Meta's Stock Decline, TikTok's Power, and Funding Transparency — With Frances Haugen

Frances Haugen is an ex-Facebook employee who shared thousands of internal documents with journalists in 2021 that revealed much about the company's decision-making and values. Haugen joins Big Technology Podcast for a live conversation at Unfinished Live in New York City that covers social media product, policy, funding, the stock market, and TikTok's power. Join us for a lively interview featuring debate, nuance, and a few good laughs. Learn more about your ad choices. Visit megaphone.fm/adchoices

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

• 0:00Alex Kantrowitz?

  This episode is brought to you by Genesis. What does it actually take to put Agentic AI to work across an entire enterprise? I recently spoke with Genesis chairman and CEO Tony Bates about why orchestration is emerging as a competitive advantage in customer experience and what companies need to get right in the AI era. Then I sat down with Adam Mitchell, head of enterprise business solutions at Voya Financial, to go inside a large-scale Agentic CX transformation, including the business decisions, trade-offs, and lessons along the way. You can watch both conversations now on my YouTube channel today. It's never been more challenging to get frontier compute. How do you know where to look, who to trust, or how much to pay?

• 0:43Alex Kantrowitz?

  Akash is an open cloud that provides easy access, flexibility, and peace of mind. That's how companies like Razer were able to cut their inference cost by 15 times. Visit akash.network/alex for a free consultation. That's akash.network/alex for a free consultation.

• 1:03Speaker 01

  LinkedIn Presents.

• 1:19Alex Kantrowitz?

  All right, everyone. Welcome to the Big Technology Podcast, a show for cool-headed nuanced conversation of the tech world and beyond. We are live here at Unfinished Live in New York City. And just to show you that there's actually a live audience before us, I'm going to ask them to let us hear it. You guys gotta be loud because these microphones won't pick it up unless you give it your all. Let's hear you guys.

• 1:45Alex Kantrowitz?

  All right. We're going to keep that energy up through this whole thing, all right? I don't want to see anyone walking out.

• 1:49Speaker 02

  You don't always get a laugh track. So you've got to lean in while you have one.

• 1:53Alex Kantrowitz?

  That was good. We're going to, we'll put that, any good moment along the podcast, we'll put that. Now, everybody boo, no, I'm kidding. We're here with Frances Haugen. You might know her as the person behind the Facebook files, the Facebook papers. I don't quite know which name to use, but they're the series of revelations that showed a lot of things about Facebook's business that really were not great. And those revelations happened a year ago, and they've sparked a series of discussions and much more informed conversation about the platform than I think we had previously. As a journalist, I'm really grateful to have been able to comb through some of the material that you put out there because it finally allowed us to put some data

• 2:37Alex Kantrowitz?

  and some actual internal research behind these things that people were trying to, you know, shoot from the hip and say were happening, and we didn't really know whether that was true or not. So first of all, thank you for revealing all that stuff. Second, let's just take a look at Meta's stock price right now. I took a look, about September 13th is when the beginning of the Facebook File series of articles came out in the Wall Street Journal. Since that week, Meta's stock, it's now Meta, Meta's stock is down 59%.

• 3:08Speaker 02

  Wow.

• 3:09Alex Kantrowitz?

  So.

• 3:10Speaker 02

  I only check it occasionally. I have one Twitter follower who messages me every time it goes up, and he's like, "Oh, no, the stock is recovering." Then I look at it again, I'm like, "Oh, it's actually down a lot."

• 3:20Alex Kantrowitz?

  Yeah, that's called unnecessary paranoia.

• 3:22Speaker 02

  Yeah.

• 3:25Alex Kantrowitz?

  So I guess, looking at the way the stock has performed, mission accomplished?

• 3:31Speaker 02

  You know, sometimes people ask me questions like, you know, "Do you think you won?" And I feel like that really misses a large part of the criticism I've given, which is, I feel like I'm fairly characteristic for whistleblowers in that when you look at it statistically, overwhelmingly whistleblowers are conscientious people who actually care a lot about their companies. And I really believe that Facebook has problems, that the only way it can successfully solve them is by working on them collaboratively. And so my intention was never to have the stock price go down. My hope was that Facebook could get the support it needs to be long-term successful.

• 4:12Alex Kantrowitz?

  And?

• 4:13Speaker 02

  And unfortunately, the stock has gone down because I think the public has looked at how they responded. Like, the biggest thing they did fresh out of the gate was they're like, "Actually, we're a metaverse company now." And in general, I think investors get quite scared when you say you have a giant problem, and they're like, "Ooh, shiny." You know, it's not really a great forward-looking indicator.

• 4:33Alex Kantrowitz?

  And this is also coinciding. It was the top of the tech market cap era. It's crashed. We've had, there are other factors involved. We also have had the Fed raise rates and large declines.

• 4:44Speaker 02

  But it's fallen substantially more than the other social media companies.

• 4:47Alex Kantrowitz?

  Right. Do you attribute that to your revelations?

• 4:50Speaker 02

  So in our SEC complaint, we said Facebook's evaluation. So I actually went and looked back at five years of Facebook's stock price. So I'm a data scientist, and so I view the world through data. And I looked at every time, I went and actually went and Googled, like specifically to target Googling, for every time Facebook's stock price declined more than 5% versus the NASDAQ over a five-year period before I came out, or before the first article came out. And there's a very limited number of times that happened, and there's about 27. And overwhelmingly, the things that happened when the stock price declined more than the NASDAQ declined were Facebook had a decline in users, Facebook had some kind of security incident where they had to come out and say,

• 5:33Speaker 02

  we're going to have to spend more on safety, or it was implied that we're going to have to spend more on safety. And something like on the order of 65% of the time, when you had those deviations from the NASDAQ. Our SEC complaint has said the price was overvaluated, was overinflated, because users used the product who would not have used it had they known the truth, advertisers advertised who would not have advertised had they known the truth, and Facebook underspent on safety more than they would have then if they had been transparent. And since then, we've seen a decline in users. We've seen them having to increase the amount they spend on safety by two to three times, which is huge compared to the past times they've increased safety.

• 6:14Speaker 02

  And so we can, we'll never be able to know exactly what fraction of this is because of the transparency. But I think there's a good chance that the information did cause the stock to correct back to something closer to its natural valuation.

• 6:29Alex Kantrowitz?

  A skeptic would say that Apple's anti-tracking moves probably had a larger impact. What do you think about that?

• 6:36Speaker 02

  I think unquestionably they had an impact as well. I would never want to assess what is the relative contribution, because as quantitative people are naturally

• 6:47Speaker 02

  I always try to be very precise in what I say, and I have no data to indicate what the relative contribution is.

• 6:53Alex Kantrowitz?

  Okay, it'll be interesting. I mean, I know you can't speak about the SEC or discussions. Kara Swisher tried that at Code, didn't get anywhere. So I won't try to press that.

• 7:03Speaker 02

  I actually just don't like that.

• 7:04Alex Kantrowitz?

  Yeah.

• 7:04Speaker 02

  Like, they've talked to me once.

• 7:06Alex Kantrowitz?

  Okay.

• 7:06Speaker 02

  And apparently, that's fairly standard.

• 7:08Alex Kantrowitz?

  Right.

• 7:09Speaker 02

  They usually try to not talk to witnesses too many times, because if you ever don't say exactly the same thing, that can actually make it harder for them.

• 7:19Speaker 02

  And they usually go dark for long periods of time while they're building their case. And so I genuinely don't know much about what's going on.

• 7:27Alex Kantrowitz?

  Okay. So as Facebook struggles, whether it's because its stock price has declined due to these revelations, or whether Apple is kneecapping it, or maybe Apple's having more success because of the revelations that you brought out, there's this other, there's a countercurrent to that. And that's that competitors get stronger.

• 7:46Speaker 02

  Mm-hmm.

• 7:47Alex Kantrowitz?

  The one that really seems to be doing quite well right now is TikTok.

• 7:51Speaker 02

  TikTok.

• 7:51Alex Kantrowitz?

