Tech Won't Save Us - Why We Need a Democratic Approach to Data w/ Salomé Viljoen

Episode Date: January 14, 2021

Paris Marx is joined by Salomé Viljoen to discuss existing proposals to expand individual data rights or treat it as a form of labor, why we instead need to see data governance as a collective democratic project, and how that would give us the power to decide what data is collected and what it’s used for.Salomé Viljoen is an affiliate at Berkman Klein Center for Internet and Society at Harvard University, and a joint postdoctoral fellow at NYU School of Law’s Information Law Institute and the Cornell Tech Digital Life Initiative. Follow Salomé on Twitter as @salome_viljoen_.Tech Won’t Save Us offers a critical perspective on tech, its worldview, and wider society with the goal of inspiring people to demand better tech and a better world. Follow the podcast (@techwontsaveus) and host Paris Marx (@parismarx) on Twitter, and support the show on Patreon.Find out more about Harbinger Media Network at harbingermedianetwork.com.Also mentioned in this episode:Read Salomé article about data egalitarianism for Phenomenal World.People who write about informational capitalism: Shoshana Zuboff and Nick Couldry on one side, and Jathan Sadowski and Julie Cohen on the side that Salomé prefers.People talking about data as property or labor: Andrew Yang through the Data Dividend Project, Eric Posner and Glen Weyl in “Radical Markets,” and Jaron Lanier.Proto-data egalitarian examples: Andrea Nahler’s proposal for a civic data trust, Barcelona’s civic data trust, the US Census, and learning from libraries’ management of public information.Support the show

What is democracy? Democracy is we all have a right of recognition in setting the rules that we are all mutually subject to. Hello and welcome to Tech Won't Save Us. I'm your host, Paris Marks, and this week my guest is Salome Filiun. Salome is an affiliate at the Berkman Klein Center for Internet and Society at Harvard University and a joint postdoctoral fellow at the NYU School of Law's Information Law Institute and the Cornell Tech Digital Life Initiative. Salome published a great piece last year in Phenomenal World looking at data governance and in particular making an argument for what she calls data egalitarianism. In this conversation, we dig into the efforts to create property rights for data and to give us greater individual rights to data. After reviewing those, we get into

Salome's argument for a more collective, democratic way of governing data that ensures social benefits, recognizes the kind of relationships between all of our data and how by treating it individually, we might not necessarily get the benefits that we might expect, and why a collective approach is superior than putting the responsibility on each of us as individuals to manage our own data. This is a really interesting conversation and I think a really important one as we consider what these future data governance frameworks are going to look like and how to ensure that we craft them in a way that is going to work best for the public good instead of simply helping corporations retain their power

and ensure maximum profits. Tech Won't Save Us is part of the Harbinger Media Network, and you can find out more about that at harbingermedianetwork.com. If you like this episode, make sure to go to Apple Podcasts and leave us a five-star review, and share the show with any friends, colleagues, or on social media to get that social proof and hopefully bring in more listeners. If you appreciate the work that I put into making this show every week and you want to support it so I can keep doing that, you can join supporters like Neil, Eric Copps, and Waking Rufus by going to patreon.com slash techwontsaveus and becoming a supporter. This week I'll be putting in the order for some stickers that will be going out around the end of the month or in early February. So if

you want to get in on that, make sure to go to the Patreon, become a supporter at $5 a month or up, and you can get some of those stickers as well. Thanks again and enjoy this conversation. Salome, welcome to Tech Won't Save Us. Thanks. So great to be here. It's great to speak with you. You had a fantastic article in Phenomenal World recently discussing kind of the governance of data, kind of digging into these proposals that we often have about the way that they should be regulated by kind of making them property or giving people more rights toward them, and laying out a different vision for how we should see data in a way that would position it as a collective project.

But before we get into that, you know, obviously, I want to kind of dig into your critique of the way that things are working right now. And you start by outlining a concept of informational capitalism, and how that kind of leads to a certain definition of the problem when we see data and the way that capitalism as we know it today is working. So can you give us a brief explanation of what you see informational capitalism as being, and then the problems that presents when we try to think about how we should deal with data as a result of this kind of orientation? Yeah. So it's really interesting to kind of start at it from that point of view. And coming to informational capitalism, it's important to say that I'm a legal scholar.

