Pod Insights Decoder Recode Decode: Bruce Schneier About this episode Security researcher Bruce Schneier talks with Recode's Kara Swisher about his recent book, Click Here to Kill Everybody. He also explains why the internet of things is a “dumpster fire," what regulations need to be implemented to keep people safe, and why the European Union and a few US states may determine the future of tech regulation.
Featuring:
Bruce Schneier (@schneierblog), author, Click Here to Kill Everybody: Security and Survival in a Hyper-connected World
Host:
Kara Swisher (@karaswisher), Recode co-founder and editor-at-large
More to explore:
Subscribe for free to Reset, Recode's new podcast that explores why — and how — tech is changing everything.
About Recode by Vox:
Recode by Vox helps you understand how tech is changing the world — and changing us.
Follow Us:
Newsletter: Recode Daily
Twitter: @Recode and @voxdotcom
Learn more about your ad choices. Visit podcastchoices.com/adchoices
Transcript 11,940 words · 7 speakers
0:00 Speaker 00 You're busy. We know that, so we'll keep this quick. Introducing AMEX Corporate, now featuring the Corporate Cash Back Card. For one annual flat fee, get 1.5% cash back on eligible business purchases, plus intelligent expense management with the unmatched expertise and backing of American Express. Built to simplify, designed to scale. Get started at go.amex/ammexcorporate. Terms and cap apply.
0:33 Speaker 01 You could try to find the perfect coffee to water ratio, or you could press a button. With Keurig, each K-Cup pod is precision-measured for a perfectly delicious cup. No effort, no grind, no math required.
0:50 Speaker 01 Great coffee without the grind. Keurig.
0:56 Speaker 02 Hi, I'm Kara Swisher, editor-at-large of Recode. You may know me as someone who had decided the internet is just too unsafe, which is why all future episodes of this podcast will only be available on vinyl and 8-track. But in my spare time, I'm just a reporter and you're listening to Recode Decode, a podcast about power change and the people you need to know around the tech and media industries who are part of the Vox Media Podcast Network. Today in the red chair is Bruce Schneier, the well-known security researcher. I'm a big fan of him. He's a cryptographer and a writer. He's also a fellow at the Berkman Center for Internet and Society at Harvard Law School. He's also the author of several books, including Applied Cryptography and Liars and Outliers, and his most recent book, which is my favorite title of any book ever, Click Here to Kill Everybody: Security and Survival in a Hyper-Connected World.
1:41 Speaker 02 Bruce, welcome to Recode Decode. Thanks for having me. So I'm a huge fan of yours. This is an area I want to talk about a lot, because I'm going to go into your background in a second, but talk about what you're doing, public interest.
1:52 Speaker 03 Technologist.
1:53 Speaker 02 Technologist. Explain what that is. I want people to understand how important this is.
1:57 Speaker 03 So, we're in a world where technology is fundamentally affecting society.
2:04 Speaker 02 Yes.
2:04 Speaker 03 We can talk about democracy, we can talk about the market economy, we can talk about all sorts of social systems. And the people doing policy don't understand the technology. We had a senator, when Mark Zuckerberg was there for a hearing last summer, who asked, "How does Facebook make money?" Which is, you know, crazy on two levels. One, he didn't know. And two, he didn't realize it was a stupid question. And we're expecting these people to make policy that governs tech, that governs us. And now there's this increasing knowledge that we need people to understand both tech and policy.
2:44 Speaker 02 Right.
2:44 Speaker 03 Policymakers who speak tech, techies who speak policy, people who speak both. And technologists need to get involved in public policy, or lousy public policy is going to be made.
2:56 Speaker 02 Or none at all.
2:57 Speaker 03 Or none at all. Now, I don't mean techies should take over public policy, I think that'll be another disaster. But we need this coming together of these two worlds, these two ways of thinking, these two ways of regulating, of governing, if we're going to solve the problems of the first half of this century. And I think of that as public interest technologists, trying to get people, from my world, to either serve in government, serve on the press, doing tech-driven journalism, work at NGOs, work for the government, maybe for the military or law enforcement, that doing tech in a more public-focused environment. It's not going to make the big bucks, it's not going to have the big glory, but it's how we're going to change society.
3:42 Speaker 03 It's not going to be through the next app.
3:43 Speaker 02 All right, so we're going to talk about how to get there. So you call yourself a public interest technologist, which is something, you've had a long history in lots of things, including cybersecurity, all kinds of things. So let's talk a little bit about your background so people understand what kind of role you've played. Go through sort of how you got to do, because you've been, you've worked at IBM, you've sold several companies, you've done a lot of stuff. The typical technologist's, the journey that many technologists like yourself have been on.
4:10 Speaker 03 Yeah, although I'm really non-standard. I think of my career as an endless series of generalizations. You mentioned my first book, Applied Cryptography.
4:17 Speaker 02 Right.
4:18 Speaker 03 Which was written in the early 90s, and it's basically a math and programming book.
4:22 Speaker 02 Right.
4:23 Speaker 03 And it caught the wave of the public internet and the need for security on the internet. And after that, I write about computer security, network security, the general technology of security.
4:36 Speaker 02 Or lack thereof.
4:37 Speaker 03 I've written about the economics, psychology of security, the sociology of security. You mentioned a book of mine, Liars and Outliers, about trust in society. And really now, the political science and policy of security. And here I am, teaching at the Harvard Kennedy School. This is a public policy school, and I'm teaching internet security policy. So here I am, building out from the tech to society. Along the way, I did, I had two companies. One got sold to BT, one got sold to IBM, so I got to be in a large corporate environment for a few years until we all got tired of each other. I've done a lot of consulting, I do a lot of writing, I do a lot of speaking.
5:19 Speaker 03 Really what I'm good at is explaining tech to non-techs. And that term, public interest technologist, isn't mine. I think it came out of the Ford Foundation, trying to look at this bridge between tech and policymakers. I am not unique, I am one of a few, and there are... A bunch of us that do tech and policy. What there really isn't is this career path.
5:42 Speaker 02 Right.
5:43 Speaker 03 Where there's a sort of a well-worn path that you can go to university and learn this and then get internships and then get jobs and increasingly.
5:52 Speaker 02 Because technology are often the secondary people in the room. Let's bring them in afterwards or before policy. But first I want to get to your, you got interested in cryptography how? What was your path to that? Why did that interest you over a lot of things?
6:04 Speaker 03 Cryptography was math and I was this math kid. And you know, it's kind of ironic that I majored in physics in college, which is really mathematics. And I'm doing these things because I don't actually want to deal with people. And then here I am throughout my career, becoming more people, right? Economics, psychology, sociology, policy. I'm really in the people's side of technology, where I started out very much as the anti-people side of technology.
6:31 Speaker 02 Right.
6:32 Speaker 03 And I think there's something we've all learned, at least I think the better technologists have, that technology doesn't stop at the keyboard and screen, that it goes to the person and to the people behind them, that we are affecting society. And we're not just building programs or building apps. We're building much more complex, what we call socio-technical systems.
6:53 Speaker 02 Right.
6:54 Speaker 03 And not understanding the sociology of it leads to some pretty big disasters.
6:58 Speaker 02 Well, let's talk about how that happens. And you act on your own career. You were, you studied, the way you studied was typical of the way others studied or came up through, which is that you make technology and the implications of it don't matter.
7:09 Speaker 03 I think that's right. I think the idea is we're building tools. In the beginning, I think we were. We're building a word processor, we're building a drawing app. And what's written and what's drawn is not our purview. And if you think about the way we would write user interfaces, we would write commands, there'd be users manual, and it'd be your job to read the manual, it'd be your job to learn the programming language. That there wasn't this real thought given to how it was used, or even the users were sort of laughed at. The user can't figure it out, aren't they dumb? And the technologists felt that they were in charge. They're building the future.
7:47 Speaker 02 Right.
7:50 Speaker 03 It's common in technology to not understand how your technology is being used and more than not care, to be really thrilled if someone finds a use for your technology you haven't envisioned. We built this tool and look at all the cool things it's doing. We never thought of that. That's the opposite of policy.