  Is TikTok. And, you know, a while ago, people would say, okay, well, the fact that Chinese companies could take the place of Facebook is just a big tech talking point that they're trying to, you know, divert attention to their own actions. But now, it really is happening. And so what have you, what do you make of the fact? You know, is there any credence to the fact that as a company like Facebook gets weaker, a company like TikTok can emerge and actually leave us in a worse position than we were previously?

• 8:20Speaker 02

  So I think TikTok's emergence is a great example. So throughout, you know, I think basically all of capitalist history, we have seen time and time again that if consumers get to choose between products that make them happy and products that make them sad, as a general rule, people choose products that make them happy over products that make them sad. And when I look at the things that make TikTok sticky, you know, I think there are genuine questions we need to ask, right? So part of the reason why TikTok makes you happy is TikTok emerged, the fact that the only successful competitor to Facebook emerged from the only place Facebook wasn't allowed to play. Right, like Facebook went into all the other corners of the world and sometimes even bribed the locals to say, "Let us become your internet," right?

• 9:05Speaker 02

  They paid for people's data to get people to converge onto it. The fact that it emerged from China meant also, the only competitor that could emerge was one that met Chinese goals. And China created a social media platform that focused on dancing, on comedy, on lighthearted things, because those things are often not the same as politics, right? And at the same time, they've created a product that makes people feel good. And so if you have a situation like that, someone was eventually going to come along with a product, or even be real. Being real makes people feel better than Facebook does. And so I think that's the real thing.

• 9:44Alex Kantrowitz?

  You looked at a lot of Facebook research. Some of that research said that people feel worse if they scroll and don't participate. Just passive consumption makes people feel worse. If that was the internal Facebook research, how do you think it is that TikTok, an app that literally has people scroll passively, is making people feel good?

• 10:04Speaker 02

  But they actually don't. Like, you have to choose to move on with TikTok. TikTok, if you are truly passive, like you're not engaging at all, if you're not choosing to do your next gesture, it just replays the video, right? And I think it's also a thing of TikTok's algorithms, because it is, like, I think I keep trying to flag for people, is TikTok is a very manipulative experience algorithmically, right? They bias towards happiness, they bias towards humor, like, they don't want you talking about sad topics, because political topics often are sad topics, right? And so, you know, funny dances, if you can be passively, you can be not really engaging, not having a conversation, and watching funny things, and you're going to feel laughter, laughter makes you feel better.

• 10:49Alex Kantrowitz?

  Is that what your TikTok feed is, mostly funny dances?

• 10:52Speaker 02

  You know, I get a lot of home restoration. I, like, TikTok has decided I really like cake decorating. I don't know why, I can't have gluten, who knows? They really think I really like cake decorating, and they really think I like renovating homes. So that's like a large fraction of my feed.

• 11:11Alex Kantrowitz?

  I don't know what this says about me, but mine is all, like, couples finding their partner cheating.

• 11:16Speaker 02

  Really?

• 11:17Alex Kantrowitz?

  Yeah. They're great videos.

• 11:19Speaker 02

  You know, you could thumbs down some stuff.

• 11:23Alex Kantrowitz?

  I like them.

• 11:24Speaker 02

  Oh, well, you know.

• 11:26Alex Kantrowitz?

  So anyway.

• 11:27Speaker 02

  Go with God, my friend.

• 11:29Alex Kantrowitz?

  Indeed, I will. So let's get back to that question, though, about whether we're worse off with TikTok dominating. I mean, yeah, OK, it's making us happier. The other side of it is...

• 11:38Speaker 02

  Feel happier, yeah.

• 11:39Alex Kantrowitz?

  Feel happier.

• 11:40Speaker 02

  I don't know if it's like existentially happier.

• 11:41Alex Kantrowitz?

  That's true, yeah. We could go back.

• 11:43Speaker 02

  I think sucking our time in, we'd probably be happier gardening, but you know.

• 11:47Alex Kantrowitz?

  That's true. So, but speaking about that, OK, let's say feel happier. But it's doing keystroke logging, or it's at least alleged to do that.

• 11:58Speaker 02

  In its in-app browser.

• 11:59Alex Kantrowitz?

  Yeah. And, which is a pretty big deal. And also, we don't fully know the extent of the oversight that the Chinese government may or may not have.

• 12:09Speaker 02

  Yeah.

• 12:09Alex Kantrowitz?

  Over that app. So if American... I know we're going to get into global stuff. Let's talk about US. If American social media users... You know what? Let's just go global. If global social media users are in an app that has these major liabilities, isn't that worse than being inside Facebook?

• 12:29Speaker 02

  Inside Facebook? Well, I think it's this question of, you know, like what we... I think we need to have a... I think we should have Facebook of 2008. Like, if I was going to... If we're going to do any kind of magical hypotheticals, you know, social media that is about our family, our friends, our communities, connecting with people is better than anything where you have hypervirality. The things that are good about TikTok are TikTok acknowledges they have power. And Facebook really doesn't want to acknowledge they have power, right? They like to be like, oh, no, we can't do anything. Like, we have no ability to do anything. The secondary thing is, you know, if we had to choose between, you know, an app that has had two genocides,

• 13:14Speaker 02

  so, you know, we have 200,000 people dead in Myanmar. This is not my opinion. The UN has a multi-hundred-page report detailing what Facebook did that was negligent. If we see the same thing happen again in Ethiopia with lots of warning, if we see this happen over again, I think there's this question of, at some point, we have to decide, when do we have to hold Facebook responsible for being an app that seems to have a pattern of causing genocides? Because I have not seen any TikTok genocides yet.

• 13:42Alex Kantrowitz?

  Well, not yet.

• 13:43Speaker 02

  Yeah, not yet.

• 13:46Alex Kantrowitz?

  There have been genocides before Facebook.

• 13:48Speaker 02

  That's true.

• 13:48Alex Kantrowitz?

  So do we now, in the age of Facebook, where do we now put the blame for all genocides on Facebook?

• 13:56Speaker 02

  No, I think there's ways of acting responsibly in this space. So for example, you know, in 2002, excuse me, 2020, 2020, 2020, in April 2020, in preparation for the US election, they held an internal working group, it's like 60 people, I was one of the people on it, who were tasked with coming up with, quote, "soft interventions" to prepare for the election, right? So a soft intervention is something like you, instead of taking down a count, taking down a post, you say, "Hey, maybe you should pop up a thing saying, hey, you know, you didn't click on a link, you didn't actually click through on this link, do you want to click on it first?" Because like, putting a little moment of intentionality before someone reshare something

• 14:40Speaker 02

  actually substantially decreases misinformation. You haven't censored anyone, it's like a soft intervention. And one of the things that was flagged was, there was a single parameter in meaningful social interactions, so back in 2018, Facebook switched from just seeing how long they could keep you on Facebook, to seeing could they drive you to do interaction? And they called it "meaningful social interactions." So stuff that got more interactions was considered better content, even though six months after they did this change, they polled people and said, "Hey, is your home feed more meaningful?" And people said, "No, it's not, it's actually less meaningful." There was a single parameter in, we had noticed for at least at that point, 18 months,

• 15:21Speaker 02

  when this report came out, that a single parameter, which was if you don't just give content credit for generating interactions, like let's say I'm trying to decide, should I show this to you, you know, my little virtual model view says, "How likely are you to click? How likely are you to reshare? How likely are you to, you know, put a comment?" But you know, if you were to reshare that, all your friends might click on it, all your friends might reshare it, all your friends might comment. We should give it credit for that. We should give it downstream credit. The only problem was, they knew that people didn't like reshares. That they've done these beautiful studies. Like, one of my favorite graphs from Facebook shows that you like original content from your 20 closest friends. And Facebook has a ranking of all your friends. They know which ones are your favorite friends, at least on Facebook.

