech giants like Google will actually want to make Siri more capable. Google makes most of its revenue from ads; if Siri can surface your Gmail information directly at the top of your screen, Google loses that engagement and ad revenue. That said, Google is already pivoting to AI overviews, so it’s clearly preparing for a post-AdSense world.
The primary incentive for Google to adopt full Siri AI support might just be consumer choice: if a rival email app supports Siri AI fully and Gmail doesn’t, users might switch. Whether Gmail's stronghold is strong enough to retain users despite a crippled Siri experience remains to be seen.
All of this remains a work in progress. Not only is this a beta OS, but it’s also a beta version of Siri—and only a partial one at that. It’s impressive, but the true test comes next. Developers need to adopt Siri AI for it to reach its full potential, and the AI itself needs to route and research requests flawlessly far more often than it currently does. So far, however, I’ve been genuinely surprised by how much it can do.
The concert was great, by the way.