So I'm really looking at kind of how the law structures and facilitates particular kinds of relations between people, whether those are interpersonal relations or in the case of capitalism, which is to say that I think that writers writing about informational capitalism are really talking about how information and information production has really been coupled to kind of core logics of production in the digital economy. And so what that means for data is, I mean, information has always played a role in how people price things and how they produce goods and how they identify consumers. But that kind of value of information has transformed under what people call informational capitalism to really being a

core productive imperative of the digital economy. And so for me as a legal scholar, the question then becomes, okay, how does our law facilitate that form of production? And how should the law think about the social benefits and the social harms that might arise from that production? And where is it legally relevant for us to kind of intervene in a particular way? So the perspective I'm really interested in bringing is saying, you know, we've had privacy law for a long time. We've had data governance law for some degree of time. But how does our role as lawyers or more particularly as legal scholars kind of change when that information production and information governance task kind of evolves from one of governing interpersonal relationships to governing core forms of economic production. Yeah, I think that's a really good point. And so, you know, if we were thinking about informational capitalism in terms of the arguments that we've had out there,

are there certain examples of kind of how this is framed, just to give listeners an idea of kind of the people putting forward these notions and the way that they kind of frame what's going on here? Definitely some of the most prominent people putting forward ideas of informational capitalism. You have Shoshana Zuboff, who talks about informational capitalism or platform capitalism, or she calls it surveillance capitalism. Again, really kind of linking these surveillance practices to core forms of production. You have people like Nick Coldry, who talks about data colonialism, also really interested in forms of informational capitalism. I quite like the work of Jathan Sadowski in this space. And of course,

another legal scholar whose work I engage with very closely is Julie Cohen. And I'd say all of them are looking at how information production, again, becomes a sort of core accumulative practice under the digital economy and what that means in terms of the digital economy and how it's ordered and also what that means for in those scholars of what the harm is there or why we should be concerned. And that in turn informs, I think, where they go in terms of what we should do about it. So I would put kind of on one side of that category Shoshana Zuboff. I think that I would really characterize her diagnosis as one of the legibility harm. So she has a lot of very vivid language comparing surveillance and surveillance capitalism to almost a colonial enterprise, that our inner lives are this pre-contact colony that is now being invaded and strip mined for the profit of these colonial invaders, which is the Googles of the world that are now undermining this inner life of ours and turning it to profit. That's a very different picture than I would say people like Julie

Cohen or Jathan Sadowski on the other side. Julie Cohen sort of draws on Karl Polanyi, which of course was a famous anthropologist and historian who looked at capitalism changing relations of production. And she points out in a lot of her work that we now have this new relation of production, which is sort of that data relationship that I talk about. Less kind of that vivid colonialist imagery in her work and more just kind of tracing the change and the shift. And I think laying out for all of us that this is kind of a new relation of production and requires of us as legal scholars, very new sorts of responses. So yeah, I would just sort of say there's a range.

No, definitely. And I think that's a really good quick summary of things to give people an idea of who's talking about these things, what they're talking about, and then how that, you know, as you mentioned there, kind of leads to a certain idea of framing the problem and also framing the solution to the problem, right? And so in your piece, you put the broader kind of idea of what the solution should be into two main categories. And so I want to get into those kind of separately to dig into them really well. And the first one that you talk about is the propertarian reforms. So can you give us an idea of what these propertarian reforms are and how they understand the problem that is being created by the way that data is extracted right

now? In the piece, I lumped together a series of proposals or almost like conceptual approaches to what's wrong with data extraction under the theory of proprietarianism and proprietarian responses. And what a proprietarian response would be is to say, okay, we're confronting this world where all of this data, this valuable resource is being extracted from people and it's being extracted from them for free. They're not getting any payment back from this incredibly valuable resource that's making companies all of this money. And the kind of intuitive response to that is to say, well, it's our data and we should be paid for it. And taking that kind of intuitive response, what that means as far as legal reforms

is that you look at a series of reforms that propose assigning a property right or a labor right to data. So there are really prominent people who advanced this reform. Andrew Yang included a property right to data in his presidential bid here in the United States, and he's actually started the Data Dividend Project. Even Alexandria Ocasio-Cortez had a tweet where she mentioned that people should be paid for their data. And there are, on the data as labor side of things, people like Eric Posner and Glenn Weil proposed in their book, Radical Markets, one of the proposals that they suggest there is that people should be paid for their data labor.

