The future of Facebook and democracy with Casey Newton and Alex Stamos

Facebook's former chief security officer Alex Stamos joined Casey Newton onstage at SXSW to discuss the difficult issues that plague Facebook and democracy. Subscribe to The Interface, Casey Newton's newsletter about social platforms and democracy. Learn more about your ad choices. Visit podcastchoices.com/adchoices

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  This episode is brought to you by On Investing, an original podcast from Charles Schwab. Each week, hosts Lizanne Saunders, Schwab's Chief Investment Strategist, and Colin Martin, Head of Fixed Income Research and Strategy for the Schwab Center for Financial Research, bring you fresh insights on what's happening in the markets and why, and what the implications might be for your portfolio. Join Colin and Lizanne as they explore questions like, how do you evaluate corporate bonds that look interesting? And what sectors are on the move right now? Download the latest episode and subscribe at schwab.com/oninvesting or wherever you get your podcasts.

• 0:45Speaker 01

  Growing teams need a corporate card built for more, something that doesn't sacrifice speed for control. Introducing the American Express Corporate Cash Back Card. Get 1.5% cash back on eligible business purchases, advanced spending controls, and real-time visibility, plus intelligent expense management. Get started at go.amex slash amexcorporate. Terms and cap apply.

• 1:14Nilay Patel?

  Hey everybody, it's Ian from The Vergecast. Real quick, before we start, do me a favor. Can you go to voxmedia.com/podsurvey? And just tell us how you listen to podcasts, what you like, what you don't like. You can do me two favors and actually just answer every question in a way that favors Vergecast. That would make me really happy. Anyhow, it's voxmedia.com/podsurvey. We're trying to make everything better, so your feedback is greatly appreciated. Okay, this week on the interview episode, I'm actually taking a step back. Casey Newton interviewed Alex Stamos live at South by Southwest. Alex is Facebook's former head of security. Obviously, platforms are in the news. The relationship between our democracy, our culture, and these huge social platforms is more in question than ever. Casey writes a newsletter about that stuff every day called The Interface. You can go to theverge.cominterface, check that out.

• 1:57Nilay Patel?

  But Casey interviewed Alex live at South by Southwest. It was a pretty amazing conversation. Alex has deep insight in how Facebook thinks and how Mark Zuckerberg thinks, and how Zuckerberg is thinking about this big pivot to privacy that he says Facebook is making. It is well worth your time. You might have to slow down because they both talk really fast, but check this out.

• 2:18Speaker 03

  Hey, everybody. My name is Casey Dune. I'm Silicon Valley editor of The Verge. I'm really excited. Also, I want you all to give yourselves a round of applause for caring about democracy. Just quick. Wonderful. The title of this panel is "Saving Democracy," so we've given ourselves 30 minutes. Hopefully, that will be sufficient. And to help us do it, I'm going to welcome onto the stage Facebook's former chief security officer. He's currently at Stanford. We're going to talk about the work he's doing there as well. Please welcome Alex Stamos.

• 2:45Speaker 03

  All right. So, the week's big news. Mark Zuckerberg wrote a very long blog post where he pledged to bring encryption and ephemeral messaging to the entire suite of Facebook messaging apps. And you had a couple tweets that I'm going to read quickly here. You said, "There are fundamental trade-offs that have to be made here. Openness versus data protection. Privacy versus safety. Platform power versus individual freedom. Government power versus citizen privacy. Authenticity versus anonymity." And you described these as kind of a series of dials that, for a long time, Facebook has been turning these between 3 and 8. And now Zuckerberg has decided he wants to slam a bunch of these dials to either 1 or 11. Talk to us about these trade-offs and what will be the ramifications for the democracy in the world we live in.

• 3:28Speaker 04

  Yeah, so, one of the big challenges all the social media companies face is there's a bunch of fundamental trade-offs that are very poorly recognized by the government and media and other critics. So, the classic trade-off in engineering is you can have something done correctly, quickly, or cheaply. Pick 2 of 3. And it turns out that, if you want to do something like keep people safe online, to do so, you need to know a lot about people to know if they're bad guys. It's very, very hard to build a social network where you only know stuff about bad guys and you ignore the good guys. If you want to respect people's individual rights of free expression, then that means you have to be very careful to have rules around how you do it. Facebook has, like I said, tweaked back and forth. Started as this thing just at Harvard, very closed.