• 16:06Speaker 02

  You like original content from your 20 closest friends, but you only like original, uh, reshares from your five closest friends. Because people don't like reshares. So content that can actually get someone to engage after a reshare has to get over the bias of not liking reshares. And so it turned out to be a scoring parameter that overwhelmingly rewarded only really extreme content. And the study came in and said, "We've known this for a year and a half. We know that if you take this factor out, you get way less hate speech, way less, like, literally violence, violent content, bloody content, violence inciting content. Maybe in places that are considered at-risk countries, we should not include this scoring parameter." It doesn't decrease how long you stay on the site. It doesn't decrease in the sessions. It doesn't decrease ad dollars.

• 16:50Speaker 02

  But it will hit our core metric of meaningful social interactions. Which people's bonuses are tied to, by the way. And literally in the notes, her remark, it says, "With regard to going broad, with taking out the scoring parameter, that we know increases violent content, hate speech, all these things. If it hits core MSI, we're not going to do it." Should they be responsible for that?

• 17:13Alex Kantrowitz?

  Yeah, I think they should.

• 17:14Speaker 02

  Yeah.

• 17:16Alex Kantrowitz?

  Okay, then there's the question of how much do they contribute to the genocide? But I think even contributing a little bit to a genocide is something you want to stay away from.

• 17:23Speaker 02

  Or to do it that shamelessly. Like, do you really want a company where you get said, "You're not going to lose any money. You're not going to lose any sessions. All the metrics that are seen externally, they're not going to go down. And you know there's going to be less hate speech, less violence, less violence inciting content, it's a separate category. You know all those things. But you might have to, like, figure out a different way to compensate people. And your CEO doesn't even entertain, "Can we figure out a way to do it?"

• 17:48Alex Kantrowitz?

  Yeah. And this is, by the way, something I've been reporting on and speaking about for a while, which is the harmfulness of the share. Cher, I think, and I'd love to hear your perspective on this because we're talking about what I think is the deep reshare, reshares of reshare.

• 18:03Speaker 02

  No, this is just single reshare, single reshare.

• 18:05Alex Kantrowitz?

  Okay.

• 18:06Speaker 02

  I have a deep reshare.

• 18:07Alex Kantrowitz?

  So, yeah, well, we don't even have to go.

• 18:08Speaker 02

  I feel strongly about deep reshare.

• 18:09Alex Kantrowitz?

  Okay, yeah, we can go, we can go shallow reshares and deep reshares. But I always felt that the ability to share something on impulse, that's news, never leads to good information distribution because it goes from, takes people from being thoughtful, "Hey, I read this story, I think you should, you know, check it out," to seeing in your feed that your political opponent is, you know, a space alien Nazi child killer, eater, and being like, "Oh, yeah."

• 18:43Speaker 02

  They're also a lizard person.

• 18:44Alex Kantrowitz?

  Lizard person.

• 18:45Speaker 02

  Yeah.

• 18:45Alex Kantrowitz?

  And you hit share.

• 18:46Speaker 02

  Yeah.

• 18:47Alex Kantrowitz?

  And never, if you have to put this under your own identity and copy and paste that link, you're gonna be much more reticent to do that than to, you know, just hit the share button and have somebody else's, you know, avatar show up in your friend's feed and be like, "Aha, I just did my part for the cause."

• 19:06Speaker 02

  My parents are both life scientists, so my, my mother's a cell biologist, my father is a pathologist, but he studies viruses. And my mother, you know, you can see the wisdom of my childhood. My mother used to always say, "Be careful what you select for because you'll get it." Really, that's the story of cell cultures. You have to be very careful about keeping things clean because like whatever you select for, that's what you're gonna get. You know, when we started talking about freedom of speech, like when the Constitution was written, if you wanted to distribute an idea, it cost money. You know, you had to buy paper, it took a lot of time, it took a lot of effort. There was a natural filter that if you wanted to say something, you had to invest in saying it. I think asking people to have a moment of intentionality,

• 19:50Speaker 02

  like a copy and a paste, is not a huge amount of cost, but it adds some cost. So you don't get unintentional sharing.

• 19:59Alex Kantrowitz?

  Yeah, and every social network seems to know this. Clearly, Facebook knows this. The research that you shared showed it. Twitter knew it inherently. That's why, before the 2020 election, added the speed bump that you had to, you know, "Do you want to quote tweet this tweet? Do you want to click the link before you reshare?" Yet they're so addicted to the virality. Why?

• 20:21Speaker 02

  I think it's a secondary thing. So I'm founding a nonprofit called Beyond the Screen, because right now, we are limited to what we see on our screens, and we need to see beyond them if we want reform. You know, I think this question of like, "Why did Twitter do it and Facebook didn't?" is such an interesting question, because it's like a really cheap way to get like 10 or 15% less misinformation. I think the reason why Twitter did it and Facebook didn't was Facebook's average user is substantially less literate than Twitter's. I know it might not seem like that. But like, given the global distribution, you know, there's huge numbers of Facebook users that are literally becoming literate to use Facebook. It is the internet where they live. You know, 80 or 90% of all the content

• 21:06Speaker 02

  in the majority of languages in the world is only on Facebook. And when you look at the outlier countries, there's some really good graphs in there that show, you know, there are countries where 35% of everything that's in the newsfeed, every impression, is a reshare. And so, when you have such a large fraction of your content is just reshares, any friction you add, any moment for pause or intentionality, substantially decreases the content in the system, which decreases consumption. It's a money thing.

• 21:36Alex Kantrowitz?

  And by the way, if you pull the reshare back or put this speed bump that causes less resharing, you don't need to engage in these messy content moderation issues that get you accused of censorship.

• 21:49Speaker 02

  That's my whole point, yeah. That's my only point since my first Senate testimony, contrary to the conspiracy theories.

• 21:56Alex Kantrowitz?

  I mean, it's so obvious.

• 21:58Speaker 02

  Yeah. Anyway. You spend a lot of time thinking about this, so it's obvious to you, but it's not necessarily, because we don't study these things in school, it's not necessarily obvious.

• 22:07Alex Kantrowitz?

  No, it's good to be talking about it in public, inside the companies.

• 22:10Speaker 02

  Yeah.

• 22:11Alex Kantrowitz?

  I mean, the person that built, and I'll get off my soapbox on this in a moment, but...

• 22:15Speaker 02

  But I'm enjoying it so much.

• 22:16Alex Kantrowitz?

  Okay, well, I'm gonna keep going then. The person that built the retweet inside Twitter, I spoke with him, and he said, "Look, it's like I handed a machine gun to four-year-olds."

• 22:27Speaker 02

  Wow, that's a line.

• 22:28Alex Kantrowitz?

  Sort of like what it feels like. Anyway, that's my hope is eventually, because of revelations like yours, and because of, I don't know, maybe conversations like this that have people inside the company listening, we start to revisit this. But, you know, we may, we may end up having product evolution before we even need to get there. I mean, we were moving to these short-form videos. And by the way, I'm curious what you make of this. We just had Brandon Silverman on the show, who was the founder of Crowgtangle, which is this great software tool that let you see what was moving inside Facebook, public links in particular. And he made a great point, that Facebook actually is deprioritizing

• 23:08Alex Kantrowitz?

  news within its newsfeed. And, you know, now we don't see as many news links. I'm curious what you think. Isn't, is that something to be applauded?

• 23:19Speaker 02

  So, part of what frustrates me so much about that is, like, the reason why they're deprioritizing news is because they're scared, right? Like, they're getting all this criticism about things like, you know, ethnic violence. And they're like, they're kind of, you know, taking a card from TikTok, where they're like, you know, TikTok, if you're not a funny dance, you're not, like, humorous, if you don't like, aren't, like, you're not gonna get distributed. And the thing that's so sad for me is it really illustrates, you know, you can either choose to design for human-scale, human-modeled communities...

• 23:55Speaker 02

  Right, so I always say like, is it a dinner party? Is it a cocktail party? Is it a church parish? Is it a conference? Is it a university? You know? We have, we have models on how people can exchange ideas. They've lasted for thousands of years. Um, I, or you end up with AI censorship. And the problem is, uh, if you must insist on content moderation, I think you should have to be transparent about it. And that means things like, you should have to put out samples of the output of your scoring systems. So we can see what's getting labeled as any of these categories and what the consequences of that is. Because when we take down, when we don't let news get distributed, that's news by the definition of Facebook's AI.

