Hello, everyone. Thank you so much for coming to this evening's Commonwealth Club World Affairs program. My name is Mike Isaac. I'm a technology reporter for the New York Times.
Now, let me tell you a little more about Emily. She is a technology reporter for Forbes, where her TikTok coverage has won awards. A Harvard Law School graduate and a former criminal defender, she previously led the Plain View Project, an investigation into police misconduct on Facebook, and covered TikTok for BuzzFeed News. She's also a friend of mine, and I've known her for a long time, and I'm jealous of a lot of her reporting, so that's good too. Please join me in welcoming our speaker, Emily.
So I would like to start with what I think is kind of your untraditional background getting into journalism. I met you years ago, and I remember when we have an editor friend in common who became your boss. And I remember when he announced your hiring, I was like, wow, that's crazy. I can't believe she's doing this. Why would you get into journalism right now? But it worked out pretty well. And I'm just curious how you made it to this. Like, what compelled you to do this, especially with your background and really interested?
I thought I wanted to be a professional violist when I grew up, so. Well, I went to Oberlin, double degree in the college and conservatory. Spent a lot of time in the practice room, and realized I didn't want to play the works of old white men for old white men for the rest of my life. So I decided I want to go to law school, which I did.
And my first job out of law school was defending guys on death row. I loved that job. Thought I was going to do it for the rest of my life. It was really impactful. It was really meaningful. And in the course of doing that, in my first year out of law school, I was investigating a police brutality claim, which was a normal part of the job. And some of the police officers in this particular case that I was working on had public Facebook pages where they were saying all manner of things about their work, about the case that could have been important from a sort of post-conviction challenge perspective.
And some of the stuff that they were posting was clearly viral. There were these sort of angry cop memes, and I was like, whoa, what is this? How much more of this is out there? And so I ended up spending two years-ish of my life researching that question and building an interactive database of things that police officers said on Facebook, on public Facebook pages, that could impact public trust in policing. That was actually journalism when I was doing it, but I didn't know that. And so I was just this like lawyer who had a research project. But what I was doing at the time was investigative journalism. And it was a big, important project that raised a lot of important questions about policing.
Yeah, this was like I was working on it 2016, 2017, 2018. Project came out in 2019. So that was more than two years, actually.
Anyway, what that project made me obsessed with was not police misconduct. It was the effect that Facebook was having on our discourse. Because in order to do this project, I went through the Facebook posts of many, many, many of America's cops, thousands of America's cops, and most of the posts weren't about policing. They were just about these people's lives, right? And I was watching the ways that people were interacting with political and civic discourse over the course of thousands and thousands of posts about these people, like I was watching a diary of their lives. And that was really weird. It was probably inadvertently public, which was also really weird.
And I became obsessed with the question of like, how Facebook was shaping our discourse and our realities and how it was shaping. Policing was just a small part of that. So, I became obsessed with this question. After this project comes out, Facebook sort of could have tried to sue me or try to hire me, and they hired me. And so I went and worked in Facebook's policy department, which was a fascinating experience about which I cannot say all the things, but NDAs are great. Like, Mike covers Facebook, so he knows. But it was a fascinating experience.
And I got to think about the questions I wanted to think about. I got to think about the effect that Facebook was having on society and how it was leading people to have certain conversations or not have certain conversations, and why certain things were going viral. But at the end of the day, I was thinking about those questions on behalf of the Corporation, for the benefit of the Corporation. And what I really wanted to be doing was thinking about those things on behalf of the public, in part of a public conversation. And that's eventually how I got into journalism. So when John, our mutual friend, announced that he had hired me, I had gone on to do similar work at Spotify after Facebook, but I eventually applied and, you know, threw caution to the wind, wrote a really passionate cover letter and said, I want to do journalism. And John took a big risk and hired me. And here I am.
And it worked out. That's amazing. And so, like, so let's move to TikTok then. Facebook is obviously super interesting. We've both written about it, covered it, or thought about it, at least for a long time. And now it's kind of this interesting zeitgeisty shift around, like, what, one of the... I mean, don't get me wrong, Facebook is obviously very important and it covers much of the world, but TikTok seems to hold a very specific type of engagement with folks across the world, how they use it, and it came on like very quickly. But I want to shift back to the beginning of the book, 2014. There's this conference that's happening in China.