And Jaron Lanier, who's been writing a long time about this stuff, also proposes that we should think of data as labor. And again, conceptually, what that means is that we sort of think the solution to the injustice of data extraction is to give people this right to this good that they own, that they can then take to a wage market or just a traditional exchange relationship and command a price for their data. And so what that proposes as a solution is a market-based solution where I'm a data producer and then I sell my wares to the large company. And what that suggests about that theory of injustice is that what's wrong with data extraction is that none of the value that's being created from data is being

redistributed back to me, who's a core producer. And that once I've achieved or gotten my kind of like fair price for my data in the data market, we've kind of achieved this meaningful redistribution and all as well. Yeah, no, I think that's a great description of it. And I will be completely honest that before I had read very much about these topics, like, you know, I was just vaguely knowledgeable about the data issue. I did see it through that data as labor lens, right? And that we weren't being paid for the data that we produce. And obviously, that just seemed like, I guess, one of the obvious ways to see it. But then, as I read more, including your piece, I began to see that, okay, maybe I wasn't really understanding this properly, because I had a very limited understanding of what the issue was, what the possible solutions

were, right? And so I think your piece does a really good job of kind of laying out the issue at that position and other positions and why we should be looking at a different way of approaching it, right? And I think that that's right. I mean, part of why I think it's important to write about these things and to give those theories their due is that there's a lot that seems really intuitive, appealing about that approach. And I try to really make this clear in the piece and in all my writing on this, which is to say that I think that that intuition is touching on a core theory of injustice that we do have here, which is a redistributive one, which is an egalitarian one, which is to say that it is the fruit of we as data subjects are being exploited or extracted

to create incredible amounts of wealth. And there is an inegalitarian flavor to that problem. And I do think that the propertarian responses offer us sort of like a ready-made and intuitive kind of egalitarian response to that. Again, I think there's value to understanding conceptually exactly what does make data extraction wrong, which will help us get around that. But I think you're totally right. And I think a lot of people who are on the progressive side of things and want to reform and democratize these technology companies start there intuitively. I want to do justice to that intuition. And I think you definitely do. I think you did a great job. I think you explain really well that just giving people this property right or this this labor right to data and then expecting that you know we'll get paid for the data that we produce doesn't necessarily give us much power in

this equation right because you know as you observe there are a small number of very large companies who are kind of controlling this data or extracting this data. And then all of us as these individual data producers will have very little power to demand, you know, say a fair price for our data labor or whatever it would be, right? Yeah, generally, the critiques of propertarian approaches can be, I kind of think about them as falling into two categories. The first is just kind of like empirical realities of the digital economy, which is to say, yes, you're talking about going up against incredibly concentrated and powerful companies that have a ton of information about you. And so they really know

they're coming into that transaction with like a huge relative advantage. And also just the fact that data producers were not all together in a factory toiling under conditions where we could organize a union, right? We're highly scattered across the world, engaged into these micro transactions with these companies all the time that are designed to feel as seamless as possible, which also takes away that rub of exploitation that might actually motivate people to try to organize for better conditions and better exchanges. But beyond the empirical critiques, I also think that there's a deeper conceptual problem with propertarian relationships. And I touch on that a little bit in the piece, but that's to say that, I mean, like if you take Marxist and other social theories of relations seriously, this might generally be the case, but it's certainly the case in data relations that

I don't really know what it means to say that data is just Salome's data and that's purely mine. Because the way that it's used and produced by companies is to make all kinds of inferences about people that share relevant population level features with me. So information that I'm uploading or I'm making a personal choice about could go on to have all of these downstream impacts on a lot of people who weren't parties to that exchange. And that's a big problem. As a legal scholar, if you're going to assign that wage right or you're going to assign that property right, you want as many of the consequences from that exchange to be contained to that exchange. And that just isn't the case with data. I think that's a great point. And it's one that we'll

definitely, I think, get back to in a know, in a little while, when we talk about your kind of vision for how we should be approaching data, right. But I think another really important point is one that I've been thinking a lot about is what would be required to kind of administer this system is to create like this vast new technical system to handle the micro payments and transactions that would be responsible to pay people for their data, right. And, you know, I think when we're in this moment and thinking about how we want to build technology and what technologies we want to exist and things like that, I think it's worth considering like whether it even makes sense to kind of build this massive new technical infrastructure to handle these payments when it's not even clear that we would