• 4:12Speaker 04

  News Feed made it very open. The big change of privacy was a huge deal. And so, that was kind of like turning the privacy way down. And has tweaked it back and forth. But I think... From Zuck's perspective, his problem is he's getting it on both sides.

• 4:26Speaker 03

  Yeah.

• 4:26Speaker 04

  That whatever decision he makes, he will have people who will criticize it for both. In some cases, the same newspaper, the same week, will make two criticisms. You know, an example of that, the New York Times like to criticize Facebook not keeping people safe from committing suicide on Facebook Live, which I think is actually, that's the correct take, is that Facebook, "We did not do enough," and I was there at the time, "to properly prepare for the impact of allowing teenagers to stream themselves live." And a lot of the anti-suicide stuff we had done to that point was completely inapplicable to video. But then later criticized Facebook's being creepy by preventing people from committing suicide and calling 911 when it looks like people might kill themselves. And those are both legitimate takes. You just can't have both at once. And I think what Mark has done is he's decided, "I can't be in the middle anymore."

• 5:09Speaker 04

  The middle is where you lose continuously. So we're gonna, we're gonna make this, this trade-off incredibly stark, and we're gonna slam as far as to one side. And the truth is, if he's going to do that, he only has one, one choice, which is to slam towards the side of technology, because the other side of like, keeping everybody safe is probably actually not practical. So, but you can keep everybody's information private.

• 5:30Speaker 03

  Right. So what are the ramifications of a world where all of our messages are encrypted?

• 5:34Speaker 04

  There's two kinds of bad things, two classes of bad things that happen on social media. One is the bad things that people do to each other where one person is a victim. Right? And so that is, if I, if a bad guy reaches out and harasses somebody, if somebody doxes somebody, if they send a death threat. In the, the case of children, it's to reach out to children and to groom them and to sexually exploit them, which is actually, of all the things, that is the worst thing that happens online every day. We never talk about it. Our society has just kind of moved on. But that is the worst thing that happens every single day, is that the sexual exploitation of children. Those have victims.

• 6:06Speaker 03

  Right.

• 6:07Speaker 04

  And so if you have a victim, then you have a participant who does not want to be part of it. And so in a world where things are encrypted in ephemeral, there are options to keep those people safe. Those options are definitely reduced. You know, we had to deal with this when we rolled out end encryption as an option for Messenger, is things that we had put in place to detect when people are reaching out to... Teenage girls online and trying to groom them, that could be detected. But some of it was able to be replaced through analysis of metadata and that kind of stuff. And then also by really beefing up the ability of the victim to report it. And then to beef up the ability to respond quickly.

• 6:42Speaker 03

  Right.

• 6:42Speaker 04

  The other class of issues online are the things that every participant is a volunteering participant. And this ranges from terrorists who want to organize terrorist attacks by talking to each other, to anti-vaxxers who want to talk about being anti-vax, right? And that is a, that is a class of issues that will now, for the most part, disappear from the ability, because you are, you know, adults generally are now opting in to be part of a discussion where there might be real world ramifications. But since all of them have privacy and all of them have consented to it, for the most part, that will be invisible. And that is the entire class of issues that now Facebook is effectively saying, that is not our problem. And that's how I'm reading it. I don't have any insight in here. But when I read that that's, I see him punting on that class of issues.

• 7:27Speaker 04

  And that's the class of issues where he gives, you can't win. And the anti-vaxxian, I think, is probably the best example of that. That we've gone to the point of where educated adults, you know, people in the media want their decisions about what data they're able to access and what they're able to say to each other to be controlled by these non-democratically accountable half trillion dollar companies.

• 7:44Speaker 03

  Right.

• 7:45Speaker 04

  And it's like, if you're at that point, anti-vax, we all talk about it because we're all like educated urbanites. I mean, I hate anti-vaxxers. I have three kids, one of whom's here. And the idea that one of them is going to get a disease that should have been wiped out in the 19th century makes me sick, right? But I also know the crappy part of freedom is other people having freedom.

• 8:02Speaker 03

  Sure.