And it's sort of like a global summit of how the world leaders kind of think about the internet, essentially. And this is a different time in the era of the internet, where the idealism of, like, Tim Berners-Lee saying this is going to, you know, information is going to be free, everyone should be able to use it. This is like, you know, nations can come together and people can learn things. This was like, this sounded like normal right now. And now we're in this era of, like, balkanization. Let's say that's kind of like not how people think, at least the same way. But how we're, you know, as these kind of social apps were rising to power, kind of party officials across the gamut, whether it's in China or Americans kind of thinking about it. And maybe talk a little bit more about this summit that I'm referencing.
Yeah. So if we go back to the platforms of 2014, or just a few years before that, we saw the uprisings across the Middle East, the Arab Spring.
And we saw Facebook in the United States trying to carve out, like, how we are making a thing that brings people together in a sort of, per se, good for democracy and expression and freedom of democracy and expression. And I think a lot of people in the US and a lot of people in tech in the US felt that way. And they pointed to things like the Arab Spring to say tech is an engine for democracy.
And, well, Twitter's... Twitter made it.
Exactly. And world leaders in autocratic states were very nervous about that. They didn't like that. And they wanted to be able to control the internet. They wanted to be able to censor. And so at this conference, the Chinese government tried to get a bunch of private sector tech leaders from around the world to sign on to a sort of document that endorsed an idea of digital sovereignty, which is basically, here in our country, we write the rules of the road for the internet. Over here in your country, you write the rules of the road for the internet. And that was the Chinese government trying to get these private sector leaders to say, yeah, yeah, you can control technology here and you can control speech here.
And the private sector leaders balked. They wouldn't sign it. They were nervous. They really weren't really on board with what you're doing over here. And so the resolution never went anywhere. And the Chinese government just never talked about it again. But that sort of set the scene for a turning of the tide from this sort of optimistic view of tech platforms as engines of democracy to a question mark about whether they would ultimately be that or something else.
I mean, that's, yeah, I don't even know if you were at Facebook when they had Free Basics, that program where they were trying to go into India and provide internet to everyone. And like...
I knew about it, I don't remember. Yeah, I wouldn't work on it, but yeah.
But it's one of those things, right?
Well, it's one of those things where it's almost the wide-eyed, optimistic view of statecraft, right? Of like, okay, as long as we can provide cell phones and internet to people and like this connectivity is a de facto good for everyone, right? Or maybe that feels naive. A darker version of that, where we're going to bring all these people online and they're going to be reliant on us.
And I don't think it's wrong to say Facebook was thinking about it that way at that time. I'm not even imputing that to Facebook. I'm just saying that is true too, right? You can have wide-eyed optimism, and you can also have the convenient fact that wide-eyed optimism is also a bunch of people being reliant on you.
I mean, yeah, both things can be true. I agree.
And I feel like there's always like the sort of, I don't know, I think a lot of the time tech leaders out here can be like, all right, there's like self... I'm forgetting the word, but like a kind of altruism that's self-reinforcing, if that makes sense. And there's nothing that like young, idealistic tech people want to believe more than that. You can do good while doing well.
Yeah. Yeah, I think that's a thing that people want to have happen out here.
That's right. Enlightened self-interest is the phrase I was talking about. And that's probably like the nicest possible way of putting it. But so that was 2014-ish era. That was a long time ago. Things are obviously very different right now, but I want to kind of talk about the intro. Does anyone know what the app Musical.ly is? When I say that, a few people, some nodding. Okay. So that was... can you tell us what that is?
Yes. So Musical.ly is an app that you probably knew about if you were a middle schooler in like 2017, or if you were the parent of a middle schooler in 2017. It was the center of preteens' world for a couple of years there. It was an app that you would go to to record lip syncs and dances to top 40 hits. And if you were like old enough to drive a car, this thing was not cool. You did not go, right? It was not for you. But even kids as young as like 8 or 9 were obsessed with it. And that app was eventually purchased by ByteDance, which is TikTok's parent company today, and revamped and re-released as TikTok in 2018.
And that's one of those things where I'm really interested in Musical.ly, because at the time, exactly what you said, like the big tech companies, the Facebooks of the world and Twitter and Google were all focused on like long-form, highly produced shows and things like that. Like short-form video kind of hit them out of nowhere, right? Am I wrong with that? Like, it felt like a very different thing than what Silicon Valley was doing.