8:08 Speaker 02 So why is that? Why is that in technology training? When you think about that?
8:11 Speaker 03 I think it's a way of thinking. I think it's the mindset that we're building the tool and then we give the tool to the world.
8:19 Speaker 02 Here's a hammer.
8:20 Speaker 03 Here's a hammer. And what you do is you're going to build a cathedral. Isn't that amazing? I don't even know what a cathedral is. You're going to build a cathedral. On the other hand, you know, you might build a tank, an artillery piece. You know, so here we are, I mean, today, we're making facial recognition technology and you can use that to find relatives you didn't know existed. It also could be used to identify protesters in Hong Kong. So here we are, building these technologies that inherently have these wildly different uses. And I mean, think about, thinking back to almost the nuclear age, the notion that, you know, I'm building the missiles, where they come down is not my department, right? That's a very famous Harvard professor and singer who said that.
9:04 Speaker 03 And that is the way of thinking. I like to think that we're beyond that, but in a lot of ways, we're not.
9:10 Speaker 02 So when you started your career, you were focused on cybersecurity. Over the way these networks were created were badly made, like badly architected essentially in terms of how safe people were. Because that was some of the early stuff you talked about, the idea of it.
9:25 Speaker 03 I mean, cryptography was the first thing I was working on, but very early on, cryptography is intertwined with privacy.
9:31 Speaker 02 Right.
9:32 Speaker 03 And here we are in the mid 90s, EFF, Epic are starting up.
9:36 Speaker 02 Right, which you're involved with.
9:37 Speaker 03 Right now, I'm a board member of EFF and I've been involved with Epic since the beginning as well. So these are tools that are viewed as tools for privacy, tools for democracy, tools for giving individuals more power than they had over their governments, over other powerful entities. So we very quickly, in security, saw the politics of what we were doing. I think sooner than a lot of other areas of computer science.
10:06 Speaker 02 Explain that. How so?
10:08 Speaker 03 Because it was, you can't Divorce security from who is security is from and who it's for.
10:14 Speaker 02 Mm-hmm.
10:15 Speaker 03 In a way, you can sort of divorce a word processor. It does italics, you know.
10:18 Speaker 02 Right.
10:19 Speaker 03 We don't have to think anymore. When you think about privacy and encrypted telephone and encrypted message app, very quickly, who is using it? How are they using it? And whether it works or not is more than, do you see italics on the screen? Whether it works or not is, did the person get privacy? Did he get security?
10:40 Speaker 02 Right.
10:41 Speaker 03 Is he being arrested or not?
10:43 Speaker 02 Mm-hmm.
10:43 Speaker 03 So, the real world effects are much more visceral.
10:46 Speaker 02 Mm.
10:46 Speaker 03 And I think that gave us a preview into what we're seeing today with so many other technologies, AI and this automatic decision-making, the things we go look now and say, "Man, you people are breaking society. You should watch out."
11:01 Speaker 02 Mm-hmm.
11:01 Speaker 03 And we sort of knew that early on.
11:02 Speaker 02 Right. And so, when one of the things that you talked about, I want to start with a bunch of these topics of facial recognition, cybersecurity. Talk a little bit about where we are with cybersecurity right now. How do you, how do you look, how people are looking at it? Obviously, there's been incursions everywhere. Google just announced, a big cybersecurity person called me and said, "After this Google thing, there will be no cybersecurity because of these computers they've created, and we have to have quantum cybersecurity."
11:28 Speaker 03 And then...
11:29 Speaker 02 Talk to me about that. I love, I don't even know.
11:31 Speaker 03 Okay.
11:32 Speaker 02 I don't know. I don't know.
11:33 Speaker 03 A lot of buzzwords floating around here.
11:34 Speaker 02 Yes, I got that. Yeah, I got. So, let's talk first about cybersecurity. I want to go talk about facial recognition, the whole bunch of them right now.
11:40 Speaker 03 So, we tend to call internet security, cybersecurity is the military term.
11:45 Speaker 02 Mm-hmm.
11:46 Speaker 03 And I think it's coming into our world.
11:48 Speaker 02 It is.
11:48 Speaker 03 Politically, we've called it internet security.
11:50 Speaker 02 Right.
11:51 Speaker 03 And that is security of your computers, of your network, of your data in the cloud. It's a much more personal way of thinking about it. Cybersecurity is more of a nation-state way of thinking about it. And I tend to look from the person out. How are we doing? We're not doing great, but we're not doing bad either.
12:07 Speaker 02 Mm-hmm.
12:08 Speaker 03 And most of us have our data in the cloud.
12:11 Speaker 02 Right.
12:11 Speaker 03 And, you know, honestly, Google's going to do a better job keeping your email safe than, than you would. So, that's, that's a win for most of us. We are being spied on by pretty much every corporation out there. Surveillance, capitalism is the business model of the internet.
12:25 Speaker 02 Shana Zuboff talks about it.
12:26 Speaker 03 Right? And, you know, that means these systems are designed for surveillance.
12:31 Speaker 02 Right.
12:31 Speaker 03 There's a... A phrase, a joke you can think of, that everyone wants you to have security except from them. Right. And that's true. I mean, Google will give you a lot of security, as long as they can read everything. And everybody, Apple's the same way. Everybody's, everybody's the same way. There are breaches everywhere. We hear a lot of them in the news, a lot of them just don't make the news because they're just another breach. So our data is not in our hands anymore. Now, we didn't do a terribly great job with the "in our hands," but if you think about it, your photos are on Flickr, your emails on, on Google. And if there is a data breach at Marriott Hotels, your personal data, including your credit card number, is stolen, and you had nothing to do with that. That was Marriott Hotels' security.
13:13 Speaker 03 So there's a lot of, of data out there being breached. On the whole, we're not doing great. We really aren't designing secure computers. We aren't designing secure phones. The Internet of Things is gonna be a dumpster fire because things are much cheaper. And those are also sensors about us. There are cameras, there are, even refrigerators are, are learning things about, are habits. And these are all on the Internet, and the data's all vulnerable. On the other hand, we're not seeing these cyber apocalypse disasters that have been talked about. What would that look like? You know, I, I tend not to like to talk about scare scenarios because they're just scary, but you can think of large-scale data breaches.
13:57 Speaker 03 We hear about our power grid being insecure. Sure. And there hasn't been any real disasters. We've seen a couple of attacks by Russia against the Ukraine, but, but nothing else. We haven't seen cyberterrorism. I mean, we're really seeing cybercrime. We see- Stealing pennies. Well, data theft, but also data hijacking, the ransomware, against lots of cities, lots of organizations. We often see it when it's a city or a hospital. We don't see it when it's a company. That's because they don't often make the news. Right? That is sort of burgeoning area of crime. So there's a lot of cybercrime out there. It's big business, very international. Very hard to prosecute because it's so international.
14:41 Speaker 03 There's a story a couple of weeks ago about ransomware. A couple of individuals in the Ukraine.
14:47 Speaker 02 Explain ransomware.
14:47 Speaker 03 So ransomware is when someone encrypts your computer and charges you to get your data back. It's actually very clever. If you think about your data, your photos, your email, the person who it's valuable to is you.
15:01 Speaker 02 Right.
15:02 Speaker 03 And actually, nobody else. So it makes no sense to steal it. What it makes sense for is to hide it from you and charge you to get it back. And if someone encrypts your photos and they're important to your life, you'll probably pay a few hundred dollars to get them back. And there's the crime.
15:19 Speaker 02 Right.
15:20 Speaker 03 So there's a couple of ransomware perpetrators in the Ukraine, and they're randomly attacking systems. They happened to lock up the Washington DC camera system a week before the last inauguration. On the one hand, wow, that was impressive. On the other hand, super bad luck because now the world is.
15:41 Speaker 02 Paying attention.
15:42 Speaker 03 Is paying attention. And they got caught. They never would have gotten caught otherwise because they would have been going all these sort of low-interest networks. So that's just an example. Now, I think things are changing. My last book is about this, that a lot of what we're talking about is data.
15:57 Speaker 02 Right.
