The Vergecast - YouTube’s bad week: pedophiles, hate speech, and Steven Crowder

Episode Date: June 7, 2019

There was so much news this week that we had to add another Vergecast episode to cover it all. Besides Apple’s WWDC, the most important story in The Verge’s coverage has been the various crises YouTube is facing over its moderation policies — problems with hate speech and harassment, pedophiles on the platform, and white supremacist content. Verge editor-in-chiefand Vergecast co-hostNilay Patel invites Silicon Valley editor Casey Newton and senior reporter Adi Robertson to the show to explain what is happening on YouTube, how YouTube is handling it, and the outrage cycle surrounding it.  Learn more about your ad choices. Visit podcastchoices.com/adchoices

Support for the show comes from Retool. Too many companies run critical operations on duct taped spreadsheets, Slack workflows, and whatever else they could cobble together. Not because they want to, but because building internal tools means weeks of waiting on someone else's backlog. That's where Retool comes in. Build custom internal tools just by describing what you need. Prompts something like,

Build Me a Revenue Dashboard on our Salesforce data. And Retool actually builds it on your company's data, in your cloud with enterprise security built in. Go to retool.com slash Verchcast. We all need to retool how we build software. What's up, y'all. I'm Skyler Diggins, seven-time WMBA All-Star, Olympic gold medalist, and mom. And I'm Cassidy Hubbard, host and reporter for nearly 20 years, covering the biggest names and stories in sports and mom. And this is Am Mom, a community for athletes, game changers, and moms of all kinds.

dropping May 14th. Tap in with us. Hello and welcome to Vergecast, the flagship podcast of the Vox Media Empire. I am Neal. I am your friend. This is a little bit different week for us on the Vergecast. There's just a lot going on. This is actually a second episode of the Vergecast chat show this week.

We did a really long episode recapping WWDC and all the Apple News with Dieter and Paul. But I'm being honest, there's like a ton of YouTube news happening. and I personally could not figure out how to talk about the Mac Pro in iOS Dark Mode against YouTube hate speech policy

in the same show. So we did the WWC show. That's out. You can listen to it right now. And then I asked Casey Newton. Hi, Casey. Hey, Meade. And Addy Robertson.

Hi, Adi. Hey. To join me so we could talk about what the hell happened with YouTube this week. So if you're not interested in YouTube, you just want to hear about Apple, go listen to the other episode.

But if you are, and I think you should be because it's really important. For a lot of reasons, this episode of our chat is all about YouTube's basically horrible week. So I would say there are three interlocking crises at YouTube right now. There's a pedophilius scandal, which is related to the recommendation algorithm. And that has been ongoing, and it got worse this week. There was an update to YouTube's supremacist content policy that resulted in a bunch of channels getting taken down.

And that kind of overreached and some legit. and some legitimate channels are being taken down as well. And then there was hate speech and harassment policy, I don't know what you'd call it, disaster. And I'll just be very honest and disclose it up front. It was definitely related to one of our colleagues at Vox Media, Carlos Mata, who is a video host at Vox.com. I don't actually know Carlos.

We've tweeted at him before, but he works at our company. He's part of our company. We're supportive of him. So I just want to disclose that from the jump. That's a true thing. He works at our company. But it's all interrelated.

It's all chaos time at YouTube. And I just want to add this too. Next week, Susan Rochisicki, the CEO of YouTube, is going to be on stage with Peter Kafka at the Code Conference. So a good moment to catch up on what's going on with YouTube. And I can't believe I'm saying this. It's an insane sentence. But I think we should start with the pedophilia crisis.

So, Casey, what is happening with YouTube in this recommendation engine and a kid's content? Because it seems horrible. It also seems kind of under the radar. Yeah. So it relates to the recommendations that you get. when you watch videos. And the New York Times did some investigating, and they talked to some experts.

And it turns out that in some cases, if you watch videos that are sexual in nature on YouTube, one of the things that you may be recommended are videos featuring children, like children in bathing suits. And in some cases, it appears that the algorithm was essentially auto-generating playlist that would appeal to pedophiles. And so naturally, a lot of people are very concerned about that. And that has gotten all the way to Congress. Yeah. So three senators have called for YouTube to take more direct action to protect children from this kind of thing. And it is a bipartisan group, Josh Holly, Richard Blumenthal, and Marsha Blackburn.

