The Institute for Digital Public Infrastructure

@ UMass Amherst

Reimagining the Internet

Talia Stroud, Civic Signals

October 27, 2020

Talia Stroud from the University of Texas joins us to talk about her project Civic Signals, a project reimagining the Internet as a public space. She walks us through what’s wrong with the type of speech currently rewarded by Facebook and Twitter, and what it might look like to promote civic speech instead. Recorded August, 2020.

Check out Civic Signals’ website and subscribe to their newsletter.

Watch Eli Pariser’s TED Talk.


Ethan Zuckerman:

I'm thrilled that we have with us today professor Talis Stroud. Talia teaches at University of Texas at Austin in the School of Communications and also in the School of Journalism. She is one of the co-founders of the Civic Signals Initiative, which in my opinion is one of the most interesting projects out there looking really explicitly at this idea of how we can tell whether an online social space is having conversations that are useful and helpful for civic health. Talia, thank you so much for being with us today.

Talis Stroud:

Such a pleasure.

Ethan Zuckerman:

So what we're doing in this series is talking to a wide variety of people, about two really simple questions. What do you think is wrong with social media as we know it? And what steps are you taking to try to make it better? And given that Civic Signals is really deeply focused on this question of how to make it better, you seem like a great person to start with. So let me put you on the spot and ask you for a diagnosis. What's wrong with social media at this particular point in the middle of 2020?

Talis Stroud:

Where to start, right? So this is something that my co-founder of Civic Signals, Eli Pariser and I have been thinking a lot about. And we've really been focusing on your second part of your question, which is how to make it better. But let's start with why we need to make it better in the first place. And I think that in the popular world, there are so many things that are going wrong with social media right now. We have the in quotations, "fake news" or the spread of mis- and disinformation. We see people behaving poorly in these spaces and the proliferation of hate speech, et cetera. So I think those are kind of the surface level indicators that something is greatly amiss on social media.

Talis Stroud:

And if you dig a bit deeper into that, I think that part of the reason that we see this happening is because the way in which social media is structured, they have a profit motivation. And the way to gain profit is to make sure that you have lots of eyeballs and people returning to the social media platforms as frequently as you possibly can. And that they spend lots and lots of time there. And that's how the algorithms for many social media platforms are created.

They figure out how is it that we keep them there as much time as we possibly can throughout the day. And I think that that's actually really a problem because it creates scenarios where the loudest voices are those that are highly partisan, they are saying terrible things. And that gets people coming back again and again. And there's a reward structure there for social media companies because they get advertising dollars as a consequences of that. So I think that this building for profit and doing so in a way that creates algorithms that reward bad behavior is a really problematic aspect of that.

And if I'll add one more thing to this, although I could probably go on the rest of our time on aspects that need to be addressed, when the social media platforms have been called out for the problematic aspects of what they're doing. So, Facebook facing scrutiny or Twitter facing scrutiny for the proliferation of again, quotes "fake news" or the spread of mis- and disinformation, the response there in my mind has really been more of like a whack-a-mole strategy. Like, "Okay, there's the bad thing today and I'm going to hit it that mole. And then, Oh, there's the bad thing tomorrow."

And yeah, they're developing teams and coming up with some infrastructure to address this, but that's really focused on responding to crisis rather than proactively figuring out, "How do we move this forward?" And as you said earlier, how do we think about this from a, what are some of the civics signals or ways in which we can signal good behavior? Which is actually a really different way of thinking about what social media could and should do.

Ethan Zuckerman:

So just understanding the diagnosis, you in many ways, sort of a single diagnosis that explaining two symptoms, right? One symptom is this wave of mis- and disinformation that seems to be making it harder and harder to figure out whether we can take anything as fact that we're encountering online.

The second is incivility, abuse, hostility, all these sorts of things that potentially silence people because they simply don't want to deal with that behavior when they're interacting on the platform. And you're linking both of them really to two things, one is to a business model where these companies that are running Facebook, are running Twitter, are running YouTube, these are for-profit companies. At the end of the day, they make money through advertising. And so maximizing eyeballs, maximizing the people looking at the content is to their fiscal advantage. And then we're hypothesizing somewhat, but there's certainly a good arguments why this might be the case, that the algorithms that bring content to the front, that bring it to the top of the Facebook news feed, or that bring it into YouTube's recommender are favoring outrageous, eye-catching, highly emotive content because it's consistent with the business model for these platforms though, not necessarily consistent with what's good for us.