see significant benefits from doing that. And by creating that infrastructure, you know, there are potential labor issues, environmental issues that all kind of arise from that. And I feel like those kind of broader aspects of this also need to be kind of considered when we consider what a possible solution should be. Yeah, absolutely. I mean, when I kind of just contemplate what such a micropayment system would look like, I imagine that it would just bring to the front end of consumer relationships what's already happening on the back end with ad auctions and ad exchanges. And those are incredibly fine-grained and incredibly detailed surveillance architectures that we're talking about implementing for the sake of getting, you know, fractions of pennies for impressions and things like that. So no,

absolutely. I mean, I think, again, asking the question of whether or not the ends justify those means is really important. I completely agree. And now, so, you know, we've discussed these propertarian reforms, the proposals for them, the issues with them, but then you lay out a second category as well, right? And that's the dignitarian reforms. So what are those? And kind of what do they propose or what do they see as being the issue with data extraction? I kind of lay them out chronologically because I think in a lot of ways, the dignitarian response is what you see. Well, one, it's already encoded in certain legal regimes, most prominently the European General Data Protection Regime.

But it's also typically kind of the primary counter response to propertarian solutions, which is to say paying people at the point of collection isn't enough. There are all of these ways in which the downstream effects of data are just as much of concern as just whether or not I've been paid a fair amount for my data. And that we really need are these rights that give data subjects far more control over to what extent they are rendered legible and for which purposes they are rendered legible. And so getting back to Shoshana Zuboff, I really think kind of in her critique is this core theory of data extraction is wrong because it renders people legible to these systems to a degree that is kind of like a violation of their inner self and their autonomy,

and for purposes that then can be used to undermine their autonomy and their dignity. And so in response to that, you have sort of a set of what I call dignitarian concerns. And those really look at rights that really impose on data extractors the obligation to get really meaningful consent from data subjects. So not just casual click-through consent, but like actual meaningful consent, as well as a series of rights that aren't kind of just exchange rights so that I can trade my data, but that don't extinguish at the point of exchange. So just because I've exchanged my data or I've given Google my data, I still have rights that carry with that data as it goes through Google systems and that prevent Google from using that data about me in ways that

violate my rights downstream. And so in a lot of ways, these are a more robust solution, right? It's sort of saying these aren't alienable rights. they're inalienable rights. They don't extinguish at the point of exchange. And they impose far more obligations on Google in terms of they don't have to just offer me a price and see if I accept it. They have to impose certain meaningful conditions of consent, make sure that I understand how my data is going to be used, make sure that I have downstream rights. And this is a lot of what you see in the general data protection regulation. So I should caveat, it's not the only aspect of the GDPR. They have other more substantive provisions as well. So, you know, I think, again, this is kind of intuitively seen as the more robust response than the propertarian response.

But, you know, I think it still has certain issues. So one, again, if we kind of get at the intuitions of what makes datafication wrong in the informational capitalist sense, what makes datafication wrong as an economic mode of production isn't just that it's violating my autonomy. It's that it's materializing a social relation of production that is unjust. And so giving me as an individual data subject, all of these strong individual rights doesn't necessarily do anything about the sort of methods of production that we're sort of drafted into that are themselves exploitative or may sort of be used to marginalize or oppress or engage in systemic violence against others.

And that's kind of where I see the limitations of the dignitarian response. Definitely. And, you know, when you were describing that one, I think another aspect of that that really stood out to me was how you described that, at least this was my interpretation of it, that it seems like it's kind of a broad brush. And while, you know, it definitely recognizes the negative potential uses of data and kind of protecting people against those, it doesn't really account for the fact that in other ways, data can be used for positive uses that benefit people.

And that that is lost when we position it in this particular way that is focused on this really negative interpretation of data use, data extraction, and whatnot. Yeah, this is a classic critique you'll see legal scholars levy against things. Part of my problem with the dignitarian solution is that it is simultaneously under inclusive and over inclusive. So it is under inclusive of ways that data subjects can give their consent for purposes and for conditions that they may find perfectly agreeable, but their information is then used in ways that violate core rights or core kind of interests of other people, again, because data is relational.