15:58 Speaker 03 When we start talking about things that actually hurt people, cars and medical devices, thermostats, I live in Minnesota, so I'm gonna take over my thermostat in the winter. My pipes can freeze, I can have a lot of property damage, that that's gonna change the security landscape because it'll be no longer about data, it'll be about life and property.
16:19 Speaker 02 Right.
16:20 Speaker 03 And I think changes are coming, regulatory changes, because of that.
16:23 Speaker 02 That people can take over a thermostat or take over a car or that kind of stuff.
16:27 Speaker 03 That cybersecurity doesn't mean your spreadsheet crashes and you lose your data. It means your embedded heart monitor crashes and you lose your life.
16:34 Speaker 02 Right, right.
16:35 Speaker 03 That change.
16:36 Speaker 02 I'm literally just talking to someone who's like, someday, Google will run your pancreas and then you don't pay them and then they say, we're turning off your pancreas until you pay us, kind of stuff. It was fascinating. I was like, oh, I never thought of that.
16:45 Speaker 03 And that's actually not unreasonable.
16:47 Speaker 02 I know that.
16:47 Speaker 03 So, in addition to all these criminals, and we talk about nation states, I worry a lot about the corporate power.
16:54 Speaker 02 Right.
16:54 Speaker 03 That there's a lot of corporate power over our data and over our ability to do things. We're in Massachusetts now. There's a right to repair bill that is going to be on the next election. It'll be voted for as a...
17:09 Speaker 02 Explain that, right to repair.
17:11 Speaker 03 So, the notion is that we have a right to repair our stuff. And if you have a car and it breaks, you can open up the hood and take a wrench and do car things. I'm not a car guy, but presumably people know how to do that stuff. As these things become computerized, there are license agreements about the software...
17:26 Speaker 02 That you're not allowed.
17:27 Speaker 03 That you're not allowed to repair your stuff. And that seems counter to the way society works. And different states are trying to enact bills to give people the right to repair, to prohibit companies from denying you the right to repair your thing because of license agreements. John Deere is in a big battle with farmers because farmers are used to repairing their tractors. That's one of the things they do.
17:53 Speaker 02 Right, and they're highly computerized.
17:54 Speaker 03 And now they're... Not that they don't have the skill to, they don't have the legal ability to. That the license agreement from John Deere says you're not allowed to look at the software and repair. And there actually is black market tractor software available on the internet for people who want to do this. And we want to make this legal. There's only two bills in Massachusetts, one general and one about automobiles specifically.
18:18 Speaker 02 Right, to be able to do this.
18:19 Speaker 03 To get people right to repair.
18:20 Speaker 02 Now, we'll get back on track, really, and then we're going to get to facial recognition and the other... in the next part. But distinguish you from cyber... from internet security to cybersecurity. This is the military. Because you're here in Harvard.
18:32 Speaker 03 They're both basically the same. There are different ways of looking at it. Cybersecurity was always a military term and a little bit derided in the internet security community because we are thinking more about people and communities. And cybersecurity is more about national security and governments and nation-state actors. They've kind of merged at this point.
18:53 Speaker 02 They have.
18:54 Speaker 03 I mean, everyone is using the same tools, the same techniques, the same tactics. And, you know, in some ways, it's surprising. And remember 2014, October, the attack against Sony?
19:05 Speaker 02 Mm-hmm.
19:05 Speaker 03 In my community, there's a serious debate about whether this attack is perpetrated by nuclear power with a $20 billion military budget, or a couple of guys in a basement somewhere.
19:16 Speaker 02 Mm-hmm.
19:17 Speaker 03 And that, in a sense, is extraordinary, that you can't tell the difference.
19:20 Speaker 02 Whether it's the North Koreans or just some people.
19:22 Speaker 03 Or just random people.
19:24 Speaker 02 Mm-hmm.
19:25 Speaker 03 And that's true. A lot of these attacks we see, we don't know who does it. We can get good at attributing when we pay attention. The United States has, in the past few years, attributed attacks to Russian actors, to Chinese actors, North Korean actors, to Iranian actors. But sometimes, it's not known who did it. I mentioned the Marriott hack. We're pretty sure that was the Chinese. It's not possible. It could have been criminals.
19:53 Speaker 02 Right.
19:54 Speaker 03 Office Personnel Management, we know, is the Chinese. The DNC servers, we know, is the Russians. A lot of stuff we know. But there's a democratization of tactics that you don't see in conventional warfare.
20:07 Speaker 02 Right. All right. We're going to talk about that when we get back, because right now, there's just, this was just in the news yesterday about this exact thing. We're here with Bruce Schneier. His most recent book is called Click Here to Kill Everybody: Security and Survival in a Hyperconnected World. We're going to take a quick break now. We'll be back after this.
21:19 Speaker 02 We're here with Bruce Schneier. His most recent book is called Click Here to Kill Everybody. Why did you call it that? Like, talk about that. Like, I love that name.
21:28 Speaker 03 I think it's the greatest title ever.
21:30 Speaker 02 It is the greatest title ever.
21:32 Speaker 03 So, right, I like-
21:33 Speaker 02 Except for Steal This Book, but go ahead.
21:35 Speaker 03 Steal This Book is a great title, that's true. I like evocative titles. I mean, I think of books in sort of a waterfall model. A title gets you to read the subtitle. Subtitle gets you to read the inside jacket copy. And that gets you to read the book. So I want a title that makes you stop and say, "What is that about?" But I'm trying to speak to something real. The notion of class breaks being brought to the real world. And this has to do with the way things fail. Think of cars. Cars have parts, parts break, and we know how cars fail. In all of our communities, there are going to be these auto repair shops. Because there's a steady stream of broken cars. Computers do not fail that way. Computers all work perfectly until one day when none of them do.
22:19 Speaker 03 And that is a notion of a class break. That suddenly, every copy of Windows is insecure. That all of your iPhones have a vulnerability.
22:27 Speaker 02 Right, that's an excellent point.
22:29 Speaker 03 And these are now things that affect the world in a direct, physical manner. So I'm worried about crashing all the cars. Or more realistically, all the cars of one make and one model a year. That we can have these large-scale computer failures that affect the real world.
22:50 Speaker 02 As opposed to, say, some screw on a tire coming off.
22:53 Speaker 03 Right, which would just be one tire. Or maybe the cars have to be recalled because now there's a chance the screw will go loose. And we know that over the course of the year, you're going to make this up. Every four or five days, one of the cars will have a screw loose and somebody will crash. Not the same as, "All screws are now loose and get those cars off the road in the next 10 minutes, hurry."
23:13 Speaker 02 Well, part of the appeal of it is that they all do work together. I mean, on the flip side of that, someone was at Google, just recently telling me, "Look, when someone gets in a car," they were talking about self-driving cars, "when someone gets in a car accident, only that person learns, if they learn at all. When a self-driving car gets, or autonomous car gets in an accident, all the cars learn."
23:33 Speaker 03 And that's true. And I, you know, I work in security, but I understand the upside of all this technology. I mean, these are great technologies in the main, and they do an enormous amount of good to the world, but I tend to look at the downside. An example might be Omedy locks. So hotels nowadays, they often have keyless entry systems. You get a key card, you wave it in front of a reader. One of the companies that makes those locks is Omedy. And a few years ago, a vulnerability was found in Omedy locks, such that if you didn't have the right key card, you could still break in. And the way you had to repair that was to go to every hotel room door manually and change the ROMs and change the code in the lock.
24:16 Speaker 03 And hotels, by and large, never did it, because their way, their mental model of way hotel locks fail is, you have a locksmith on staff, every few weeks, someone can get into their room, and you call the locksmith. Not that we have 400 hotel rooms and they all have to be upgraded. They're just not used to that failure. And that's that difference. That's the notion of a class break.
24:39 Speaker 02 So when you're thinking about the idea of class breaks, it iterates everywhere in terms of what you're using, whether it's going to be sensors, or cameras in a home, or whatever, to all these products. How do you think of that differently, of this is how, this is, they are architected that way, that they all break at the same time, or they all have, one vulnerability means vulnerability for all of them?