Josh Holly, really, he's just really in it on the tech policy stuff. Josh Holly's people send me a daily email about something Josh Holly would like the tech platforms to change. I mean, he's a Republican. He was a, he was in a, he was in a house before. He's a center and he's just going for it. It's a new name. If you're listening to Richast, you've heard his name a lot. But they've, YouTube's already done stuff, right? I mean, this is not a new probably turned off comments on kids content, right? What is different now? Well, so the, yeah, the previous pedophilia scandal was that like pedophiles were leaving comments on videos involving scantily clad children. And so advertisers, you know, freaked out.

after reports and YouTube got rid of the comments, but they haven't made any changes to the recommendation engine. You know, and we should say, like, this is, these are videos that many people would look at and not think of as sexual at all, right? I mean, it's like children like splashing around in a pool, but, you know, some people are basically like purving out on these videos that are being suggested to them by algorithms, which, of course, are not taking pedophilia into account when they're making their suggestions. They're just saying, well, you know, people who like this video like that video. But, you know, YouTube has not been able to account for that as it builds their recommendations. So what can they do? Well, the, I guess, nuclear option is turn off recommendations

around videos involving children, which is what I think Josh Hawley, you know, would like to see. I don't know. Like, building the right algorithm, like, it's going to require, like, YouTube engineers to step and decide the mind of a pedophile and then, like, sort of, I don't know, reverse engineer that. I understand why technically it seems difficult, but of course if you're on the side of like, you know, preventing pedophiles from, you know, preying on children on online social media platforms, like you actually probably do want to overcorrect

and just, you know, eliminate any possibility of it and then, you know, work from there toward what is possible as opposed to like, you know, putting some sort of bandaid on it. It does seem like the algorithm here is bluntly doing what it is designed to do, right? 100%. Like you watch a video, you watch another video, there's a high correlation between people who watch the next video. So it just serves that up to you. Yeah.

By the way, a story that I feel very dumb about writing, even though I was excited to do at the time, was a story that I wrote about how good the YouTube algorithm was. Because my interest that I watch on YouTube are like cooking videos and video games and pro wrestling. And I was amazed at how good the recommendations were getting. Like all of a sudden I was loading up the YouTube homepage several times. a day because it was finding people who liked other cooking videos and saying, hey, like, everybody, you know, likes this creator too. You might like them and I clicked and, you know, sure enough, I did. We should say that, you know, most people's interests that they're exploring on YouTube are extremely non-problematic. And as YouTube's recommendations, engines have gotten

better, you know, it's benefited people like me who get to see more cool stuff. It's benefited advertisers who, you know, can now reach larger audiences. And it's benefited the creators who are benefiting from the larger audiences. Like, that is the backdrop against, you know, which all this is happening, but tech platforms typically never think of the worst case scenario because it is not profitable to them. So, you know, they very much operate in this mode of ask for forgiveness rather than permission, and that's where YouTube is now. And it just seems like they need to recognize, hey, we're generating a pattern that is inherently problematic and they just haven't done that work. Yeah. Addie, from the sort of regulatory side, it's not even

speech, right? The videos themselves are not the problem. Like, how do you go into a regulatory posture on this. I mean, in terms of whether the things are speech or not, it's weird because historically there have been cases where you have had sites making things that kind of collect and shape user responses and then that thing is considered content that's being produced by the site. So there was, I think, roommates.com sent out a, like, questionnaire that maybe, that allowed people to discriminate. Like, it allowed people to note discriminatory preferences. And they were like, No, no, you actually are producing that thing. So you could maybe consider it that, but it is so complicated,

and I don't know how you would sue it for this. Actually, the Holly Bill is just, it's going to prohibit video hosting websites from recommending videos that feature minors. I'm not totally sure how they're defining recommending. Like, so much of this seems just based on the idea that YouTube is the only video site. I mean, that's the other backdrop here, is that YouTube has all these problems because everyone feels like they don't have a choice.

Why don't people think of Vimeo as a choice? A, I don't, I mean, Vimeo is owned by IAC, so they're also owned by a big company. IAC, like, owns Tinder. They could definitely, like, flow some of those Tinder profits towards making Vimeo good. I mean, that's a big company. It certainly is big and stable as any of the big ones. But they, I think Vimeo wants to be this, like, R-D platform. And so they've specialized away from this broad general interest site.

My choice is that we should go all in on Quibi. And Quibi is the future and Quibi will be the solution to all these problems. I don't know if that's going to work out. But it does seem like everyone has taken it as a default position that YouTube is the only choice that you have. And thus we must regulate YouTube directly. Whereas Holly Bell says video platforms cannot recommend videos featuring minors. That creates just an ever expanding array of problems. for other services.

Like, Nickelodeon creates many videos with miners. Should they not recommend inside of the Nickelodeon app the next video for you to watch? No, that's the exception they have is that they say professionally produced videos, quote, like primetime talent show competitions would be exempt, which opens up its whole can of forms there too. Yeah, so all of this seems very hard. It touches on the same problems as the other chaos is going on, which is how do you define and categorize and regularize and regularize.

the very, very fuzzy boundaries of speech. Clearly, the pedophilist stuff is deeply, deeply important, but it got overshadowed this week by the hate speech and harassment controversy. So, Casey, do you want to just very quickly, I mean, it's going to be very hard, but try to summarize what has gone on? Yeah, so separate from the harassment issue, which we will get to, YouTube updated its policies around what it calls supremacist content. So believe it or not before this week, you could put up a video saying Nazism is a great idea and there should be more of it and, you know, click to like and subscribe.