• 24:39Speaker 02

  We don't know what doesn't get through or what does get through. And when I talk to activists around the world that work on issues like, um, women, like violence against women, gay rights, they say across the board, our accounts all get taken down. Because these systems that are meant to keep us safe, because they're not done in a very high quality way, counter-speech looks like violating speech. And, and so I, I, I a hundred percent am against badly done content moderation. And I think we need to have some pretty large societal conversations around like, you know, if we must insist on doing content moderation, we need to be doing it in a different way than we're doing it today.

• 25:16Alex Kantrowitz?

  Yeah. And as many, I, I feel like as many system changes we can make-

• 25:22Speaker 02

  Yeah.

• 25:23Alex Kantrowitz?

  -before we get to that decision-

• 25:24Speaker 02

  Totally.

• 25:25Alex Kantrowitz?

  Stay up or-

• 25:26Speaker 02

  Yeah, I love it.

• 25:27Alex Kantrowitz?

  -that's probably better.

• 25:28Speaker 02

  Yeah.

• 25:29Alex Kantrowitz?

  So, let's talk about another, I, I love how we're going into product.

• 25:33Speaker 02

  Hmm, I always love-

• 25:34Alex Kantrowitz?

  I've wanted to talk to you about product for a long time, so I'm glad we're doing it. Um, another thing that we should talk about is, um, algorithms. Algorithms on the news feed. And just a little interesting history. I was reading the documents that you made available to reporters. And there is one internal study that said they removed the ranking algorithm from the news feed.

• 26:00Speaker 02

  And then made it worse?

• 26:01Alex Kantrowitz?

  Made it worse.

• 26:02Speaker 02

  And he wrote about this. He said, "I got it wrong."

• 26:04Alex Kantrowitz?

  I didn't say you got it wrong. Did I say you got it wrong?

• 26:07Speaker 02

  You said that she didn't see this document, or she didn't like blah, blah, blah.

• 26:10Alex Kantrowitz?

  I don't think I said that, but I could be wrong. Anyway, we're here to heal.

• 26:14Speaker 02

  Yeah, we're here to heal. Good. I'm all about truth and reconciliation, so I'm down.

• 26:19Alex Kantrowitz?

  I appreciate you sitting down with me. And I think that at the time I heard from you saying that, like, "Look, you're not looking at this."

• 26:27Speaker 02

  Politically.

• 26:27Alex Kantrowitz?

  Yeah. Not, yeah, you didn't say I was wrong, just you said, "There's some context you might want to take a look at." And just to set the story straight. It was a study about what happens when you put the ranking algorithm on the feed, which means that Facebook's algorithm is making a decision of what you see first, as opposed to the algorithm showing you what's happening in a reverse chronological order, and just showing you the most recent first.

• 26:54Speaker 02

  Let's actually, let's like make that a little more concrete for people who... There's a lot of big words in there. So, when we open our email, the thing that we see first is our most recent email, unless we have some kind of fancy AI email inbox. We could have a social network where we got content from, you know, things we chose to get and it showed up, you know, in reverse time ordering, just like our email.

• 27:20Speaker 02

  And or I say have a time order for email, but same kind of thing. Like, there's different ways we can be doing this. And one of the things that's true, there's a document in the disclosures that say, "Hey, we tried getting rid of the ranking, like the algorithmic ranking, you know, where we try to decide what's going to be better for you." And guess what? We see more bad things. We see more violating content. We see more hate speech. And I think one of the most important things for people to understand is that back in 2008, when we had chronological ranking, you know, you saw things in the order they came in. It was kind of like your inbox or a reverse of your inbox. So, when you knew you'd run out of new things to see, we had no big groups. Right, there were no million-person groups on Facebook.

• 28:04Speaker 02

  Facebook had not been pushing you for years into groups that, you know, you may not have been that interested in, or those groups were intentionally selected because they were groups that got a lot of interactions. For context, like in 2016, they found out that 65% of people who joined neo-Nazi groups in Germany, where they are illegal, like you can't have a neo-Nazi Facebook group in Germany, joined them because Facebook suggested that group to people. Over time, though, Facebook needs you to stay on the platform for longer and longer. And your friends and family let them down. They didn't make enough content for you to get longer and longer sessions, so you could click on more and more ads. We cannot have million-person groups that are basically just like fire hoses without having algorithms.

• 28:49Speaker 02

  Because that group makes 1,000 posts a day, and you flood your feed. And what the article found was if you get rid of the algorithmic ranking, people's feeds, their feeds get flooded by these high-frequency groups. You no longer see your friends and family. And so you can't really think about the idea of taking the current product that has all of these baked-in assumptions and just removing the algorithm. You have to really design for safety holistically.

• 29:12Alex Kantrowitz?

  And so your point is that if we were to go to the most recent first, as opposed to the algorithmically-selected stories, that the company would have to redesign its product.

• 29:25Speaker 02

  Mm-hmm.

• 29:26Alex Kantrowitz?

  It wouldn't be, you wouldn't run that experiment on the current product.

• 29:29Speaker 02

  Exactly.

• 29:30Alex Kantrowitz?

  You would rebuild the product, which I found interesting.

• 29:33Speaker 02

  Yeah.

• 29:33Alex Kantrowitz?

  And I think that's good context.

• 29:35Speaker 02

  You'd have groups that were like Discord servers. So you can have a Discord server that has 100,000 people in it, and when the conversation gets too loud, you break off into little subgroups. You can imagine a world where maybe your group's content never got inserted into your feed. You had to navigate to your group. So you can still have your cancer-support group. You just wouldn't have it insert all the content into your feed.

• 29:55Alex Kantrowitz?

  And you're not worried that the feed would be overrun by spammers who just like... Post every five minutes and try to get stuff in front of people?

• 30:04Speaker 02

  So I like to call, my favorite ranking, no one else calls it this, but I got to brand it because I made it up. I call it chronological plus plus. So in computer science, the way you add one if you don't want to write like n plus, or n equals n plus one is used to n plus plus. Is, imagine if you had a little bit of demotion by frequency. So it's largely chronological. But if you send 20 posts in a day, you know, each progressive thing demotes you down a little bit. So it's loosely chronological. And I think the better way to describe like what I would want is describable ranking. Right? That right now we have unintelligible ranking.

• 30:49Speaker 02

  We have a black box neural network and we don't know what the model really is optimizing for. You could imagine having a publishable algorithm where you said it's basically time-based. We try to stay as close to time-based as possible. But if you post five times a day, if you post three times a day, you know, we might move you down a little bit. You'd be less likely to get seen.

• 31:06Alex Kantrowitz?

  And that's the description. But there are people who are saying we should remove protections from these platforms that use ranking algorithms to show feed, to show you content. And even using chron, what did you call it, reverse chronological plus plus?

• 31:21Speaker 02

  I just said chronological plus plus.

• 31:23Alex Kantrowitz?

  You are ranking in some way.

• 31:26Speaker 02

  But I'm doing it in an intelligible way, a way that we can have a conversation on it.

• 31:28Alex Kantrowitz?

  That we can.

• 31:29Speaker 02

  Because right now, we never get to have a conversation on how it's prioritized.

• 31:32Alex Kantrowitz?

  Yeah, I understand.

• 31:32Speaker 02

  Yeah.

• 31:33Alex Kantrowitz?

  And then, but because that's still going to be ranking, it would still fall under the, these companies can be liable, so.

• 31:40Speaker 02

  But my complaint on the current way we rank

• 31:43Alex Kantrowitz?

  Right.

