Tech Brew Ride Home - (TWTR SPC) Facebook's Fortnight Of Controversy

Episode Date: September 25, 2021

While I was off observing my wedding anniversary, your host @chrismessina spoke with @justinhendrix and @AuthorPMBarrett about Facebook's recent controversies as well as their own research into social media and political polarization. Learn more about your ad choices. Visit megaphone.fm/adchoices

On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco. Hey, who did this to you? What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. Okay, we are going to just basically get started today. This is going to be a little bit of a different format because Brian is celebrating his wedding anniversary. So he is not going to be joining us today, which gives me a whole lot of space to stretch out and go deep into a topic that I'm very personally interested in, have thought a lot about.

and have actually kind of been, I don't know, engaged in an ongoing conversation with Justin over months and months through his tech policy press, which is a podcast and a publication that he operates, and I'd love to learn a little bit more about that, and that he has recently also collaborated with Paul and Grant on a writing report. But before I get there, I want to say first, welcome everyone to the TechMeme Ride Home Experience for Wednesday, September 22nd. I can't believe September is almost over, but here we are. And today, we are, as I mentioned, going to dive right in. We're going to be talking about Facebook, about polarization, about social media, about who's to blame, who's responsible, what to do about this, I don't know, this mess, this reality, this, this humanity, society, etc. But first, Justin Ball, why don't you guys come up, introduce yourselves, give us a little bit of background on where are you from, the work you do, and how you're coming to this conversation today. Justin, you want to go first? I can do, Paul. So I'm Justin Hendricks,

and I am the editor of a site called Tech Policy Press, which has been around for about a year focuses on the intersection of tech and democracy. I'm also a research, associate research scientist at NYU, where I do some work related to looking at disinformation. And I teach a course called Technology, Media, and Democracy that looks at the intersection of those issues. And I look forward to being part of the conversation tonight.

Awesome. Paul? Yeah, hi. I'm Paul Barrett. I wear a couple of different hats, also at NYU. I'm the deputy director of the Center of the Center for Business and Human Rights, which is part of NYU's Stern School of Business. I also am an adjunct professor at NYU's law school, where I co-teach a seminar every year called Law Economics and Journalism. And as Justin says,

He was kind enough to join us in working on a research report we just put out about the role of social media in intensifying political polarization in the United States. And this report is kind of typical of the work we do. We do extended research reports that I would describe as white papers that try to synthesize information. synthesize research and other information has been done, provide analysis and then recommendations for how problems that we identify might be addressed. So I'm very glad to be here with you tonight, and thanks for inviting me. Awesome. Very helpful. And just to sort of, I guess, set the groundwork for how I imagine this conversation will go, since I am rolling solo and Brian is not here, I think mostly we're going to have kind of a conversation between the three of us.

Diving in a bit deep, for those of you, of course, I'm just jumping in who forgot or don't know who I am. I'm, of course, Chris Messina, and I have previously worked both on social media platforms when I worked on Google Plus, and I've also worked on the other side of the aisle on the open source and open standards side of the house. Specifically, my contribution to social media is the hashtag, but I've also worked on the aisle. on other technologies in that space. And I've just been, I'm just super interested in the connection between technology and society and people and behavior.

And one of the things I actually want to get to, and this conversation is something I've been thinking a lot about, which is incentive structures and decision architectures. I won't get into that too closely right now. But I guess maybe Justin Paul, why don't you guys start with where your report stems from? because you guys dropped this report, and I put it up in the nest in the space, before the Wall Street Journal published their sort of Shark Week file drop,

enumerating a bunch of stories, which I think we're going to get into, about how Facebook has, over the years, sort of systematically not taken action against problems that were going on in the platform when it might have compromised growth, engagement, the bottom line, et cetera. So what was the work or what was the impetus for what you did? Is this just sort of like a regular thing that you do every six months? Or where did this report and information come from? Sure.

I'll take a stab at that and then Justin can fill it out. Yes, in answer to your question, is this the sort of thing we do every six months? It's a little bit more frequent than that. But the Center for Business and Human Rights looks at human rights issues as they pertain to companies as opposed to more traditional human rights focus on governments. In connection with that mandate, I've been writing reports for the last four years on various aspects of the social media industry and its role and responsibilities in a democracy. And this particular topic, you know, I think the way to think about is, you know,

we decided to do this in the wake of the January 6th, the institution, insurrection on Capitol Hill. And a short time after that dismaying event, Facebook, in a variety of contexts, made strong assertions that the political polarization or divisiveness or partisan hatred that so many people were even more concerned about than usual had nothing to do with Facebook,

that there was no, the evidence. evidence out there in the social science journals did not support the idea that Facebook played a role in high levels of political polarization. And that, in a sense, was kind of almost a challenge. And Justin and I had been talking about doing something together for some time, and we decided to collaborate on this and to look at the social science research. figure out what it says, talk to some of the people who have done that research firsthand, ask them what the significance of their work, and basically provide a response to what Facebook

had been serving. And our goal was not only to answer Facebook, but to clarify this really important question, namely, what contribution, if any, does widespread use of social media play in terms of heightening political divisiveness. So that's kind of the background to what we did, as far as I'm concerned. Justin may have other perspectives on it. Yeah, I just add three specific things, Chris. One was, you know, Mark Zuckerberg's testimony to Congress in the House in March where he made some very specific comments about these issues. in his opening statement, which I'd refer people to, and kind of offered his diagnosis for what is to blame for the division in American politics at the moment, which he laid at the feet of political

and media elites who had sown division, which, you know, of course, is not wrong. And then went on to talk about, you know, social media and technology being part of the solution. And then Nick Clegg, wrote a medium piece, which got into some of these matters in detail, and in particular, advanced the argument that, you know, that is not a not settled science, and, you know, which is true, and yet, you know, seemed to kind of obfuscate some aspects after our review. And then I would add, too, that BuzzFeed published a report about an internal memo at Facebook that dealt with some of these issues and, you know, appear to kind of offer a sort of playbook for employees in terms of how to think about these issues. So all those things together

kind of served as the impetus for our report. I mean, there's so many aspects of this. There's so many directions that I want to go in. One of the challenges of having this conversation is to piece apart kind of what we're talking about, why it's important, and maybe just, like, how to structurally think about it, and then what to maybe. we advocate for. And so, man, how do we even start this? However, I do want to say a few things. One of the things that I realized last night as I was going through all these materials and, I guess, doing a little more background research is there have been recently some, I guess, analogies or comparisons from what Facebook is doing and the playbooks that they're using for their

PR and marketing outreach with big oil and big tobacco. And I think one of the reasons why it's important for us to remember those examples and how historically these, you know, literally cancer-causing industries have been able to persist for decades is because of the way in which they talk to the public, talk to regulators, talk to the rest of us, tried to downplay, whether it was like climate change or whether it was the cancerous consequences of their products. And so that's a well-worn playbook. This is not exactly the same context. But what I noticed was that Facebook, I believe, is the sixth, I think it's the sixth, most valuable company in the world. It is in the echelon of having a trillion dollar valuation.