Now, YouTube would not promote that video. They actually would withhold that video from recommendations. If you tried to watch it, there would be an interstitial video that would, you know, I guess try to scare you out of watching it. But you could post it. And what they said this week was that is no longer true. So if you put up that video, they're going to remove it. And if you put up a bunch of those videos, they're going to remove your channel. And this applies not just to the stuff you'd expect, like Nazism, but they actually introduced a couple of new categories of protected classes.

So cast, for example, which is, you know, very important in India, is now a protected class. You can't make a video saying one cast is superior to another. I'm using C-A-S-T, not C-A-S-T, in case that's confusing on a podcast. And then also, much more interesting to me, as somebody who has written about this a lot, is sort of survivors of, like, mass tragedies. So if you had a child who died in the Sandy Hook Elementary shooting or a relative who died in 9-11, you're now considered a protected class. And one of the things that that means is that no one can now make a video saying 9-11 didn't happen or the Sandy Hook shooting didn't happen. I regard those as positive steps.

Those are the biggest changes they made. There were a couple of other ones. They have this extremely hard to describe category of videos called borderline content, which is a video that sort of comes up to. the edge of violating a policy but doesn't quite get there. And what they've said is that if you repeatedly make this kind of content, which they insist they can identify primarily with algorithms, then they are going to either demonetize your channel, prevent you from making money, or they might get rid of it altogether. So those were kind of the gist of the changes that they

announced. And when they were sort of planning the rollout of this announcement, the thinking was like, you know, everyone was just going to pat them on the back and say, oh, you know, thank goodness, you finally take an action, and then they got blindsided by everything else that happened. You know, it's funny that you say that they can, they insist they can recognize these borderline videos based on algorithms alone. One, because that is the root of the other crisis. And two, they can't even recognize when their own algorithms are doing something

that's very obviously borderline already. Yeah. So that, that to me is, is very difficult. I want to, before we move on to the other thing, because these are very deeply interlocking. We should just talk about them kind of at the same time. But you're using this specific phrase protected classes, which just jumps out at me. That's a phrase that you normally apply to the government. Right. The United States government has designated protected classes. So, for example, as an employer, you know, I'm not allowed to discriminate based on age.

Right. It's a protected class or sexuality or gender. those are traditionally how we think of the words protected classes. YouTube, is that a YouTube phrase or is that your phrase? That is the phrase they use with me. So YouTube itself is now using this phrase where they're overlapping the terminology that is used by the government. And I just,

that is nuts. Like, I just want to be really clear about that. That's another symptom of how big and dominant YouTube is. they're they're now making policy decisions that map to how we make governmental policy decisions. And I think many, many people are confused about the difference between YouTube, its policies, its ability to enforce those policies and rules on a private platform and the government. Like that is the root of this.

We cannot tell the difference between Google and the government is nuts to me. Yeah, it is very much the case. Like YouTube, like Facebook is a quasi-state. right? One of the things Facebook is doing is trying to create effectively a Supreme Court, what it calls an oversight board, to handle the most difficult questions of content moderation. And, you know, there are issues to discuss about it. But one thing that is good about it, I think, unequivocally, is that Facebook is trying to devolve some of its power and to create some sort of structure where it can be held accountable for its decisions, you know, where it is not under, you know, the sole control of Mark Zuckerberg. We'll see if they get there. It's still in the planning stages. YouTube has made not even a gesture toward this. And so that's why you see people go absolutely insane when their channel gets even a temporary strike, right? Because it feels like an existential threat from a government that is totally unaccountable to you. And I do think that it is that kind of like that size and power and unaccountability that is at the root of so many of the problems with our biggest social platforms.

I think the other thing is, like, Mark Zuckerberg has said, yeah, across the world, there are all these different community standards. It's very complicated. We have to deal with this. So far, it seems like, and I think you were saying this a little bit in chat, nilite, in Slack that YouTube kind of just thinks you should create one giant blanket policy for absolutely every video. Yeah, it seems completely unworkable to me that the policy against supremacist content that they rolled out. Yep, they got rid of a lot of videos. They also overshot. And there's like World War II historians who are like,