But it's also over-inclusive in that it can give people really strong autonomy claims and dignity claims that kind of prevent their information from being shared for purposes that we might think are like core quintessentially socially beneficial functions. And so again, I think it comes down to what is the theory of what makes datification wrong at the heart of dignitarian claims. And at the heart of that, those claims, I think, are a series of legibility harm. And what I really want to reorient is to say, it's not whether or not I'm being rendered legible, it is for which purposes am I being rendered legible and under which conditions is that legibility occurring? So yes, I mean, I'm happy to talk about all the kind of ways that dignitarian

reforms foreclose socially beneficial data production, but I also think they allow plenty of socially harmful data production as well. I think that's a really good point. And I think we'll get to some of those kind of like positive aspects when we talk about your next proposition, right, the more positive one that you would like to see. But before we get to that, I did want to ask you about one other point that I've been seeing a lot recently, and I'm sure you have as well. And it probably actually relates back a bit more to the propertarian reforms. And that's this notion that these digital infrastructures and this data extraction is kind of creating a kind of digital feudalism,

where now we're all kind of sharecroppers on the platforms, plot of land or whatever. What do you think about that proposition and how it kind of frames the way that data is used and kind of the economic relation that is associated with data. The interesting thing is, is I see a lot of people talking about the digital feudalism thing as a sort of a setup to get to data as labor, which makes sense. It's just like a neat historical analog. You know, you had feudal relations and then capitalism happened and we've had industrial

ways relations. And so we just need to do that. But for data and it's maybe not wrong. I mean, there are definitely some compelling ways in which the analogy holds. But, you know, and I kind of end the piece by asking this, but, you know, I think we have hundreds of years of legal innovation and social theoretical innovation after the imposition of the wage relation as a response to feudalism. And by the way, the wage relation as a response to feudalism. And by the way, the wage relation as a response to feudalism didn't abolish exploitation. I mean, that's what the last

300 years of social theorizing about capitalism has been exploring. And so my response would be to say, sure, we may have digital feudalism, but let's understand this as an opportunity to theorize and put into place new social relations that aren't the old forms of social relations that we already have and that we already know are exploitative and are limited. Even in the ways that they might not be exploitative, they are easily kind of turned to the project of exploitation. So I think data relations offer us plenty of challenges. They may be futile at this point, but we also have an opportunity to put into place forms of relating to one another that might be meaningfully better than some of the responses we've had to IRL feudalism in the past.

Definitely. And obviously, one of the main reasons I wanted to have you on is because you also did have a proposal for the way we should move forward, right? And so I think that you've really explained really well how these kind of propertarian and dignitarian responses kind of position data in this really individualist way, are really focused on kind of granting people these kind of individual rights to kind of control their data or be paid by their data or what have you. And, you know, you kind of argue that we should maybe be looking at data in a more collective way instead, because of the way that it's so important for data to work together. That's really how it has its value, right? So you lay out this idea of data egalitarianism.

What is data egalitarianism? And why do you think it's superior to the two other approaches that we've already discussed? In a lot of ways, this piece is really just me applying theories that political philosophers have kind of worked through in terms of what makes relations in a society just. And I'm just basically saying, what would make data relations just? I'm just applying political philosophy, which I get to do because I'm a legal scholar. I don't have to come up with the theory. So what is data egalitarianism? Well, again, if we sort of take the view that data relations are relations of production, then what data egalitarianism is, is creating more equal relations of production

in this economy. So, you know, if you are part of a tradition that's committed to democratizing forms of private economic governance, then this points towards an agenda for data governance that is about democratizing these private economic relations of digital production. And so that's kind of the conceptual goal here. Normatively, what that means is that we need to kind of evaluate the relationships under which data production is occurring and asking, are these relations of equality or are they relations of oppression or exploitation or marginalization? And so that's kind of the diagnosis and the standard of data egalitarianism. I think that's a really important

way to frame it and to look at it, right? And so obviously we've talked about how these other solutions see data in this individualistic way and see the need to provide individual rights for people to control their data, right? And so what do you see as kind of the collective benefits that can come of, you know, this form of governing data in a more collective way? And then how would we go about, I guess, kind of exerting democratic control over the way that data is then used and governed? The core goal of talking about data egalitarianism for me, again, as a legal scholar, is moving from conversations about data subject rights to conversations about data subject and just generally people with a stake in data production

having rights of recognition in the ways that we kind of produce and govern data. What that looks like will probably be different in different settings. So one could imagine, for instance, if we're looking at like gig economy worker data right now, you could say that each individual worker, maybe we could give them stronger rights to like know, know what price they're getting when they like do each individual ride or have greater insight into like their individual routes and all of that. And that's like giving them a right to access the information that the Uber or the Lyft or the Seamless is collecting about them. But recognition would be to say, okay, what are the forms of surveillance that they're being subjected to? And can they have sort of like a meaningful collective capacity to set those interests as in a legal setting would be to say that they have as much of a say in setting the terms of that data collection production practice

