He Fathered Android, but Robots Were Always His First Love | A Curious Person

Without him, there would be no Android as we know it.

Who: Andy Rubin (1963–), creator of the Android operating system and of the Sidekick, one of the first smartphones.

His curiosity: When Andy Rubin has a hunch about the future, he asks why nobody is building it yet. Having made Android the world's largest operating system, he now wants to connect every electronic device.

Last month, Andy Rubin's secretive company PlayGround revealed its plans in a company blog post: it will help startups build hardware, and it will build the infrastructure to link all electronic devices together.

It was the first time the father of Android had set out his plans in detail since he left Google three years ago to found PlayGround.

Nobody who knows him was surprised by the idea. Andy Rubin has loved robots and automation since he was a boy.

At 15, the Star Wars fan built his own R2D2 robot and programmed it to roll into his younger brother's room.

"Kids who love programming computers live in their own world," Rubin said. "But if you build an R2D2, it takes you out of your world and into real life."

This boy, at home both in his own head and in the world, went on to found two companies named after robots. One was Danger, named after the robot that sounds the alarm in the sci-fi TV series Lost in Space. The other he simply called "robot": Android.

The first made a smartphone four years before Steve Jobs did, and failed because it was too far ahead of its time. The second was bought by Google and became the world's largest smartphone operating system.

After smartphones, Rubin, now 53, went back to what he loves most. Three years ago he left Google and founded a new company, PlayGround. He hired a crowd of experts in chips, supply chains and industrial design, and invested in a string of startups in data, healthcare and drones, helping them make products, especially hardware.

He wants to build a framework, as he once built Android for the phone makers. Only this time it isn't for phones but for every kind of gadget: smart ovens, fridges, even your electric toothbrush. In short, smart robots everywhere.

Ask Rubin why he is doing this, and the answer may simply be that he thinks it's cool.

"He does things because they're cool, with a childlike innocence," Zarko Draganic, a longtime colleague from Apple, once said of him.

Before the iPhone, Rubin put the internet and an app store on a phone

Rubin is best known for creating the Android operating system.

It did not come overnight. He had worked in tech for years before Android. In 1989 he was a software engineer at Apple. He later moved to General Magic, an Apple spin-off that made handheld personal computers. In 1998 he joined WebTV, a maker of interactive internet television. Microsoft soon bought the company, and Rubin left Microsoft a year later.

That was 1999. Out of a job, Rubin rented a storefront in Palo Alto, California, called it "the Lab," and sold robots he had brought back from Japan. The shop never did much business, but it became a hangout for Rubin and his engineer friends, who often stayed up all night brainstorming new and interesting things to build.

The idea they arrived at: a device that cost under $10 and was no bigger than a wrapped piece of candy, yet could browse the internet.

Nobody wanted to invest in it.

So Rubin and his friends reworked the idea. The new device was the size of a bar of soap, with a radio receiver and converter, a screen that flipped up, and a keyboard. It could make calls, install apps and go online. They called it the SideKick.

To run the venture properly, Rubin set up a company called Danger. The dot-com bubble had just burst and money was scarce, but Danger still raised funding from Greg Galanos, an investor with a taste for cutting-edge technology, and teamed up with the carrier T-Mobile to bring the Sidekick to market.

Danger's technology was far ahead of its time: the phone could download games, browse the web, even work with a headset. Yet it never earned what it should have. Sales never took off, and it stayed a niche product loved by geeks and teenagers. A year later, the investors pushed Rubin out.

"Maybe one in ten engineers today has heard of Danger, yet a lot of what modern smartphones do, we did first on the Sidekick. We were clearly five years early. We knew that if we had launched it in 2005, we would have owned the world."

So said Chris Desalvo, an engineer at Danger who later followed Rubin to Google.

The Sidekick did not own the world, but it did catch the eye of Google co-founder Larry Page.

From an unloved team inside Google to the world's largest mobile operating system: Android took six years

Pushed out of Danger, Rubin founded a new company, Android, in 2002 to keep making phones. It struggled. By 2005 Android was barely hanging on, and even the $100,000 Rubin had borrowed from his old friend Steve Perlman to tide it over was almost gone.

So he laid out the company's whole plan to Samsung, hoping for a buyout or an investment. Instead, he hit rock bottom.

"What army are you going to build this future with? You have six people. Are you high?"

That was what Samsung's CEO told them after Rubin pitched his vision for Android at Samsung's headquarters in South Korea. Rubin recalled that they left the meeting room almost to laughter.

Then the story turned fast. Two weeks after Samsung's rejection, Rubin met Larry Page.

"How about Google buys Android?" Page had made up his mind before the meeting began: $50 million plus stock incentives to buy Android, or rather, to buy Andy Rubin.

Inside Google, too, Android was unloved at first. It was Page's secret project: heavily funded, but hidden away. That cut against Google's culture of openness and sharing. To prevent backroom dealing, any Google engineer could read the code of any product. Android was the exception.

That alone caused plenty of friction inside Google. Worse, the Android team's first smartphone, the Sooner, died before it was ever launched.

The reason: in June 2007, Steve Jobs unveiled the first iPhone, with a full touchscreen, a virtual keyboard and a strikingly new way of using a phone.

When Apple announced it, Rubin was on his way to a meeting. He was so stunned by what he saw that he had the driver pull over so he could watch the whole livestream.

The Sooner could do things the iPhone could not yet do: run several apps at once, offer an app store, and work without regular syncing to iTunes. But next to the iPhone it was ugly: a small screen, a full keyboard, a primitive look. Nobody would want to buy it.

