Eshu Marneedi

Gemini, write me a QVC segment

The Google Pixel 10 Pro. Image: Google.

Google’s Wednesday Pixel 10 announcement live stream struck me as one of the oddest technology events in recent years, not because of the smartphones, but because of how it was presented. While the Pixel 10, Pixel 10 Pro, and Pixel 10 Pro Fold are incremental updates over last year’s largely fantastic devices, the event was anything but typical, featuring Jimmy Fallon and a host of other celebrities who know nothing about any of the new phones. I usually write about Pixel events the night of the live stream, as the phones speak for themselves, but this time, I just needed to digest what I watched for a few days.

The Phones

Seldom are Pixel launches exciting because they’re largely just good Android phones. I’ve said that they’re the best non-iPhones ever since the Pixel 6 line because Samsung’s One UI Android skin annoys me as a purveyor of good software design. Google has historically marketed the Pixel lineup of phones as the “smartest smartphones,” and I generally agree, though I think the gap has diminished in the post-Gemini era. Samsung’s phones have the same Gemini artificial intelligence voice assistant and other smart features that have distinguished Pixels since their conception. Yet, no company has a better Android skin than Google itself, and all of its Pixel features work remarkably well. It’s as if Apple were competent at AI software.

The new Pixel 10 builds on Google’s effort to build the best iPhone competitor, and I think it seals the deal. The iPhone has never been the most technically impressive smartphone, but it provides the greatest user experience of perhaps any consumer technology product in the last 20 years. That’s why so many people love it — the iPhone, as the cliché goes, just works. The Pixel pastiche functions much in the same vein, as they’re nowhere near as powerful as Samsung’s finest, most expensive flagships, but they’re so much nicer to use, all at a reasonable price. Google knows this and sells the Pixel line not as a direct competitor to Samsung’s phones, which come out eight months earlier, but to the iPhones, which launch just weeks after the Pixel. (More on this later.)

Google this year introduced Pixelsnap, a blatant knockoff of Apple’s iPhone MagSafe¹ feature first introduced in the iPhone 12 series. All of the new models support Qi2 wireless charging at up to 15 watts — or 25 watts on the Pixel 10 Pro XL — and Google now sells magnetic accessories to attach to the back of the new phones, including its own version of the MagSafe charger and a nightstand dock. The Qi2 standard includes specifications for magnetic wireless chargers, as Apple helped the Wireless Power Consortium engineer Qi2 after its learnings from MagSafe, but it appears Pixelsnap is Google’s bespoke system with its own provisions. Mostly, though, MagSafe and Pixelsnap are indiscernible, and I largely think that’s good for the consumer.

MagSafe is terrific on the iPhone, and it has spawned a whole ecosystem of accessories, from tripod mounts to car phone holders to docks and cases. While I don’t imagine the Pixelsnap ecosystem will be so vibrant, taking into account Google’s minuscule smartphone market share even in the United States, there ought to be a few accessories that make Pixels more interoperable with a host of add-ons. MagSafe, in hindsight, should’ve been what wireless charging was all along, solely because it prevents coil misalignment issues leading to inefficiency. It’s just such a neat feature I couldn’t live without on my iPhone — I charge using a MagSafe charger every night, and whenever I travel, I miss not having it.

The Pixel 10 Pro Fold is also my favorite foldable phone due to its design and aspect ratio, which still trumps Samsung’s Galaxy Z Fold after this year’s update. The primary update this year, aside from a new processor and the typical minor camera upgrades, is the IP68 ingress protection rating. IP ratings are convoluted and largely too obscure for most people to follow, but they comprise two discrete measurements: dust and liquid protection. The first number after “IP” refers to the level of dust and sand protection the device has, and in the Pixel 10 Pro Fold’s case, it’s at Level 6, which is standard across most flagship smartphones whose makers can afford the laborious certification process. The second number refers to liquid ingress protection, and the Pixel line, like every other modern flagship, has been certified at Level 8.