Yeah, I think that's right. And I think part of the reason is that they didn't... I think a number of Silicon Valley companies were worried about making something for people so young.
But Musical.ly became immensely popular with this very young user base. And they were, by default, people talk about how TikTok is so much more authentic. Nine-year-olds are by default authentic, among many other things. They are authentic. And I think that kind of came through, and through a lot of detailed work on... Musical.ly was the startup before it was acquired, on the startup point, to really try to understand the users and give the users what they want. And even when those people are like 12 years old, I think paid dividends because they were able to really create a unique thing. It was very narrow, right, but a unique thing that everybody knew. Oh, that's a Musical.ly, right? It's not just a video that you recorded. It's like there was a distinctiveness about it that sort of gave it an identity.
And so I'm trying to lay the groundwork here because this stuff leading up to where we are now, I think, is really important. And this is where ByteDance comes in. And let's talk a little bit about Zhang Yiming and ByteDance, this company that, let's say, ten years ago, maybe no one over here would have understood, maybe a few people probably have known it. But like most of the United States, I would say, had no idea. And tell me a little bit about him and maybe this company and where it kind of came out of, basically, and why we had to start hearing about it.
Yeah. So I think today a lot of people would say that Zhang Yiming is one of the founding fathers of the modern recommendation algorithm, and he was obsessed with the idea of recommendation algorithms really early, back when the first ones were starting to pop up. And he had this lens of thinking about algorithms. He wanted to replace the idea of you going to the computer and searching for information with the idea that information could search for and find people instead of people searching for information. And the idea was, if you have indexed all the information and you understand it, and you understand all the people and what they're interested in, what they like, you can serve them stuff. You can show them stuff that they don't even know that they should be looking for, but that they really want to know.
Yeah. And like in an idealistic way, that's an interesting idea. Maybe that's a more efficient way of transmitting information. It also has a kind of dystopian cost to it, where why would I cede to some corporation or some math problem the agency of deciding what information I want to know? That's kind of crazy, but we've all moved remarkably in that direction since 2012 or 2014 because we do less of going to the computer, looking for a thing, getting that thing and leaving, and do a lot more of open the computer and see what comes to you.
Right. Like we used to use the thing. Now the thing is using us a little bit more.
That's, I mean, I really think that's important because it does underscore what getting online used to look like, right? Like even going to Google was a very intentional... or if you remember AltaVista, if you remember like search portals from AOL and Yahoo and things like that, right? It's a super intentional act where you... and like search strings weren't all perfect. You could find like funky stuff. Search results were not great all the time, but like that was very much how we navigated the internet, right? And this shift to, I say the algorithm with like big scare quotes, but like shift to like algorithmic curation and serving things, I think like it seemed to kind of, you know, we talk about it with like scare quotes now, but there was some pathway there, right? Like there were, I guess you could call Facebook was definitely like a precursor to this, or the News Feed, right? You went to people's walls and then you didn't go to their walls anymore and stuff just came to you.
Yes. And that even then, this is also why it's sort of shocking to me that Facebook felt very late to what TikTok is doing, right? Because they saw that just a reverse chronological timeline wasn't good enough. Twitter was not competing in the same way that Facebook was.
But like that's the... so was he the one that was kind of instrumental in seeing this is the way the world is going? It seems super prescient.
Yeah, yeah. He was very prescient. He was early to this idea. He was passionate about this idea, and way before ByteDance was on the radar of most people, even in Silicon Valley, it was massively impactful and important in China. And it became known as a sort of app factory, right? They had tons of apps. They still have tons of apps. They launch apps all the time. A lot of them don't work, and then they discontinue them. But they throw stuff at the wall all day, every day, trying to get the next biggest thing. And their first biggest success in China was not short video. It was a news aggregator app called Toutiao, meaning 'total' with today's headlines. And this news aggregator app was also one of the first really aggressively, algorithmically curated, sort of personalized bottomless feed type of products.
And that was kind of scary to the Chinese government because Zhang Yiming was out there saying, we're going to serve you the news that you want to see, we're going to figure out what you want to see, and then we're going to serve it to you. It's all going to be personalized, it's going to be great. And the Chinese government is like, wait, we want you to see the news that we want you to see. And that sort of early on put him and his sort of ideas about recommendation algorithm into tension with a government that was nervous about that.
