Where the Idea of Amazon Web Services Came From

  WE WORKED ON Amazon Web Services (AWS) behind the scenes for a long time, then finally launched it. AWS has become a very large company by reinventing the way companies buy computation. Traditionally, if you were a company and needed computation, you would build a data center, and you’d fill that data center with servers, and you’d have to upgrade the operating systems of those servers and keep everything running, and so on. None of that added any value to what the business was doing. It was kind of a price-of-admission, undifferentiated heavy lifting.

  At Amazon we were doing just that: building data centers for ourselves. We saw it was a tremendous waste of effort between our applications engineers and our networking engineers, the ones who run the data centers, because they were having lots of meetings on all these non-value-added tasks. We said, “Look, what we can do is develop a set of hardened application program interfaces—APIs—that allow these two groups, the applications engineers and the networking engineers, to have roadmap meetings instead of these fine-grained meetings.” We wanted to build in a service-oriented architecture, where all of our services were available in hardened APIs that were well documented enough that anybody could use them.

  As soon as we hatched that plan for ourselves, it became immediately obvious that every company in the world was going to want this. What really surprised us was that thousands of developers flocked to these APIs without much promotion or fanfare from Amazon. And then a business miracle that never happens happened—the greatest piece of business luck in the history of business, so far as I know. We faced no like-minded competition for seven years. It’s unbelievable. When I launched Amazon.com in 1995, Barnes & Noble then launched Barnesandnoble.com and entered the market two years later in 1997. Two years later is very typical if you invent something new. We launched Kindle; Barnes & Noble launched Nook two years later. We launched Echo; Google launched Google Home two years later. When you pioneer, if you’re lucky, you get a two-year head start. Nobody gets a seven-year head start, and so that was unbelievable. I think that the big, established enterprise software companies did not see Amazon as a credible enterprise software company, so we had this long runway to build this incredible, feature-rich product and service that is just so far ahead, and the team doesn’t let up. This team, led by Andy Jassy, is innovating on the product side so rapidly, and they’re running everything so well. I’m very proud of them.

  Alexa, AI, and Machine Learning

  ALEXA IS THE agent in the cloud, running on the internet. Echo is the device with a multitude of microphones so it can do far-field voice recognition. From the time we started working on it in 2012, our vision was that—in the long term—it would become the Star Trek computer. You could ask it anything—ask it to do things for you, ask it to find things for you—and it would be easy to converse with in a very natural way.

  Working on Alexa and Echo was very challenging from a technical point of view. There are thousands of people working on Echo and Alexa, with teams in many different locations, including Cambridge, Massachusetts; Berlin; and Seattle.

  With Echo there were several different things that had to get solved. One of the key insights we had when we started planting that seed for Echo was an always-on device, a device that was plugged into wall power, so you didn’t need to charge it. It could sit in your bedroom or in your kitchen or in your living room and play music for you, answer questions, and ultimately even be the way you might control some of your home systems, like lighting and temperature control. Just saying, “Alexa, please turn the temperature up two degrees” or “Alexa, turn off all the lights” is a very natural way of interacting in that kind of environment. Before Echo and Alexa, the primary way people interacted with their home automation system was bad: an app on their phone. If you want to control your lights, it’s very inconvenient if you need to find your phone, take it out, open a particular app, and find the right screen to control the lights on that app.

  The devices team has just done an amazing job, and there’s so much progress still to come. We have a fantastic road map for Echo and Alexa. We have a big third-party ecosystem now of other companies who’ve built what we call skills for Alexa, so it’s kind of expanding what Alexa can do.

  We—as humanity, as a civilization, as a technological civilization—are still quite a ways away from making anything as magical and amazing as the Star Trek computer. That has been a dream for so long, kind of a science-fiction scenario. The things we’re solving with machine learning today are extraordinary, and we really are at a tipping point where the progress is accelerating. I think we’re entering a golden age of machine learning and artificial intelligence. But we’re still a long way away from being able to have machines do things the way humans do things.

  Human-like intelligence is still pretty mysterious, even to the most advanced AI researchers. If you think about how humans learn, we’re incredibly data efficient. So when we train something like Alexa to recognize natural language, we use millions of data points. And you have to collect what’s called a ground-truth database. It’s a huge, expensive effort to collect this ground-truth database that becomes the training set that Alexa learns from.

  If today you are designing and building a machine-learning system for a self-driving car, you need millions of driving miles of data for that car to learn how to drive. Humans learn incredibly efficiently. Humans do not need to drive millions of miles before we learn how to drive. We’re probably doing something called “transfer learning” in the parlance of the machine-learning field.

  Humans have already learned so many different skills, and we’re able to map those skills onto new skills in a very efficient way. The AlphaGo program that recently just beat the world Go champion played millions of games of Go. The human champion has played thousands of games of Go, not millions. And they’re almost at the same level, the human champion and the computer program. Plus, the human is doing something fundamentally different—we know because we are so power efficient.