25:01 Speaker 03 Well, we have to think of different ways to fix them. I mean, you think about your computers and your phones, the same is true for them, but there is this ecosystem of upgrade. So we don't know how to design secure software.
25:14 Speaker 02 Right.
25:15 Speaker 03 No idea, so instead, we are agile. So every Meaning dealing with brakes. Dealing with brakes. So if you have a Windows machine, second Tuesday every month, you get about 90 different security upgrades in that Patch Tuesday. You're on a Windows machine, it happens more ad hoc, but we can quickly repair and improve security when we find vulnerabilities. And that works great. We've gotten a lot more security from that process. That doesn't exist for a lot of low-cost embedded systems. Right. Those OmniLocks had no way to do that. Right. But a lot of these systems are designed offshore by third parties. Teams come together, write the code, disperse. They don't have the ability for people to write patches. They don't have the engineers on staff.
25:58 Speaker 03 Even worse, a lot of these devices can't be patched. Right now, the way you patch your home router is you throw it away and buy a new one. That's the mechanism. There isn't another one. Right. Now, that works. That's actually a recent decent mechanism. It's terrible for the environment, but it does get us security. But we expect our devices to last a lot longer. You think about your phones and your computers. You replace them every three to five years. And every time, it is more secure, it's better designed, and it's a better object. The systems in our world have a longer lifespan. You're going to replace your digital video recorder every 10 years. Your refrigerator every 25 years. I got a new home thermostat two years ago.
26:43 Speaker 03 I expect to replace it approximately never. And yet, we're going to expect these devices to be on the Internet for that long. Apple and Microsoft, they depreciate their operating systems after like eight years because it's hard to maintain the old stuff. So, I don't know. Imagine a car. You buy a car today, software's two years old. And say you buy it, drive for 10 years, sell it, somebody else buys it, they drive for 10 years. At that point, they sell it, and we're in the US, so somebody else buys it, puts it on a boat, sends it to Latin America, where someone else buys it, drives another 10 to 20 years. So we're talking about the equivalent of taking a computer today from 1977.
27:27 Speaker 03 Trying to turn it on, put it online, making it work, making it secure. We have no idea how to maintain 40-year-old consumer software.
27:36 Speaker 02 Right.
27:37 Speaker 03 And a lot of these are not going to be cars. They're going to be toys, they're going to be appliances, small appliances, internet gadgets, companies that are out of business, where nobody knows what the code is, where they can't be maintained.
27:49 Speaker 02 You know, it's so funny you say that. Years ago, when Facebook was allowing all those third-party apps on Facebook, I kept saying, "What if that goes out of business? What happens to the data that you have collected via Facebook?"
27:59 Speaker 03 And what we know now is that's often sold as an asset.
28:04 Speaker 02 I thought it was a good question. Yeah, exactly. Who protects it? What happens to it?
28:08 Speaker 03 Nobody does.
28:09 Speaker 02 So where do we find ourselves in that way? Because this is the way development is going now, right? It's sort of the develop and then junk, and then move on to the next thing.
28:16 Speaker 03 Well, some of it is our consumer culture, develop and then junk. Some of it is the way technology, when it's just software, we iterate all the time. And we're used to it when it's just software.
28:28 Speaker 02 Right.
28:29 Speaker 03 We're not used to it when it's
28:30 Speaker 02 Integrated.
28:31 Speaker 03 Things, when it's objects. Some of it is our unwillingness to regulate technology at any level.
28:39 Speaker 02 Mm-hmm.
28:40 Speaker 03 That's sort of given us this world where short-term corporate interests are the interests of governoral decisions.
28:48 Speaker 02 Mm-hmm.
28:49 Speaker 03 And when you have that, this is what you have.
28:51 Speaker 02 Well, let's get into that. Talk about where you see regulation going, because now everybody's talking about regulation. And whenever I give speeches about this, everyone's like, "Well, what are the laws in place about technology?" I'm like, "There aren't any. There hardly..." There are some very light regulations, and most of the regulations are pro-technologies, such as Section 230 or whatever. A lot of them are facilitate development of technology. But in general, there aren't any laws. There's one, there's a few privacy bills scattered around the world, there's one in California, but otherwise, it's pretty light compared to most industries that are regulated in our society, like finance, chemicals, cigarettes, cars, things like that.
29:27 Speaker 03 So the US, especially, there's almost none. And some of it is because the industry was so new, because it was so complicated, no one knew how to regulate it, because it was so valuable.
29:39 Speaker 02 Right.
29:40 Speaker 03 People were afraid to touch
29:41 Speaker 02 To touch the innovation.
29:42 Speaker 03 This enormous commerce engine that generates so much value for society.
29:50 Speaker 03 Some of it is the fact that it didn't matter. That this is a word processor. This is a photo app. And these things became part of our society infrastructure, sort of slowly through the back door. So automobiles are regulated. Medical devices are regulated. But now that those are computers, existing regulations are pushing down. So airplanes, automobiles, medical devices, they are all regulating the computer aspects of those things. Consumer devices, not so much. That's probably going to change as we start seeing more toys and small appliances.
30:33 Speaker 02 Talk about the regulations you think need to happen.
30:35 Speaker 03 So I think we need basic consumer protection regulations. It's a combination of things. Let me sort of step back a little.
30:43 Speaker 02 There's a security aspect, there's a privacy aspect.
30:45 Speaker 03 And it's never one thing. So software has been long excluded from normal product liability laws, which like makes no sense. So just making software more in line with everything else. That if a product is defective, the victims of that defective nature can sue. I think we need more consumer protection, rules like we have in any other industry. I think we need product safety rules. If Ford made a car that as a side effect of driving around the country would destroy democracy, they would not be allowed to sell that car. And they couldn't say things like...
31:25 Speaker 02 What kind of car would that be?
31:26 Speaker 03 I don't know. Maybe it had signs on the side. I'm not making this up.
31:29 Speaker 02 It might look like Elon Musk's new truck.
31:32 Speaker 03 And the notion that a company can say, wait, our product is harmful, but if you tell us to make it not harmful, we won't be profitable, is just not a valid argument in any other aspect of society other than software right now. So I think bringing software in line with everything else would make a lot of sense. We are missing both prescriptive...
32:01 Speaker 03 Requirements. And we're starting to see some of this. So California has a new law coming into force in January for Internet of Things. The one thing it says is no default passwords. And that is basic consumer protection.
32:13 Speaker 02 Right.
32:13 Speaker 03 Now, that is like, on the list of 100 things we have to do, that's one of them. But at least it's one of them.
32:18 Speaker 02 Right.
32:19 Speaker 03 And the fact that it took us that long to get there sort of shows how far we need to go. But government is-
32:26 Speaker 02 Oddly enough, I just had a Schlage lock and they had a default. I'm like, why does this here at all? Well, that's the default password. I'm like, why? I was like, it shouldn't have a default password because most people don't change it. It was like, it was the longest argument with the installers. And I was like, do you understand why this is a problem? Like, anyway, but go ahead.
32:41 Speaker 03 Is it physical lock?
32:42 Speaker 02 Yeah.
32:42 Speaker 03 The physical locks forever came with the, the basic password.
32:47 Speaker 02 No, I, right, this was an electronic lock.
32:49 Speaker 03 Electronic lock, right. You'd think would have a unique password.
32:51 Speaker 02 Right.
32:52 Speaker 03 Now, I think most padlocks you buy, the-
32:54 Speaker 02 1234, or something like that.
32:55 Speaker 03 Nah, I don't, I don't remember. But while I was up, what a padlock.
32:58 Speaker 02 0000.
32:59 Speaker 03 There, there have been ones that, and you're supposed to change them. Some, like the Masters, do have a random one.
33:05 Speaker 02 They do.
33:05 Speaker 03 And it's written on the package.
33:07 Speaker 02 It is, but they're random.
33:08 Speaker 03 They're random, but you're right. Most people will not change the default.
33:11 Speaker 02 Right.
33:11 Speaker 03 And we just have to accept that and, and, you know, make it so we protect the consumer from themselves, like we do in so many other aspects. It's, you know, what is the, the example that-
33:23 Speaker 02 Don't eat that tainted meat. Go ahead, but go, sorry.