It’s difficult for a folding phone to earn any level of dust ingress certification due to its hinge design, which adds an enormous amount of mechanical complexity to a device that otherwise would be “solid-state” throughout. There are hardly any moving parts in a modern smartphone, but the hinge on folding phones is a major one. If dust or sand wedges its way in, it could render the hinge useless. Samsung accomplishes some level of dust protection — IP48, specifically — using an array of brushes in the hinge to sweep away any grit or detritus whenever the hinge is moved, but Google’s design uses a dust-tight seal somehow. Either way, this level of dust protection diminishes, if not eliminates, a primary concern most foldable phone owners espouse. Now, the goal should be to make plastic-like screens more immune to scratches and scuffs.

If these updates sound minor, they are. The standard Pixel gains a new telephoto lens at the (slight) expense of some visual fidelity in the standard and ultra-wide sensors, and the new Tensor G5 in-house system-on-a-chip underperforms Apple’s five-year-old iPhone 12 in certain benchmarks. None of that is particularly remarkable, and neither are the new handsets at large, but that’s just standard-issue for a software company that happens to make halfway decent hardware. Prices remain the same, and frankly, if they increased due to tariffs, that would finally be something intriguing to write about.

The AI (I’ll Make This Quick)

Everything I wrote about Google’s AI image editing features holds up today, and I don’t have the patience to castigate Google yet again for misunderstanding, if not intentionally debasing, the value of a photograph.

The new AI feature this year, Magic Cue, is one Apple ought to replicate after it sorts out its Siri shenanigans. It works much like Siri Suggestions on iOS today, but powered by Gemini, as it should in the 2020s. An example Google provided onstage was digging through a user’s Gmail inbox for flight information as Gemini notices they’re calling an airline. What makes this example so futuristic is that it’s a promise of ambient computing, where a computer is doing some work on behalf of its user without additional intervention. This, in my eyes, is the true future of AI — not just large language model-powered chatbots, and the closer we get to a society without busy work, the more productive, creative, and stress-free humans will be overall.

The way most artificial intelligence companies, like OpenAI or Microsoft, accomplish this level of ambient personalization is by summarizing everything the company knows about a person and providing the model a copy before a chat. Google’s approach is different, supposedly tagging important emails, text messages, and other content to come back to later. If someone has a flight reservation in their email, it’s probably important, more so than a takeout receipt or a newsletter. Google’s AI has known this even before the advent of LLMs because it sorts important emails in Gmail, and using that prowess to power LLM features is truly something only Google has the wherewithal to do. The only other company that has enough data to personalize its product this deeply is Apple, and while it tries, it has just never been as good as Google.

What’s great about Magic Cue is that it isn’t particularly hallucination-prone, despite using Google’s highly inferior Gemini Nano on-device LLM. A few days ago, I posted about how Google should replace Gemini 2.5 Flash Lite with the standard Gemini 2.5 Flash in its AI Overviews in Google Search because it hallucinates too often, but the opposite is true here. People expect Magic Cue to be quick, and there’s very little room for error. Magic Cue doesn’t generate new text as much as it decides what to copy and paste and when to do so, and that makes it a perfect fit for less-accurate, smaller models. It isn’t a generative artificial intelligence feature as much as it is one rung above the usual Gmail and Android machine learning features from a few years ago. It works with and produces a limited amount of data.

The same goes for Camera Coach, which (as the name suggests) coaches people on how to take good photos. Some of the most common amateur photography mistakes include not choosing a predefined focal length (2×, 3×, etc.), not leveling the camera, and not cleaning the lens. A gentle reminder telling people how to take better photos as they’re doing it would drastically improve people’s experience with the camera on their phone, as most people don’t even use it “right.” It’s a harmless AI feature that can genuinely empower people to do more with the tools they have, and, moreover, respects the sacrosanct nature of a photograph. It seems like Google finally hired some people who see photography as an art of human expression, no matter how inconsequential that photo might be.

The ‘Show’

Google has always been one to lean on celebrity “endorsements” (advertisements) over substance because its products largely don’t sell themselves. Most people in the United States, by far Google’s largest Pixel market, buy their phones through carrier stores, where they either ask for the newest iPhone or the newest Samsung phone. Google is not only relatively new to the smartphone industry, having begun less than a decade ago, but it just doesn’t have the brand equity Samsung and Apple do. The same goes for Nothing, OnePlus, and the other Chinese smartphone manufacturers that haven’t been banned in the United States (yet). If Google wants to sell any smartphones, it needs to get into people’s heads, and celebrities are in everyone’s heads.