And the only non-tech company that ranks in the top, I think, seven companies is Saudi Aramco, which has, I think, $2.3 trillion of market cap. The rest are tech. And so that's why this is so important. You know, I was thinking, like, why do we fixate on Facebook so much? What difference does it make? But it's because of, one, the value that accrues to these platforms and, two, how they do it. And then the third part is the, I don't know, even know if I want to call them externalities, the consequences of the business model and of how they actually produce all of that value for shareholders.

And so what we're talking about, roughly, I think, is whether or not the sort of second order effects, you know, which I think, Boz, the new incoming CTO has said, you know, sometimes you're going to break a few eggs, I kill a few people, you know, when you're building a global social network, and that's okay, maybe it's not okay. And maybe we think that actually these platforms should have and bear more responsibility for some of the consequences and cleaning up their superfund sites from a cultural perspective. So that's one thing that I just want to put out there in terms of why we fixate on Facebook for anyone who's been wondering that or is frankly exhausted by it. The other thing that I wanted to point out, and Justin, you alluded to this, and Paul, you also talked about this.

You guys did a great conversation on the Tech Policy Press podcast was that polarization won, and I'd love for you guys actually to provide a working definition of this in a second. Polarization has been increasing over time. Polarization has also existed throughout the history of the United States and has amplified at different moments where cultural moments are occurring, you know, where there's a civil rights movement, for example, or other things like that are going on. The other thing that I feel like has been missing from this conversation, and this is the part that you alluded to, Justin, is that specifically talk radio and conservative talk radio and Fox News multiplied by the effectiveness of social media to amplify through virality,

messages that lead to certain types of behaviors, ideologies, or outcomes. And specifically, one of the things, things that I think, Paul, you mentioned in the research, was that the demographic, who is 65 and older, have seen an outsized increase in their polarization in the last 10 years or so, which sort of, I forget if it's 10 or 20 years, but it sort of excluded social media as being the prime driver of that. So I want to, I guess, put those things out there as topics, and I want to help, I want you to help us kind of create relationships between them. So first, if you guys can provide kind of a definition for polarization as we understand it, why that's important.

And then those other factors that might be contributing thoughter, perhaps, for the polarization that social media amplifies. Okay. Want me to go first? Sure. Okay. So you've implied correctly that polarization itself is a very complicated concept. It can be a good thing.

It can be a bad thing. It can be an inevitable thing. in a democracy, particularly a two-party democracy, some degree of polarization is unavoidable and can be good. It can sharpen policy. Right. Two sides of different issues. Right. Yeah. And it can also be a byproduct of righteous movements for change, whether, you know, racial or gender-related, cultural movements, you know, if there's a just movement to give a certain group, African-Americans say, you know, equality under the law, that may well

and has, in fact, provoked backlash, and there you have polarization. You wouldn't say we want to lower polarization and get the African-American people to stop protesting. That wouldn't make any sense. So that's one thing. It's complicated. Second, there are several stripes of polarization, two main ones that are probably worth mentioning. One is divisiveness over particular issues. And somewhat confusingly, the social scientists tend to call that ideological polarization. So you and I disagree about abortion or about the appropriate level of income taxation or what to do about immigration. That's ideological or issue polarization. Then there's something else, effective polarization, which is... Is that an E or an A? A. A factive. So effective, okay.

Yeah. And that's a form of divisiveness that I think you could think about as partisan hatred. I not only disagree with you because you're in the other group, the academics would call it the out group. I think that by intent of belonging to the other group, you are an enemy of democracy, you are abhorrent, you are evil. and pretty much anything would be justifiable to keep you out of power. So this is pure us versus them Trump-style hostility. Vilification. Yeah, vilification. You're the enemy of the people.

An argument that it's very difficult to actually have a rational response to. Because who are the people in that example? Also, how do you prove that you're not the enemy? Sure. How do you prove that you're a patriot, not a traitor? That kind of thing. So when you get very high levels of effective polarization, you run the risk of a series of really deleterious consequences. And this is what's really most important.

It's not the abstract measure of polarization that's so important, although a lot of of, you know, academics focus on that and chart it over time. What's really important is the consequences of extreme partisan hatred. And those consequences include, you know, a loss of trust in democratic institutions like, you know, elections, a undermining of respect for, you know, mutually held facts, for example, you know, public health responses to a lethal pandemic. And experts stand up and say, you should do the following things, get a vaccination, wear a mask when you're close to other people and so forth. And you reject that because they're in the output.

You say that that's a product of some conspiracy as opposed to listening to those facts and trying to respond to them, adapt to them. And ultimately, you know, the most potentially dangerous consequence is what we saw. on January 6th at the Capitol is actually political violence, that because you're unhappy with the outcome of the election, rather than beginning to plan your campaign for the next election, you go and try to take over the Capitol building and search for the vice president in order to hang him. You know, that's kind of, you know, a nightmare scenario, and it is a complete undermining of the very basic democratic notion that elections count. And one of the one

there's an election, there's a peaceful transfer of power, and that's how democracy works. Yeah. Okay, I want to hear from Justin two. I want to, I think this is the moment where, as you're describing this, what I'm imagining is the feeling or sense of disenfranchisement and on two fronts. You know, one is that democracy doesn't really include me. It has nothing to do with me. It's a bunch of elites. It's a bunch of other people who I don't connect to, who I don't know, I don't know anybody who's a representative, and therefore they don't speak for me, and yet I live in that context. And then on the other hand, there's the erosion of trust in institutions because many large institutions have proven to be untrustworthy over time. They've taken advantage