33:26 Speaker 03 The example Elizabeth Warren used, we talked about financial regulation, that it is illegal to sell you a toaster that will catch on fire.
33:33 Speaker 02 Right.
33:34 Speaker 03 Right, that, we can't sell you that. Yet we can sell you a financial product that would metaphorically catch on fire. We can sell you software that will metaphorically catch on fire. And we just kind of need to fix those things.
33:44 Speaker 02 Where is that right now? Where, where are those? Who's pushing for these things?
33:48 Speaker 03 Not, there's not much. I mean, I tend to look towards Europe, who really is the regulatory superpower on the planet right now, and to the States, specifically California, New York, and Massachusetts. No appetite in the federal government, not just because the federal government is pretty dysfunctional right now, but in general, there was just not an appetite to regulate the tech industry. It's just too powerful and, and too valuable.
34:13 Speaker 02 Do you think that's changed?
34:14 Speaker 03 I think it is changing. I think the social media and democracy is starting to change things, but it's only changing things in the perceptions of people. Has it changed things in the perceptions of rich people before the US government notices? So I really, if we want change, I think it's going to be coming out of Europe first. And the neat thing about technology is that it's right one cell everywhere.
34:35 Speaker 02 Right.
34:36 Speaker 03 And right now, the car you buy in the United States is not the same car you buy in Mexico. Environmental laws are different, and the manufacturers tune their engines.
34:42 Speaker 02 Right.
34:43 Speaker 03 But the Facebook you get in the United States is the exact same Facebook you get in Mexico. Because it's easier. So California has a law coming into force in January, no default passwords. The maker of a thermostat is going to change their software to meet California law. They're not going to sell two versions, one for California and the rest of us. We will all benefit from that law. Similarly, Europe has GDPR, a big privacy law, and a lot of us not in Europe benefit from that. I used to work for IBM. They made a decision early on that they would implement GDPR worldwide because it was easier and cheaper to do that than to figure out who a European citizen was.
35:23 Speaker 02 Right.
35:24 Speaker 03 And I think it's a really interesting feature of software that will benefit us, that well-written regulations in any large enough market move the whole world.
35:35 Speaker 02 Like California or Europe or any world.
35:37 Speaker 03 California, Europe, right, the UK even, will move the whole world.
35:42 Speaker 02 So when you think about this is moving through, that means a very slow version of public policy and not led by the country where these companies started from a federal point.
35:52 Speaker 03 I think that's what we're stuck with.
35:53 Speaker 02 Right.
35:54 Speaker 03 I just don't see the US government having appetite to do this anytime soon. We might see some movement in the judiciary where court cases, I think, will put this industry in line. We're seeing some movement in some of the federal agencies. I mentioned the FAA with aircrafts, Department of Transportation with automobiles, whoever regulates medical devices, I'm blanking on their name.
36:18 Speaker 02 Oh, FDA.
36:19 Speaker 03 FDA.
36:20 Speaker 02 FDA, yeah.
36:21 Speaker 03 Right, the Federal Trade Commission.
36:22 Speaker 02 Yes, the FTC.
36:23 Speaker 03 Is doing a lot with consumer protection. Although their mandates are a little weird, there are things they can do about unfair and deceptive trade practices. So I think we are going to see some push there, but primarily, I am looking to Europe and to some of the states. It's, New York is trying to regulate cryptocurrencies, California has a data privacy law, and now Internet of Things law. Massachusetts is pushing some things, we see a digital right to repair.
36:49 Speaker 02 But that creates a real... Problem in that they're all different, that for these companies too, in terms of they're all different. So they just have to pick the one that's most powerful or the most stringent.
36:59 Speaker 03 But countries are already doing that. I mean, yes, it's a problem. It'd be nice to have harmony, but the big worry about national laws that preempt state laws and the reason a lot of the tech companies push them is they know they're going to be lousy. It is much easier to have a lousy federal law than to have a few good state laws.
37:17 Speaker 02 Right.
37:18 Speaker 03 And companies will scream that they can't meet everything in order to try to get that watered-down federal law. So be careful about the complaints. I mean, yes, it's a problem, but industries have been doing this since forever.
37:32 Speaker 02 Right, on every other thing.
37:33 Speaker 03 On every other thing, and, you know, it's been working okay.
37:36 Speaker 02 It is interesting, the mentality that this should not be regulated, given that every other major industry has been.
37:41 Speaker 03 It's hard to regulate. I mean, I'm not going to say this is easy. It's a very fast-moving industry. And I think we need to figure out, we don't actually know how to legislate in the 21st century. If someone invents, I don't know, the telephone, you know, 100 years ago, and it takes 30, 40 years before it permeates society for us to figure out the law of the telephone. Someone invents a drone, and I remember this, there were the years we said, "We can't regulate drones. It's too early. You'll destroy the industry."
38:13 Speaker 02 Right.
38:14 Speaker 03 Then suddenly, everyone gets one for Christmas, and then it's too late to regulate drones. They're everywhere.
38:19 Speaker 02 Right.
38:20 Speaker 03 With no intervening middle ground. We have to do better than that.
38:22 Speaker 02 Right. All right, when we get back, we're talking with Bruce Schneier. His most recent book is called Click Here to Kill Everybody: Security and Survival in the Hyperconnected World. When we get back, we're going to talk about some coming technologies that also are probably not going to be regulated and what we need to do. He is a well-known techie, I guess, on lots of areas, but he's focusing on the public interest now here at Harvard.
38:44 Speaker 04 Priceline Negotiator. It's me, the. Priceline Negotiator. We don't need the jingle twice. What about a third time? Stop it. This is about vacation inflation and how Priceline negotiates amazing deals on hotels, flights, and rental cars. Seems like they just started. Yeah, but I didn't mention that you can save up to 60% off hotels in the Priceline app. Time to read the tagline. Fine. No one deals more deals than. Priceline. Please stop. Priceline. Touche. Priceline. Priceline.
39:15 Speaker 05 How about a latte?
39:16 Speaker 01 Maybe dark roast?
39:17 Speaker 05 Vanilla?
39:18 Speaker 01 Caramel?
39:18 Speaker 05 Blueberry?
39:19 Speaker 01 You could get them to agree on one flavor or give everyone what they want. With Keurig, there's over 500 K-Cup pie varieties of fresh, flavorful coffee. No effort, no grind, no settling. Great coffee without the grind. Keurig.
39:42 Speaker 02 We're here with Bruce Schneier. He's the author of Click Here to Kill Everybody: Security and Survival in the Hyper-Connected World. He's also a fellow at the Berkman Center for Internet and Society at Harvard Law School, and he's big in cryptocurrency, security researcher, all kinds of stuff. Talk a little bit about what's coming next, some of the technologies you're focused on of the next quantum. I'm gonna go through them. First, let me, actually, I'm gonna start with facial recognition. Where are we with facial recognition and where should we be? Where should we be focused on?
40:10 Speaker 03 So facial recognition, I think, is really interesting technology because you really can't separate the tech from its uses. And all these technologies were pretty good, getting better every year. I mean, we all know how tech works, that every year it gets better.
40:22 Speaker 02 Sure.
40:23 Speaker 03 And there are two kinds of facial recognition. There is the one you use if you have a new iPhone. And that is, that is answering the question, is this the person I think it is? Is this Kara Swisher's phone? The other type of facial recognition is, who is this? And that's the kind you might see deployed by the police in the United States, in Hong Kong, in stadiums.
40:52 Speaker 02 And is this a dangerous person?
40:53 Speaker 03 Well, it's two questions. Who is this? Because once you know who it is, then you go back to your database and you make whatever decision you might want to make. If you are a department store, you might decide to give them better customer service or worse customer service. If you are a police department, you might decide to arrest them or not. It really depends on where it's being deployed. These are technologies of ubiquitous surveillance. Now, that might be a good thing in some circumstances.
41:25 Speaker 02 Right, just as being recognized, getting on a plane, it might be faster, easier, that kind of thing.