It seems super Western. Like, did he have any Western education, or was he just like... he seemed pretty... if that's your framework, if you're like, you know, this is, you know, the state-enforced point of view or whatever, those seem like pretty radical ideas under that framework, you know.
I don't know if I just... he certainly had a sort of radical devotion to using math and logic to advance the ways that we think about stuff. I think that is really, really core to who he is and how he thinks. And it always has been.
Yeah. And yeah. No, I mean, maybe it's just like a... I don't want to, you know, romanticize this guy too much, but maybe it's like, this is the internet, right? Like, how does the internet work? How should information be found or find people or whatever, you know? Like, I don't know, I think like a lot of, non-pejoratively said, tech bros...
I think he's just... it's just a noun now.
Right, right, right. I don't mean anything by that.
But I think like a lot of tech firms, he's really into efficiency.
He's really into like, how can I get the optimal result out of this? Well, let me play out all of the, you know... like he's very sort of scientific about it. I think like many of the tech founders that we see right here in Silicon Valley.
Yeah. Like, how can we make this work better.
Right. And to the extent that something like censorship is inefficient, like that's annoying. That's in the way of this, you know, more efficient, more rational solution. So okay, so we're in the... let's say we have moved pretty quickly into the age of the algorithm or whatever, you know, like our Americans... I always try to figure out, it's hard for me to get a grasp on how Americans like feel about this. There seems to be like a level of ambivalence. Like on one hand, it's like, I don't want to be controlled. I don't want, you know, someone to tell me what to do. And then I'll go home and watch like five hours of TikTok or something in bed, or like something my wife will walk in and like just see a blank blaring screen while I'm trying to sleep. So like, how... I know it's very embarrassing, but how do you feel? Like, let's just stick to Americans, like just sort of like picked up on this pretty quickly. Like, did we adopt it en masse? It sure seems like it. Are people... like maybe we had time with it and now we feel a little weird about it, you know what I'm saying?
Yeah. I think Silicon Valley talks a lot about friction, like removing friction from products. Like the Amazon Buy Now button is so beautifully efficient.
If we all buy now, more things than we used to.
And the idea of removing that much friction, I think as consumers we walk through the world and we love the frictionless experience. And then we regret the frictionless experience. And I feel like that's what's happening, right? Like you sit there and you open TikTok or you open Netflix or whatever, then three hours later you're like, that didn't feel good. I'm not happy about how I just did that. But you love the frictionless experience in the moment. And so I feel like now we're in this place where like many people will put timers to like limit themselves on certain apps.
People will lock their devices. You can get a brick thing for your phone. Now you can get a pouch that will turn your phone off.
I was on one of those the other day.
Like they're all kidding. It's very important. There are all kinds of things that we're now doing to introduce friction back into our lives because we realize that actually we want some friction. The Buy Now button is not the best.
And so I think it is now sort of the onus is now on the consumer to stop and introduce friction, whereas it used to be less efficient to buy things, to spend the time and watch the ads.
Right. And that's a win for Silicon Valley or tech in general, no matter where it's from.
No, that's a really... I mean, the shift to consumer is super interesting because normally regulation would play a part in the whole thing. And that's kind of... it's a whole other... it's another part of the interview, I guess.
But so here's the... okay, here's my experience with TikTok, and I'm trying to divorce it from my early experience with Facebook, because I've been on Facebook so long, I don't know how many things I've clicked yes to just to skip and keep going. But when I would sign up for TikTok, the point I'm getting to is around data and data collection. When I would sign up for TikTok, the number of notifications asking me for my contact list, for like the... I need to be able to send you a zillion notifications at a time. Like, it just feels more encouraging to me than, let's say, some Silicon Valley apps that I'm familiar with. Is that me overstating that, or would you say it's on par with that? Because I don't know. I've had Facebook so long I can't remember what permissions I've given it, but TikTok feels particularly intensive.
I think another thing that we're experiencing is a reduction in time. Like, how many permissions have you given Facebook over ten years? And then TikTok grew so fast and happened so fast that it may be a similar number of permissions that are just like reduced. I don't think TikTok's types of data collection are fundamentally different from most other tech giants. One thing that TikTok does do that I think makes it feel different sometimes is the extent to which it relies on our revealed preferences instead of our stated preferences.
