I've been reviewing mobile phones for over a decade, but in 2024 I find myself doing all sorts of things I wouldn't normally do when testing a mobile device: coming up with photo ideas that don't exist yet, calling people who speak a different language and having an AI translate the conversation in real time, and lining up group photos using holographic-like digital representations of my friends even when they're not there.
I'm talking about features in Google's Pixel 9 series and Samsung's Galaxy phones, which are the result of both companies aggressively building AI into their flagship products. Apple, too, is set to release a beta this fall of its own suite of AI-powered tools, Apple Intelligence, which will let you do everything from create custom emoji to summarizing your text messages.
The above scenarios may sound futuristic and exciting, and in some ways, they certainly are. But that doesn't mean it's worth upgrading your smartphone. With their new AI features, Samsung and Google offer novel solutions to seemingly common frustrations, like not wanting to hand your phone over to someone to take a group photo, or not being able to speak another language. But in practice, these tools can feel clunky and unnatural.
Read more: Why Gemini is more than just Google Assistant 2.0
Still, Google and Samsung's AI features feel like a glimpse into where things are headed, which piques our curiosity and excitement. For now, the most important new AI features on the Pixel 9 and Samsung's Galaxy devices feel siloed and built for specific situations, like editing photos or transcribing calls. They're certainly useful, but we long for the day when AI can make the whole phone smarter and more intuitive to use. It's entirely possible that virtual assistants and generative AI can make smartphone software more efficient, reducing the time we spend digging through apps and menus.
That mission is already in full swing, as evidenced by Google's Gemini helper, Circle to Search, Apple's upgraded version of Siri, and new (but seriously flawed) products like the Rabbit R1 and Humane AI Pin.
That vision is not yet a reality: A lot needs to happen in terms of feature development, usefulness, accuracy, and safety before AI becomes a reason to upgrade your phone.
AI capabilities today still feel indispensable
Samsung, Google, and Apple have all introduced (or are planning to introduce) new AI features for tasks like photo editing, creating images from scratch, summarizing and rewriting text, etc. These features have potential, but they're not fundamentally game-changing enough to dramatically change how we use our smartphones.
Take the newly launched Pixel 9, for example: some of its flagship AI features include the ability to search for content within screenshots, a tool to change an object in a photo into something else, call summaries on the phone, weather summaries, and a new app that generates images based on prompts.
By themselves, these tools are only that useful in certain niche situations. For example, in my review, I could think of only a few specific scenarios where Pixel Studio would be useful, even though it was fun to play around with. Unless you're taking photos with your phone for a project and need advanced editing tools, or you just want to create funny photos for fun to share with friends and family, I don't see Reimagine (the editing option that changes objects in photos) being that useful on a day-to-day basis.
Read more: Samsung and Google's mysterious mixed reality plans: What we know so far
The Pixel Screenshots app allows you to search for content within your screenshots.
James Martin/CNET
The other features mentioned above, like screenshot search, call and weather summaries, are certainly useful, but they're too minor to really influence a phone-buying decision. I've been using many of Samsung's Galaxy AI features on the Galaxy S24 Ultra and Galaxy Z Fold 6 for the past few months, and I feel the same way about them.
Like Google, photo editing and content creation are key focus areas for Samsung, as evidenced by features like Generative Edit (remove or modify objects in a photo) and Sketch to image (create a polished image based on a rough sketch). Samsung also builds some translation features into its Galaxy smartphones as part of Galaxy AI. With Live Translate, for example, Samsung's AI stands between you and the person you're talking to and translates the phone for you. Samsung smartphones can also translate entire text message threads into another language. However, unless you travel a lot or regularly communicate with people who speak another language, this isn't a reason to buy a smartphone.
Samsung's Sketch to Image feature can work wonders, and sometimes the results can be surprising.
Lisa Airdicicco/CNET
Apple Intelligence, currently only available as part of the company's developer beta program, includes similar features, such as image creation and tools for summarizing and rewriting text, though not all features are enabled yet and you'll need to spend some more time using it to determine how useful it is.
AI features lack the human element
But perhaps my bigger criticism is that, while some of these tools are certainly useful, they just don't feel very… human. The suggested text I created using Samsung's Composer feature, which creates messages based on prompts, felt impersonal and generic, which is to be expected considering it was written by an AI.