41:29 Speaker 03 But these are technologies that allow a level of surveillance that just isn't possible because they replace human eyes. We can always recognize people in public. That's not nothing new. But to have a million cameras in the city of London, all of whom can recognize everyone in public all the time, allows the people who control those cameras to map where people are in a way they haven't done before. Now, I think it's real important we separate the technologies from the uses. We spend a lot of time talking about these technologies as different thing. I don't want to say this. Facial recognition allows us to identify people. So do many other things.
42:13 Speaker 03 Every one of us carries the most exquisite surveillance device ever invented in our pockets every day. It's our phone.
42:22 Speaker 02 Right, it follows us around.
42:23 Speaker 03 And it knows where we are at all times, knows who we're with, knows when we go to sleep, when we wake up. No one ever lies to their search engine. It knows more about us than our intimates. And facial recognition is another one of these technologies. And there is gate recognition, there's fingerprint recognition at a distance, eye scanning, I mean, so many technologies of identification. And really step back and think, what does it mean for someone to be identified throughout their day, throughout their life, as they move through society? First question to ask is, who owns that data?
43:03 Speaker 03 The data from your phone is owned by the phone company, by whatever apps you've given permission to access your GPS. The data from the cameras is owned by whatever private companies are running those cameras, whatever governments are running those cameras. Very much the power dynamic. Who has access to that data? And we're moving to a world where, in general, the notion of ephemera is disappearing.
43:33 Speaker 02 Right.
43:33 Speaker 03 Already, many of my intimate conversations happen over text. No longer happen over voice, no longer happen in person, where the words are said and dissipate. And I don't know about you, but I'm willing to bet that my relationship will not be improved by being able to go back and read the transcript of five-year-old arguments to know that I was right. I think there's a lot of human value.
43:58 Speaker 02 In forgetting.
43:59 Speaker 03 In forgetting. And I think we're going to lose that.
44:01 Speaker 02 You know who doesn't text? Donald Trump.
44:04 Speaker 03 It's, you know.
44:05 Speaker 02 Only phone.
44:06 Speaker 03 Texting is dangerous, because texting leaves an audit trail. And a lot of journalists will not text with sources for the same reason. And a lot of dissidents will not. But a lot will, because we as humans need to communicate.
44:19 Speaker 02 Right. So facial recognition, where are we in terms of regulation? We aren't at all, right?
44:24 Speaker 03 There's no regulation.
44:25 Speaker 02 Right.
44:26 Speaker 03 And.
44:26 Speaker 02 And it's full of errors.
44:28 Speaker 03 It's full of errors, it's full of biases. You know, these are likely to shake out over the years and, you know, maybe decades.
44:35 Speaker 02 It's my fear that there's not going to be a lot of errors. It's going to be very good at what it does.
44:39 Speaker 03 That's right, and that's the worry. And now this depends, the errors depend on the use. Now, if I'm misrecognized as I'm walking by a billboard, they show me an ad for a Chevy I don't want to buy. If I'm misrecognized, you know.
44:52 Speaker 02 Crossing the street or a bank.
44:53 Speaker 03 Not crossing the street, think of me in Afghanistan. Someone's up to drone on my house.
44:56 Speaker 02 Right.
44:57 Speaker 03 So, right, the cost of the error depends a lot on who you are, how much power you have, whether you're marginalized in society. And disproportionately, the mistakes fall on the unempowered.
45:14 Speaker 02 So who should be regulating this? The government should be regulating this?
45:17 Speaker 03 The government should be regulating this. But unfortunately, the government also is using this.
45:20 Speaker 02 Right.
45:20 Speaker 03 And we see this even with something, this is another technology, just Facebook advertising. I mean, I'd love if the government regulates with Facebook advertising, but every politician who uses it says, "Whoa, neat, that was fantastic. I want that forever."
45:34 Speaker 02 Behavioral targeting.
45:35 Speaker 03 And they don't want to regulate that.
45:37 Speaker 02 Right.
45:37 Speaker 03 Because they want to use it too.
45:38 Speaker 02 Right.
45:38 Speaker 03 Because in a sense, politics is the same as any other kind of advertising.
45:42 Speaker 02 Let's talk about that behavioral tracking. That's another one. The ability to track you based on your behavior, which will be girded by facial recognition because it will see you as you move through the world.
45:54 Speaker 03 Facial recognition and other things as well. And again, there's still a lot of errors here, but the basic idea is that we can look for certain kinds of behavior. Someone standing when they should be moving, someone moving when they should be standing, someone who looks suspicious, is wearing the wrong clothing for the weather. I mean, we can make up all these behavioral characteristics. And yes, I mean, your question is perfectly reasonable. You know, what are the error rates? Who does it disproportionately fall upon? And how well do these systems work? Sometimes you don't care.
46:24 Speaker 02 Right.
46:25 Speaker 03 You're a country like China, you get a high error rate, you kind of don't care. Half of your aim is to cause terror in the population, so, you know, a few errors actually make your system work better for that intended purpose.
46:39 Speaker 02 All right. So then when you get to that quantum computing.
46:42 Speaker 03 All right.
46:43 Speaker 02 These new Google things.
46:44 Speaker 03 Quantum computing, everyone should calm down.
46:46 Speaker 02 All right, good.
46:46 Speaker 03 Okay, so this is going to be maybe a little more mathy than you want, but I'll try to...
46:51 Speaker 02 Go right ahead. Go mathy.
46:52 Speaker 03 All right, I'll try to stay on top. It's basically a new way of doing computing that is not constrained by some of the linear constraints of our normal computers. Very experimental, and it has the potential to...
47:07 Speaker 03 Solve problems that can't be solved by conventional computers, any human lengths of time. And two of the big problems are problems that we use in cryptography to keep things secure. So you hear people talk about the cryptography apocalypse. Yeah. I think it's overblown. I've literally gotten 26 messages. Super overblown. I'm asking you, Bruce. So, so a couple of things. One is that we are already working on quantum resistant cryptography. NIST has a program in it, we have algorithms, we will have algorithms that will resist quantum computers before we have working quantum computers. That can solve, that can actually solve these problems at any real scale. The challenges to a quantum computer that can solve serious problems are significant. And it's hard to surmount those.
47:51 Speaker 03 We truly don't know if by hard, we mean go to the moon hard, or faster than light travel hard. It could be either. Right. So we could have a quantum computer that can solve real problems in a decade or more, or we could have it never. And we don't know. My guess is a decade or more. My guess is these problems are solvable. What do you mean have it never? Just... That if it's faster than light travel hard, we're not going to do it. That'll forever be science fiction. Right. And these are problems with coherence. Basically, the problem is how to get the data out of a quantum computer to read it. That, that the answer will decay so quickly, and that it's quantum, of course, like measuring it destroys it, you know, that sort of problem. And, and, you know, a lot of companies are working on this.
48:35 Speaker 03 There are quantum scientists who agree with me, who disagree with me, there's no, we don't know. What, what is the problems it solves for the layperson? What is the big deal here? The, the problems it solves are optimization problems and search problems. So if you want to figure out the traveling salesman problem. So you have 50 cities in your territory, you need to visit them all, what is the shortest driving path that will get you to all of them? That's actually a really hard problem. It is. It's easy to get a good one, but to get the best one is super hard. Right. A quantum computer can solve that easily. Someone was telling me self-driving cars is a perfect example. I don't think self-driving cars is a problem that... Well, in terms of keep...
49:16 Speaker 02 Keeping them all with each other.
49:18 Speaker 03 See, I don't think so, because you don't need the best answer, you just need a good answer. And if you can get an answer that's 1% better for 10 times the money, no one's gonna do it. So it really is times you need the perfect answer. And breaking encryption is one of them. That's why you had the NSA, even as early as 2014, you saw in the Snowden documents, had a budget for quantum computing. Because they saw the cryptographic benefits of it.
49:43 Speaker 02 Right.
49:44 Speaker 03 But I don't see a quantum arms race, I don't see a quantum apocalypse. We will figure out cryptography before the quantum computers. And I don't think that's controversial.
49:55 Speaker 02 What about AI?