of their position in society or culture, whether that's the church, or whether that's big, you know, business Enron, you know, whether it's Occupy Wall Street. Like, those are, some of the reactions that we saw to institutional power that failed us or failed some part of our culture and society. And the lack of equity and inclusion across society seems to have exacerbated over the last decade or so with just the way that the financial system sort of operates and works. So that seems to have created this vacuum or space for a new narrative or story to be propagated by, let's say, you know, Fox News or conservative media that kind of connects with that sense of being left behind. And then you can kind of make up whatever articulation of truth

or facts that you want, because whatever it is that's on the other side is, as you said, sort of vilified or is falsified, it's fake news and so forth. Because there's no way to sort of disprove in one's lived experience, which is the only real reality that a lot of people, I think, can trust and rely on in their own experience. I got to imagine that a lot of people that went to the Capitol on January 6th had never been to DC before, had never actually seen the White House or any of these places. And they'd only sort of existed in this imaginary idea where the swamp needed to be drained and all these crazy things were going on that Q&N said were happening.

So it feels like when you expand polarization to the breaking point, you sort of of fall off a cliff, go down the rabbit hole, and you're willing to, or are very vulnerable to be exposed to all sorts of other new mythologies that cause you to behave in those violent ways. So help me understand. I mean, is that the fault of social media? Is it effective use of social media? Where does the responsibility lie, given all that? And tell me if, of course, I've got anything wrong. So, Chris, maybe I'll just, I'll add one thing to your kind of diagnosis there. Sure.

What's happening in the States right now? And then maybe Paul, if you want to kind of take on that question that he just asked about, you know, is this the fault of social media? So one of the things, I mean, you know, you're kind of referring to like, what are the underlying cleavages here that are driving people apart? It just feels like we don't causing people to behave. Okay. The thing is we're not coherent as a, as like a culture or as a, you know, as a people, as the American people. And so we kind of talk about each other in these polarized terms. And it feels like that rift is like really fundamental. And one of the things that's happened is social media has exposed us to the other.

And we look at the other and we say, I don't know you and you're not like me. And now we're fighting over what is real. and whether democracy exists for us or for the other or for both. You know what I mean? Yeah. I mean, I would certainly, you know, say that based on the review of the social science research, we looked at, I mean, certainly there are a variety of different cleavages in American society, but one that you didn't point to, which is important, and that we took, you know,

some pains to look at and to cite empirical research on, of course, is race. And we quote in particular, Lillianna Mason, Julie Ronski, John Kane, who looked at surveys across thousands of respondents across seven years. And I'm quoting, a wellspring of animus against marginalized groups in the United States that can be harnessed for political gain. We go on to point out other results that are similar that point to a concern. about the ideas that non-whites will have more rights than others in future as compared to 16% of non-republicates who hold that belief.

Just before we got on this, I saw a tweet that Tucker Carlson's on tonight going on about replacement theory and similar ideas. So this is a very, you know, very concerning piece of what's happening in America right now. this is the reality of the American political moment. And one of the things that we wanted to kind of get across in the report is to paint a picture of the situation we're in. Because, you know, to some extent, the question isn't what is the role of social media's, our social media in some kind of, you know, clinical, utopian notion of democracy. But what is the role of social media in the democracy we've got? And what do you, so, I mean, that leads to a very interesting question and maybe starts to get into some of your report.

What do you think a healthy culture or civilization that is using social media in a healthy way looks like and behaves like? Well, maybe before we get there, I'll kind of let Paul answer the flip side of that question, which is the question you just asked. Okay, great. So that question was, who's to blame? Yeah, I guess it seemed to me, yeah, Chris was asking, is social media primarily to blame for the problem? Well, before venturing an answer to that, and I don't think there is a certain provable answer to that, primarily to blame. I think the first thing to say is that a wide variety of factors contribute to political divisiveness and pushing political divisiveness to the extreme that we see in American society today.

And we've mentioned most, if not all, those factors. There's right-wing talk radio. There's Fox News and partisan cable television. There's continuing animus across the race line. There, you know, there are political elites, leaders, you know, and the recent, you know, period that's probably relevant would probably on that score begin at the early 1990s, when the style of leadership in the Congress began to change very discernively. The rise of Newt Gingrich and the throwing out of traditional, you know, sort of niceties and, you know, sort of niceties and. courtliness in government that had, at least for some generations, been the rule. So you've got all kinds of factors, and they all feed into it.

And I think, you know, beginning sometime relatively recently, social media became one of those factors, not in the sense that it is the prime cause of polarization. It obviously, you know, wasn't the cause of polarization in the 70s. 80s and 90s as divisiveness was growing in this kind of right so you know and we'll look we should come back to that point in connection with the particular study you cited the age group comparison study we'll come back to that in a second um but once once social media but became a significant factor in our politics which probably started in the 2008 2009 yeah i mean sort of post-obama right right yeah well obama's people harnessed social

media for the first time in a presidential election, and that had an influence. But then, you know, much more so beginning in the 2010s. And it really heated up once Donald Trump, a master manipulator of social media, the disinformation purveyor in chief, you know, used that tool, social media to, you know, to continually and consistently heightened divisiveness and sort of enrage and inflamed people. And, of course, he wasn't the only one. The really important distinction, I think, is between primary cause and accelerant. Primary cause and a heightening factor and intensification factor.

And Facebook, when they fairly disingenuously talk about this, repeat over and over again, there isn't evidence that we're the primary cause, that we started all this. No one is actually saying that they started all of it. What they do is they provide a medium that accelerates the process, makes it far more intense. So one of the things that I've been noticing, and I'm, I don't know, I'm going to bounce around a little bit probably, at least when it comes to, actually, I'm going to pause and I don't want to do that yet,

because we're still talking about the polarization piece. We're talking about the lead up to this. I'm just like, yes, I mean, the question that I'm kind of, of trying to understand is culpability and is the underlying societal aspects that are already there that were present and that by creating a medium that was bi-directional. You know, I guess like what I'm trying to turn on is my Marshall McLuhan hat, you know, the medium is the message. And as we shape our mediums, our mediums then shape us.

