In late 2010, an up-and-coming Google executive named Sundar Pichai took the stage in front of a bunch of reporters and explained that Google had a big new idea about laptops. In the future, he said, the only thing you’d need on your computer was a web browser — that was where all your apps, data, and connections would live going forward. Your computer didn’t need much storage, Pichai explained, or any fancy hardware, because the cloud would do the work. The software designed for these new laptops, Chrome OS, was an enormous bet that the web would eventually overtake everything.

Chrome OS (which eventually became ChromeOS) has been, in many ways, a huge hit. But the computer currently on Rick Osterloh’s desk isn’t a Chromebook. Osterloh is Google’s head of hardware and devices, and the sleek, slim device he’s tapping on as he joins our Google Meet is a 13-inch Dell clamshell with decidedly un-Chromebook specs: 16 gigs of RAM, a cutting-edge Qualcomm processor, a light and thin and MacBook-ish metallic look. It’s called the XPS Googlebook, and it’s one of five new devices running Googlebook OS, Google’s latest big idea about computers.

At first blush, Googlebooks feel like classic Google in a bad way: yet another operating system, yet another brand, yet another thing destined for the Google graveyard. But Osterloh sees it differently. He greenlit the whole project in part because he saw it as “the ultimate focusing move” for Google. For a long time, he says, Google has built, maintained, and shipped two operating systems. ChromeOS stayed on laptops, while Android conquered phones and tablets and TVs and cars and everything else. Slowly but surely, Google has built Android into a familiar interface and a large app ecosystem across all those devices. Except for laptops.

Meanwhile, Google became convinced that the advent of AI might change the way we use computers forever — but might not need to change the computers themselves. Osterloh says he has tables full of wheel-reinventing prototypes, and of course the company is betting big on things like smart glasses, but high-end PCs, which have the power needed to run AI models for coding, dictation, and more, have been a surprise winner of the AI revolution so far. (Just ask Apple, which has seen AI enthusiasts buy high-end Macs in huge volume.) “I think laptops are one of those resonant form factors,” Osterloh says. “It’s an established category that’s incredibly useful for productivity and apps, and it’s going to be really hard to replace that.”

Those two ideas together — Android’s dominance and the rise of AI — tell the whole story of Googlebook. There are a lot of bets inside of the Googlebook project, both about what we do on computers and how we’ll do it going forward, but maybe the most surprising is Google’s commitment to building kickass laptops. A generation ago, then-CEO Eric Schmidt argued that the future was “disposable” hardware, that devices wouldn’t matter as the internet took over. Googlebooks represent Google’s growing belief that devices matter more than ever. And that a good laptop is worth paying for.

The Googlebook project started in earnest a couple of years ago, but some of the ideas have been floating around much longer. Rumors and reports of a ChromeOS-Android merger have been around for more than a decade, and Google has tried various ways to bring the two operating systems into harmony. The company spent years programming ChromeOS to be able to run Android apps (and even Windows apps), but never nailed the integration. For years, a project called Fuchsia looked like the combination users had been waiting for, but that petered out into a smart home-specific project.

In early 2024, though, a couple of Googlers asked a new question: What if, instead of trying to bring Android apps to ChromeOS, we brought the best of ChromeOS to Android? They weren’t the first people to ask that question, but they were the first to time it right. “The key thing that had changed,” says Prajakta Gudadhe, Google’s VP of engineering for laptops and tablets, “was just the rapid pace of evolution that Android itself has gone under.”

For so many years, Android was a phone operating system that was sometimes hamfistedly shoved into other devices. But Gudadhe felt like Android had finally matured into something truly cross-platform. Developers had begun to embrace multiple screen sizes and orientations, and the Android team at Google had changed the way they built the operating system. “Android at that time was serving multiple form factors, from phones to wearables to TVs to autos,” Gudadhe says, “and because of that they just had to invest in a much more modular stack that could support this diversity.”

What that meant in practice was that it might finally be possible to build an Android laptop that could natively run the universe of Android apps, without looking and working like a grossly oversized Android phone. By combining base Android technology with some of the best features of ChromeOS — excellent multitasking and multi-window support, fast boot times, easy software updates — Google could have the best of both worlds.

The idea became a memo proposing a new laptop OS, which was excitedly received by Google leadership, and Project Aluminium was born. (To be clear: It is A-lu-min-i-um, not aluminum, and there was a swear jar in the office for people who dared to pronounce it the American way.)

In addition to the technical underpinnings, the Aluminium pitch included a couple of big ideas about the future of laptops. The first was that more and more users were starting their computing journey with Android devices, and thus were expecting all their other computers to both interact with and even look like their Android phones. They wanted widgets, homescreens, notification shades, and customization options. They also wanted all their apps on their laptop, and their data in perfect sync all the time. That became one of the core tenets of the operating system — perfect integration with Android devices. (iPhone owners were, and are, mostly out of luck on this front.)