Talis Stroud:

Two things that I would quickly add is I think that what we see in terms of mis- and disinformation and people questioning what's real or not actually goes beyond online. And it can make people question those things then offline and think, "What do we know in any context?" Which I think is really tricky.

And then on the algorithms favoring outrage, we've done some research where we were looking at just for instance, on the New York Times, what of comments get the most recommendations. And it's those that talk about partisanship and use uncivil language. And if you have an algorithm that rewards things that get lots of interactions, like recommendations, those do percolate to the top. So I think absolutely fair to say that that's something that's happening there.

And then in terms of civic signals, what we're thinking about there is from platform behavior, from anything that happens on a platform, we send all sorts of signals. Right now, the way the architecture of the platforms is set up is that the signals that we're sending that are feeding algorithms are things such as liking a comment on Facebook or replying to something on Twitter. Those send signals. But they're signals that are really based on interactivity and engagement. And those are being used by algorithms to elevate content.

Well, we actually send a much broader array of signals when we interface with these platforms, the words that we write, the words contained in the articles that we share, the impression left by an image. All of these are also signals. And if we think about those as being signals, then there might be some content that's really civically rewarding, and that really helps to create a more public friendly space.

So we've been really thinking about these signals as we've been focused for so long on what's user-friendly, thinking very much at the individual level. And that's what it means to hit like on something. That's your individual behavior rewarding someone else's individual behavior. But what if instead of user-friendly we thought about this as public-friendly. So it's odd that we haven't really gone down this path and it's been so individually focused. But those sorts of signals are so much easier. It's so easy to add up the number of likes and use that to be an algorithm. But figuring out what are the indicators of civically beneficial behavior and do people get these civic rewards when they're taking part in social media? That's really what we wanted to uncover in the Civic Signals Project is thinking through, can we identify signals that platforms are actually having a beneficial influence at a societal or public level?

Ethan Zuckerman:

One of the things that you seem to be implying is that there's civic content out there, and we just have to work to figure out how to surface it. Is it that easy? Are people really using these tools for civic purposes? And if they are, how do we find it? How do we find the content that would be healthy and helpful for us as a public?

Talis Stroud:

So I do think that there is civic content out there and we see it. There are fundraisers where people are raising money to deal with a crisis in another country, for example. That's a very civic act. And we see people increasingly doing that, at least on my Facebook feed, for example, for their birthdays. "I'm raising money for such and such, a nonprofit organization." What a wonderful civic act.

Then you also see people trying to encourage others to sign petitions or to get involved in a variety of ways. Increasingly we see more and more hosts encouraging people to vote in the upcoming election. So I think that there is civic content out there and there's content that helps you humanize another person who you may never have really seen in that light. So I think it exists.

Do I think that that's all that has to happen here is just surfacing what's already existing? I don't think that's the case. I think that there's more to it than that. And I think essentially we have to think about the architecture and we keep using this metaphor of space because we find it so interesting. Eli really honed in on this. But it's the architecture that prioritizes your behavior in certain ways over others. And there has been research for instance, looking at if you design a space and you give all sorts of signals that you should be thoughtful there, like make it look like a library do all of that. People actually act very differently in that space than if you just say, "Hey, leave a message," or something more casual. And so I think the architecture of these spaces also has to be considered, or is the architecture promoting people's actions in a civic way or not?

And as part of that, I think that that's a call to action essentially to platforms is thinking through not only algorithmically, how you might surface some of this content, but how could you create products and experiences within these platforms, given the opportunity to be citizens in these spaces when we still need these public spaces, particularly now in the middle of coronavirus where there's no space to have public conversations other than in digital life. So I think it's a "yes, and" sort of a situation of surfacing this content that does exist. And making it possible for people who want to have those experiences to find the architecture that supports them.