  I don’t remember the exact figure, but AlphaGo is one example that uses thousands and thousands of watts of power. I think it’s over one thousand servers running in parallel. And Lee Se-dol, the human champion, uses about fifty watts. Somehow we’re doing these unbelievably complex calculations incredibly power efficiently—we’re data efficient and power efficient. So when it comes to AI, we in the machine-learning community have a lot to learn.

  But that’s what makes it such an exciting field. We’re solving unbelievably complex problems and not just in natural language and machine vision but also in some cases even the fusion of those two.

  Privacy organizations take claims about privacy invasion related to devices or services and attempt to reenact the invasion claims. It’s actually pretty easy for privacy organizations to do this, and they do it all the time. They reverse-engineer devices to see if their privacy claims are true. And that’s a very good behavior, and I’m grateful for all those privacy organizations that do that. And they have uncovered honest mistakes that companies have made—sometimes maybe companies just weren’t careful enough.

  Our device is not transmitting anything to the cloud until it hears the wake word, “Alexa.” And when it hears the wake word “Alexa,” the ring on the top lights up. When the ring is lit, the device is sending what you say to the cloud. It has to do that because we need access to all of the data in the cloud in order to be able to do the full range of things that Alexa can do—check the weather for you and so on.

  Hacking is one of the great issues of our age, one that as a society and a civilization we have to globally figure out. And some of the solutions will become laws. Some of it is nation–states doing things that you wouldn’t want them to do, and it’s not clear at all how that’s going to be controlled.

  With most devices and the technologies we have today, nation–states can easily listen in on any conversation by bouncing a laser beam off one of your windows in your home, or they can put a piece of malware on your phone and turn all the microphones on. A typical high-end phone today has four microphones. So we’re going to have to figure out as a society that it’s probably easier to control certain institutions like the FBI because we can come together and decide what the rules should be, what the laws should be, and how courts should enforce them. But when it comes to nation–states cyberhacking and so on, I consider it an unsolved issue. I don’t know what we’re going to do.

  I don’t know the answer to the question of whether an internet-connected society can ever be made really secure. We’ve lived with these technologies for a long time. People want to carry a phone around, and I think that the phone phenomenon is here to stay. And that phone is completely controlled by software. It has multiple microphones on it. The microphones are controlled by software. The radio in that phone can transmit the data anywhere in the world.

  And so the technical capability is there to turn any mobile phone into a listening device surreptitiously. With Alexa, the team made a very interesting and, I think, noteworthy decision. I hope other companies might emulate this decision—to include a mute button that turns the microphone off on the Echo. When you press the mute button, it and the ring turns red, and that red light is connected to the microphone with analog electronics. So it’s actually impossible, when that red light is on, for the microphones to come on. You can’t do that remotely through hacking. But phones are not like that.

  Physical Stores and Whole Foods

  WE HAVE BEEN very interested in physical stores for years, but I always said we were only interested in having a differentiated offering, something that’s not me-too, because that space—physical stores—is so well served. If we had a me-too product offering, I knew it wasn’t going to work. Our culture is much better at pioneering and inventing, and so we have to have something that’s different. And that’s what Amazon Go is. It’s completely different. The Amazon Bookstore, completely different. And we have ideas about how to merge Prime and Whole Foods to make Whole Foods a very differentiated experience.

  Amazon buys a lot of companies. Usually they’re much smaller than Whole Foods, but we buy a bunch of companies every year. When I meet with the entrepreneur who founded the company, I’m always trying to figure out one thing first and foremost: Is this person a missionary or a mercenary? The mercenaries are trying to flip their stock. The missionaries love their product or their service and love their customers, and they’re trying to build a great service. By the way, the great paradox here is that it’s usually the missionaries who make more money, and you can tell really quickly just by talking to people. Whole Foods is a missionary company, and John Mackey, the founder, is a missionary guy. And so what we’re going to be able to do is take some of our resources, some of our technological know-how, and expand the Whole Foods mission. They have a great mission, which is to bring organic, nourishing food to everybody, and we have a lot to bring to that table in terms of resources but also in terms of operational excellence and technological know-how.

  Buying the Washington Post

  I WAS NOT LOOKING for a newspaper and had no intention of buying a newspaper. I had never thought about the idea. It wasn’t like a childhood dream. My friend Don Graham, whom I’ve known for over twenty years, approached me through an intermediary and wanted to know if I would be interested in buying the Washington Post. I sent back word that I would not because I didn’t really know anything about newspapers.