And as you said, like Trump was a very effective media manipulator, but he also started on television. And so the other thing that changed that Obama really didn't have the benefit of was, I don't know if it was around 2012 or 2016 or something, maybe it was 2018. When Facebook started to notice that there was a lot more kind of zumbification of news feed consumption happening as a result of highly produced video becoming popular on the platform. And this was something that Facebook Watch wanted to build for. They wanted to go after YouTube ad dollars. and they put a lot of money. I remember, you know, I have tweets about how my Facebook watch icon was always,

it always had like a red number one on it because they wanted me to go and watch the thing and try to get rid of the unwatch count. And you couldn't. It was literally designed so that you had to go to Facebook watch because they wanted to amplify that kind of content. And then what they found was that, I suppose, people were just turning on videos in the background and not really paying attention and just becoming less active and less engaged. and therefore less valuable users.

But Trump did something really, you know, talking about meeting the moment. As a television communicator, he brought that style of bombastic communication to a platform that previously was far more static, you know, whether it was just static images or static text to become much more like a, I mean, the McLuhan sense, a cool medium, but more interactive, more engaging, more demanding of your attention. You had to know what he had to say. So that just seems like that was a mastering of the medium built upon the fissures that already existed in culture. And Facebook sort of just sat there and was like, this is great. Growth is good. What am I missing? Well, I mean, I'll just

jump in, Paul. Maybe you can take in there too. You know, one of the things that, you know, is clear in the literature is that there are you know there are a variety of different behaviors that are tied to the issue of polarization and its relationship to social media and media consumption generally yeah so it's not necessarily any one thing you know per se and you know researchers have looked at this from a variety of different angles in fact I was looking today at another meta study that's been released that looks at, you know, well, really dozens of social science studies on this issue over the course of a period of years. And there's a lot of different ways to think about how the various ways that people interact on platforms and specific platforms.

You know, it's not the same across different platforms. How their behaviors, you know, may differ across different cultures and different media. environments and different geographies. Because, you know, what happens here in the U.S. is not the same as what happens in, well, Bosnia and Herzegovina where, you know, a recent study found a slightly different effect than what some of the studies that focus on the U.S. have found, for instance. So there's a lot of nuance here. You know, I guess one of the things we can point to, though, is what we know about with regard

to Facebook itself, both from the. company and then from information that has been reported about the company. So you can kind of look at announcements the company's made. It's, you know, head of integrity, Guy Rosen just a couple of years ago, issued a statement about what the company was doing to reduce polarization, which, you know, kind of implicitly suggests that the company understands that there's a connection. and the investments that the company was making towards that. We, of course, know about various internal research that's been revealed more recently by the Wall Street Journal, which points to a variety of different phenomena which are related to the idea of sewing division or helping to exacerbate division in society.

Well, in that particular case, and I think this is really important, right? Because one of the ways in which we're trying to understand how to either regulate or change the functioning of these platforms is to look at what levers they have at their disposal. And the thing that you're specifically talking about is the meaningful interactions initiative, I suppose, where they looked at what they tried to do was actually to correct for the video zom bifurcation that I was just describing by correcting the optimization of the news feed to focus more on close ties and on promoting. promoting comments. And they just kind of let the algorithm run and what they found and Jonah Priddy, you know, opined about this was that the things that inspired conversation between close ties on a social network were the things that were the most vitriolic, the most enraging. But, you know, when your crazy uncle says something about Q&ON, you're going to comment on that and tell him he's your crazy uncle. And to the algorithm, that looks like things are going

great. But in fact, that's what's causing these negative consequences. And So from a levering perspective, Facebook made a change in an attempt to improve things and had the exact opposite effect. So how do you sort of square and think about these things when it comes to either regulation or what can be done? Well, you know, I think what Justin raised and you've just expanded on is a very important point because if we're first trying to figure out whether there is a phenomenon here that deserves attention, if there is some relationship between social media.

a destructive form of divisiveness, one important place to look is what is going on at the company itself. And just to elaborate on the story you were just telling, everything you said, I think is completely correct, but the story goes further. In-house, you know, the data scientists and engineers came back to management and said, we've now studied what the company has done, and our conclusion is this backfired, rather than accomplishing what we wanted to. accomplished, we're kind of accomplishing the opposite. And what followed from that was potential fixes were proposed all internally, and those were turned down because, as you indicated earlier, there was a danger they would cut into engagement and ultimately the bottom line. So that story, which the Wall Street

Journal, I think, laid out effectively and told in several different contexts, is a story of a company that actually studies itself pretty diligently, comes up with solutions to the inadvertent problems they're causing, and then sets the potential ameliorating steps to one side because they threaten the business model. And that is an important fragmentary piece of evidence that there is a problem here. I mean, Facebook would not be repeatedly studying problems, concluding that we have problems, proposing how to fix them, and then turning the fixes down if there was nothing going on. Okay, so let me push on that a little bit. Because in terms of evaluating, you know, progress, and I'm really not on one side of the other of this. Like, these are really deep,

hard problems with tradeoffs. That's kind of what it comes down to, for me, I guess. And so I don't think, like, you're never going to have cars that don't crash, you know, as I think Maseri said on Peter Kafka's Recode Media, you know, once you invented the car, you invented the car, you invented the car crash. And granted that is a highly regulated environment. And so, yes, let's apply similar regulations to social media. But what do those regulations look like? And so had Facebook publicized the result of those internal studies, which maybe is a thing that could be forced, what would you expect regulators to then do with that information? And one of the things I think that was reported on was if they removed the reshare button, a lot of the polarization and

negative consequences would instantly diminish. Now, that's interesting and also incomplete, because, for example, Instagram doesn't have a reshare button and yet has different problems, you know, associated with body image and things like that. So no matter where social media exists and where humans go, there seems to be some problems. So perfection isn't really something that seems attainable, but it seems like they're not doing enough or willing to, as you said, sacrifice some of the bottom line. And so where do you, and maybe this is an impossible question, but where do you see a rebalance or a more appropriate balance or measure of what success could look like if perfection isn't the standard?

Yeah. Justin, I think it's your turn. I'm happy to answer that, but I also don't want to monopolize it. Sure. Well, you know, I'll just say that maybe to your question of, you know, what should we do? You know, the very first thing that we should do, and I see a handful of people who are are academic researchers who are listening to this conversation, of course, is create some mechanism to make sure that the type of data that the Wall Street Journal revealed is possible

for external independent researchers to come across in a more reasonable fashion than we have in place at the moment. This is a complicated thing. There are legitimate issues. to do with trading social media data, making it available to researchers in a way that respects privacy, respects free expression, doesn't create problems of its own or have its own issues. But this is not a question that has not been discussed, has not been, policies haven't been theorized or come up with. There are literally pieces of legislation that are in consideration, you know, unfortunately don't seem to be advancing very quickly, that would address some of this. And so, you know, that's one of the first things is to get to a bit of transparency. And I would add, too, that that type of transparency would help to understand the linkages between different aspects of our public sphere. What does the media?