I post videos of what the Nazis said. And in my video, my channel is getting removed. And there's just no way that I think you can write a policy, which is effectively a speech policy that covers beauty influencers, Dieter talking about the iPad, conservative comedians and World War II historians all at the same time that has any coherency. You know, Casey, you are the one who writes about how these companies moderate that then you can disson. to an army of like minimum wage contractors to apply its scale. Like it's just obviously not possible. Well, yeah. I mean, you know, you were sort of saying earlier, you know, that you doubted that an

algorithm can do it. And like, I'm here to tell you that a human can't do it, right? Like every moderator I've ever spoken to has talked to me about some edge case, you know, and by the way, they faced near daily edge cases where two human beings who have both been trained on the same policy reach a different conclusion. You know, this should not be surprising given that like this is also true. of our judicial system, right? Like, this is why we have nine Supreme Court judges, you know, rather than one is because sometimes you have to reach some sort of consensus to create the

appearance of justice. And when there is no analog in these platforms, and it is at the root of so many of their troubles. So that really does lead into what happened with Carlos Mata and Stephen Crowder. So once again, just because I know that people are going to tweet at me, It is true. We work at Vox Media. Carlos works at VoxMedia at Vox.com or sister publication. It is further true that Vox Media did not exist until the verge started. So we are very biased in favor of Vox Media. It's our employer. It's our company. And it is true that once upon a time I was managing editor of Vox.com. So those are all the disclosures. You can read into them as you wish. I don't, I'm not trying to hide anything. But Casey, tell us, the chronology here is so

broken, but we try to go through it, what happened between Carlos and Crowder. Yeah, so, you know, Carlos is a media critic and he often criticizes cable news and conservative media and he does it on the Vox YouTube channel, which is a big channel, right? It's got, I don't know, five or six billion subscribers. And there's another big channel that is operated by a conservative pundit named Stephen Crowder who would frequently attack Carlos's videos, right? And the sort of very, you know, large tradition of YouTube response videos. The issue was that when Crowder would go after Maza, he did it in the manner. of a schoolyard bully. So he would use a lot of name calling, right? He called Carlos a Lispy queer repeatedly or a gay Mexican and would do so in the tone of like, you know,

every schoolyard bully whoever called me similar names in elementary school. You know, Carlos is, you know, dealing with this. You know, it's not just that these videos are hurting his feelings. Crowder has three million subscribers. Many of them would get Carlos's phone number and text him or just kind of, you know, shout abuse. at him on every social platform. And so for Carlos, this became a real doxing and harassment issue. And even though he would report the individual videos to YouTube, YouTube would never do anything. So Carlos, because he's a video editor, finally took all of these videos that he's been dealing with for two years and he put together a supercut of all the times that Crowder had called him a name.

And it went hugely viral on Twitter. And all of a sudden, people started paying attention to this conflict between what YouTube's stated rules were about harassment, right? The rules say, you're not allowed to say hurtful things about another person and the actual lived reality of enforcement, which is that no one does anything. So that's kind of the start of it. And I will maybe pause to see if anybody wants to add anything. So many, many things. The first one, I don't think anybody does a good job of connecting the, hey, there was something in this video that was said and then the outsized harassment experience. Right. So this is, I'll just be honest, This has personally happened to me where I've lived through a multi-day harassment campaign and had to lock everything down because of a video, right? And there's literally no way to connect those dots. There's no way to point at those things. But that's the heart of what Carlos is getting after, which is it wouldn't be so bad if it was just like a video. Like, sure. But what's happening is this army of trolls is then escalating this in uncomfortable and in sometimes threatening ways.

Addie, that's like happened for a long time. Have we gotten any better at connecting those dots? I think that in a sort of broad, we understand this culturally way. We understand the idea that there are dogpiles and, I don't want to say economies of scale, but like harassment economies of scale, that the internet just makes everything very, very large in a way that is sort of qualitatively different in the same way that huge amounts of metadata become qualitatively different. If we're talking legally or even policy-wise, I think this kind of shows that YouTube clearly can't really do it. There's not really a great legal strategy for doing it.

There have been cases where people sued and said, look, like, you were clearly inciting harassment against me. Those are kind of happening. But it's just there's a lot of plausible deniability where you can have a bunch of horrible people who follow you and you know full well what they're going to do. But you also can claim that it's not happening. And conversely, you can be a person who is fine and you're a nice person on the internet and you will still have people who follow you who decide that they will defend you to the death and do terrible things. I mean, aren't we all, we're all just in it for stands. Absolutely. So that's honestly my first reaction is culturally we understand that like fandums can be toxic, right?

Like Marvel fans can be toxic. Like that they're not, they're not even in the realm of this kind of political speech. But here, there's a deep connection between what you say, what you do, who you're attacking, and then how your very passionate followers will emulate your behavior that we, at scale, that no one is, like, deeply considering. Casey, you said this. It's not that his feelings were hurt. It's that his safety was threatened. Exactly. And, you know, I think just a truth is that our law and our, like, law enforcement system does not take online harassment seriously.

or if it does, it doesn't take it seriously enough. The counterpoint, though, is that if you are a very powerful person, you will then use the moments that it does take it seriously to sue some random nobody on Twitter like Devin Nunes Cow. So if you're a member of Congress, you will use any law to shut down your critics that was meant for a person on YouTube who's being harassed. Right. And so that's one reason why, like, speech laws distinguish

between public figures and non-public figures, right? And like, I think we're going to need to introduce some kind of, you know, gradations into our understanding of it, right? Like, you know, I wrote a piece this week where I said, you know, if somebody who has 100 subscribers makes a video saying, like, you know, that they don't like me, I don't care. But, you know, if somebody with 10 million subscribers says that I'm an existential threat to speech on the internet and some of them come after me and my life is threatened, then I do wonder, you know, whether we have a different conversation about that. because it seems like this sort of thing. And the reason why this debate is important beyond Carlos is we're just going to see a lot more of this. They're going to be a lot more channels with millions of subscribers and they're going to be criticizing each other. And their followers are going to do wild and hurtful things.