What does that mean? Can you touch on that?
On Facebook, you friend people, you follow accounts, you basically tell it what you want to see. Same with, you know, something like YouTube. You rely on the search bar much more, right? On Instagram you have your friends. There's like explicit, this is what I want to see, this is what I want, or whatever. And even if you like, like or share stuff, right, that is a more explicit signal. TikTok, you open and it happens to you. You do nothing. Except the thing is, you are doing things because you're deciding exactly how long before you scroll. You're deciding what to skip, you're deciding what to come back to. And the platform famously relied much more heavily on those implicit signals. It's like, oh, you linger for two more seconds on videos about addiction. Or like I said, sometimes they're not things that you thought you were telling it, but you ended up telling it. And so people would say, oh my goodness, TikTok knew I was pregnant before I was pregnant, or TikTok knew I was gay before I was gay. And it's like you are inadvertently revealing things just with where your attention is held or not held. And I think that makes it feel more encroaching sometimes.
It's where it actually goes back to my favorite conspiracy theory of Facebook is listening to me on the microphone all the time because I got this ad when I was like not expecting it or something. And like the point I always make to people is like, it doesn't need to listen to you on the microphone because it already has a bunch of collection that's going on all the time, and the implicit signals.
A super interesting way of putting it.
So, slowly building to our current moment. But I'm curious when... So I remember a New York Times column a long time ago saying TikTok is like the last fun place on the internet or whatever. And there was a period where that was how people felt, right? Like it was like, oh my God, this is delightful. I'm seeing people dance. I'm seeing dogs and cooking, whatever. And I felt the same way. And then all of a sudden, or maybe not all of a sudden, it turns into this is the tool that is going to infiltrate and destroy America. And maybe or maybe that's not true, but like, how did that shift start happening? Was it a DC thing? Was it other companies, sort of like lobbyists sort of coming in? Like, what did that progression kind of look like?
It was a combination of things. There was a group of China hawks in DC who felt that way for a long time before it blew up.
And like Marco Rubio was on this train, like basically as soon as Musical.ly rebranded as TikTok, he was like making these posts being like, TikTok is awfully fun, isn't it? It's also a Trojan horse. Like he has a long track record of like, he was really about that. But that viewpoint was sort of confined to a relatively small group of hawks until at least after the pandemic, when TikTok became way, way more central to American life. TikTok took off during the pandemic in a way like growth that you've sort of never seen in any other company in Silicon Valley. And that sort of speed and scale.
And that's just we're all bored and stuck in our houses. And it was different. And it was like, you know, you could do it socially distanced. It was great.
But I think that's part of it, right? That all of a sudden this thing that really was just kids like dancing and lip-syncing became it. It became much more a place where people were talking about everything, right? It became much more a place where adults were talking. It started to be a thing that politicians used carefully, like to campaign and like promote themselves.
And there was Donald Trump's first attempt to ban it, and I don't know...
if that's before or after the moment when you feel like all of a sudden, there was this sort of huge anti-TikTok movement. I think it was like leading up to that.
Yeah. I mean, so a bunch of mostly teenage K-pop fans decided to reserve tickets to a Donald Trump rally in 2020, okay, and not attend that rally. And I don't know that that's when Donald Trump first learned what TikTok was, but it seems definitely possible that that is the time that Donald Trump learned what TikTok was, and soon thereafter, he announced that it was a huge national security threat and he was going to ban it. I'm not going to allege causation there, but correlation is strong.
I mean, never anger the K-pop fans, right? Is that like the rule?
It's definitely the rule of thumb. And that's actually like in 2020, this was not the only incident where people angered K-pop fans and that had serious repercussions. I mean, during the Black Lives Matter protests, the K-pop fans flooded the hashtag White Lives Matter with dancing memes. They were just like, no, if you're going to search for White Lives Matter, if you're going to try to do that, all you're going to see is K-pop dancers. And like, these were like 13 to 15 year olds essentially running like low-key influence operations on the internet for fun.
Oh my God. I mean, that is not violent resistance. I guess that's pretty amazing. So like, let's talk about you, because you end up playing a role as this stuff starts happening. When did you decide TikTok was going to be a reporting target for you, or how did that sort of get on your radar as a reporter?