49:57 Speaker 03 AI is gonna be interesting to watch, right? AI is very nonlinear. AI, easy problems turn out to be hard, hard problems out to be easy. When I was a student, we learned that Go, the game, would never be solved by a quantum computer. Because it was computationally infeasible, blew up too fast. And that was true until Google says, "Oh yeah, we did that last month because we were bored." And now Go can be solved by a machine learning.
50:23 Speaker 02 Right.
50:23 Speaker 03 And if you play Go, anyone listening who plays Go, I recommend
50:27 Speaker 02 This is a Chinese game.
50:27 Speaker 03 Going to look at some of the Go games played by AlphaGo. And the way it learned is it learned the rules of Go, then got every human Go game ever recorded, learned from those, then it played itself for a gazillion times and got really, really good. And it plays Go phenomenally well. Then Google did another experiment. They didn't feed the new program any human Go games. And said, "Figure it out yourself." And that program plays Go like an alien. It's really fascinating. It makes moves that a human would never make. And I think that's really a way to think about the future of AI.
51:10 Speaker 03 It'll do things we don't imagine in ways that'll be completely strange. Some things that we think are impossible will be done. Some things that we think are easy, which can be very, very hard and not be done. It's very hard to predict. Insecurity, we'll have AIs attacking, defending. We don't know who will have the advantage.
51:33 Speaker 02 Right.
51:33 Speaker 03 And again, this is going to be a thing about power. I don't see a data race. You'll see a US-China story. Because China spies on its people more, they have more data to train their AIs. I think that's largely nonsense.
51:48 Speaker 02 Okay, tell me why.
51:49 Speaker 03 It's not about the amount of data. It's about the type of data and how it's used. And the open science of how to use data is really where the work is. And not the, I have a big data set of, you know,
52:01 Speaker 02 People going to school.
52:01 Speaker 03 Chinese citizens going to school that I can use. That's not going to separate the AIs from each other. It's going to be how they're built. And a lot of this is being built in the open. This is public science.
52:13 Speaker 02 Right, okay. So the idea of that, this is an argument I get from Silicon Valley a lot. We have to be doing this because the Chinese will beat us with their giant.
52:21 Speaker 03 Yeah, I think they just want to spy on us, and that's a good excuse right now.
52:24 Speaker 02 The Chinese, yeah.
52:25 Speaker 03 No, the Chinese, but also Silicon Valley.
52:27 Speaker 02 Yes, oh yeah. We need to be this big so that we can fight them. I call it the she or me argument.
52:32 Speaker 03 Yeah, I don't buy it. And I think we're at the point where the companies are too big. And a lot of our problem is the fact that these are all monopolies.
52:41 Speaker 02 I want to get to that last, but robotics and automation.
52:44 Speaker 03 From a security perspective, I think these are also really interesting. And it's definitely, automation's the right word. It's automation, autonomy, and physical agency. Those are the things that I think about. Robotics really epitomizes all of them.
52:57 Speaker 02 Right.
52:57 Speaker 03 And it's going to be things in our world doing stuff.
53:01 Speaker 02 Right.
53:02 Speaker 03 And things that have all the computer flaws that we're used to. All of the vulnerabilities we're used to, that are going to be doing things. And things that are increasingly critical, increasingly life-critical.
53:17 Speaker 02 Right.
53:17 Speaker 03 And these are aspects I worry about. And again, this goes back to the title of my book, Click Here to Kill Everybody.
53:23 Speaker 02 Right, because they all can be told to do one thing.
53:26 Speaker 03 Right, we've seen people program Roombas, I mean, this is still academic exercise, to follow people around and try to trip them.
53:33 Speaker 02 I get one of those.
53:35 Speaker 03 You kind of laugh, but...
53:38 Speaker 02 I didn't know you people did that, but of course you do.
53:40 Speaker 03 Of course you do, because it's the Internet. People are haters. That's the way things work.
53:44 Speaker 02 I want one of those. Now, please send me a rope with the truth, people. All right, so robotics, but in terms of the fail points is what you're talking about? That there's multiple points of failure here?
53:54 Speaker 03 Right, and I worry about who owns the data. Back to surveillance, capitalism. Maybe these robots will collect enormous amount of data about us, and who's going to own that data? What happened to that data? I worry about who's going to control those robots. Talk a little about the right to repair. If that robot is owned by Google, by John Deere, by a company, then you might not have the ability to change it. We're seeing that in medical devices. There are patients who are trying to get access to their automatic blood insulin level data, which is being collected by these
54:25 Speaker 02 The companies.
54:26 Speaker 03 Computerized systems, and a company owns that data. And it's like, it's my body. Why isn't it my data?
54:32 Speaker 02 What about humans becoming more robotic, rather than robots becoming more human?
54:37 Speaker 03 Yeah, I don't follow the sort of body augmentation. I think we sort of see the harbinger of that when we think about our devices. In a lot of ways, our phones and computers are extensions of us.
54:47 Speaker 02 Right.
54:47 Speaker 03 My email is my offline brain. There are things that I remember because I know it's there and I go get at it. I mean, a lot of us really can't function without the calendars and and the Internet, and email, and messaging. And it'll be those same type of things more integrated. I don't follow body augmentation or any of that sort of thing. Again, I worry about the software, about who controls it, whether it can be taken over. I don't want to have robotic legs and suddenly, you know, some hacker from Sub-Saharan Africa is walking me around, and now there's ransomware against my legs. And this is not unreasonable.
55:24 Speaker 02 It's not, it's not. I just think that's kind of a fascinating idea. It's true, though. It's not unreasonable whatsoever.
55:29 Speaker 03 No, we've seen demo ransomware against cars.
55:32 Speaker 02 So, lastly, the power of these companies. I want to finish up talking about that. One of the arguments I hear a lot in Zalima is we need to be this big, as I said, in order to combat all these problems that you just talked about. What is the dangers of this size? To me, it's the, one, ownership of the information and who gets to make decisions about that. Two, it's that there's no way, once they get that big and they get into monopolies, you can fight them. Even the government can fight them. And three, it's innovation dies when you have that.
56:02 Speaker 03 So I think it's all of those things. I mean, a lot of it is the generic problems of monopoly that we've known for decades and just forgot since the eighties. That monopoly stifles innovation. That monopoly leads to complacency. That monopoly just leads to substantive outcomes where we're paying more and getting less. And, you know, we're now in a world where, really, the vanguard of capitalism is, in competition, is in Europe and not in the United States. They pay much less for, see, even things as simple as internet access, because they have actual competition and we don't. And I think you lose a lot when you have a monopoly. And these tech monopolies are hurting us in many ways. And even something as simple as forcing Facebook to open up their API for compatible
56:47 Speaker 03 apps would go in enormous way towards fixing all the problems Facebook is having
56:53 Speaker 02 Yep.
56:54 Speaker 03 with their platforms and society. We have to be big, this big, to fight these threats? Makes no sense. I think a diversity is better at fighting these threats.
57:05 Speaker 02 And in the next, given that situation, because all these companies are American companies, is this where the next internet age is? If you had to look around where innovation is coming from? Because to me, that's what will dominate. Like, people talk about Facebook or whatever, or any of these companies, and to me, you see it over and over again, the same thing, is the replacement of these. Where do you imagine innovation is coming from right now? How do you push it as a country, in this country at least?
57:33 Speaker 03 I think innovation is still coming around the edges. Facebook and Google are still buying companies left and right that might out-innovate them.
57:42 Speaker 02 Right.
57:43 Speaker 03 And, I mean, they're still accreting. Apple as well. So, you are still seeing a lot of innovation. I think the innovation can be helped by keeping things open, by getting rid of the monopoly behavior. But using something as simple as net neutrality, to stop a monopoly in one area from finching on another area, does an enormous amount for innovation. I don't think we see a lot of innovation in social media right now. Because there's no point. You'll never get a competitive Facebook public funded.
58:15 Speaker 02 No.
58:15 Speaker 03 It makes no sense.
58:16 Speaker 02 I just talked to a big VC.