The team was also convinced that Gemini and other AI tools were part of the future of computing, and that they should be built into the operating system and not just a separate app. The advent of vibe coding meant people might want to build their own software, in a terminal or with the help of on-device agents, and would surely want easier ways to get it running on their devices. But even little things about computer use felt ripe for AI change. “The mouse pointer? That thing seems pretty stupid,” says Alexander Kuscher, Google’s senior director of product management for laptops and tablets, just naming one example. Your cursor represents your attention, and moves around with your brain and your eyes, so why can’t your computer just see and respond to it? “I’m copying and pasting stuff all over the place, and typing to chatbots that don’t give me the ability to just express my needs with the mouse pointer.” That thinking became one of the most novel features of Googlebook OS, Magic Pointer, which lets you invoke your AI bot just by wiggling your cursor over an image or a bit of text. The idea, Kuscher said, was to “not change the ergonomics [of a laptop], but to make sure we take care of people coming from the ecosystem, living in an AI world.”

As the Aluminium team started working on an AI-centric, Android-powered laptop, they began discovering problems — and other teams at Google trying to solve them. “One is Android Automotive, which cares a lot about multi-user,” Kuscher says. “They need fast user switching because different people get in the car. We were like, oh, me too!” They found folks on the Android Wear team working on improving the Bluetooth stack for better accessory connectivity, and people on the XR team thinking about multi-window support for Android.

This is the focusing that gets Osterloh excited. The well-earned stereotype about Google is that nobody ever talks to anybody else, and so the company ships a thousand slightly different messaging apps, reminders systems, and AI subscriptions. In that vein, the Aluminium team could have come up with new ideas about everything and baked them all into ChromeOS; what they did instead was integrate so many disparate ideas about Android into a single platform.

Even the ultimate name of the devices and the software, Googlebook and Googlebook OS, hint at the extent to which Google hopes Aluminium brings everything Google-y together in one place. “We realized pretty quickly that this is about the whole Google experience,” Kuscher says. “If you have a phone that runs Google, and a car, and a laptop, that’s what this is about.”

(The naming here is a delightful bit of Google chaos, by the way. Googlebooks run Googlebook OS, which is mostly Android, but with some ChromeOS trappings thrown in. Get it? Me neither. They’re Android laptops.)

It became very clear to the Googlebook team that this operating system was going to ask a lot of the hardware it ran on. All the system-level AI would need enough processing power and memory to feel seamless, of course, but it went further than that. “We didn’t want to have Magic Pointer, and then it’s really hard on your trackpad to actually use it,” says Kuscher. “That’s just incongruent.” Likewise, what good is letting you build software with a prompt if the keyboard is awful to type on or the battery dies before it’s done? If the screen is too small, or too low-res, nobody will want to use it instead of their phone. For a Googlebook to be even a little bit worthwhile, the hardware would need to be great.

From the beginning, Osterloh says, Google wanted to build an ecosystem of Googlebooks. The company began reaching out to PC makers, chip companies, and even phone manufacturers, all with roughly the same pitch: You’ve already been making either great laptops, great Android devices, or both, and now we’re putting the whole puzzle together. For a PC industry always looking for Windows competition and increasingly interested in building ecosystems of products to compete with Apple, the pitch worked: Google is starting with five devices from five partners — Acer, Asus, Dell, HP, and Lenovo — and Osterloh says there are many other partners waiting in the wings. Potentially including Google’s own Pixel team. The first five devices, he said, are merely the ones that are ready now. And he says partners are clamoring to be next.

The harder group to convince of this whole thing? Users. For all their success in education and elsewhere, Chromebooks never really caught on as personal computers. That’s largely Google’s fault, and largely for two reasons: because Google never successfully convinced people that all they needed was a web browser, and because Google and its partners spent so many years making and marketing crappy computers. Early Chromebook commercials were, in fact, explicitly anti-hardware: They showed people lighting their laptops on fire and throwing them into rivers, because who cared about the device when everything was in the cloud? As a result, even when Google and its partners have made high-end Chromebooks, like the terrific Pixelbook, they failed to convince anyone that they were worth the money.

The first lineup of Googlebooks will cost between $899 and $1,299, which puts them squarely in competition with devices like the MacBook Air. Osterloh says he thinks Googlebooks will eventually reach other form factors and lower prices, but seems unfazed by the task of convincing people to spend top dollar on a Google laptop. For one thing, he thinks the increasing popularity of AI has made people more willing to buy high-end computers in general, and has maybe even increased the importance of laptops and desktops. “I still fundamentally believe it will always be your central computer,” he says. “The thing that connects you to networks, that has the primary brain, the important screen for entertainment and productivity.” But more than that, he agrees that this kind of hardware, and this kind of price tag, are simply what it requires to do Googlebooks well. And Google definitely won’t get a third chance to convince you it’s right about laptops.

When Google first announced the Googlebook concept in May, it was heavily focused on AI and Gemini. But the work that led to Googlebooks goes back much further than the company’s current obsession with LLMs, and is about more than Google’s chatbot. In some ways, Osterloh says, the whole history of Android has led up to the launch of Googlebooks. Their first job, as he sees it, is to “run the applications you know how to use and sync really well with all your Android usage and applications and devices that you’ve connected to it. And it just happens to be an awesome laptop.”

Sixteen years ago, Google shipped a laptop for the internet age. Now it’s trying to build one for the mobile era that has already come, and the AI-centric one that’s still forming. Getting it right will require marshaling nearly everything Google does well — and the best devices it has ever made. It will also require proving to users that they need a computer that’s more like their phone, and more connected to their AI ecosystem. None of that is a given; it might be an even bigger leap than trying to move your laptop to the web. But if Google’s bet pays off, our cursors might never wiggle the same way again.