• 8:02Speaker 04

  Me having freedom is easy.

• 8:03Speaker 03

  Yeah.

• 8:04Speaker 04

  Other people having freedom is hard.

• 8:05Speaker 03

  It's the worst.

• 8:06Speaker 04

  Right, it's the worst. And so, and I think that's the class of things that he's like, I can't win. So we're gonna wipe our ability out to, to deal with any of those.

• 8:12Speaker 03

  Sure, okay, so, so I write a newsletter about these issues. And I, I feel like I mostly write about unintended consequences. Facebook is at this enormous scale, it makes a decision, and then we just sort of, you know, wait to see what happens next. You know, what you've just described is a world in which life might become much easier for somebody who wants to exploit a child or, you know, like plot terror. Um, and how, like, how do you feel about the fact that this platform is, um, like, going to have all these ramifications? And we just sort of have to wait and see? Like, do we have any sense that this is an informed decision? Like, can data inform this? Or is this one where Mark says, well, because I can't win on this front, like, let's just encrypt everything and see what happens next?

• 8:48Speaker 04

  Yeah, so again, those are two different, I think on the child stuff, there's a lot that can be done.

• 8:53Speaker 03

  Okay.

• 8:53Speaker 04

  And so my hope is that they, I was really happy to see that he explicitly, so first, he explicitly talked about child exploitation and terrorism. Facebook traditionally, one of the things that drove me nuts about the company, is the company doesn't want to talk about the fact that the world's a horrible, dark place.

• 9:05Speaker 03

  Yeah.

• 9:06Speaker 04

  And so if you don't talk about that, then you can't engage on the issues. And so the fact that he used the magic word terrorism was shocking to me. Right, like that just jumped, the fact that that made it through like the entire comms filter, of all the people who tried to edit that document before it went out, I thought was a good sign.

• 9:20Speaker 03

  Yeah.

• 9:20Speaker 04

  And so the fact that they're going to focus on that is great. But you're right, on the other things where people are able to do, but this is the funny thing is, I think Facebook's in this position where they get blamed for things that are the, this is the emergent property of the fact that every adult in the world is about to have a pocket supercomputer with always on connectivity. This is just the truth, right? The truth is, is that anti-vax is going to exist. Terrorists are going to be able to communicate with each other. The question is, can you provide a service that is as responsible as possible? And where you mitigate as much of the harm as possible? That even if harms are happening elsewhere, that at least within the place that you're, you are operating, that you're doing the best you can? And I think on the direct stuff, they can, but I think on a lot of the content moderation stuff, they're not, and the place where that is going to be most harmful is probably not in the United States. It's going to be places like India, right?

• 10:05Speaker 04

  Like you already see this in India, with violence that has been sparked from rumors that are being spread on WhatsApp. Now, communal violence between ethnic sects in India is as old as India itself, right? The country was born into violence and a horrible split with Pakistan. Like, it's not something, but the internet makes this a bigger issue. But this is the flip side. What would it take? I want people in the media to start to engage with the issue here. Because you have to think, well, what would it actually take to stop people from spreading horrible rumors?

• 10:34Speaker 03

  Oh, I have such a good idea. So one of the things that these platforms could do is to stop providing viral sharing mechanics that enables someone who is, you know, participated in this centuries-long conflict from, you know, being a suggested group on Facebook, right? And we just saw this with the anti-vaxers. You join a new mom's group and algorithm says, hey, why don't you join an anti-vax group too?

• 10:54Speaker 04

  Yeah.

• 10:54Speaker 03

  And like, that was true until this week, right?

• 10:56Speaker 04

  Which is, and that is where there might be an upside to this move, is that this move probably eliminates a bunch of those different amplification. The two things that cause the most dangerous amplification on Facebook are the advertising platform and the recommendation engine. Partially because of just the Size of the amplification, partially because those are the two products that allow you to put information in front of people where they did not ask for it.

• 11:18Speaker 03

  Yeah.

• 11:18Speaker 04

  Especially advertising, like you can find strangers and put something in front of them. And so that's the riskiest, but depending on the structure of this, it could evolve in a way where that kind of stuff is much more difficult. And so that might be the safety upside. That's though why exactly the product design is going to be critical. And what I would like to see is I'd like this to happen in public.