• 31:44Speaker 02

  Is right now, we have these ranking algorithms that are optimized for business objectives. And you could imagine a world where you said it's going to be chronological, or it's chronological, or you have to give us a defendable reason why improved safety. Right? You can't tell us you made more money on it, like that's why you did it. You have to like show us the thing and show us like what was the safety objective. And you have to do it publicly and show us consequences so we can have a conversation on it. Like that's a very different world than one where, um, you know, they, they, they, uh, one of the problems that's outlined in the documents is that safety teams would work for months and months and months to figure out a way to make the product safer but not hurt ads. And because the model was unintelligible,

• 32:27Speaker 02

  people are just like throwing new factors at the model and would often undo changes that have been made. Because you could represent the thing that had been pulled from the model or the fix that had been done by munching together four or five other features because people just didn't understand how they interacted.

• 32:42Alex Kantrowitz?

  Yeah, and I like this proposal. So, but does that, does that then mean, um, companies should, should still not be held liable? What do you think about this meme? Because this, this whole idea of go completely reverse cron, pure reverse cron, without any filtering at all, even, well, explained or not, and otherwise you're liable for what you suggest?

• 33:02Speaker 02

  Can I give another counter, counter offer?

• 33:05Alex Kantrowitz?

  Yeah, yeah, let's hear it.

• 33:05Speaker 02

  So, so let's imagine instead of like demoting people who post too frequently, imagine if you posted more than three times a day or maybe even twice a day, your post got collapsed down to a single line that said Frances Haugen, posted two times already today, do you want to read the post?

• 33:20Alex Kantrowitz?

  Okay, I like that.

• 33:21Speaker 02

  Because then it's like a little shame, we're still doing chronological, but you're like, hey, you kind of talk a lot. Have you ever thought about like listening a little, you know?

• 33:29Alex Kantrowitz?

  I was trying really hard to, to have you say that this, um, making, making platforms liable for going pure reverse cron is an unworkable idea, but we got, we got pretty close.

• 33:40Speaker 02

  I think, I think, I think the way we should talk about liability is the way we talk about it with cars.

• 33:45Alex Kantrowitz?

  Okay.

• 33:45Speaker 02

  Right? Where it's this question around, you know, um, people should have to demonstrate that they are actively pushing towards safety. And, you know, if you have a hundred chances at bat, And when we look at, you know, like one of the things I've suggested is right now, every time they run an experiment, there are these dashboards. And the dashboards have hundreds of metrics that can be sliced in a ton of different ways. Imagine if they had to publish the dashboards. They don't have to tell us where the experiment is. They don't have to give up the IP. You know, they don't have to tell us about their magic secret sauce, but they have to show us every single time they run an experiment and release the data so that we can analyze whether or not there's a trend. Where, you know, consistently, they had chances to make us safer and they chose not to do them.

• 34:29Speaker 02

  And they had a whole bunch of chances that made it more dangerous for us, but it made them more money and they shipped those. Like, show us what experiments you run and show us what you ship or don't ship. And if we see a pattern of behavior, we should hold you responsible for those consequences.

• 34:43Alex Kantrowitz?

  I think that's much better. I'm all on team transparency. And for context, I appreciate you unpacking this. I will link this podcast underneath that story.

• 34:55Speaker 02

  Oh, thank you.

• 34:56Alex Kantrowitz?

  So whenever someone reads it, they'll get the broader discussion.

• 35:00Speaker 02

  You can even do it at the top of the article so they'll actually see it.

• 35:02Alex Kantrowitz?

  Consider it done.

• 35:05Alex Kantrowitz?

  All right. Let's take a break. We're here with Frances Haugen. She's the founder of Beyond the Screen. And the person behind all these big Facebook revelations that we're speaking about will be back right after this. This episode is brought to you by Avpoint. Everyone's racing to roll out AI right now. Co-pilots, chatbots, agents doing real work. But here's the part nobody loves talking about. All that AI runs on your data. And most teams have no single way to see it, secure it, and prove it's under control. That's exactly what Avpoint does. For 25 years, they've been the trusted layer beneath the world's most demanding data, now extended across your entire AI estate.

• 35:46Alex Kantrowitz?

  Your data, your cloud, and the agents acting on your behalf. It's how more than 28,000 organizations deploy AI with confidence. So innovation scales without scaling risk. Avpoint, the unifying trust layer for AI. Head to avpoint.com to see how enterprises deploy AI with confidence. Learn more at avpt.co/bigtechnologypodcast. That's avpt.co/bigtechnologypodcast.

• 36:18Alex Kantrowitz?

  Violent threats against executives are growing at an alarming rate. When a company experiences backlash, executives are the first people blamed. When their address and other personal details are sitting online, this backlash can lead right to their front door. The team at Ironwell knows this better than anyone. They've protected some of the most targeted executives and individuals on the planet for almost two decades. As a wide-glove enterprise security service, they protect your people with continuous personal data removal, proactive prevention tools, and human-led emergency support. So when someone goes looking for your executives, Ironwell ensures they hit a dead end. So here's what to do. Go to ironwell.com/bigtechnology, fill in the quick form, and request your free risk assessment. Their team will show you just how exposed your executives are

• 37:03Alex Kantrowitz?

  and how to lock it down before a threat reaches their front door. That's ironwell.com/bigtechnology. Learn how to stop online threats before they become real-world attacks. The link's in the show notes.

• 37:23Alex Kantrowitz?

  And we're back here for the second half of Big Technology Podcast. We're here with Francis Haugen at Unfinished Live. By some miracle, or actually probably the draw of what Francis has to say, our second-half audience is actually larger than our first-half audience.

• 37:37Speaker 02

  Substantially larger.

• 37:38Alex Kantrowitz?

  Much bigger. This has never happened for a podcast that I've done before live. So, anyway, let's hear it from you guys. Thanks for showing up, because it's gotta be loud.

• 37:55Alex Kantrowitz?

  Thanks, everyone, for being here. All right, let's get into the second half. One interesting thing you said in the first half that caught me by surprise was that TikTok acknowledges its power and Facebook doesn't. Really?

• 38:07Speaker 02

  Yeah, I mean, TikTok is designed to be censored, right? Like, we've had lots of scandals around this, right? So, how many people remember, I don't know if it was a couple of years ago, there's like a makeup tutorial where she's like, "This is how you do a smokey eye," and like, blah, blah, blah. "Oh, and by the way, the Uyghurs are being killed." Does anyone remember that?

• 38:25Alex Kantrowitz?

  I do.

• 38:25Speaker 02

  Yeah, good, good. So, the reason why she did the smokey eye for like 15 seconds and then started spilling the tea on, you know, they're killing a million people, or they have a million people in concentration camps, is TikTok manually censors all their highly popular content.

• 38:43Speaker 02

  And, you know, I don't agree with any system where they take down lots of content and they don't tell us why or how, where they don't show us the biases in what they take down or don't take down. I highly don't support that. But they acknowledge the idea that social media has power and that social media has social implications. And I think the fact, like, they've publicly talked about the idea that they watch for people who fall into rabbit holes and they try to pull them away from those. Do they actually or not? I don't know. They're not transparent either. But they understand the idea that they have an impact in society. And so are they rooting for our team? I don't know. But acknowledging power, I think, is better than pretending, you know, the world is flat and, you know, everyone's

• 39:28Speaker 02

  the same and you're just a mirror and everything bad has always been bad and, like, now you can see it and that's why you're mad.

• 39:33Alex Kantrowitz?

  Would you rather have a platform acknowledge power but actually, like, work to do things that undermine our society or one that doesn't acknowledge its power but

• 39:44Speaker 02

  Undermines our society.

• 39:46Alex Kantrowitz?

  Is trying? I don't know.

• 39:48Speaker 02

  I would be highly hesitant to say that Facebook is trying, right? Like, Facebook, if Facebook earnestly was trying, you know, when they got asked, "Please share the research with us that you have internally on whether or not kids are killing themselves because of your product," they would not have given back a three-page letter saying, "What research?"

• 40:13Alex Kantrowitz?

  Yeah.

• 40:14Speaker 02

  They would have given over their research.

• 40:15Alex Kantrowitz?

  They lied to the Oversight Board also.

• 40:17Speaker 02

  Yeah.

• 40:17Alex Kantrowitz?