What is its role? What is the role of political elites? What is the role of, you know, other actors? What, how do we think about this thing a little more holistically so that we can have broadly policies, whether there are media policies or social media or tech policies, communications policies, that help us to produce a more healthy democracy? And then I'll actually, I just want to add maybe one thing and then I'll turn it over to Paul, which is, you know, the title. of this conversation is Facebook polarization inevitable? I mean, I would say no, you know, it's not inevitable. In fact, there's a lot of interesting conversation these days around what and to what extent social media could play a role in depolarizing or reducing division in society. And I think

that may be a very fruitful area of discussion. They're definitely academics that have done very interesting work on what affordances and what changes could be done that might help create space for deliberation, help create space for people to, you know, at least not exacerbate. Let me throw some out there, right? Because there's been some examples of product decisions that have been launched. So, for example, in WhatsApp, they reduce the number of reshares to prevent this type of virality. on Twitter, if you attempt to retweet, which again is the reshare action, without reading the article first, then you're asked, hey, are you sure you want to retweet this? And at least, you know, initial research found that, yes, that actually helped people sort of slow down. You know, it added a little bit of friction and made people be a little more thoughtful about what they shared. In other cases, there was another example that I was going to cite. Well, for example, on Instagram, Instagram has actually added a lot of,

features and functionality to help with bullying, with blocking different words, with different types of controls. So those things are happening to some degree. Do you think that those things are effective? I would offer that, you know, I think those are examples of things that should be experimented with. The platform companies should be given some degree of praise and reinforcement for trying those things, but they're probably not enough. And, and And, you know, beginning with the kind of, you know, significant increase in transparency, providing enough information about how the algorithms, you know, rank and recommend and remove content, so that outside people, whether they are researchers at universities or policymakers on Capitol Hill,

or the public at large, can see how these decisions get made. That would be a big first step. But I think you can also gain insight by looking at things that companies are already doing. On a number of occasions, Facebook has actually announced that in anticipation of unrest in society, we're about to modify our algorithms. We're going to turn the dial, as people say, break the glass, whatever. And then a few days later, they actually literally say,

and now we've turned it back. So there was a version of Facebook. Yeah, a version of Facebook with apparently some employees refer to as the nicer Facebook or the nicer news feed. And then there's the not-so-nice news feed. speed, which is the norm. Well, one big question that should be discussed publicly is why not do that more systematically? Actually, Monica Bickert, the vice president for policy,

content-related things, was asked that question at a congressional hearing in April. And she said the answer was, well, when we do that, when we basically make the filters stronger and comb out more potentially incendiary material because, for example, there's an environment like the days after the election in November 2020 or the days before the anticipated verdict in the trial of the Minneapolis former police officer who was ultimately convicted of killing George Floyd. They turned the dial in anticipation of that verdict. She said, we don't do that all the time because that would cut off too much free expression. Okay.

Worrying about false positives and taking too much stuff down. is a very valid concern. You want to promote, you know, free political and cultural debate as best you can. But that's not the, that shouldn't be the end of the discussion. That's like the beginning of the discussion. So, you know, tell us what you are doing to fine-tune the algorithms to diminish the amount of false positive material that you're filtering out. Keep working at it.

Refine everything. And that should be sort of like a never-ending process with the goal of a change. achieving whatever you thought you were going to achieve when you turned the dial and made the filters a little stronger. I mean, that's the kind of thing that I think if it was debated in public and the companies had to really explain themselves more thoroughly would lead to greater insights as to how we could, you know, reduce the amount of harmful content and reduce the secondary effects, you know, the externalities that these companies are causing. Well, it's interesting because like the image that I have is how, you know, the police will send out like a police force or something ahead of, you know, when there's going to be, you know, riots or there's going to be some protest or something. And the police are there largely, at least ostensibly to keep the peace or prevent conflagrations of violence breaking out when there is this polarization in real life. And then afterwards, because the police work for us, you know, and are paid for by tax dollars, they are accountable to us. But because Facebook is paid by the advertisers, that type of accountability doesn't really exist because it's a private company.

It's a private actor. And so the dichotomy here seems to be a bit that, you know, the public square is actually run by a private actor. And so that is why they are really not that accountable to us and why, like, I understand what you're asking and proposing and suggesting. But that dynamic feels like it's not really, we almost don't have the concept in capitalism for that type of. It's a very profound point you're making. How do you accomplish accountability? It's not like police forces have all been completely accountable, but as you say, the theory of it is, and sometimes the reality is there is accountability.

The police chief gets fired. The city council announces a wave of reform, and we hope we muddle forward in some constructive way. But it's not a hopeless situation. Social media as an industry is anomalous in. connection with government regulation. Unlike the equity markets, which are overseen by the SEC, or a broadcast and radio, which is overseen by the FCC, you know, the farmers who are overseen by the agriculture

department to some degree, there's no one doing sustained regulation, sustained oversight of social media. The one really important law, Section 230, is actually a deregulatory law. It's protective of the social media industry. Well, that could all change. I'm not saying it necessarily will with our current politics. The very polarization we're talking about right now would probably preclude what I'm about to say. But in theory, there's nothing stopping Congress, as we recommended in our report, from, for example, empowering the FTC, the Federal Trade Commission, to basically write, perhaps with cooperation from industry, a code of conduct that would.

require, as a legal matter, the companies to exercise a certain degree of transparency to, for example, explain themselves in terms of how they, you know, modify their algorithms and why, if they only do it for short periods of time, they, you know, don't try doing it for longer periods of time. And that code of conduct could then be enforced. And the enforcement mechanism wouldn't have to be that crazy complicated. You would set up a system whereby the obligations that the companies are required to adhere to would be included in their terms of service with their users. Okay, let me pause you for a second.