• 11:38Speaker 03

  Right.

• 11:38Speaker 04

  Like, I don't think they should just go in a room and two years from now, it rolls out. Like, my suggestion to Facebook is there needs to be a series of conversations, not just in the United States, not just in the West, where you have all these people, because there's a lot of people are going to flip out. Law enforcement's going to flip out, but they're going to be against people in civil liberties groups. You're going to have various governments who want content moderation, various governments who think there's too much moderation. You've got to have this in a public manner. That can't just all be lobbying on the back end and then we see this weird Frankenstein come out and then nobody understands why they made the decisions. We should understand what trade-offs they made.

• 12:10Speaker 03

  Would all this just be easier if these companies were a lot smaller? Like Elizabeth Warren just said, like, why don't we just break these companies up?

• 12:15Speaker 04

  Well, so I got to give Elizabeth Warren credit in that her paper on this explicitly talks about antitrust being used for competition policy, which is what antitrust should be used for. So we've gone to this weird place where people with completely diametrically opposed motives all believe antitrust is the right deal. So you've got progressives saying, "I want more moderation on Facebook, break them up." And then you have the Breitbart crew saying, "There's too much moderation on Facebook, break them up." They can't both be right. And the truth is, is neither one of them is right. All of these dials exist, no matter how big your social platform is. The question is, is it one big guitar amp or multiple small amps? And I think there are trade-offs here. If it's multiple smaller companies, then you end up with more choice, right?

• 12:58Speaker 04

  And so you're like, this is the Disney World safe place, and then this is the wild, crazy 4chan place, and I can choose which one I want. The downside is, and when you talk about organized actors... The going smaller means the chance of catching them is much smaller. So, we're living in this world where we look at the 2016 election and Russian interference through the lens of the fact that we only know about Russian interference that was found and caught by companies that care, right? So, everything in the Mueller indictment about activity on Facebook, our team found and voluntarily turned over to the special counsel's office. That was not found by the government. That was not found by outside researchers. That was found by us. Other companies have not looked, or they've looked and not talked about it.

• 13:43Speaker 04

  And the truth is, is I was able to have a team that had ex-NSA Russian-speaking intel analysts. We had the ability to build a machine learning algorithm to look at every single advertisement run in the 2016 election. Reddit does not. Pinterest does not. 4chan doesn't want to, but they wouldn't, you know, even if they did, right? And so, that's the other thing that you trade off is if you have more, you can have more choice with smaller guitar amps. God, we're pushing that metaphor, right? You can have smaller companies and more choice, but the flip side is then the capabilities of any one company. And so, I think if we're gonna end up with smaller companies, then the other thing we have to do is we have to facilitate them working together in a way that they are pooling their resources in a way that's not happening right now.

• 14:21Speaker 03

  All right.

• 14:22Speaker 04

  But that's actually a very complicated legal issue, it turns out, too.

• 14:25Speaker 03

  And maybe we'll talk some more about what the government can do at the end. You know, there's also been this debate that I'm fascinated by. So Zuckerberg has this blog post, and some folks like me say, "Wow, this seems like a really big deal." He's saying the future of the company is around encrypted messaging. And then you have other people who say, "Whoa, whoa, whoa, let's not take this quite so seriously. This is basically just a product roadmap for messaging products." You seem like you took it pretty seriously. Like, you called this a "burn the boats" moment to me.

• 14:47Speaker 04

  Yeah, yeah, yeah, to quote from my favorite movie, right? You know, when Cortez reached the New World, he burned the ships. I avoided doing a Sean Connery impression there. You're welcome.

• 14:56Speaker 03

  But come up afterwards, and I'll do it.

• 14:58Speaker 04

  Well, I'll do it when it's not being recorded. Yeah, so I do see it as him being serious. I also don't think he knows what the outcome is, right? Mark Zuckerberg is sitting on more data about what kind of stuff people want to do online than anybody else on the planet, right? So it's somewhat hard for us to reverse engineer his incentives. But knowing that he knows exactly what people are doing of every category of people on the planet and what direction it's going. What this indicates to me is a couple of things. One, he's giving up on newsfeed and giving up on public, right? So apparently, his data from this decision shows that the "I want to be public or semi-public and to have lots of people be able to read my stuff", that that is a declining desire of consumers, right? He's giving up on the web. And so, doing end-to-end encryption in a web browser, effectively impossible.