  Like blatantly lied.

• 40:18Speaker 02

  Blatantly. So for context for people, when they asked the Oversight Board, "Hey, like, you know, was taking Trump down the right thing? Can we have a conversation about it?" The Oversight Board came back with, like, 15 or 16 questions and they were on a variety of things. And Facebook wrote back with answers. And one of them was, they asked, like, "Hey, this crash-check program, this idea that VIPs get a special pass, like, you say you treat everyone the same, but then you also say, like, the reason why Trump hasn't got in touch is because, like, he's in crash-check. Can you give us some more context on that?" And they said, "Oh, we've already explained what crash-check is," and they linked to a blog post from 2018. And that blog post said things like, "This is

• 41:01Speaker 02

  not a two-tiered system. This is just, like, getting a little extra double-check to make sure our policies are accurately being enforced."

• 41:10Speaker 02

  And in reality, because they weren't willing to spend a very small amount of budget, they didn't actually put that second check in for most safety systems. So for most safety systems, people got whitelisted. So if you were in cross-check, those safety systems no longer applied to you. And even for the ones where it did apply, because they were willing to allocate so little budget to this VIP program, it would take like three days before a human would look at your content. And so people would get about 75% of all the distribution they would have gone anyhow because of that underfunding.

• 41:44Alex Kantrowitz?

  Okay. What do you think of the Oversight Board? I mean, the meme about the Oversight Board among Facebook critics is that it's a public relations stunt that's a shield that, you know, will make it look like Facebook actually cares what the public has to say, but it doesn't. Do you subscribe to that?

• 42:01Speaker 02

  I find the name deeply, deeply, deeply manipulative. So if we were to ask, like, what does the Facebook Oversight Board do? You know, if you believed in the theory of what's on the labels in the can, you'd be like, oh, I mean, they do oversight, and they look after Facebook as a whole. It really should be called the Content Moderation Appeals Board, because that is the entire scope of what they can do. They can't ask about the performance of the content moderation systems. They can't even ask, like, what language are the content moderation systems in? They can't ask about how systemic or like, are there other examples of the ones that get appealed to them? They can't do anything. They can't ask about the algorithms in general.

• 42:44Speaker 02

  And so my heart really goes out to the Oversight Board, because it is full of a number of very conscientious people, who I think genuinely try very hard. But because they are so limited by what Facebook allows them to do, they get put in a very hard position.

• 43:02Alex Kantrowitz?

  What do you think about the Metaverse?

• 43:04Speaker 02

  Ah, the Metaverse. So the Metaverse will make certain problems easier and some problems harder, right? So on the side of reducing risk, you know, we talked about before, your friends and your family are not the problem on Facebook. Hyper-virality, million-person groups, five-million-person groups, in a lot of African countries, the single largest news source in the country, or maybe the only really mass news source will be like a five-million-person Facebook group.

• 43:36Speaker 02

  That's the problem with misinformation, that we have a selective fire

• 43:42Speaker 02

  megaphone where 1,000 posts come in, but only two get delivered to you, and they're the most extreme two, right? That's the problem. In the metaverse, it's much more about one-on-one interactions or small group interactions. And so you're going to see probably less things, less things like misinformation problems. The big problems with the metaverse are things like, you know, studies out of China have shown it's more addictive, or more habit-forming, like substantially more so than phones, because it's immersive. I think there's going to be a bunch of psychological things that we're going to be unpacking over the next few years, things like you have a socially-isolated child who, you know, maybe they are of a lower socioeconomic tier, they can't afford after-school activities, maybe

• 44:25Speaker 02

  they have a single parent, they don't really like their life, and they come home and they put on their headset, and they have a way cooler apartment, their friends are more beautiful, have cooler clothes, they do these fun things. But at the end of the night, they take their headset off, and they look in the mirror, and they no longer like, they don't like the person they see, right? The idea that you would, the only person you liked being was a virtual avatar, I think is just, it's heartbreaking to me. And I worry that the metaverse is going to become this stopgap for us, where we say, instead of like, making sure our old people, our elders are like integrated into our communities and have adequate resources, we can like plug them into the metaverse and leave them alone. Or instead of building community centers and after-school activities, we can, you know,

• 45:09Speaker 02

  leave our kids to go into there. And so I worry a lot about these kind of systemic problems. And we know, we know that they haven't invested in safety again. So when it comes to things like hate speech, we're creating these spaces where, you know, if you can hear 40 people speak, but you can't, you know, pause the tape and roll back two minutes and realize the person who's been shouting racial epithets at you is that one. You basically can silence the ability of minorities or women to be in these spaces because they just get yelled at and they get pushed out. And so I think there should be upfront investment in safety and shared systems across the games because it is very clear that the reason why Facebook wants to go to the metaverse is

• 45:51Speaker 02

  they want when someone comes to them and complains about getting groped or complains about getting harassed, that they can say, "I'm so sorry, that's really tragic, but we didn't write that app, we just made the hardware."

• 46:05Alex Kantrowitz?

  So you think, it sounds like you think the metaverse is going to happen because that's still an open question.

• 46:09Speaker 02

  I am a follower of Moore's Law, so I think the headsets are going to get cheaper, I think they're going to get faster, they're going to get lighter, they're going to get higher resolution. I have trouble imagining that 10 years from now, we won't spend more time than we currently do in the metaverse. Do I think Mark's vision of spending 12 hours a day in the metaverse will come true? No. I think Mark does that, Mark thinks that because Mark lives that. But not all of us have to avoid cafes because people glare at us when we walk in. So, you know, we have different incentives.

• 46:44Alex Kantrowitz?

  Another reason why Facebook wants to go to the metaverse, okay, maybe there's the, you know, we don't need to be responsible, but the real reason is that they are an app on Apple's phone, an app on Google's phone.

• 46:55Speaker 02

  They want it on the phone.

• 46:56Alex Kantrowitz?

  They want the operating system. No doubt about that. An interesting thing has happened where Apple has asked its users if they want to cut off Facebook's ability to track them off the app, and people have overwhelmingly said yes. Now, an interesting thing has happened after that, where Apple is in the middle of doubling its ad sales team, or its advertising team, not sales, the team that builds ad products there. I gotta ask you about this. If the business model drives the problems that we're seeing, and Apple is kneecapping Facebook's business and, in some ways, emulating it, don't they put themselves at risk of falling into some of the same problems

• 47:41Alex Kantrowitz?

  that we see Facebook facing today?

• 47:43Speaker 02

  That is such a great question. I do think that they are playing with fire. Like, I think it's always hard when you have an economic incentive that is not... So, right now, Facebook's or Apple's economic incentive is safety, right? They make their money selling new devices. So, if you get more privacy and having more privacy makes you want to buy their phone more, they can make more from their phone. I think they are mixing their incentives. And I have not sat down and, like, fully wargamed out that. But I... So, I would want to be cautious to speak with too much confidence, but I can totally imagine that might become a problem in the future.

• 48:24Alex Kantrowitz?

  I think it's a major problem for them. And we're just scratching the surface there.

• 48:28Speaker 02

  Yeah. It's definitely worth talking about.

• 48:29Alex Kantrowitz?

  Yeah. Let's talk a little bit about your nonprofit. I want to ask this question beforehand. I read this interesting quote from you that said, like,

• 48:40Alex Kantrowitz?

  "Your greatest hope is that you're not relevant..."

• 48:42Speaker 02

  I won't be needed anymore.

• 48:43Alex Kantrowitz?

  Not needed anymore. But, at the same time, you're working on a book. You signed with CAA, which is a big Hollywood agency. So, how do you square those, your hope and the action?

• 48:55Speaker 02

  Well, I... Right now, I don't see very many people who are on the stage who are proposing and pushing for things like, you know, "Let's solve the product, not the content." At the same time, I'm someone who's almost died, right? Like, I spent two months in the hospital when I was 29. I am fully sensitive to the idea that, like, life is fleeting and precious. And the way we solve real problems in ways that have the highest chances of working is you cannot be relying on individuals, right? We have a hit-by-the-bus problem right now. Like, we should be wary of, like, Facebook employees driving buses around New York, right? Like, you can imagine nothing that would be more convenient for them.

