Inside the mind of the man behind Google's most important products
If you use any Google product, chances are Google’s SVP Sundar Pichai has had some say in how it was created. Pichai joined Google in 2004, and over the course of just a few years, Pichai has extended his domain beyond Chrome to encompass Google Drive, Maps, apps, Chrome OS, Android, Search, Ads, ATAP, and more. Pichai’s expansive role was formalized back in October of 2014, when CEO Larry Page handed him the day-to-day responsibilities of running the tech giant.
We sat down with Pichai to hear his vision for the Google of the future. He laid out a plan to improve Google’s products through machine learning — but more importantly, he sketched out a grand effort to deliver computing capabilities to billions of people around the world. Both in the way he manages his internal teams, and in his belief that technology can change people’s lives for the better, Pichai advocates an egalitarian ethos.
After Page and co-founder Sergey Brin, Pichai is almost certainly the most powerful person at Google. Executives in Pichai's position tend to lead their teams with vision, charisma, bombast, or (usually) some combination of all three. But there's no bombast to Pichai. His office is clean to the point of being spartan, and that simplicity is reflected in his demeanor. Pichai is thoughtful and friendly in person, nodding carefully as he listens and then responding with real empathy. After the first day of I/O, I watched as Pichai walked into the hungry press gaggle, giving everybody who approached him equal attention. Amidst the tumult, he even sought me out to finish a conversation we’d had the previous week.
Outside the media scrum, Pichai is slightly more relaxed, slightly more willing to dive into technical detail. When we briefly discuss the new Internet of Things communication standard that was announced this week, he immediately grasps my (low) level of technical understanding and comes right down to it gracefully and without a hint of condescension.
Pichai’s approach toward other people is reflected in his attitude toward what technology itself should do. With his larger role, that attitude is rapidly infusing Google’s new products. Now on Tap, organized photos, virtual reality videos stitched together in the cloud, and even Project Ara's first foray into the consumer market in Puerto Rico are all happening in large part thanks to Pichai's goal of making Google's technology available to as many people as possible. Under Pichai, both Google's product vision and its out-there R&D projects all seem to be heading in the same direction much more than before.
So when Pichai talks about the next billion people about to come online with smartphones, I get the impression that, for him, Google’s monetization strategy really is secondary to Pichai's stated goal: giving people everywhere the power of Google's machine learning whenever and wherever they need it. He's clearly proud of the fact that Google's products work the same whether you're a billionaire or a rural farmer in a far-flung place. And Pichai's vision is to ensure that dedication becomes a part of everything Google makes.
Coming from a different company, the claims Pichai and his product managers make would sound like clichés: a supercomputer in your pocket, a new way to systematize innovation, empowering the next billion. But if anyone in Silicon Valley has the power to transform moonshot fantasies into powerful realities, it’s Google — Sundar’s Google.
Dieter Bohn: This is the first I/O where you were officially in charge of everything being presented. Did that change your approach?
Sundar Pichai: It’s a good question. Historically we’ve used I/O as an opportunity to take a step back and help people get a sense across what Google is doing...and tended to focus more on our platforms. We’ll continue to do that this year, but maybe it’s a bit easier to do it end to end.
I think it’s been a while [since] we stepped back and told the story, not just from a platform perspective but also from a Google standpoint — what we hope to accomplish, and also focus on developers.
Back in October there was a little bit of reorganization, and you were able to take charge of more, or most, of Google’s core products. Are there specific priorities you’re better equipped to drive now?
In some ways, yes. We have these large services which users use and they work at scale. We have ongoing platforms we’re evolving. But if you want to think ahead and plan for the next two to three years as to where online computing platforms are evolving and how do you drive the experience on top of it, being able to think holistically, road map it out, and think it through — I do think that’s been unique.
For example, we’ve been thinking hard about how we help organize users' information on mobile. It’s core to our mission of what we set out to do. [With] those kinds of things, it helps us to step back, think about it, and do it better.
How are you able to do it better now?
At a high level I think of what we are doing in terms of two core things. One is what we set out and wrote in the Founders’ Letter, which is to say we really want to work on big problems that help solve big problems in users’ lives. We want to do it at scale, we want to do it for everyone. If you look at search, it worked that way. Maps, YouTube, Chrome, Android [all did too]. So, we want to continue doing that. How do you identify problems, solve [them] at scale and bring technology to bear?