• 15:43Speaker 04

  So unless they have some kind of massive crash project to change W3C standards and all this kind of stuff, realistically, they're not going mobile first, they're gonna have to go mobile only, which is pretty crazy. And he gave up on China, which I think is less about data and more about just realistically, under the Xi regime, everything is going the other direction. And there's just no practical way that, under any terms of any deal, would they allow Facebook to operate? WeChat is effectively a part of the state. The Ministry of State Security has people sitting at WeChat. You would not do any deal with any American company when you have that kind of power over your domestic audience. And so, I think those are the trends he's probably looking at. But I don't think he knows what the outcome is, because nobody's ever done this before. So we know point-to-point messaging, group messaging, and stories are things you can do over end-to-end encryption.

• 16:28Speaker 04

  Nobody ever uses it, but WhatsApp has a story, a Snapchat-like functionality. And so we know those things work. Nobody's ever built like a newsfeed, and I think just conceptually, it would be extraordinarily difficult to build something that is a full copy of Facebook that's end-to-end encrypted. But what you might be able to do is build something that's like Instagram, right? So data's available to you in an encrypted way, and all of the intelligent stuff Instagram does to figure out what you want will have to be done client-side. And then the other thing people haven't talked about that has been kind of a quiet thing is there's been a bunch of Facebook research papers on doing AI. In the ARM processor on CPUs, they actually have released a library to make this super fast. And so they've been researching, can you move a bunch of the machine learning stuff out of the super expensive Intel server into distributed into the phones? So if you

• 17:12Speaker 04

  add that plus that, you can probably get there. But from Mark's perspective, he doesn't know, and he's got 18 months of work already. So merging these name spaces, doing end-to-end encryption, my guess is they're going to focus on that. They're going to try to build everything to be extensible with open APIs internally. And then, once they know, like, we have this all merged, it's working, then he can start to push the Instagram feed, the Instagram public stuff, comments, groups, all those people to port over. The real question is what speed that happens? And as they give up this ad revenue, are they able to find other revenue to match it? Right. Right? And that's a super open question, which is why I think he just kind of committed the company to it. Because if everybody was worrying about that every day, they would get none of this done. Yeah.

• 17:53Nilay Patel?

  Hey, everybody. It's Neelai. We got to take one step away for a quick ad, and then we'll come right back to Casey and Alex Timas.

• 18:05Speaker 00

  Ask yourself, what are your best people spending their time on right now? Expense reports, receipt chasing, month-end close that takes weeks. You become what you spend on, and that's not what you're building toward. Brex is the intelligent finance platform that eliminates that work before it starts. AI agents that handle the manual stuff automatically so your team can spend their time on what actually compounds. It's time to get Brex AF. Learn more at brex.com/af.

• 18:41Nilay Patel?

  Hi, I'm Neelai Patel, editor in chief of The Verge, and Decoder is my show about big ideas and other problems. This week is part of our special series on the future of business. I talked with Matthew Prince, co-founder and CEO of CloudFlare, about how AI bots and agents are remaking the business model of the web. And how CloudFlare is trying to build a new economy around websites getting paid every time an AI crawler shows up.

• 19:02Speaker 05

  We polled all of that data and looked at it and said that, you know, it would be the second half of 2027, automated traffic will become larger than human traffic. I was stunned when, just a few months later in May, a team came to me and said, you won't believe it, but automated traffic is now past human traffic online. And that's just expanding like crazy. And if you extrapolate that out, five years from now, we think that automated traffic will be a thousand times human traffic online.

• 19:26Nilay Patel?

  This was a fascinating conversation, not just about the future of the web, but also about how one of the world's most important software companies is deciding which of its employees it can replace with AI. Listen to Decoder wherever you get your podcasts.

• 19:39Speaker 06

  This series is presented by EY.

• 19:44Speaker 06

  All right, back to it.

