Apple’s annual iPhone event is happening next week, when the company is expected to unveil a range of new devices that includes its thinnest phone to date. It’s also the first iPhone event since the embarrassing Apple Intelligence flub earlier this year, which saw the company push back AI-powered Siri features announced last year all the way into 2026.
Apple is rarely first to introduce cutting-edge new features, instead preferring to wait until it can offer a differentiated take on technology that’s ready for the mainstream. With artificial intelligence, though, Apple has essentially issued a mea culpa; it actually wanted to jump into the fray but just couldn’t get ready in time.
With that in mind, it’s worth looking at the current smartphone landscape to see what the new iPhones will be going up against. Apple made the case last year that AI would transform the smartphone experience, even if it whiffed on actually shipping anything to that end. So what about the alternatives?
Subscribe to Multicore. Multicore is about technology hardware and design. It’s written from Tokyo by Sam Byford. To learn more visit multicore.blog
Google Pixel 10 Pro
The just-announced Pixel 10 Pro phones are Google’s most advanced yet when it comes to AI capabilities, and offer arguably the most AI-forward options on the market.
Google’s AI focus with the Pixel 10 line goes right down to the decisions it made with the Tensor G5 system-on-a-chip, which was designed in-house and, for the first time, fabricated on TSMC’s leading-edge 3-nanometer (nm) process. But that hasn’t brought a huge leap in performance; instead, Google is devoting a huge amount of the die space to Tensor cores, which handle on-device processing tasks. That means that the Pixel 10 doesn’t perform all that well in benchmarks or gaming compared with other current flagship phones—Google is putting its silicon where its mouth is when it comes to the importance of AI.
This year, Magic Cue is the most prominent new AI feature, and it’s notably reminiscent of what Apple claimed the iPhone would be able to do with Apple Intelligence. It’s a context-aware feature that proactively surfaces information from across your other apps. For example, if you’re in a chat and a friend mentions a restaurant reservation that’s buried in another inbox, Magic Cue will suggest the information and relevant actions right there within your current app. This all runs on-device using Google’s Gemini Nano model.
Pixel phones have long been ahead of the curve when it comes to processing language on-device. The auto-transcribing Recorder app is a godsend for journalists, while the ability to automatically add translated subtitles to any audio or video in real time opens up whole new worlds of content. This year, the Pixel 10 adds the ability to translate phone calls in real time, meaning that the individuals on the other end of the call can hear you talking in their language in what sounds something like your own voice.
It’s hard to say exactly how reliant the Pixel phones are on their AI-focused silicon, because most of these features simply don’t work on other phones. What is clear, though, is that Google has identified AI as the biggest differentiator for its own smartphones, which historically have been a minor player in terms of market share.
This feels like more of a long-term play rather than something that will attract most consumers today. But it’s hard to deny that the Pixel 10 Pro offers the most advanced and complete implementation of AI technology on a smartphone today.
Samsung Galaxy
Samsung, meanwhile, didn’t shower itself in glory during the years when it was pushing its Bixby voice assistant on everyone, but Galaxy phones are in much better shape today when it comes to artificial intelligence. Although much of the Galaxy AI suite relies heavily on Google’s own technology, Samsung has implemented it in smart ways, and it benefits from tight collaboration with Google.
The Circle to Search feature, for example, which scans what’s on your screen for text and visual information and lets you search based on that, actually made its debut on Samsung phones in 2024, before later coming to Google’s own Pixel. Samsung has also been up to speed when launching features like real-time Gemini Live chat across its phones and watches. And all of this works alongside Samsung’s own system-level implementation of features like generative photo editing, writing, and translation tools.
Samsung’s Galaxy Z Fold7 is also the most advanced foldable phone available in the U.S., a form factor that I’ve found is particularly well-suited for AI workflows. The square-ish inner screen can be an awkward fit for many mobile apps, but ChatGPT and Gemini have handy column-based interfaces that give more real estate for responses, which often include tables or other formatting that can be difficult to read on a regular phone screen. AI apps are also especially well-suited for multitasking; foldable phones let you essentially run two full-size phone apps at once, which is very helpful when editing and cross-referencing.
Apple iPhone
Part of the reason the botched Apple Intelligence rollout was so surprising is that it felt like Apple should have had this in the bag from a hardware perspective. Apple Silicon is the envy of the entire tech industry, with Apple’s homegrown chips powering unbeatable performance across Macs and iPhones alike.
The A17 chip in last year’s iPhones does have a lot of theoretical AI capability due to the large onboard NPU, which Apple calls the Neural Engine, but it’s difficult to see how that truly improves the current iOS experience. Right now, the chip mostly seems to be deployed to speed up core functionality like photo processing and Face ID scanning—which isn’t nothing, but it might not be maximizing the potential of the silicon.
There is definitely a conceivable future in which Apple delivers the best overall AI smartphone experience due to its on-device capabilities. But despite its hardware advantage, right now we’re nowhere close. It’s unlikely that the iPhone 17 lineup will do anything to change that—at least not until Apple gets its AI software in order.
That said, Apple is much better at keeping software leaks under wraps than it is with its hardware. It’s possible the new iPhones—and the new chip that powers them—will come with some new AI features that were developed separately from the Siri disaster.
It will be interesting to see to what extent next week’s iPhone announcement does or doesn’t mention AI. But regardless of what’s in the pipeline, the bar has already been set a whole lot higher than the one at which the iPhone currently rests.
Subscribe to Multicore. Multicore is about technology hardware and design. It’s written from Tokyo by Sam Byford. To learn more visit multicore.blog
The early-rate deadline for Fast Company’s Most Innovative Companies Awards is Friday, September 5, at 11:59 p.m. PT. Apply today.