Google's Add Me feature is another example of a tool that seems useful but just feels socially weird to use: When I tried to take a group photo using Add Me, it felt like I was organizing a photoshoot rather than just casually taking photos with friends. For those unfamiliar with this new feature, using Add Me requires quite a bit of explanation.
The resulting photos don't always look natural, either: As I mentioned in my review, a colleague and I tried to take a photo where she was leaning on my shoulder, but she ended up standing too far away, and in the final photo, it looked like she didn't even realize we were standing next to each other.
This photo was taken using the Pixel 9's new “Add Me” feature.
Lisa Airdicicco/CNET
I felt the same way about Samsung's language translation feature, although it was well-intentioned. When using live translation, it can be hard to know when to start speaking and when to stop, as the AI translation works.
Samsung's conversation mode, which displays language translations on both the outer and inner screens of its flip phone, is also great in theory but, from my experience, doesn't work in practice: it's intended to use the flip phone's dual screens to allow two people in a conversation to simultaneously see a translation of what the other person is saying.
But when I tried it out at Le Marché des Enfants Rouges, a bustling food market in Paris, I found it hard to concentrate on the conversation because I was mainly paying attention to what was on the screen. I ended up ordering the old-fashioned way, by pointing, gesturing, and using what little French I actually knew. CNET's Patrick Holland also tried it out in Paris to order a coffee, but a slightly inaccurate translation left the French barista feeling awkward.
I can give Samsung and Google credit for trying to solve real problems some people might encounter, but these are features that you have to go out of your way to try out, rather than something that naturally fits into your life.
Read more: How Samsung decides whether to turn a futuristic concept into an actual product
The potential is there, but there are also big obstacles to overcome.
Circle to Search is one of Android's most promising new AI features.
James Martin/CNET
That doesn't mean there's no opportunity for AI to upgrade our phones in meaningful ways. What will make AI truly useful is its ability to understand the data on our phones and give us what we're looking for with little to no effort. Google's Circle to Search, a feature that lets you search for almost anything on Google by circling it, is a step in that direction.
The industry is moving quickly to make that a reality, and the recent upgrade of Google's Gemini assistant is evidence of this. For example, you can call up Gemini and ask it to analyze the content on your screen, including asking questions about the YouTube video you're watching. Luckily, features like Circle to Search and Gemini may not require an upgrade depending on the phone you have. Circle to Search has been available on a variety of Galaxy and Pixel Android phones for the last few years, and Gemini is available across Android as an app and virtual assistant, and within the Google app on iOS.
Apple, too, seems committed to this goal with Apple Intelligence, which includes a new version of Siri that can connect the dots by taking into account your personal context when answering your questions. All of these changes feel like they're getting us closer to the goal of turning our smartphones into true virtual assistants that can get things done without requiring too many taps and swipes.
But to get there, there are some very big hurdles to overcome: The massive language models on which many generative AI features are based are still prone to hallucinations, as we noticed when Gemini gave us the wrong answer while testing the Pixel 9. And remember when the AI hack built into Google's search engine suggested putting glue on pizza?
As Digital Trends and The Verge point out, tools like Pixel Studio and Reimagine (a feature that changes objects in a photo into something else) let you bypass Google's guardrails to create disturbing images. And, as The Verge points out, these features and apps are built into Google's phones, so you don't need any technical knowledge or special software to use them.
The Pixel Studio app allows you to generate images using prompts.
Andrew Lankson/CNET
In a statement to CNET, a Google spokesperson said Pixel Studio and Magic Editor are designed to prioritize creativity and “respect the intent of user directives,” but that “offensive content may be created at the user's direction.”
“That said, this is not a free-for-all,” the statement said. “We have clear policies and terms of service about what content we allow and don't allow, and we have built guardrails to prevent abuse. From time to time, some prompts may run afoul of the guardrails of these tools, and we remain committed to continually strengthening and refining the safeguards we have in place.”
With Samsung and Google showing off new AI features, and Apple Intelligence requiring a new iPhone to work, you might be wondering whether AI is a reason to upgrade.
The answer is no, at least for now.
The biggest thing I learned from trying out new AI-enabled phones is that these features (even handy ones like Circle to Search) force us to think about using our phones in new ways, and while technology evolves quickly, it's only our willingness to adopt it that makes it feel essential.
Weird and wonderful images from Google Pixel Studio
See all photos
Source link