• 19:47Speaker 03

  Let me ask about content moderation. Even in this world where everything is end-to-end encrypted, Facebook is going to have to pay attention to this stuff that people are posting in the various feeds as they are today. I recently wrote a story about content moderators and some of the working conditions there, which are really, really rough. How should Facebook and other platforms think about moderation going forward? Should these employees be paid more? Should this be a full-time job?

• 20:08Speaker 04

  Yeah, so you can always pay people more. I think specifically for content moderation, you have to think about what the mental health impacts are. I thought your story was great and really helped outline those. I didn't have any content moderators working for me. I had a child safety team, I had a counterterrorism team, and the emotional and psychological impact on those people is pretty extreme, right? We were able to do a bunch of things to support them because they're full-time employees. The fundamental issue here is that the companies, I think they pretty much all do this, they use companies like Accenture to provide their content moderators, and you can't provide that kind of like intense mental health support through the contractual barrier of a different employer, right? Like Accenture is never going to do more than what is minimally contractually required to help their moderators. And so, in the long run...

• 20:52Speaker 04

  This will reduce the need for growth of content moderation. Maybe this change, but my guess would be 10 years from now, Facebook still has the same number of content moderators. They're doing different kinds of things. I think they're gonna have to bring most of them in-house so they can provide them with support, especially the people that work on like the really high-risk stuff. The harassment and bullying, the child stuff, terrorism, like looking all day at beheading videos really starts to mess you up. And there's actually some, there's been some psychology research in that people that work in these fields, it's sometimes gender biased, but like, men will get violent, there's more violence at home for people who work in these kinds of jobs. Women sometimes become cutters. Like, there's, there are outcomes, and it has been reasonably studied because you have the same problem with police officers, social service workers, for people who have to work with these horrible things.

• 21:34Speaker 03

  Yeah.

• 21:34Speaker 04

  And so the company, I think, does have a responsibility to do that, and they can't do it while these people are contractors.

• 21:38Speaker 03

  Talk about some of the things that you did for your people who were doing these really kind of high-risk jobs.

• 21:41Speaker 04

  Yeah, so we, when I was there, we built like a new resiliency function between a number of different, you know, my team didn't do it by ourselves, but we had people who participated in this. And so we brought in like a mental health professional. It turns out the laws here make it hard. Companies can't just employ psychologists and psychiatrists because of HIPAA in a law called a RISA. And so because of the way the laws work, you can't just hire like a PhD, have them on staff to see people as doctors, you have to come up with like a crazy insurance plan, and then you have to offer the insurance plan to all of your employees. It's kind of, it's, it is messed up. I think what I would like, honestly, and I'd like the company to come out and say that, because if Congress is looking for things to do, making small changes to employment laws to make it easier to give mental health services to people is the kind of thing that's actually positive.

• 22:22Speaker 03

  Yeah.

• 22:22Speaker 04

  You know, we did that kind of stuff, but you can't do any of that for contractors because it becomes what's called a co-employment situation, and so legally you just can't do it, and I think that's why they gotta bring them in.

• 22:30Speaker 03

  Right. Okay, so taking a quick setback from Facebook, last year you went and you joined Stanford, where among other things you're working on the Stanford Internet Observatory. Give us a sense of what you're working on right now.

• 22:39Speaker 04

  So, one thing's going on is effectively everybody in the political sciences and the social sciences wants to study what's going on in the internet. But the problem is, is, you know, let's say you're a specialist in the East Asia Studies section at Stanford, and they're studying what happened a couple of weeks, you know, over the last several weeks between India and Pakistan. You know, I know we were all distracted by the news, but we came close to nuclear war again, which is like a regular thing in the region.

• 23:03Speaker 03

  Yeah, which is like a Page 5 story in America.

• 23:05Speaker 04

  Right, yeah, yeah. So, if they wanted to go study this thing that's immediately happening, the activation energy to understand what is happening on Twitter, on WhatsApp, on Facebook is so high, they can't do it. And so our thesis is... I am not qualified to do that work, right? I am not a social scientist, I'm not political scientist. What I might be qualified to do is hire a bunch of engineers and data scientists, put the contracts in place for the API access and stuff, build the scraping systems for people that don't have APIs, build a data analysis pipeline, and then build a team of data analysts who don't need specialization in the social sciences, but can provide the help. And so if a professor comes to us and says, "Man, what's going on in India on Twitter?" We could say like, "Oh yeah, Sally can go help you with that." And Sally has access to an iPython notebook,

• 23:49Speaker 04

  and she's able to pull all this stuff and to do all the work, and then we can bind up. And my thesis is, if you make the activation energy low, if you make it simple to access those kinds of resources, that we will massively expand the use of it by social scientists, political scientists. And we want to do this first at Stanford, but maybe it's a model that other universities can do.