The second is our core mission statement, which is to organize users’ information. We’ve seen this with things like Google Now, where we try hard to tell you, "Okay, maybe you want to leave earlier to go to a particular meeting" or when you land and reach an airport line we say, "Here’s your boarding pass." When we think about organizing the world’s information in the context of mobile, people are trying to get stuff done. So we need to go a step further and be assistive where we can.
Underlying all of that is we’ve invested a lot in machine learning. Today, if you open a phone and you say, "Hey, what does a tree frog look like?," we are understanding voice, going through all images and trying to understand how a tree frog is different from other types of frogs, salamanders, et cetera. Sometimes [that information] needs to be translated. All these are huge advances in computer science, and are primarily driven by machine learning. Just in the last three years, we have taken things like error in word recognition down from about 23 percent to 8 percent. And it’s because of what we call deep neural nets. We [have] made great strides.
[But] how do you use all this technology and [these] advances to be more assistive to users? So, you will see us talk about how we can take Google Now and make it work better in the context of your phone. If you’re looking for information in the context of using applications, how can we do that? To me, doing things like that are core to what Google promised users it will do, so that’s the second big part of what we do.
I’ve been talking to a bunch of people in Google over the past couple of weeks, and a phrase I’m hearing over and over again is "the next billion." There’s a billion people using Android — how do we get the next billion? Is there a coherent strategy behind that goal?
The entire PC industry reached about 1.7 billion people. [But with mobile,] we are truly dealing with the first computing platform [that] is going to touch people at scale. I can see a clear path to getting over 5 billion users one day. I think just in the last 12 months there [have been] over 600 million users who for the first time have had access to a modern computing platform. A smartphone, effectively. So we deeply care about it.
At our core, we want to build products for everyone. At a basic level, at a foundational level, [that means] both providing computing and making it accessible, which is why we are interested in Android One [and] affordable Chromebooks. And thinking about connectivity over time, [that] is why we are doing things like Project Loon.
The second is from our services standpoint: we realize for those users, the experience needs to evolve. So we are thoughtfully, across our products, thinking about what it means for a user in India or Brazil or Indonesia or Africa with limited connectivity to use products like YouTube, Google Search, Google Maps. How do we make their experience better? That’s a big part of it.
I’m sure you’ve seen there’s been a bunch of debate about Internet.org and net neutrality. Does Google, or do you personally, have a stance on whether it’s important to try and have some net neutrality standard in emerging markets?
Net neutrality has served us well, so we are pretty committed to it. I’m not fully familiar with all the issues around Internet.org, but … it’s pretty obvious to me that users in these markets want the same as users in the developed world. They want data to be plentiful and affordable, right?
I think we have a model which works well. Which is why I am a big fan of more foundational efforts. A lot of carrier partners are investing [in] a lot of efforts. To me, this is not a physics problem … We know how to provide connectivity at scale, so how we get that done in developing markets has got to be the test of the effort.
You get asked this question frequently, but I’m going to ask you again: Android seems like it’s getting more and more popular in China, but Google services aren’t. Is your philosophy still "we hope to get there but we’re going to wait and see"?
We are going to continue providing Android to the fullest extent possible in China, so we are incredibly excited. The market [there] is a substantial market; it’s unique. So we deeply care about investing there. We would love to serve Chinese users with Google services as well, obviously. I think it will be a privilege to do that, but we need to be thoughtful in how we do it. We are open to newer approaches. We’ll have to wait and see.
In addition to providing information and its core products, Google also does a bunch of really crazy shit: Project Loon, Calico — all these moonshots, large and small. A lot of this stuff is really interesting, but it’s hard to see a path for it to be anything more than cool experiments. Should these divisions keep doing crazy stuff or should they be pushed to become productized, core parts of Google?
I would draw a distinction between things like research and ATAP, which are research efforts with different flavors of how we approach research, versus areas like our core products, self-driving cars, or Project Loon or even our Iris contact lens. Can we apply technology and bring a differentiated and a unique way of solving a problem? Does it apply to millions or billions of people? Those efforts are done under that framework, and while the effort may seem ambitious or crazy, we take a very disciplined approach.