Ethan Zuckerman:

I like the spatial metaphor a great deal. And I was actually on with a team of designers from Facebook who were thinking about some of these civic challenges. And the explanation that I gave was a church and a pool hall and a library are all public spaces. But they have really, really different affordances. They have different things that you can technically do in them. Most churches don't serve beer, most libraries lack pool tables. But they also have norms. The speech that is acceptable in a pool hall might be different than the speech that's acceptable in church, or that's acceptable in a library.

One of the things that is so incredibly challenging about a space like Facebook is that it is trying to build one room, one single type of room for every sort of frankly, public and private interaction. And it's incredibly hard to think about whether one room could ever meet those needs.

Let's get concrete on that for a second. What's the most surprising civic signal that someone has brought up? The most sort of unexpected or interesting one that's sort of come up in this conversation with a hundred or more experts?

Talis Stroud:

The one that has been the most vexing, I would say, and it's not perhaps surprising, but really thinking through is actually tricky. And so I would consider this to be surprising is how tricky it is, is there's a theory that dates back several decades called Agenda-Setting, which is the basic idea that if the media cover a particular issue, then that issue is seen as important by the public. And so there was this sense in talking to these variety of experts that that's kind of a responsibility in some ways in social media is to surface these important issues for the public to recognize.

But as soon as you get underneath that a little bit, it's actually much more complicated than that because you can have different subgroups that might have very, very different issue priorities. And how do you then think through what issues to prioritize if one subgroup has quite a different idea of an important issue compared to another?

and where we've settled that in chatting with again, many different people who had perspectives on this is that one important civic signal or one important contribution that social media can make is by surfacing the interests of a variety of different major subgroups. So that if you have one particular subgroup and I'm using subgroups a bit loosely here, but thinking about this in terms of demographic or major political parties, is if an issue of the importance of one of those, not only surfacing it for those people, but surfacing it for those that might actually not be part of that subgroup.

And that I think was, it was surprising to me how a theory that's one that I've taught in my classes for a decade and a half now that I thought was almost a done deal in some ways, turned out to be one of the most conceptually challenging ones to think through because there are so many diverse interests. And I don't think that social media typically consider a responsibility to different divisions within society, which goes back to what you were saying earlier of creating one space that caters to everyone. Well, everyone doesn't always have the same shared interests. So that was one that I found surprising in its complexity.

Ethan Zuckerman:

I had a funny example of that recently. I've been driving my neighbor up to doctors appointments in Southern Vermont. And I live in Western Mass and weirdly enough, the politics even going in that sort of 30 mile drive can change quite a bit. And here we are surrounded by Black Lives Matter signs. And I drove 30 miles up the road to the nearest big city. And it turns out that there are signs in people's yards that say Black Rifles Matter, which is certainly a very different sentiment. And it's one that it's a useful reminder that it is shared by some of my neighbors.

And it's certainly true that social media, for me, has surfaced an enormous amount of Black Lives Matter content. It has been less good at surfacing the Black Rifles Matter in part because I'm sure on some level, the algorithms know that that is not a political point of view that I support. But it was this nice reminder that we have the ability to inscribe our opinions into physical space, but we can often end up in these really badly circumscribed social spaces, instead of remember that that view out there.

You also mentioned that you got some feedback about the platforms doing a good job with these signals and affordances. What's a case where someone felt like the platforms were doing a good job?

Talis Stroud:

So in our focus groups, we ask people to tell us about experiences that they had had on social media. And some of the kind of surprising things from doing that, or people's incredible stories that they had about what they gained from being on social media. So it's the story of the grandmother that gets a chance to actually know her grandchild. It's the isolated individual who suddenly reconnects with their high school classmates. There was one story about a woman who didn't have the money I believe to buy a prom dress or something like that. And then everyone rally behind her to make sure she could get it. It's just these beautiful stories of people using social media to these incredible ends that have repercussions for individuals and for us socially. And I think that we tend to forget that a little bit.

So in our survey, we actually asked people to tell us whether they thought that social media made things better for them, worse for them, or didn't really have any effect. Most people think that these in general made their lives much, much better. And even though there are certainly worries and concerns, for the majority of people, these are a net good in their lives overall.