58:17 Speaker 03 And we want those. We would like competition. I think it would keep Facebook honest. I think we'd all benefit. But as long as Facebook is allowed to be as big as it is and buy whatever it wants, you'll just never have that kind of innovation.
58:33 Speaker 02 So what do we do? What are your solutions? Give me three solutions that we need to do right away in the short term, maybe middle term and long term.
58:41 Speaker 03 So I think going after the monopolies would be an enormous...
58:44 Speaker 02 Breaking them up.
58:45 Speaker 03 Breaking them up. And I think we have to do that. Yeah, and I think I'd do that for, I mean, I think it's better for society in almost every possible way. I don't even think it's bad for the tech companies. I don't think Mark Zuckerberg...
58:58 Speaker 02 You create more shareholders.
58:59 Speaker 03 If he has, you know, right, if he has majority interest in 17 companies instead of one, I don't think he's worse off. I think he's going to be better off. And I'm not sure why he's not embracing this. I think we need to get governments serious about regulating technology. That the missing ingredient in the way the market is supposed to work is the well-regulated part. And that is harming all of us, including the companies. I mean, the whole point of regulation is to establish a playing field in which the market can operate. And when you don't have that, the market doesn't operate efficiently and doesn't operate well. And I think the companies are hurting because of that.
59:43 Speaker 03 And they might have more near-term profits, but it's not going to last.
59:45 Speaker 02 Okay.
59:46 Speaker 03 So I think, and these are all very, very general things.
59:48 Speaker 02 Right.
59:49 Speaker 03 And I think we need to really get serious about international agreements and norms and way to think about cyber conflict and the way the governments use the internet and cyberspace. Very long-term, very nebulous, a lot of talk about cyber war, now a lot of talk about cyber peace. But flipping that would be nice. And these are kind of decades-long processes, slow and iterative. You know, I work at the Harvard Kennedy School, and they are still working on nuclear deterrence, right? You'd think they haven't even licked that yet. But, you know, we do a lot better than we were decades ago. And whether it's nuclear, or small arms, or human trafficking, or bribery...
1:00:34 Speaker 03 Or, you know, we sort of know how to do this. Even in an international world where there are pariah countries. And the more we can do that on the internet, the better. The internet is global. The internet affects everything. It is critical infrastructure for the planet. It has ramifications in human rights, in climate change, in everything that matters in this century.
1:00:59 Speaker 02 Right, absolutely.
1:01:00 Speaker 03 And if we don't get it right, it's gonna be real bad.
1:01:04 Speaker 02 So, let me end on this. This election's coming up. How do you feel about the security of our systems compared to last election?
1:01:13 Speaker 03 I don't think we're any better. Well, I'll separate two things. I'll separate voting security, which is security of the machines and the systems. Actually, it's several things. It's the election rolls, who can vote. It's the machines, whether the vote is accurate. It's the tabulation systems, and it's the reporting systems. All of those have vulnerabilities, and they have not been addressed. And I'm very worried about election security.
1:01:36 Speaker 02 Right, the rolls.
1:01:37 Speaker 03 Right. And then there's the security of the greater election process. And that's propaganda, disinformation, both foreign and domestic. I think you looked at the last election in Governor Kentucky. You saw an example of what's gonna happen in two years from now, or next year, whenever it's gonna. I forget when the election is. That you had, and this was all nothing, you can tell nothing was foreign influenced. But memes are going around, election was stolen, and suddenly there's all of this.
1:02:07 Speaker 02 Sure.
1:02:08 Speaker 03 All of this fake news. I hate the term, but that's what it is, out there. That's being pushed by domestic actors with an agenda. And Facebook and Google do not have a good handle on this. Twitter doesn't. I think the things we're seeing in Brazil and India, where a lot of this moved to some of the messaging apps, to WhatsApp, to direct messaging from Facebook and the social media, is what's gonna happen in the United States. And I think we have to worry about these fake issues and agendas hijacking the real agenda. I think the mass media, the New York Timeses, haven't owned up to their.
1:02:52 Speaker 03 To their part in 2016, I mean, Facebook might have done some pretty bad things, but at least they said, "We did this. We're going to try to do better." I don't think the mainstream media has really acknowledged their role and how they were played by the Russians, how they're going to do better. So I do worry about this at all levels.
1:03:10 Speaker 02 Yeah, absolutely. All right, Bruce, what's one positive thing about technology right now? What's your most favorite thing that you've seen?
1:03:16 Speaker 03 I might sound like a downer, but I think technology is mostly good, is almost entirely good.
1:03:23 Speaker 02 Yeah.
1:03:23 Speaker 03 I mean, we're spending our time talking around the fringes. And as a security researcher, it's real easy to think that everything is bad. But I am just as-
1:03:32 Speaker 02 Well, one nuclear bomb can ruin your day.
1:03:33 Speaker 03 One can ruin your day. But, you know, one nuclear power plant can, you know, save a whole lot of fossil fuel burning.
1:03:40 Speaker 02 Right.
1:03:41 Speaker 03 And there are a lot more nuclear power plants that have done that than nuclear bombs that have gone off. So by and large, like all technologies, I think we're doing really well. It's, we're right about the edge cases, we're right about the fringe uses. We just don't want them to take over the whole.
1:03:57 Speaker 02 Right, right. And please, I hope there's nothing you can click to kill everybody.
1:04:01 Speaker 03 Not yet.
1:04:01 Speaker 02 Not yet. Let's hope not. Anyway, thank you so much for coming by the show, Bruce. You can follow me on Twitter at Kara Swisher. My executive producer, Erica Anderson, is at Erica America. My producer, Eric Johnson, is at HeyHeyESJ. Bruce, where can people find you online?
1:04:14 Speaker 03 So I am on Schneider.com. I do tweet, actually, I mostly blog. I have a Twitter feed, Schneider Blog, that mirrors my blog. And so Schneider.com is where to go for my books, my writing, everything I've done is there.
1:04:27 Speaker 02 There's also something called Schneider Facts. He hates them, but I love them. It goes, "Bruce Schneider doesn't use Tor, he just accesses your site."
1:04:35 Speaker 03 All right, so this is a series of dorky math and security jokes that have nothing to do with using my name, and they are pretty funny.
1:04:42 Speaker 02 Very funny, especially with pictures of you in a thing. It's quite funny. They're all math jokes, but they're very, very funny about security and other things. So anywhere else they can find you?
1:04:51 Speaker 03 That's where I am.
1:04:51 Speaker 02 Okay, all right.
1:04:52 Speaker 03 Other than that, I'm keeping a low profile.
1:04:55 Speaker 02 Yeah, I love the blog. It's really great. If you liked this episode, we'd really appreciate it if you shared it with a friend. And make sure to check out our newest podcast, Reset. Just search for it in your podcasting app of choice or tap the link in the show notes. Thanks also to our editor, Joel Robbie. Special thanks to WHRB at Harvard University for hosting us. Thanks, Sergio. Thanks for listening to this episode of Recode Decode. I'll be back here on Monday. Tune in then.
1:05:23 Speaker 03 In this life, you can't win.
1:05:25 Speaker 06 When the world gets you down.
1:05:26 Speaker 03 No!
1:05:27 Speaker 06 Escape to Pluto TV.
1:05:28 Speaker 03 This is objectively fantastic.
1:05:30 Speaker 06 Relax into the movies and shows you already know and love, all for free.
1:05:34 Speaker 03 Schwing!
1:05:35 Speaker 06 Even our news is uplifting.
1:05:36 Speaker 00 You stay classy, San Diego.
1:05:38 Speaker 06 Watch School of Rock, Crazy Ex Girlfriend, Anchorman: The Legend of Ron Burgundy, The Ex-Files, Wayne's World, and so much more.
1:05:44 Speaker 03 We're not worthy!
1:05:45 Speaker 06 The best things on TV really are free.
1:05:48 Speaker 03 You just have to know where to look.
1:05:49 Speaker 06 Escape to Pluto TV. Stream now, pay never.
Previous Recode Decode: (Bonus) How To Save the 2020 Election Next Recode Decode: The Swisher family Page updated October 13, 2026 · transcript and summary are machine-generated