Those [projects] are thought through like businesses. It’s just that we are willing to take a long-term view. But we run them in a very disciplined way. We have a good notion of, "Okay, here’s how much investment it will take, here’s how long it will take, what are the use cases and applications, and how will this whole thing play out?"
Where we see traction, we will double down. Otherwise, we will course-correct. We will do both. We need to be thoughtful about these things. We do them because we believe that software is at a stage where [it is] increasingly playing a more and more critical role in solving [problems] it didn’t before.
You’ve just announced Project Fi, and you’re also announcing Android Pay. Both of these feel like departures from the early days of programs like Nexus, when it seemed like there was an attitude of going at it alone. With Pay and Fi, you’re partnering with carriers. Is there a shift in Google's attitude toward working with other companies?
By the way, the same is true for Loon. We are doing it with Vodafone, Telefonica, and Telstra. Our goal is to reach everyone.
Out of [that goal] arises a lot of what we do. Because we have decided we want to solve it for everyone, inherently that forced us to think hard about a partnership approach. That’s what helps us reach everyone with Android. We do work with 400 OEMs, [and] over 550 carriers to make 4,000 distinct devices.
We shipped over a billion Android phones in 2014. When I look at the scale of how it happens, all the way from the low end the high end, that matters to us.
[Search, Chrome, Maps, and YouTube] are all products available for everyone. For something like [Project Loon] to work across the world, you will inevitably end up working with carrier partners. For self-driving cars to have an impact, you will see us take a partnership approach there as well.
"We shipped over a billion Android phones in 2014."
What do you think of Microsoft’s effort to get Android running on Windows mobile? Are you guys working with them on that? Do you have a dog in that hunt?
My understanding is based on what I read — probably [from the] Verge summary of the Build Conference. Today, there are many efforts underway like that, right? Others have efforts, Amazon does it for their products, so it’s not new to us. It’s a testament to the scale of the app ecosystem on top of Android. A lot of what we do in Android is open source, and things are designed for it to be used easily elsewhere, so it’s good to see all that.
If I bring up Microsoft, I have to bring up Apple. Apple has entered a slew of new markets: they are rumored to be trying to get on TV, they’re rumored to do music streaming, they just released a watch, they have Apple Pay. They’re directly competing with you in lots of areas: you’ve got a music streaming service, you’ve got a TV product, you’ve had Wallet for a while. What unique advantage does Google have over Apple across all of these different product categories?
Look, Apple and [Google], we [get] talked about together, [but] we two are trying to do different things. To users, it arrives as a choice of an iPhone or an Android smartphone.
But [Apple is] thinking through end-to-end. They are vertically integrated, and they think about an end-to-end product. We are trying to build a horizontal platform and enable many people. So it’s a bit tough to compare and contrast, right? Having said that, we do approach things a bit differently. We generally think about whatever we build [as] cross platform, trying to reach everyone. We think deeply about them being cloud services and leveraging the power of the hard work we have done over the years to build several large cloud services which are building users at scale.
Those are the things we tend to go deeper on. It’s good for users that there are companies who are approaching it differently. It gives rise to choice and over time I would say healthy innovation, which pushes all of us further along. Apple is doing really well, [and that] forces us to work harder. Hopefully it’s a virtuous cycle.
Is there anything personally important to you that you wish Google as a whole were doing more of?
The thing that attracted me to Google and to [the] internet in general is that it’s a great equalizer. I’ve always been struck by the fact that Google search worked the same as long as you had access to a computer with connectivity, [whether] you were a rural kid anywhere or a professor at Stanford or Harvard.
I want Google to strive to do that — not just build technology for certain segments. For me, it matters that we drive technology as an equalizing force, as an enabler for everyone around the world.
Which is why I do want Google to see, push, and invest more in making sure computing is more accessible, connectivity is more accessible. And going back to our core mission, when we do things like machine learning and assist users, I view that as a huge game changer. Because over time, someone who has [access to] just a smartphone hopefully has...the same [capabilities] as someone who is more privileged. That’s what’s very exciting about what we are doing.
Photography by Cara Robbins