And so we need to understand whether we're going to hold those people accountable at all or whether we're going to throw our hands up and say, well, that's the Internet for you. Say that you, like, if you tell your audience that this person's going to do a really horrible thing, they're going to go harass them. That's a lot of what's going on here. How do you account for situations in which someone is actually doing something very hurtful and you call out, call them out to your audience and you're like, this person is doing something that I think is dangerous. They're trying to pass a law that I think is going to lead to these terrible consequences. And I just want to make sure you're aware of this threat. How do you end up accounting for that? Well, that seems like a clear example of political speech, right? Like somebody who is like trying to pass a law, that just seems to fall so, so purely in the

realm of the First Amendment. You know, I'm more interested, like, you know, to try to lighten the mood a little bit, you know, there was that extremely engrossing controversy a few weeks back between James Charles and Toddy Westbrook in which some vitamin gummies, you know, triggered one of the greatest friendship breakups that YouTube had ever seen. On the other hand, though, the thing she was claiming was that he was in a very veiled term, a sexual predator. That is true. Now, you're right. Subtrack that. Because let's say, like me, you think this was mostly an issue about vitamin gummies. Like, it really felt like a business. deal that had gone bad, right?

You know, if, and she has, I don't know, now like 10 million subscribers and he's got 13 or something. So if they really start going after each other and their lives are threatened, I don't know. Again, like, I'm like you, Addy. Like, I don't want the law to intervene every time that, you know, crosswords are exchanged on YouTube. But, you know, we do have a prohibition of inciting violence, generally, like, in the law. and on these platforms.

And I think one of the things that's happening that's really challenging is that some bad actors are getting really good at inciting kinds of violence without directly calling for them. And so are you going to kind of continue to let them be edge lords forever

or are you going to build systems that hold people accountable when they are dog whistling to their audiences to kind of go out and, you know, take care of business? So I think it's a good inflection point for the next turn of the story,

which is there is a con, conflict between dog listen, your subscribers and being accountable for it. And here you had Carlos saying, okay, here's, here's evidence of the dog whistling. Here's evidence of the harassment. Here's YouTube's rules. Here's me reporting it. And here's nothing happening. And then an outrage cycle.

What happened next? So YouTube put its staff on, on the issue because like a sad truth of YouTube support is that the more retweets you get, the more seriously they will take you. And so they watched all of Crowder's videos, and the determination they made was that while the videos contained some hurtful language, the bulk of the videos was just a response to, you know, essentially political speech. And so they were not going to weigh in. And they responded to Carlos's tweet storm with a four-part tweet of their own that was written in this kind of bloodless, like boilerplate language. It was kind of like an automated, like, letter from your insurance. company basically. And of course, this sort of triggered a new cycle because it offered no rationale, right? It just defaulted to this content does not violate our rules, which, you know, feels very arbitrary, right? Like, there's not a rationale that's offered. And so after that, that's when it actually became a national story because, you know, multiple media outlets kind of,

you know, looked at the facts of the case and said, well, gosh, like, you know, if calling somebody a Lisp be queer repeatedly for two hours and mobilizing an army of followers against him, like isn't harassment than what is. And I will say this is where we enter the story. Like I am personally a piece of this story. Not that I want to be, but YouTube offered a bunch of media outlets an explanation, but they refused to be on the record with it. They said, here's what we think.

And specifically what I said was you can paraphrase this. I know this because Gizmodo just published it. But we actually, I made the call that we would not print it because I am fierce about the fact that if you are going to run something that looks like a legal system, you have to be accountable for your decisions. You should not go to reporters and have us try to explain it on your behalf as though we understand a secret because you told us on background. So that's just like my call. And we made that call. That's probably the most popular tweet I've tweeted in a while. Like, I think it's like 9,000 retweets or something now, or likes.

So we're in it. I don't want to back away from the fact that we're intermixed in this story. I'm just trying to be as transparent as possible. But I really feel like YouTube blew it right then and there by being anonymous. Then they said they were going to investigate more, right? Like even in that tweet storm, they're like, but we're still looking at it. Yeah. And then the next day they announced the supremacist policy.