  Don, however, over a series of conversations, convinced me that was unimportant because the Washington Post already had so much internal talent who understood newspapers. What they needed was somebody who had an understanding of the internet, and so that was the first thing. That’s kind of how it got started, and then I did some soul-searching. And my decision-making process was definitely intuitive and not analysis driven. The financial situation of the Washington Post at that time, in 2013, was very upside down. It was a fixed-cost business and had lost a lot of revenue over the previous five or six years, not through any fault of the people working there or of the leadership team. The paper had been managed very, very well. The problem was a secular one, not cyclical. The internet was just eroding all of the traditional advantages that local newspapers enjoyed. It’s a profound problem across local newspapers all around the country and the world. So I had to do some soul-searching, and I asked myself whether this was something I wanted to get involved in. If I was going to do it, I was going to put some heart into it and some work into it, and I decided I would only do that if I really believed it was an important institution. I said to myself, “If this were a financially upside-down salty snack food company, the answer would be no.” I started thinking about the Post as an important institution, the newspaper in the capital city of the most important country in the world. The Washington Post has an incredibly important role to play in this democracy. There’s just no doubt in my mind about that.

  And so, as soon as I had passed through that gate, I only had one more gate I had to go through before telling Don yes. I wanted to be really open with myself, to look in the mirror and think about the company and be sure I was optimistic that it could work. If it were hopeless, that would not be something I would get involved in. I looked at the Post’s situation, and I was super optimistic, but it needed to be transitioned into a national and a global publication. There’s one gift the internet brings newspapers. It destroys almost everything, but it brings one gift, and that is free global distribution. In the old days of paper newspapers, you would need to build printing plants everywhere. Your logistics operations, to have a truly global newspaper or even a really national newspaper, meant super-expensive, heavy-capital-expenditure investments. That’s why so few papers actually became national or global. But today, with the internet, you get that gift of free distribution. So we had to take advantage of that gift, and that was the basic strategy. We had to switch from a business model where we made a lot of money per reader with a relatively small number of readers to a tiny bit of money per reader with a very large number of readers, and that’s the transition we made. I’m pleased to report that the Post is profitable today. The newsroom is growing. Marty Baron, who leads the newsroom, is killing it. I think he’s the best editor in the newspaper business. We have Fred Ryan, the publisher, and Fred Hyatt on the editorial page. They’re killing it. Shailesh Prakash, our head of technology, is a superstar. So it’s working. I’m so proud of that team, and I know for a fact when I’m eighty or, let’s say—I always project myself forward to age eighty, but as I get older, I’m starting to do ninety—so I know that when I’m ninety, it’s going to be one of the things I’m most proud of, that I took on the Washington Post and helped it through a very rough transition.

  If you’re the president of the United States or a governor of a state, for example, you don’t take that job thinking you’re not going to get scrutinized. You’re going to get scrutinized, and it’s healthy. The president should say, “This is right. This is good. I’m glad I’m being scrutinized.” That would be so secure and confident. But it’s really dangerous to demonize the media. It’s dangerous to call the media lowlifes. It’s dangerous to say they’re the enemy of the people. We live in a society where it’s not just the laws of the land that protect us—we do have freedom of the press; it’s in the Constitution—but it’s also the social norms that protect us. It works because we believe those words on that piece of paper, and every time you attack the Constitution, you’re eroding it a little bit around the edges. We are so robust in this country. The media is going to be fine. We’re going to push through this. By the way, Marty Baron will always make a super-important point when he meets with the newsroom: “The administration may be at war with us. We are not at war with the administration. Just do the work. Just do the work.” I’ve heard him say it many times. I say it myself when I meet with journalists at the Washington Post.

  Trust

  THE WAY YOU earn trust, the way you develop a reputation is by doing hard things well over and over and over. The reason, for example, that the US military, in all polls, has such high credibility and reputation is because, over and over again, decade after decade, it has done hard things well.

  It really is that simple. It’s also that complicated. It’s not easy to do hard things well, but that’s how you earn trust. And trust, of course, is an overloaded word. It means so many different things. It’s integrity, but it’s also competence. It’s doing what you said you were going to do—and delivering. And so we deliver billions of packages every year; we say we’re going to do that, and then we actually do it. And it’s also taking controversial stances. People like it when you say, “No, we’re not going to do it that way. I know you want us to do it that way, but we’re not going to.” And even if they disagree, they might say, “We kind of respect that, though. They know who they are.”

  It is also helpful to have clarity. If we are clear that we are going to do this and we aren’t going to do that, then people can opt in or opt out. They can say, “Well, if that’s Amazon’s position or Blue Origin’s position or AWS’s position on something, then I don’t want to be part of that.” And that’s okay. We live in a big democracy with lots of opinions, and I want to live in that world. I want to live in a place where people can disagree. What I want, too, is to live in a place where people can disagree and still work together. I don’t want to lose that. People are entitled to their opinions, but it is the job of a senior leadership team to say no.