Right there in the car, Rubin told a colleague: "Man, I guess we're not going to ship that phone."

Android's first launch event, in November 2007, was only four months away.

Worse than the deadline was the pressure inside the company. Google had been on good terms with Apple, and when the iPhone came out, Google's team building apps for Apple was still growing fast. If Google entered the phone business itself, that partnership would suffer badly.

"If we wanted Android to succeed, we had to stop giving our best stuff to Apple." Over this, Rubin clashed hard with Vic Gundotra, the executive then pushing for more investment in the team writing software for Apple.

At the 2007 launch, Google had no hardware to show either, only the Android operating system and a global alliance.

The alliance began with 34 handset makers, software developers, carriers and chipmakers, and grew into the Open Handset Alliance of 84 hardware makers, software developers and carriers working together to develop and improve Android.

Put more plainly, these companies joined because Google paid them. That included the first Android phone to go on sale, the HTC G1 in 2008, a deal Google secured for several million dollars.

But in the rush to ship for Christmas 2008, the G1 and its Android 1.0 had plenty of problems; you couldn't type with the phone held upright, for one. To keep the team going, Rubin paid extra year-end bonuses to Android staff out of his own pocket. One engineer's bonus doubled that year.

The G1 sold poorly, but because Android was more open, many phone makers were willing to sign on, including Samsung, the company that had once turned Rubin away.

From there, Android became the fastest-spreading operating system in the world. In 2009, Nokia's Symbian still held most of the market and Android had just 3.9%. A year later Android's share reached 22%, overtaking both Apple and BlackBerry maker RIM.

Today 80% of smartphones run Android, and the systems of its rivals, Microsoft, Nokia and BlackBerry, have shrunk to almost nothing.

Andy Rubin never gave up on robots

The Android story is better known, but Rubin never gave up on robots. Since building his R2D2 as a boy, he has been hooked on making fun things with his own hands.

In 2004, when he was too busy with Android to build robots himself, he slipped his longtime friend Sebastian Thrun $100,000 to fund Thrun's driverless car research.

Only later did Thrun learn that Android was still losing money and Rubin was close to broke at the time.

"I've always been confident I'll be able to make money. I'd rather spend it on things I'm interested in, or on people I think would be great to work with," Rubin said with a shrug, recalling it.

He was also the man directly behind Google's self-driving car project. In 2008, Rubin got Larry Page to hire Thrun, who went on to lead the project.

In a 2007 New York Times feature revealing the story of Android and Rubin, reporter John Markoff described the pile of toys in Rubin's home: drones, robots, an automated security system, most of them built or modified by Rubin himself.

His front door has an iris-scanning lock that opens only for the owner. "It works great against ex-girlfriends," Rubin joked.

The doorbell is just as elaborate. When a visitor arrives, a mechanical arm inside the house swings a hammer against a large gong.

A few years ago he helped his wife open a bakery in California called Time Traveler. He coded its point-of-sale system himself, so it could take payments and record sales with no staff involved. He also built a private meeting room in his backyard, fitted with a homemade magnetic lock.

So it was no surprise that when he stepped down as head of Android in 2013, he was named head of Google's robotics division. Nothing could have been more Andy Rubin.

Once the division was set up, Google spent $50 million to $90 million buying robotics companies, among them Schaft, which had posted the top score in the DARPA Robotics Challenge, and Boston Dynamics, which had earlier unveiled the dog-shaped robot Atlas.

Rubin talked most of these companies into joining, yet he himself left in a hurry in 2014. He could not stand a job that consisted of "sitting among a bunch of robotics experts and constantly asking whether things were done yet."

He is impatient, and always wants to build things with his own hands.

Former Apple engineer Steve Perlman described Rubin this way:

"Silicon Valley entrepreneurs today worship network effects. But Andy belongs to a generation of founders who held a soldering iron, wrote code and worked out a business, all at once."

Now, Rubin says, he is planting seeds

Rubin believes the next computing platform will be artificial intelligence. Over the next 10 years, he says, AI will grow powerful enough to perform complex tasks we can hardly imagine today, and it will soon become a cloud service powering thousands upon thousands of devices and machines.

Just as Android runs on all kinds of devices, it will one day be hard to buy a device without AI built in.

That is what his new company, PlayGround Global, is working on.

On the surface, he is investing in hardware startups. PlayGround gives them money and advice, and Rubin also lends them an all-star engineering team of former engineers from Google, General Magic and Apple.

But PlayGround Global is not tied to any one product or startup. Rubin wants to build an open platform of basic software and hardware modules.

If you are a hardware founder, Rubin wants you to stop worrying about which chip to use, which one to use when you update the product, or what can go wrong in the supply chain. He has experts to handle all of that. You only have to think about the world-changing product you want to make.

This is the reverse of the approach Samsung, Apple and Google are taking, which starts by setting Internet of Things standards. PlayGround first gives startups the tools to develop their technology, and then builds the ecosystem.

The future Rubin imagines runs on artificial intelligence, and it may be decades away, he says. For a man as impatient as he is, waiting for it feels like being stuck in traffic.

"When I have an intuition about the future, I try to figure out why it doesn't exist today."

Everything he is doing now, all the modules and all the hardware investments, is in his own words "planting seeds for the next ecosystem."

Just as with the 15-year-old who built an R2D2, these investments come partly from a long game for the future, and partly from Rubin's own passion: he simply loves hardware of every kind.

"I'll make you pancakes! The push-a-button kind!" That was how he welcomed visitors who came to tour PlayGround and interview him not long ago.

Cover image from Forbes; illustrations from Wired