And so everyone thinks it's the same thing. Right. Right. And then, you know, to compound matters, then they said that they were going to, like, temporarily demonetize Crowder's channels. And then there was an issue where he was selling a shirt that looked like it said socialism is for fags. But the A had been replaced with a picture of a fig, you know, which I don't think was fooling anybody. But that was, you know, one of the big ways that Crowder is making money is by selling this homophobic shirt. And so there was a follow-up tweet where YouTube said, you know, he's got to be. got to remove this.

And then everybody was like, okay, calling somebody a Lisp be queer is fine, but it's like selling like a t-shirt is, is what pushes it over the edge. So like this was the point where any like sort of logical rationale surrounding this decision collapsed, at least in terms of what was publicly visible. And so then that led to like more stories and more outrage. And all of this is, by the way, is being done in responses. tweet replies to Carlos from the at team YouTube Twitter account, which is a anonymous. Right. So it looks for all the world like a rogue like official of a banana republic.

Like just like making policy like drunkenly making policy. Well, and it and it also directed like fresh torrents of abuse to Carlos every time the account tweeted, right? Like on Twitter. Liz Lapado or deputy editor. I had a mild debate to hit about the word blunder versus disaster. Blunder sounds, it's the right word, but it sounds too fun. This is like a disaster. Right. like straight up a disaster because they're just inconsistent everything they do is causing abuse Crowder is riled up and at war and his fans are at war like Ted Cruz is like retweeting Stephen Crowder and being like this is a problem it's just like it's building it's starting to hit Fox News and then they finally said one more thing right which is they're forming a committee

they put out a statement saying that, yes, they were going to form some sort of committee and we're going to aim to revisit their harassment policies is, I think, the language that they used. So they've, they've set an intention. They're putting it out into the universe that they want to revisit their harassment policies. And, uh, you know, maybe that, that will happen. But they're like, we want academics. We want some people who've been harassed. We should talk to them. We should have some like, uh, you know, some lawyers maybe. here's and Adi I'm curious I don't

the problem is not that the policies aren't correct right the problem is there's no way to enforce them and they're inconsistently enforcing them in this case I'm just I'm really baffled at how bad this is because again they are literally using language in their tweet that is also

enlisted in their terms of services banned so it's like yeah we know it's hurtful but we're not removing it and then a bunch of people tweet that actually your policy says you remove hurtful things and then they offer kind of a statement that says, okay, but we don't do it in this instance. Like, it doesn't seem like bad policy. Yeah, it seems like they just don't,

they have this shadow policy that they're not actually explaining at all. And this is Megan Ferkman. I wrote a story for us today that's like Google employees think that they're just getting taken for a ride by far right commentators, by the right wing in general, which is much better at playing this game of insisting that they're being censored and discriminated against when the policies are. very clear. And I think that brings us to like the, the final turn and specifically the reason I wanted Adi, I wanted you here, because I know you get as irritated and frustrated about Section

230 as I do, which is everyone saying this is censorship of some kind when really this is a private company enacting its own moderation decisions. For everyone to be clear, a Crowder got this wrong. He's like, they should just say they're a media company, which is like the dumbest trope in the world. Can you, Addie, can you just quickly explain so everyone understands the difference between YouTube enforcing its rules, its obligations under 230? And I don't know, say the First Amendment. Okay, so I'll start with the version that is wrong that I hear a lot, which is, okay, there are platforms and there are publishers. If you're a publisher, then you put out material and you curate that and therefore you are liable for like if someone, someone sues you for defamation. That's on you. If you're a platform, then you don't really

moderate or you do something. I don't know. This is a wrong argument again. You're a platform somehow. You're neutral. And therefore, if someone sues a user, you are not liable. Everyone thinks this is actually a law somehow. It's not. There's not a law like this. Section 230, which is the thing people are sort of talking about is specifically talking about different kinds of content. So if you are YouTube and you put out a YouTube video and you're like, hey, we're YouTube. We made this YouTube video and it says something defamatory, then you can get sued because you are publishing that content. You are a publisher. If someone else posts a video on YouTube that is completely different and someone sues them, they're still a platform. The issue here is whether you published

a thing or whether someone else published a thing on your platform. So it does not matter what YouTube calls itself. A good example of this is Pages on the verge.com. We are liable for everything above of the comment section because we publish it. And then our comments, which anyone can participate in, are a user-generated content platform. And we're not liable for what you do in our comments. That's all on the same webpage. And we are definitely a media company. So that's, no one gets it.

It drives me insane. Right. Also, we can moderate those comments and we don't lose any sort of privilege. And that's, in fact, why the Section 230 exists, because they wanted to make sure that you can take bad stuff down off your platform without having to come up with some system by which you aren't actually a publisher. I've weirdly also seen the argument that this means that we need a fairness doctrine for the internet and we need to regulate YouTube as like a company that's on the airwaves, which is weird.

I don't know. There are a lot of very bad legal arguments around this. And again, First Amendment. The First Amendment means that you can't make a law that bans specific speech loosely. If you are a private company, you can ban whoever you want. In the first moment, just to be clear, applies to the government. And Google, thus far, is not a. not the government.