Ethan Zuckerman:

It is really interesting to think back into the distant, distant past of March, 2020, where we were sort of in the full flush of a tech lash, this sort of movement where there was a real reconsideration, "Are these platforms good for us? Do we like what they're doing to us individually? Do we like what they're doing to us as society?" And in a funny way, the tech lash to a certain extent has been put on hold because at the moment, these are the only public spaces we have. So, we continue to have a tech executives going in front of Congress, but I don't really hear a lot of people talking about boycotting online space for the simple reason that right now, online space is space. And particularly as winter comes to some parts of the country, it's going to become more and more sort of important [inaudible 00:18:14].

Let me throw out the question you [inaudible 00:18:18] involved with this project, all the experts you've talked to, all the survey work that you're doing, you're putting together a set of recommendations that could be incredibly helpful for a new social media platform. It could be a real roadmap for someone like YouTube or Facebook. How are you going to get these very, very powerful companies to take this advice? What is your feeling about how this work is going to be received and how it might be picked up?

Talis Stroud:

So there are different philosophies about how to go about doing this sort of work. And the tone that we've taken is we've actually chatted with folks within the social media platforms as we've gone throughout this process. So we've kept them in the loop. Like, "These are some of the things that we're finding." And so I think we've been collaborators in a sense, quite external to what they're doing. And it's not something that we needed to call them about every week, but we have made an effort to make sure that they're fully aware of what we're doing so that they can adopt these sorts of strategies as they were thinking moving forward. So that's part of the strategy is making them into the process as we're going throughout it.

And then after that, our full intention is that we'll have briefings inside many of the major platforms to let them know like, "Hey, here's what we found at this point. Here are the sorts of things that we're recommending." And then I do think that some of the information about how the platforms are performing, sharing that publicly, I think also has an effect in terms of how the platforms will respond to this.

Ethan Zuckerman:

Yeah, it feels like in a certain way, for this to succeed, there needs to be a bit of a movement behind it. There needs to be a bit of an insistence that the platforms take our collective civic life seriously. And for me, it's one of the places where I see Institute for Digital Public Infrastructure and Civic Signals most allied, which is that we're sharing this common vision that these places, whether we like them or not, are going to be critically important for civics, even if the pandemic ends tomorrow. These are simply the spaces in which we are most likely to have these critically important conversations. And if they don't have the norms or affordances that help us have those conversations, they're going to remain an impoverished space. And over time, they're going to impoverish our democracy more broadly.

Talis Stroud:

Yeah, I couldn't agree more. And I think that this movement part is a key part of Civic Signals. So there's a newsletter and we're working on developing a community around all of these. We've been really active in terms of seeking out advice along the way, and bringing people with us slash having people bring us with them because we certainly aren't the first people to ever think about these ideas. There really is in my sense, a community forming around this. And I think that we hope that Civic Signals can start to be a place where people come together and kids start to have these conversations, both within platforms and creating new social media outlets and with others like you and the work that you're doing. That we can all come together to say, "Look, now we're all part of this movement that we think this is a really, really important thing for democracy." To have this conversation and I think through the influence of social media outlets and to think through what the positive vision is that they should be working toward as opposed to whack-a-mole for the negatives.

Ethan Zuckerman:

So Talia, with that shared vision and the notion of a movement in mind, what are things that people can do to get involved with Civic Signals, to learn more about it? Give a takeaway, or sort of a next step that a viewer of this might be able to take.

Talis Stroud:

So in terms of learning more about it, Eli has a TED Talk that is a fantastic place to kind of hear the thinking about space and how this relates to it. And so that would be a great place to learn more. We also have a website where people can sign up for our newsletter. And it's a fantastic newsletter if you haven't seen it. And so that's a great way to get involved. And then we're also will within this next month, be sharing a number of new ways to get involved as well as. So urge you to sign up for that newsletter so you can be first to hear about all of those things. and then we'll be sharing the research also through that newsletter and in the website. And we'll also be doing some public writing about it. So much more to come. Encourage as many peoples as are interested to get involved.

Ethan Zuckerman:

Talis Stroud, Civic Signals, and University of Texas at Austin. Just a pleasure to catch up with you and hear what's going on. I hope more people will check out this project. I think it's one of the most important pieces of research being done right now to sort of understand what signals we would have to monitor to have a really positive vision for what's going on in the future with social media. Thank you so much for being with us.