And then I want to bring up Emily, actually, who can present something of the government's perspective. The FTC already has rules about ads and disclosure of ads, whether it's using, you know, hashtag-sponsored or something along those lines. And it feels like there's, one, not a lot of regulation, and two, there's just no enforcement dollars. So I understand, you know, roughly what you're describing. But how would we actually make that work logistically? Well, actually, Emily, what I want to bring Emily up here and invite her into the conversation?

Because I think she can maybe speak to some of the government side of this. Emily, do you want to introduce yourself and give us a little background? Sure. Thanks, thanks, Kristen, everybody. My name is Emily Tavilarius. I'm a fellow at the Georgetown Back Center for Social Impact. And I teach at Georgetown in Columbia. And I'm a former advisor in the Office of Science Technology Policy at the White House.

And so what I wanted to add here is, you know, this is such a fascinating conversation on so many levels. And, you know, of course, the short answer is like there's no, there is no short answer. right? It's like the most unsatisfying, yeah, the most unsatisfying answer over. But, you know, to my mind, a lot of the answer is rooted in having the right people at the table. And right now, the government does not have the right people at the table to be candid. Like they're getting the right people on the table. I would love for you to unpack that. What makes them the right people and who is currently at the table. Sorry, I must have lost you guys.

Oh, you're back. My question was, what makes the right people, the right people? And who's currently at the table? Yeah. So, this is, it's a tough one because I don't, I don't want to ruffle feathers. And it's hard to say that without, like, offending people candidly. But the unvarnished answer is that we have a lot of a very well-intentioned, well-meaning, well-read people from the policy circles in these positions of power and decision-making authority

who quite simply do not understand the technology that they are in a position to regulate. And when I say don't understand it, I mean, like, they don't understand, like, functionally what it is, how it works, and how it connects to the business model. Well, you can't send email over WhatsApp, for starters. Just let's clarify that. Yeah, it's very easy. The FTC, I think, is a really good example of this, where you have a building full of very smart people who really know their stuff, suddenly having to deal with companies that they fundamentally don't understand. So when they present a question or a problem to the company, they are met with what to them is a lot of technical jargon that goes largely over their heads. And the companies can essentially really pull a fast one on them in many ways.

They can talk their way out of almost anything because the people they're speaking with are lawyers who don't understand the product. And what we're seeing at the FTC, which I think is a really, really interesting case study here is they have very proactively brought people in who do understand product. They just appointed a new CTO, Erie Meyer. They're about to bring in a deputy CTO who is also an engineer by training. And I think what we're seeing there is the attempt to correct that and say, hey, what if we teamed out up programs? and people who deeply understand product with lawyers and help them write the, write the questions differently, write the regulations differently in a way that is in the language of the companies and of the products.

I don't know if that's satisfying, but. And who would be the right people to bring in? Are these people who work at these tech companies or who do you imagine being like kind of able to balance? Because it feels like one of the problems, and what you just alluded to is this core tension. And actually, Marissa, I believe who's listening mentioned this, is, you know, the bonuses and the, again, the rewards, the incentive structure at Facebook is not to go along with whatever, I don't know, interest of coming up with new regulations, you know, would mean. They need to keep going to grow, grow, grow. That's like the Facebook mantra.

That's the way they work. and then you have people who want to regulate and they're beholden to their base and their base is asking questions or saying, you know, I feel like I've been shadow banned and all this other stuff that just kind of is, you're like, what are you even talking about? And yet they have a perception who, you know, people who are outside the company who, like you said, maybe don't have the same level of familiarity with how these systems work and how, you know, AI as a kind of black box. Yes, you can like move some levers up and down and you should have some accountability for the outcome of your system. But it's also, at least at this stage of the game, impossible to wholly predict the

ways in which humans find ways of manipulating and outsmarting these systems. I mean, that's why it's really hard to open up these systems. So, let me give an example of this, just because I think it's somewhat relevant. There was a conversation yesterday on Twitter about NFTs, and specifically about, I believe it's beer. EnS and Steph Curry's OpenC account. And because your Ethereum address is public, I mean, that's how it works. You sort of put art or NFTs that you've purchased onto your OpenC profile. That means anybody who can mint these NFTs can sort of throw them around and, you know, put dickpicks on your, you know, profile because it's a public address. So even though it costs money to do that, people are willing to do it for the engagement, for the attention,

seeking for whatever, which is true about email. Any open system is going to be exploited by any actors, because that's the way it works. It has to function that way. So when it comes to this regulation, I guess I'm just worried that it's really an ongoing, nuanced set of tradeoffs and balances. And like I said at the beginning, the real question here is about the incentive structures and the decision architectures that we're operating within. And those are the things that I feel like we don't understand. One of the suggestions actually that Paul has is that Biden, President Biden, should come out and basically make a statement about what should, what we should do as a culture, as a, I guess a people about this problem and this challenge, that we're not operating as,

there's no unity, I suppose, or there's just this polarity. I don't know exactly what he would say, and maybe Paul has some thoughts about that, but we have to understand that it's not going to, we're never getting to a perfect place with our use of social media, because humans are imperfect. But we have to keep moving forward in some way that kind of, as you said, sort of, what was the word? You fumble along in a sense? Yeah, I mean, it's, and I think that actually, it's very similar to the way policy develops in a lot of ways. I'm curious what Paul and Justin think about this.

But I often, I talk to my students about, you know, software development as quite, quite analogous to policy. Like we software right exactly like it doesn't die right like it doesn't die until you kill it and like neither is policy But but but we treat them in a very similar way which is like we like write the thing and we put it out there to the world and we're like great cool it's done now and we don't really revisit it So I think that there's actually a lot more similarity in the way that technology is designed and developed and delivered and managed with policy, but the processes and the systems currently make it very difficult to operate in that sort of ongoing continuous learning, continuous action kind of way.

Yeah, actually, so one of the things that I want to point out, and I just pinned a tweet to Congressional Bill S-299 of the 117th Congress, which is what is, the Safe Tech Act. And this is from Mark Werner. And when I went down and I read the text of this, what I found was basically it was a diff, you know? I mean, yes, this is obvious to anyone who's like looked at, you know, congressional bills before and how they work and how it's like editing and inserting code. I mean, it literally looks like if I went to GitHub, it's like, okay, here's the before and here's the after. We're going to test it out. Now, the problem with a lot of legislation, as far as I know, is that you can't really do A-B tests.