as engineers at Seamless or Uber? And that would be kind of like a right of recognition approach. And again, what that looks like for different forms of data would probably vary. In the labor setting, what I think that looks like is to say that they have labor rights in negotiating the terms of surveillance that they're being subjected to. So essentially, it kind of repositions it not so much as people having rights to their individual data or being paid for their individual data, but having more actual power over the way that that data is extracted and used in the first place in a collective sense instead of just on

my individual Paris's data. Yeah. I mean, when I talk about democracy, I literally mean that. I mean, what is democracy? Democracy is we all have a right of recognition in setting the rules that we are all mutually subject to. And so recognizing that if you and I are both mutually subject to a system that is exerting coercive power over us, then we get a say in what forms of coercion you and I are mutually subject to. And again, I think this reorientation is really important because I think the legibility harm is the approach that says coercion is wrong. And the egalitarian response is to say, no, coercion, if one is

going to be coerced, requires that everyone who is going to be subject to that coercion gets a stake in setting the terms of that coercion. And that's kind of, again, the reorientation that I'm interested in. I think it's a really interesting way to reorient it and to see it, right? And to kind of see data in this collective way presents us with new opportunities to think about the way data extraction happens, to think about whether we even want data to be extracted in certain instances and not in others. So do you see this data egalitarianism kind of giving us more of an ability to look at this larger structure of data, I'll say extraction that has been created. And I guess to make collective decisions about whether we even want

data to be extracted and used in certain instances. And maybe it's not in some cases that could be really beneficial right now and could be like, do you see that as also being part of this larger project? Absolutely. I mean, I think the way I articulate this point quite frequently is I'm not interested in conversations about whether or not we're collecting too much or too little data. I'm interested in a conversation about changing the distribution function of the kinds of data that's being collected and for which purposes. So my aspirational hopes for this sort of theoretical reorienting is that it leads us to be able to say, well, yes, we should be collecting far less data about like Salome's

consumer preferences for particular types of shoes. We should definitely be collecting far less location tracking data about Salome that helps to train systems that are then used to monitor and detain undocumented immigrants, for instance, which is a thing that happened recently in the US. And we should be collecting far more information about, for instance, my water usage, if we want to administer smart, effective, responsible, and fair water distribution systems as we head into a world of climate crisis, like we're probably not collecting enough of that data. And we should, you know, our theoretical and our legal responses to data governance should be aware of that task. You know, I think we should be collecting a lot more data about wage violations.

We should be collecting a lot more data about international financial flows towards tax havens. We should be collecting a lot more data about individual people's CO2 emissions and water usage. And we're not collecting enough of that data. We should be collecting more of it. And our theoretical responses to capitalist data extraction should hold open the theoretical capacity for us to also do all of this extremely important, socially pressing stuff that will require data collection. When you talk about that legibility, a lot of my research recently has focused on transportation, right? And so just to give an example from the transport sector,

what we see is often that the data is collected on automobiles. And so that then encourages a greater investment in infrastructure and whatnot to promote the use of automobility instead of cycling or transit or something like that, because those things are not tracked to the same degree or not considered to be as important or whatnot. And so I think that's really important when we talk about making things legible. And maybe some things we don't necessarily want to be legible, and we can decide once we have that kind of collective power, that democratic power to make those decisions, we can decide if we don't want certain things to be tracked. But as you say, with like, say, a climate change example in water, there are certain things now that are not being effectively monitored. And then that leaves us open to decisions that are not in, say, the public benefit, because we have focused so much on, say, consumer preferences or other things that benefit capitalist corporations instead of the public as a whole, right? Totally. And I think, again, from the vantage point of

all of us sitting inside of a late capitalist crisis kind of world where digital capitalism is the latest form that that form of extraction and accumulation is taken, I understand the impulse to say legibility is in and of itself harmful or dangerous. But again, I think, is it the fact that we're being rendered legible or is it the purposes and the conditions under which that's happening? If the conditions are highly extractive and the purposes are private profit accumulation, then yes, the legibility is potentially problematic. But legibility is the first step towards us kind of forming collective ways of being and responding responsibly and fairly