• 49:36Speaker 02

  We need 100,000 people who can speak as detailedly about the choices we have with regard to social media as I can, right? A world where there's only one person talking this way is not a safe world. And I think anything, I had a lot of people pushing me really hard when I first came out to be like, "Give us the five-point plan. How do we fix Facebook?" And I have said repeatedly, any world where, you know, I am the solution is just as autocratic as the world we live in right now with Facebook. And I really believe that the process of democracy is about having a lot of different people with different interests and different stakeholder groups, having to hash out something that they're willing to feel comfortable with

• 50:21Speaker 02

  together. And right now, the people who get to sit at that table are a very, very small number of people. Like, there are hundreds of people in the world right now who understand these systems at the level of detail that I do, right? The interaction between the algorithms and the product features and how all these things feed back on each other, how they differentially impact different kinds of people and when. And that is terrifying to me. And so, we need a world where there is not, and we need to think about a continuum of expertise. Like, we need, we need, right now, it's not just that the only security people get trained inside the companies, right? It's as if SpaceX was dependent on training its aeronautical engineers in-house. If SpaceX was training its aeronautical engineers

• 51:04Speaker 02

  in-house, we would not be going to Mars. Period. The reason why Twitter can't afford safety people is because there's so few of them, and Twitter can't pay high enough. All of us are endangered by the fact that we are reliant on these companies to train the experts. But if we look on the policy side, it gets even more scary. Because just like those security people are trained in-house, the only place where you become fluent in these policy issues around social media is like working on these issues on the Hill. You're a legislative aide, your congressperson really cares about it, you start learning more about it, you get more fluent, and as soon as you start to be a little effective, you get poached off by Facebook, you get poached off by Twitter, you get poached off

• 51:44Speaker 02

  by another tech company. And in the couple of weeks before my Senate testimony, we met with an aide from Blumenthal's office and one from Blackburn's office, and shared with them, we'd given them the SEC filings and all the files associated with it. And we asked them, like, "Hey, like, you know, this is a really good conversation. Is there anyone else on the committee that would benefit from having these documents?" And I don't remember which aide it was, but one of them was like, "Yeah, you know, I get why that would make sense, and I want to say yes." And I would say yes, if we could go more than two weeks without having a staffer from the committee leave for Facebook. Two weeks, right? And that's a huge problem.

• 52:27Alex Kantrowitz?

  Did you see it? This just happened, but it's a revolving door problem across government. They've hired people from the FTC. There was a banker that was, no, bankers were testifying before Congress and one of the members thanked the bank for hiring one of their staffers. And the bank was like, "Her father already works here." It's amazing. It's a problem. The revolving door is a problem.

• 52:54Speaker 02

  But when it comes to banks, like, you know, you can get a finance policy degree. Like, if you want to go get your PhD and specialize in like, how should we regulate banks, you can go and get probably a free degree in public policy in that subject. You can probably get grants for it. You cannot take a single class, a single class, where you really understand, here's the spectrum of choices we have, you know, here's the very wide eclectic palette of colors we could paint with to build social networks, and here's the consequences of those choices. You cannot take a class, if you're a poli sci person, or even a CS person, anywhere in the United States right now, anywhere in the world. And that's unacceptable.

• 53:35Alex Kantrowitz?

  And I've heard you talk about the fact that you want to build a simulated social network so people can train on it. You also want to train lawyers who are interested in bringing the action on the ins and outs of these networks so they sound smart about it. Is this what Beyond the Screen is going to be all about?

• 53:51Speaker 02

  So Beyond the Screen has two core products, or two core things we're working on. One is a project that we got seed funding for, and we announced yesterday. So the McCourt Institute is graciously giving us our first seed funding for something around duty of care. So when it comes to any other physical product, any food you eat, any drug you take, we have certain expectations about what the floor is for safety. Like, what are the questions you should ask? Like, you should have done a certain amount to like, make sure this was a safe thing. We don't have an expectation around that. We don't have a consensus around what that would be yet when it comes to social platforms. And I say "social platforms" because I would also include things like games, or Discord, or things like that in there. And so we're doing a project that's around, I think Jonathan Haidt has done really

• 54:35Speaker 02

  good work around helping us begin to have a public conversation about the harms around kids' mental health. We want to mirror that process across a wider set of harms. Let's talk about human trafficking. Let's talk about cartels. In addition to kids, let's talk about elder issues. You know, let's do a much broader set and reproduce that process. Let's kind of do the truth and reconciliation phase. And because I'm a product manager, like the way we solve problems is we get really clear about articulating what the problem is. And then let's talk about levers that would prevent or mitigate that problem. Because often, right now, we confuse a lever for strategy for pulling that lever. I'll give you a concrete example.

• 55:17Speaker 02

  A lot of the harms for kids, a common lever is let's keep under 13-year-olds or let's keep under 16-year-olds off the platforms. Right now, because people who understand social problems often don't understand tech, they reconverge on, let's check IDs. And I don't think checking IDs works. Like, you can go, like, we can have a conversation on it. I also think there's huge collateral damage. I don't think any of you really want to have to have your ID checked. There's, like, a variety of issues with it. But if you had asked a technologist, "Hey, I have a problem. I need to find under 13-year-olds. How could I do that?" They would come back and say, "Here's 10 or 15 ways to find an under 13-year-old. We don't have to check IDs."

• 55:58Speaker 02

  If we could all look at the same list and say, "Hey, here's the menu of what's possible." Right now, the platforms are only doing one or two things. They're, like, asking you how old you are, and if you say you're 13, they're like, "Cool." You know, I'm not saying you have to do all 15 things. Like, I don't want to be prescriptive. But I do want us to sit down and say, "Hey, maybe there's six or seven things that we all feel comfortable about that we think are the expectation." And if you're not willing to do that amount of work, then you're negligent. So that's the goal of Duty of Care. And then, because a lot of the logic in Duty of Care is actually dependent on having an intuition on the physics of social networks, instead of saying, "Just trust me." You know, like, we talked about deep reshares versus shallow reshares, or, like, how long should you have to wait before you reshare, all these things.

• 56:43Speaker 02

  Imagine if we had, you know, we spend hundreds of millions of dollars every year on flight simulators to train jet pilots to keep us safe for the Air Force. Or keep our jets safe, you know, for Delta Airlines. We spend zero dollars on flight simulators for social networks. We have to train all of our safety people in-house because we don't have a lab bench for training them in academia. And so those are the two things that we want to work on over the next few years.

• 57:13Alex Kantrowitz?

  Sounds great. I read that you're trying to raise, or you were trying to raise, five million dollars for this. I have a couple of questions about that. A, I'm curious if you've raised it. B, if you have, where's it come from? And I would say, most importantly, there are, there's, I think, a plague in this world of big tech advocacy, where there's so much dark money that gets spent all over the place. And no one ever discloses where that money's coming from. And so we don't know where the agendas lie. So can you, I don't know if you have or not, but are you going to commit to share the list of your funders moving forward when it comes to running this nonprofit?

• 57:56Speaker 02

  You know, I am totally willing to entertain that idea. I want to put a quick caveat, which is I would like to check in. So we've only taken, we've only taken two checks ever. Like we, McCourt is, we're in the process of getting the check from them, but they made a public announcement of it yesterday. I would want to check with those two other funders on whether or not they would want their names publicly. But for context, prior to McCourt, we'd only raised a very low six figures. Like, I've gone by on a shockingly small amount of money. And part of my book advance is all that has funded me since I came out. The people who tell me, "Oh, it's easy. They'll just give you a check for $300,000." I don't know how that actually works because it has never happened for me.

• 58:41Speaker 02

  Not going to complain, though. But I like that idea. And I want to double-check first. But we've been public about our funding, at least in the last, after those first two checks.

• 58:53Alex Kantrowitz?

  I feel like you should make it a prerequisite.