However, we have a federalist system that allows you to do some of these tests at the state-by-state level. And that's why the federal government sort of overseas interstate commerce. But at the state level, you can do those A-B tests and see which things work and which things don't. And you can take an account local considerations that California may have to worry about that New York doesn't or that Michigan has to deal with or whatever it is. So that's a really interesting system. But it doesn't quite work the same way when it comes to software, especially the speed by which software can be tested, deployed, you know, shard it off where, you know, you do a whole experiment on

New Zealand, you know, and you run it for like three weeks and you've got enough data, you know, to make advances and change. Like, we've barely even gotten a bill written, you know, in that amount of time. So, like, is it just that like the way in which we approach, regulating the systems has to fundamentally, fundamentally change to work more like the way code, like software code is written? I mean, I think it partially does. And sorry, I've been talking a lot, so I'm going to stop here, but my short answer is, I think the short answer is yes. The long answer is no. And well, and the longer answer, which I will not tell you because I'm going to stop talking now,

it has to do with a need for more partnership between, I think, the private sector and government in thinking through the nebulousness that you were talking about, Chris, right? Like, we don't know what is going to happen with these products and we can no longer right the muddling through. That's not the case. Right. So there has to be a different, we have to have a different starting point. Yeah, Paul, Justin, like, because you guys have actually put together, and I pinned to

tweet with your recommendations in brief about what can be done. And so I really want you guys to now take this opportunity to dive into some of those ideas. Yeah, and maybe I'll turn it over to Paul to go through those in some detail. there are two pieces that really recommendations to the government and then recommendations, you know, to the platforms themselves. But just to maybe put an underline on the point you were just making, you know, we have in this country, you know, regulatory bodies, which have in the past, you know, dealt with new technologies and new media technologies in particular, and have over a period of years been able to make rules and set precedent. and then study the impact of those rules and then make changes along the way. And, you know, unfortunately, that process is bit broken, and that's a major piece of this. Unfortunately, you know, that process is suffering to some extent from the root cause polarization as well.

So, you know, that's something that we kind of thought about. But Paul. Well, yeah, I mean, we actually start, I mean, I started to talk about, some of our recommendations, we have a very ambitious recommendation concerning the FTC. I have no illusion that that's going to happen tomorrow or even next year, but in a better, more rational political environment, I think that the idea could be discussed. And that is to, for the first time, apply some type of systematic oversight to this industry. And I think the code of conduct that we refer to, you know, could

involve and incorporate ideas from industry, but ultimately it would be enforced by the government. And we've proposed that it be enforced through the FTC and its existing authority, you know, to address, you know, false and deceptive, you know, consumer practices. And the deception would come in if the company was required to make certain disclosures that part of their terms of service, And then if those terms of service were ignored, the FTC could step in. Again, you would need a more muscular FTC with more resources than it has now. Absolutely. As I said, there are a lot of hurdles here.

We also have suggested that Congress ought to consider encouraging the exploration of alternatives to the current business models. I mean, there are technologists and entrepreneurs out there who are imagining a different way to do social media. For example, to separate the sort of network membership component of the social media platform from perhaps the content moderation component and allow people who are on the basic platform to choose from multiple providers of content moderation, giving choice and giving the opportunity to explain. So let me jump in there because there's two examples of that that already exist that I think might be useful for you to be aware of and maybe you already are. But one is Block Party, which is a service that essentially kind of outsources that type of content moderation, content moderation as a service. And then the other may not exist yet, but this is something that Jack Dorsey from Twitter has talked about and maybe related to the blue sky, you know, decentralization of Twitter effort, which is the ability to choose different algorithms for your feed. So essentially, you know, for everyone. There's a bundle of or collection of content that you could be shown every time you log in. And some of that is just public content from the broader space. And then there's stuff that your friends or people you follow have recently posted. And the computer algorithm or the system algorithm, whatever, decides, you know, very, very quickly, very, very quickly, which things to show you.

And then as you scroll down, it sort of like pulls from that basket and, you know, theoretically, you may never exhaust that. you could choose different levers or different sources of algorithms that do or care about different things to show you. And, you know, I can say personally, I've been logging into Instagram recently and my Explorer page has been vacillating wildly. You know, it goes from, you know, tattoos and cocktails to the other day it was like puppies and ducks walking in the road. I was like, what is this? I was like, oh, there's a glitch in the matrix. I know what's happening. But the fact that you could choose, you know, for your own personal interest might be an opportunity for people who are building the future to think about. Now, okay, we've been going for about an hour. I really appreciate

your time and your attention in this. And I want to jump into another topic. But if you guys want to bail, now would be an opportunity. And I can open up the room to more folks. I will give you a preview of this topic and you decide if you want to stick around. So we've talked a lot about polarization specifically on, I would say, sort of news feed content. You know, Facebook as we know it, Facebook as it has existed since I want to say 2016 or so. And a lot of the regulations that we're talking about in some ways are retroactive or retrospective. They're about things that happened in the past, trauma that has occurred. How do we heal from that? How do we fix it? How do we prevent that from happening in the future? Now, prior to recording today, Shrep, the,

Facebook CTO, who actually met when he was at Mozilla, and I was working on Mozilla Firefox, has announced that he is stepping down. He's becoming a senior fellow. He's going to continue to help to recruit and sort of build out fellowships and stuff like that. But he's out and Boz is in. Now, Boz is the guy that's been working on Oculus and the Metaverse. And I feel like Zuckerberg has been teasing this for several weeks. So this is not new, I would say. I would say like Shrepp was slowly probably walking out the door and Zuck was conferring with Bos. By the way, I'm just making this all up. And essentially suggesting, okay, we're going to pivot this to become the

Metaverse company. We're going to be in VR. We're going to be in a different world. The news feed is gradually going to, I don't know if it's going to die off, but the world that we're going to be in is much more immersive, is much different than the one that we're in now. The regulations that will be written, or are being written currently are for two-dimensional media. So my question to you guys who are here now, who have just come out with this report, is how or does this impact the metaverse? What things should we be thinking about and preparing for and perhaps preventing when we

enter into that space? You know, whether or not it's going to happen, I don't know, but to me, this signals that Zuck wants to own the metaverse. And that's where we're going next. That's definitely for Justin. Well, I mean, I would, I'd point to two things, and I think they are consistent with the recommendations in our report. First again, is the transparency. You know, if you're talking about these, you know, far more immersive and tactile experiences that using devices that might have various inputs that are quite advanced compared to.