So, Google brought Jimmy Fallon, a late-night talk show host, to “interview” Rick Osterloh, Google’s hardware executive. In reality, it was a highly scripted, pre-choreographed affair where Osterloh and Fallon sat down in a “late-night” setting (midday in Brooklyn) and yapped about the new phones for an hour or so. Google even included an “Applause” light, like many talk shows, telling audience members — inappropriately, including the media — to clap when told. Tech events, live or pre-recorded, are usually presented by executives who know a thing or two about the products they’re selling. In Apple’s case, this works great because everyone knows what Apple events are. They have brand equity because Steve Jobs invented onstage tech presentations. Samsung brings celebrities and influencers onstage for the same reason — those people have equity.

Fallon has equity, but only to some extent, and certainly not in the way Google portrayed him as having. Fallon’s show mainly discusses popular culture and politics, and technology barely falls into either of those categories. Specification sheets, Gemini features, and camera updates aren’t Fallon’s shtick, and they’re best discussed by someone who knows what they are and the story behind them. Osterloh is that person at Google, much like Kaiann Drance or Greg Joswiak is at Apple, but because he doesn’t have name recognition, Google felt the need to supplement his knowledge with Fallon’s name. The result is a forced, awkward concoction that ultimately boosted the view count by six times last year’s event, but still felt awkward for the people who care the most.

The event ended somewhat unsurprisingly: a brief QVC-style segment selling people on the new phones with (clearly paid-off) audience members oohing for some reason, even though most of them had been briefed a week prior to the event. And the people truly shopping around for a new phone probably will watch Marques Brownlee’s hands-on video or The Verge’s writeup, both of which are more concise, resourceful, and entertaining than the nonsense show Google put on. The seven million people who watched the event video (a) pale in comparison to the tens of millions who watch the annual iPhone event, the magnum opus of technology, and (b) are probably just tech enthusiasts who saw the commentary about the event and decided to watch it for themselves. Normal people don’t burn an hour to watch a Google presentation.

This all makes you wonder why Google has such a tough time selling phones. Despite being a burgeoning hardware manufacturer, Google has an enormous amount of brand value. Everyone knows Google, Gemini, and all of its software, but hardly anyone buys its phones. Google has misunderstood its problem by thinking that it has a brand awareness issue, when in actuality, it suffers because it failed to break consumer habits. If you asked any random American who Apple’s No. 1 competitor is, they’d in all likelihood answer Samsung, when it’s almost certainly Google. It’s just that Samsung has made a name for itself by sneering at Apple products and positioning itself as the de facto Android market leader. In a way, it is, but Google has the home-field advantage of developing Android itself. It still, in my eyes, makes the best Android phones on the market.

For Google to succeed, and for its events to start picking up speed, it doesn’t need Jimmy Fallon or some other washed-up, spineless celebrity’s endorsement. It has to poach Samsung users by appealing to what makes iPhones interesting: their intuitiveness. There’s a large contingent of people who really believe Samsung phones take the best photos and have the best screens, but Google has to prove that the Pixel line is not only as performant as Samsung’s flagships, but adds to the experience in the same way iPhones do. Google’s phones, like iPhones, are tastefully crafted. They’re really well done, and they’re also cheaper than Samsung’s high-end flagship. Why not cater to that market of Samsung buyers eying an iPhone? Let the advertisements speak for themselves — position the Pixel as the iPhone with everything users (supposedly) like about Android.

Google, to some extent, is already doing this. The ads it airs on television are Apple-esque to their core. Yet, Samsung has a stranglehold because it has positioned itself as the central antagonist of Apple’s empire in a way Google hasn’t. I don’t think it must become the evil-spirited, spineless copycat weasel corporation that Samsung has turned to be, but it can position itself as a tasteful alternative for Android stalwarts. No amount of celebrity endorsements and cringey events will further that goal.