• 24:07Speaker 03

  Right, and social scientists are salivating for this, right? I mean, they're desperate to get this kind of information.

• 24:11Speaker 04

  Right, but it's super hard to find somebody... Like, just the structure of academia does not react well to this kind of stuff. It's very hard for them to find a grad student who goes to write R or Python, right? But they don't need it. They need a grad student who understands Indian politics and who can read Hindi and several other languages, and then we'll give them somebody who can write Python and then already have the data sitting up in whatever cloud provider wants to give us free cloud services. Please contact me later if you work at Google, Amazon, or Microsoft, that will already have terabytes and terabytes of data up there that they can go and look at.

• 24:39Speaker 03

  You know, we talk a lot about Facebook's role in saving democracy. I wondered if you would maybe speak quickly to what role you see for government here and maybe what you think the media should be doing, if you want to tee off on this a little bit.

• 24:50Speaker 04

  Right, so to get a little off my soapbox, there's three groups that really screwed up in 2016. Tech companies screwed up. The government and media also screwed up. And the way I imagine this is you've got Mark Zuckerberg in a hoodie, you've got a politician in a suit, and you've got a reporter with like a reporter press hat. And the guy in the hoodie is saying, "Oh man, we made a lot of mistakes." And the government and media people are like, "You're right, you really screwed up." Yeah, right? So the truth is, is there were three different branches of Russian interference in 2016. I'm gonna skip the attacks against infrastructure, because that's a whole nother thing. The two information operations, one of them was a pure social media information operation

• 25:32Speaker 04

  by the Internet Research Agency and a bunch of other organizations in that sphere. The org chart's actually quite complicated. That is mostly the responsibility of social media companies, to understand that, to understand that, and to put mitigations in place. There was another, and what is quantitatively, almost certainly, the more effective attack, was the attack by the GRU. So there's three intelligence agencies in Russia, main intelligence agency. GRU is the Russian, the main intelligence directorate of the Russian military. They work for the uniform Kremlin officers. They are the ones who broke into John Podesta's email, who broke into the DNC, who broke into Colin Powell's email. There's a whole interesting problem here of how do you protect, like, guys who used to be in the government, and now just grandfathers, but they have national security information in their Gmail account? So that's actually an interesting problem.

• 26:16Speaker 04

  They took that information, and then they planted it in the media, via personas on Facebook and other social media sites, and then the media did their job for them. So Politico ran a John Podesta email live log. The New York Times, the Washington Post, the Wall Street Journal, every major newspaper, covered the claims of the GRU uncritically. And the problem for the media is that claims were true. It was true that Debbie Wasserman Schultz wanted Hillary Clinton to win. It was true that John Podesta had a bunch of embarrassing things in his email. But those emails were leaked in a strategic manner, with the goal of changing American democracy, and you start to reach this, like, difficult place of, is this disinformation? Is it misinformation? It is certainly manipulation, it is certainly an information operation.

• 27:01Speaker 03

  Which we should say that journalists did not know at the time.

• 27:03Speaker 04

  Well, some of them suspected, right? And so in, like, the ninth paragraph of a Washington Post story, you might have, this might be a Russian information operation, but it doesn't matter, because people only read the headlines of the first paragraph.

• 27:13Speaker 03

  Right, so what do we do about that?