We don't know that. It just feels like one. That's like, that's like really the issue. If you ask me, the fundamental issue is that Google feels like the government. Google has so much power that you expect it to act the way the government does. And in fact,

Google itself in Facebook, to Casey's point, often do things that seem like they think they're the government. Like Facebook's like, we should have a court system. And you're like, no company in history until now, his like,

voluntarily try to build a court system for itself. This is a little bit like what we were talking about earlier, where it's hard for courts to recognize certain kinds of dog whistles, but it's easy for the culture too, which is that a lot of the problem is that Facebook and YouTube, these are websites. They're websites that are literally in some sense exactly the same as the verge.com as you putting a server in your bedroom and launching something, but they also feel qualitatively different. So everyone keeps trying to propose policies that will deal with platforms or websites, but then they end up struggling to define what is, when does something become a very large website and when does it

become something that feels like a government? I like that we're at the difference between very large websites and the government. Like, everyone's like, Yelp is a pretty big website, not the government. Like, they should just put me in a room and let me, let me sort between these entities. Well, by the way, though, that is the solution everybody actually wants is let me decide what content stays up and let me decide what content stays down. And as long as we can do that, then the internet will be safe for all.

It's true. And I should be that person. I mean, Casey, I'm going to be honest. The only solution here is benevolent dictators everywhere. And by which, I mean, there should be multiple companies that have multiple strict policies. And then people can choose the ones they want instead of trying to get Google,

which, by the way, like treats YouTube like it's over there. Like, it's not even like YouTube is dominant. YouTube is a dominant thing inside of another dominant thing, inside of a holding company called Alphabet that also does like, like flying internet balloons. This whole thing is a mess of enforcement and lack of accountability because it's just so big. The other thing with YouTube specifically is that it's big because it hosts just a ton of data. Like it's very easy to imagine creating your own forum to substitute for Reddit. But a lot of the issue is that YouTube isn't sort of just like Amazon Web Services,

where it just powers huge amounts of data that the average person could not host. So there's this layer of YouTube that's just a bucket for data that it seems very difficult to deny people access to ethically. And then there's this whole system of recommendations and channels and moderation that kind of gloms everything together. So I have become increasingly enchanted with Reddit's approach to content moderation because they let communities come together around certain issues and then decide on their own rules. There is a floor of rules that all reddits have to observe. Nobody can post child porn on a

Reddit, but they can raise the ceiling. So if they want to decide that there's no cursing in their subreddit, like, they can decide that. So if YouTube creator, if channels had to post like into communities on YouTube and those communities set their own rules and some were okay with harassment and others weren't, would we be having a different conversation? You know, and you sort of assume that, you know, you'd only ever be recommended videos. within a particular community and, you know, YouTube would stop recruiting, you know, subscribers for some of these other hateful channels. I mean, we still talk about Reddit a lot, too, we fair, and how terrible it is. Because there's a lot of...

We talk about it way less than we talked about it five years ago. And a lot of the issue is that there's cross-contamination between channels and Reddit. So yeah. Yeah. Okay. So here we are. We've identified all of the problems. That alone is a monumental task.

How does YouTube begin to try to solve these? And in Casey, you literally write the interface, which is about social platforms and democracy and effectively content moderation every day. Addie, I've tasked you with as many 230 articles as we can generate. What happens next? Oh, man. It's very hard to know in this moment because the Mazza case, which we've been talking about, it continues to go on. You know, Carl's is still tweeting every day about this. and YouTube has no real response to it.

So, like, I don't know what is going to happen in that particular case. What I do know, as I allude to earlier, is we're going to see a lot more of this. Creators are going to go after other creators, and some of it's going to lead to, like, angry, abuse and harassment. And some of it might lead to worse than that, right? Like, some of it might lead to violence. And I think we should be talking about that openly because, like, we are not far from any of that happening. And so I think it's a good time to think about, how we want these platforms to treat that kind of behavior and whether there are regulations,

policy changes that could be implemented that would still, you know, maximize the amount of free speech on the internet while discouraging actual violence against people in the real world. So someone tweeted at me in one of the many threads on Twitter that are just happening in my mentions, a private company serving the public in the name of the First Amendment is suppressing voices of those whom they are against. This is bigotry. And this is a best thing. that moderating Crowder. I just keep coming back. That's like the heart of it, right? Like, what is their obligation to serve the public? In what world is a private company operating in the name of the First Amendment? Where are those obligations coming from? Addie, is there any like policy

answer? Is the 230 debate advanced? I mean, like, we have talked about carving out exceptions to 230. We, you know, we just did it with Fasta and Sesta, which is the law that prevents, you know, people from hosting advertising for sex work. Like, what is the next step here on the policy side? I don't know because so much of this just feels like 80s cyberpunk where the government is just fundamentally ill-equipped to navigate a sphere of power that has suddenly come up. Like, people say First Amendment because that's just kind of our shorthand for a really big important thing that makes you free to say things, where now we're dealing with, there are these companies that just kind of host all speech. And it's very hard to regulate them as something that's different