to our current inputs into social media systems, which are largely about text and typing and being on a two-dimensional screen. If we're adding in neural interfaces, and we're adding in biometrics, and we're adding in heads-up displays, and we're adding in various tracking of our phobia, whatever it might be, we're going to need

more transparency and the ability for people to study those things. And in this case as well, I think study everything, not only the types of questions that we've raised in this report, but also physical and physiognomic impacts of using these things, as well as mental health ones. So transparency is going to be key. And then the second thing, which is obviously far, way decades now past due is privacy legislation. We've got to have some rules that are consistent across the nation around privacy. And that seems to me to be probably one of the biggest weaknesses right now of the American regulatory regime with regard to technology and something that we're getting behind on

with regard to other democracies that take these things seriously. Interesting. Take them seriously. Emily, I was going to say. Emily seems to be in violent agreement. Yeah, I agree with all of that. And the other thing I would add here is that, you know, a lot of what you are just kind of running through as the potential, like, future state of Facebook is so much in this,

in the futurist realm, which ties back to. something that you said earlier, Chris, about not being able to predict the downside, right, of the technology, like how it will be, what's the black mirror, right? And the question is not, we have a couple seasons of that now. Is there going to, right? Like, we're not in a place where we're like, oh, is there going to be a black mirror? Yeah. Now we've seen that there will be, like pretty much always. And product teams need to operate in a way that assumes that is the case and be structured in a manner that is looking for and ready to respond when that happens. So what I'm getting at here is that part of, if they were smart, they would want to collaborate with government

because if what they do is build something that has a really bad, scary black mirror side to it, the government is going to come for them at some point, right? It doesn't matter who's in power. At some point, someone is going to intervene in some way. And it is... What makes you so sure about that? I mean, like, it feels like it's going to be two, like a dollar short, a day late, all that. I mean, it might be, but the issue is, like, isn't it better for the government to write

regulations that are actually, like, useful and implementable rather than write, regulations and policies that you then have to like bend yourself into bizarro pretzels. I wish we could get that first thing out, right? I mean, even Facebook, I don't know, like I listen to a lot of podcasts and the number of ads that I've heard where Facebook is like calling for regulation is more than one. It's been going on for a while. And yet we still don't have that. So I don't want to be like the, I don't know, the contrarian. But it's just I, as you said, the right people are not at the table. The incentives are not there. I mean, if people were, I don't know, if it's money or like what it would take to get the right people at the right table to write the right regulations that would actually bring about the things that you're describing.

You know what I mean? Like it just feels like the tech companies are the ones that are making all the money and are offering these lavish lifestyles. And, you know, you've worked to the government and it just doesn't seem like that's the case there. Actually, well, but this is this is actually, there's a, there's a really interesting, there's interesting proof against that actually. Okay. And that is that, and I am one of them, many of my colleagues at the U.S. Digital Service and 18F and who have gone to work for the, for various parts of the government, including the FDC, are people who left those lucrative jobs. Yep. And the bottom line is once they went into government, they saw how much further their skill set could go in the public impact.

And it's not for everyone, but I think there is, I think it's safe to say that there is a pretty sizable COVID. of disillusioned technologists of all strike, right? Product, data, programmers, designers. I mean, clearly the people leaking the stuff to the Wall Street Journal. They should go get a job with the USDS. They are the people, Chris. Like those, if any of you are in the audience, like, DM me. We need you, right?

Like, so those are the people who can make the biggest difference because they've had their hands on these products. So if they're tired of what it looks like, go help people who want to do the right thing, do the right thing, because right now they're kind of bumbling along. And they need your help. Totally. Well, all right. I feel like my metaverse thing wasn't exactly bitten, which is totally fine because

it's the future and who knows. And we've got a lot of stuff to deal with today. And right now, I did pin one tweet from Justin's tech policy press on an upcoming event on October 7th. Justin, do you want to promo that? Well, yeah, I appreciate the opportunity. We're doing an event on October 7th, 1 p.m. to 4.30 Eastern, which is on this issue that you raised earlier, or that Paul raised, I suppose, from our recommendations. Looking at the idea of unbundling social media, you know, what would happen if we had more decentralized or unbundled social media networks and components.

And you mentioned Block Party. The founder of Block Party is one of the speakers, along with a bunch of other interesting things. people, Corey Doctor, Frank Fukuyama, you know, Stanford political scientists, Daphne Keller, Stanford, Natalie Merakhal, Richard Rizman, Ramesh, Shrinivasa, Joan Donovan, just a handful of folks that are going to kind of think through some of the kind of policy and idea issues around that set of proposals, but also some of the legal and kind of commercial complications and kind of look at how that might happen or how it could happen, what the obstacles are, and think that through. So, you know, that's also pinned at Tech Policy Press for anybody on Twitter.

Awesome. Well, I think that's a pretty good place to end here. I do want to re-extend at Emily Tav, TAV. If you want to talk to actually going into the government and helping to become part of the solution to the problem, I, you know, believe that she's ready and willing to help you do something, which I think is great. And I think we'll end it there. On behalf of Brian, I want to thank everyone for honoring his anniversary because I support their marriage and that matrimonial bliss by hosting this myself. This show may not go out immediately this weekend because we actually

have the second episode of the World Cup of Entrepreneurs coming out, which is a super fun, super great episode. But obviously, this stuff is timely. Anyways, I want to thank Justin, I want to thank Paul. I want to thank Emily for coming up here. We've been going at this for over an hour now and we could just go on indefinitely. I would love to have you guys back on sometime when we've got some of these things written down in terms of how we want to see this happen. And then, I don't know, the whole Metaverse thing, it's interesting.

Any parting thoughts, anything that you guys want people to check out? You know, obviously that event is one. Where actually, Paul, where can people find this report? You can go to the homepage of the Center for Business and Human Rights at NYU. And right on the homepage, there's a banner across the top. You click on and it takes your right to the report. Is there like a short link or something? Not really.

I'm not sure. I know what you mean. I mean like a short web address. The landing page for the report is not a long link. It's quite modest in length. Okay. You can find it in my tweets as well. BHR.

that stern. N.YU.edu. Okay. That works. Great. Okay, guys. Again, thanks so much for being here. Thanks, everyone for listening.

This was another episode of the TechMeme Ride Home Experience, and we will see you here next week. Thanks, everybody.