• 27:15Speaker 04

  That's a super hard problem. And first, you gotta protect the stuff. So you gotta protect John Podesta and such, you gotta protect the campaigns. So Bob Lord went over to the DNC, he's a very competent security guy, he's doing a bunch of work. I've been talking to him. People who are building campaigns campaigns are super hard in that you're like great we're gonna build an IT infrastructure for two years that's gonna be run by 17 year old volunteers and we have to stand up against kernels in the German military or the Russian military

• 27:41Speaker 03

  Maybe the Germans who knows

• 27:42Speaker 04

  Oh god, it's the Germans too. There's your headline Sarah, right there. I'm sorry, the Russian military. And then you tear the whole thing down. And so there are things that you can do there. I would like to see the media have one tenth of the examine themselves 10% how much they've examined Facebook and Twitter in this area, right? Like I'd like them to talk about internally and then publish publicly what happened in the editorial rooms when they decided we're going to massively amplify the claims of the GRU. And what I'd like to see is I'd like to see two or three newspapers and maybe NBC or like one or two TV networks. There's one, the big problem here is there's one TV network that would never do anything. But like for all the reasonable people to talk about what are our standards when

• 28:24Speaker 04

  information has been stolen and we believe we're being manipulated. And I am not a journalistic expert, but I

• 28:30Speaker 03

  But you play one on Twitter.

• 28:31Speaker 04

  I play one on Twitter. This is a lot of journalists play tech experts on Twitter. So we're all, we're all kind of pretending, right? I think there's a lot that could be done if you say we will cover things, but we won't over amplify them. We'll cover them once, we will not cover the scandal. Like newspapers love to do this thing where they're like horrible thing happened. Next day, people talking about horrible thing that we talked about yesterday, right? Now Congress is having hearings about horrible thing we talked about yesterday that we published the first day, right? Like, like you don't have to do the media a ladder of we're going to make this into a massive scandal just because it happens to be true. It's a hard issue, but it's something we got to deal with for 2020 because the companies have done a bunch of stuff around disinformation. The hack and leak,

• 29:10Speaker 04

  nothing has changed, right? If right now Amy Klobuchar's inbox was released to all the major media organizations, I don't think anything would be different.

• 29:18Speaker 03

  I think that's right. That's a perfect place to stop. You guys, thank you so much. Please give a round of applause to Alex Stamos.

• 29:24Speaker 04

  Thanks Stacy.

• 29:25Speaker 03

  Thank you so much.

• 29:28Nilay Patel?

  All right. So that was Casey Newton and Alex Damos, live at Southwest Southwest, talking about the future of Facebook. You should subscribe to Casey's newsletter. It's called The Interface. It is required reading if you are interested in social networks and democracy. We'll be back later this week with another regular Vergecast on Friday. And then next Tuesday, I'm interviewing Shoshana Zuboff, who wrote a book called The Age of Surveillance Capitalism, which is all about how our modern economy and our regular conception of how money works in this world has been totally upended by data collection. She's super fascinating. We'll see you on Friday and Tuesday.

• 30:03Nilay Patel?

  Hi, I'm Neelai Patel, editor in chief of The Verge. And Decoder is my show about big ideas and other problems. This week is part of our special series on the future of business. I talked with Matthew Prince, co-founder and CEO of Cloudflare, about how AI bots and agents are remaking the business model of the web and how Cloudflare is trying to build a new economy around websites getting paid every time an AI crawler shows up.

• 30:24Speaker 05

  We polled all of that data and looked at it and said that, you know, it would be the second half of 2027, automated traffic will become larger than human traffic. I was stunned when, just a few months later in May, a team came to me and said, you won't believe it, but automated traffic is now past human traffic online. And that's just expanding like crazy. And if you extrapolate that out, five years from now, we think that automated traffic will be 1,000 times human traffic online.

• 30:49Nilay Patel?

  This was a fascinating conversation, not just about the future of the web, but also about how one of the world's most important software companies is deciding which of its employees it can replace with AI. Listen to Decoder wherever you get your podcasts.

• 31:02Speaker 06

  This series is presented by EY.

• 31:05Speaker 07

  Your heart can tell you a lot about your health. Apple Watch Series 12 measures your heart rate every five seconds with the most accurate heart rate sensing and wearable. So your vitals app, now with heart rate variability, can tell you when something is off. And your readiness score can let you know when to rest and when to push. Hear the story in every heartbeat with Apple Watch Series 12. The features described are for wellness purposes only and not for medical use. iPhone 11 or later required. Based on Apple conducted study of heart rate accuracy, August 2026. Visit apple.com slash Apple Watch Series 12.