from lots of other companies that are somewhat smaller. I guess antitrust is the thing a lot of people are depending on. Like, maybe if we can break up the ad section and the video hosting section, maybe they'll have more incentive to actually make a place that's nice. Maybe other competitors can show up. Maybe you can have a system that's more. decentralized so you don't have this thing that seems like a government. I don't know. I don't know if there's a good way directly forward from this specific point. It really does seem like, A, the constituencies of the politicians does not really know what's going on. Because it is, to be fair, very complicated. The First Amendment doesn't protect you in my comment section is like a slash dot trope. Right. Like,

even at much smaller scales on the internet, people didn't get it. Then there is the sort of ever expanding, and I would say platform-fueled culture war that's occurring, where literally any move against bigotry or sexism or homophobia or what have you seems to have a one-to-one identity with conservative critics saying you're now going to suppress conservative speech. Adi brought up the fairness doctrine in the past. What's amazing to me now is the idea of conservatives saying we need a fairness regulation for the platforms is conservatives have traditionally hated the actual fairness doctrine, which was an FCC regulation that was done away under Reagan that required broadcasters on public airwaves to give like to be fair in news story, right,

which is as much of a speech regulation as you can have. And when that went away, that gave rise to conservative talk radio, which feeds into Fox News, which feeds into now, you know, online commentators on the right, conservatives hated the fairness doctrine. They wanted it gone. And it was a boon to their viewpoints. So what's wild to me is the sides have flipped, right? And it's because they keep insisting,

conservatives keep insisting that they're being discriminated against when what people are complaining is their indiscriminatory speech. I do not know what happens next here either. But I can tell you, like Casey said, it's not going to stop. And I think the policymakers need to get a lot smarter, not only about the internet, but literally about the laws of the United States, Ted Cruz tweeting that YouTube is censoring people makes no sense. Playing God.

Playing God. Why is like a blood-red Republican interfering with a private company? Like, it truly makes no sense. That is as backwards as I can think of. Here's where it makes sense. Conservatives regard others' power as illegitimate, right? Like Clinton was an illegitimate president. Obama was an illegitimate president. And now we live in a world where YouTube and Facebook have power that conservatives do not control. And so they are going to constantly try to delegitimize that power or get what they can for it for themselves. And so if that means adopting some sort of new fairness doctrine, the only point of the fairness doctrine is to give them some of the power that is no longer theirs. So I think you have to view it through that lens because it's the only way that any of it is coherent.

I guess it's similar to the issue that YouTube policy is not actually like we're talking like there's rules and things. things that happen when really it's just if you yell at them enough, they'll do something. And then maybe if someone else yells, they'll do something else. Like, it's very hard to have any of this discussion in good faith. Yeah. Okay. Well, I do know one thing that's going to happen next, which is that Susan Wichickie is going to be the CEO of YouTube is going to be on stage of the Code Conference next week, interviewed

by Peter Kafka. Casey and I are going to be there. Casey, I'm planning on just chasing her through the hallways. I don't know about you. But we're going to be there. We're going to report from that conversation. It, it, recode is in the code conference also owned by Vox Media. It's transparent as I can be about that stuff.

But we're going to be there. We're going to try to get as many answers out of these folks as we can. This story is not going away. Casey is going to keep writing the interface every day. Casey, where can they find that? You can go to the verge.com slash interface. Literally every day. There's a problem of this scale happening every day.

So Casey writes the interface. Addie. I was like, you say it's every day. I do not usually write it on Fridays. I want to be open with people. about that. It's true. On Fridays, I demand that you podcast with me. That's right. Go find the interface. Addie is writing about all this stuff on the verge.com on our policy section. We're obviously paying a lot of attention to it. You can look for that there. Adi, where can they find you on Twitter?

The dexterity. Casey. I'm at Casey Newton. And I am at Reckless. So thank you for joining us for part two of the Vergecast this week. Again, if you listen to this whole thing and you're like, why don't you talk about Nvidia graphics cards in the Mac Pro? That's the other episode. So go listen to that. It's not coming in this one. We're at code conference next week. We actually have a big interview coming next week on Tuesday. It's the CEO of Beyond Meat, Ethan Brown. Yeah. Yes. And that's a rocket ship of a company right now. That was a really interesting conversation. And then we'll be back. Did you have a beef with him? I did have a beef with him. Talking to him about how we've used animals in the food supply, he's like, well, they're bioreactors for meat.

It's like, it's a lot. So that's coming on Tuesday. It's true. It's what he said. And then next week on Thursday. We'll have the chat show again. That's the Vergecast this week. Three in one week. Can you believe it? We'll be back next week. Thank you so much.