• 58:55Speaker 02

  Yeah, but I was just saying that going forward, I can talk to them and be like, "Hey, like, we need to have your name out in the open." I'm totally fine with that.

• 59:01Alex Kantrowitz?

  Oh, we're running out of time.

• 59:02Speaker 02

  Yeah. Do you want 20, 10 more minutes? Or are they going to get angry at you? We're the last ones, so you never know.

• 59:09Alex Kantrowitz?

  They're mad.

• 59:10Speaker 02

  They're gonna get mad.

• 59:10Alex Kantrowitz?

  Let's do five more and then we can bail.

• 59:12Speaker 02

  We'll be shameless. We'll make them get, pull us off the stage.

• 59:15Alex Kantrowitz?

  Sounds good. This is just a curiosity of mine. In the 60 Minutes interview you did after you revealed who you were, maybe it was there, maybe it was the journal, you mentioned that you had basically been living off of Bitcoin winnings? Or something like that? I don't know. Correct the record if I'm wrong.

• 59:37Speaker 02

  I don't think there was anything in 60 Minutes.

• 59:39Alex Kantrowitz?

  Oh, but was, is that true? Because you were, I heard you were living in Puerto Rico.

• 59:44Speaker 02

  No, I think people conflate the two.

• 59:45Alex Kantrowitz?

  Yeah, yes.

• 59:45Speaker 02

  They conflate, yeah, they conflate the two. So I do hold some crypto. It is not life-changing amounts of crypto.

• 59:51Alex Kantrowitz?

  Okay.

• 59:52Speaker 02

  It is part of a diversified portfolio.

• 59:54Alex Kantrowitz?

  All right.

• 59:55Speaker 02

  But I live in San Francisco, so like I have lots of people who talk about crypto all the time.

• 1:00:01Speaker 02

  I moved to Puerto Rico because I, when I, I was so ill that I was paralyzed beneath my knees. So I went into the hospital because my leg turned purple. Because I had a foot-long clot that had been in my leg for between 18 months and two years. They don't know how long. And when I went to the doctor and complained about it, I'd be like, I'm in a lot of pain. I'm really exhausted. They'd be like, oh, it sounds like you have depression. You know, there's nothing gendered about women's pain. Clearly. But it turned out that I was also starving to death. So I had psoriasis and I was, even though I had gained weight, I was so malnourished, my protein synthesis had stopped. So I showed up in the hospital and for context, my father-in-law died last week and he had been suffering from

• 1:00:47Speaker 02

  cancer for a long time, maybe seven months. And he went through a period of time where he didn't really eat anything for four months. He ate very, very, very little. And he lost 75 pounds or something insane. And

• 1:01:00Speaker 02

  My father was reassuring me as he was looking through his lab results. He's like, "These look real bad, but to be fair, your numbers looked worse." And so I was really malnourished. And I got paralyzed beneath my knees. Because my nerves, my nerves died. And I'm much better now. I can walk on my toes. I can hike for hours. But I have constant pain. I have constant, very severe pain. To the point where, in addition to my neuropathy drugs, I have to numb my feet. Like, that level of pain. Because I can't take more, because my heart beats too slow, for context. And I was really unhappy in San Francisco during COVID. Because if you've ever been, you know how balmy and warm San Francisco is.

• 1:01:43Alex Kantrowitz?

  I used to live there.

• 1:01:44Speaker 02

  Yeah, yeah. And the buildings aren't insulated. So like, I couldn't even keep my inside warm. And I was really unhappy. And a bunch of people were like, "We're all gonna move to Puerto Rico." And I was like, "Well, we have to only socialize outdoors, because it's COVID." You know, I have nothing to lose. And I went down there for like, a three-week vacation. And I was like, "Wow. I've been complaining about being cold for years. Why did I live in San Francisco?" And it's just been a game changer. It's, I actually credit Puerto Rico with my ability to have been as successful as I was with the whistleblower. Because I had so much more energy after I moved there. Right, I was able to lower my pain meds a bunch.

• 1:02:29Speaker 02

  And it made it, it just, it gave me many more hours a day where I was effective.

• 1:02:36Alex Kantrowitz?

  Okay, so, we get to the bottom of the crypto narrative. We're out of time. Do you have, like, 60 seconds? Is there anything you're optimistic about?

• 1:02:43Speaker 02

  Oh, sure. Can I, can I, so I can, can I do my soapbox now?

• 1:02:47Alex Kantrowitz?

  Yeah, but they're gonna make us wrap. But yeah, definitely. Soapbox, so what?

• 1:02:50Speaker 02

  I'll hold the floor.

• 1:02:51Alex Kantrowitz?

  Okay.

• 1:02:51Speaker 02

  Okay, so...

• 1:02:54Speaker 02

  We talked about some heavy stuff today, right? So like genocide, like, you know, business incentives that are dystopian, kill democracy, whatever. The reality is, every single time we have invented a new communication technology, it has been really, really disruptive. Like, people like to say the printing press was disruptive. The printing press was not disruptive, because no one could read, right? It's like 60 years between the invention of the printing press and the printing press causing harm. The real thing that caused harm was Martin Luther was like, "The Catholic Church has been lying to you. You're not going to hell. Just go be nice to people, love each other. Let's teach everyone to read so they can read the New Testament too." And it turned out, when you taught people to read, it went from being 3% literate in

• 1:03:38Speaker 02

  Germany to being 30% literate in 20 years. And people started publishing pamphlets on things like, you know, "How do you know if your neighbor's a witch? You know, I hate to tell you, you might have to burn them. It's gonna be hard, but you can do it." We invented cheap printing presses. Those are newspapers. And we literally had wars over misinformation. Historians have looked at the telegraph. The telegraph did not cause the Civil War, but they believed that the telegraph influenced when the Civil War was, because for the first time, people could actually see that the opinions in DC, the opinions in other parts of the country, were very different than the ones at home, because the news all converged into real time. Radio.

• 1:04:22Speaker 02

  We've had whole genocides over radio, things like Rwanda. Or like WWII. For the first time, people had personal relationships with their leaders. And literally, Hitler was the best person in the world at radio. What we are experiencing right now is not novel. But the reason we feel overwhelmed, why this feels different, why is this so scary, is because it's actually our problem to solve, right? Like, we feel overwhelmed because we can see that we're in the driver's seat, that we have to figure out the way forward. And I want to be honest, it might get a lot worse before it gets better. But the reality is, is we have figured out how to do this every single time before. And I have 100% confidence that if we do bring 100,000, if we bring a million more

• 1:05:07Speaker 02

  people to the table, we are going to figure out a path forward where we get to live with social media, where we get to feel good about it, and it's good for us. So, I'm excited about the future.

• 1:05:17Alex Kantrowitz?

  Frances Haugen, everyone.

• 1:05:20Alex Kantrowitz?

  Thank you.

• 1:05:28Alex Kantrowitz?

  Thank you, Francis, for the chat. Thank you. What an amazing audience. You guys are amazing. This is gonna be on Big Technology Podcast feed, so if you go on your phone, subscribe to Big Technology Podcast. It's available on all podcast app. Would love to have you. We do these every single Wednesday. We've interviewed Francis's lawyer, Lawrence Lessig.

• 1:05:48Speaker 02

  Oh, Lawrence Lessig, yeah. Yeah, he's wonderful.

• 1:05:49Alex Kantrowitz?

  And a bunch of other people that agree and disagree with her. So, thanks again. Thank you to Unfinished Live for having me here, having this conversation with Francis. It's been awesome. And, well, we hope to see you again on the internet.

• 1:06:13Speaker 03

  Now, more than ever, we need someone on our side. But Rob Whitman? Whitman's votes drove up costs, cut Medicaid, and risked 350,000 Virginians' health care to pay for tax breaks for himself and his donors. He's not on our side. Shannon Taylor will fight corruption, cut taxes, and bring down costs. This November, vote for Shannon Taylor. She's for us. Paid for by HMP, the House Majority PAC dot com. Not authorized by any candidate or candidate's committee.