to like pressing social challenges. So I just think it's worth disambiguating those two things, again, as a legal scholar in our law governing data production. Definitely, I think that's a that's a fantastic point. And so I also wonder, you know, when you talk about data egalitarianism, you mentioned how, you know, when it comes to dignitarian reforms, we've seen it's in the GDPR and some other kind of, you know, when it comes to dignitarian reforms, we've seen it's in the GDPR and some other kind of data privacy legislation. When we look at kind of the propertarian reforms, we've had presidential candidates or congresspeople promoting that kind of approach to data. So when it comes to data egalitarianism, have you seen any kind of examples

of this working in practice? I don't know if I would say there are like full fledged examples, but there are plenty of what I call like proto data democratic examples that I think are really neat. So the former leader of the Social Democratic Party in Germany or Social Democrats, Andrea Nalles, suggested putting something like a limited monopoly right on data collection for these private companies so that after a certain number of years, like let's say 10 years, which is what we use in the US for pharmaceutical management. After 10 years, all of these private companies that have collected all of this consumer data, that data set gets converted into a public good. And at that point, that's a public resource that now is managed by the state. Presumably, you would have then like researchers could kind of access that information for various sorts of purposes. Ideally, in my dream world, that would be managed by sort of like a national library or a national research agency and could be sort of used again for the public good. And maybe you could even imagine other businesses taking out limited licenses in that

data set to try to develop other commercial products. So that's one example. The famous example, another fairly prominent European example, is Barcelona has a municipal approach to citizen data that is managed publicly for the good of the city. So they also require various tech companies that come and operate in Barcelona to turn over certain information to the city. So they also require various tech companies that come and operate in Barcelona to turn over certain information to the city that they then manage as a public resource for the benefit of citizens. They also sort of have tried to open up all of their data processes to sort of create a digital civic life around Barcelona city data. And so those are sort of two proto-democratic examples from Europe. I'm really actually

interested in the history of the U.S. Census. It's a very old census. It is data collected about the public for the public. Social scientists have been developing amazing insights from U.S. Census data for a long time. And there's a really strong culture of managing that information in a way that is able to derive statistical use from that information, while there are very strong protections, legal and technical, against individuals being targeted in a way that might be harmful to them in the U.S. Census. Now, there's a troubled history to that in the U.S., and it's not been perfect historically, but it is an example of really meaningful purpose limitation and sort of

responsible management of public information for high stakes public uses in the public realm. I really like that kind of historical focus and that focus on existing public solutions that kind of could be adapted or learned from in order to develop these new solutions, right? Because I feel like so often when it comes to technology, we forget that there are these public institutions that we could learn from, or that could play a really significant role in improving technology for the public good, right? Whether it's, you know, learning from the US Census to think about how we collect and manage data in a different realm or in a different way, or even, you know, looking at universities and libraries to see how they could assist us in thinking about the ways that data

are going to be managed in the future, you know, and, you know, when you mentioned libraries, my thought went to, I think it was in Joanne McNeil's book, Lurking, where she talked about, like, if we wanted to have a public search engine that was really focused on elevating different sorts of things or elevating educational things or just having a different kind of focus, libraries would be the perfect institution to manage a search engine that would be in the public good rather than focused on, say, ensuring clicks and advertising and promoting Google's own products and whatnot, right?

But that's just to say, like, I think there's a really rich history of public innovation and public institutions that can be drawn from when we think about how technology can be used for the public good moving forward. Absolutely. You know, another example, I mean, you're, you basically spelled this out, and I'm just picking this up from a bunch of social science researchers, but my ideal world kind of dream scenario of public data management would, I think, involve something like the network of municipal libraries and research libraries that we already have developed for public information stewardship, just sort of being tasked with the management of the public resource that is data. Librarians have a long professional history of being able to do public information management and stewardship in ways that are sort

of like sensitive to the ways in which that task can be fraught or difficult or complicated, yet nevertheless kind of very public-minded and public-spirit-minded, balancing sort of the sensitivity of information with a commitment to access to knowledge as like a core democratizing force. And, you know, as we move into a digital world, I think it's sort of an open question what the role of libraries are in this world, and they themselves have been sort of transforming. So again, I think that there are so many counter-narratives or counter-examples of public information management for the public good that we can draw on in developing alternatives that start to become the things that we focus on and the things that we start to sort of develop solutions around when we reorient the problem. So, yes, I'm very into the idea of digital libraries and digital data management happening in libraries. I love libraries and I absolutely love that idea.

So I think you've given us this fantastic overview of, you know, existing proposals for how we should manage or change our approach to data. And then another one that kind of repositions this conversation and this focus in a way that is more focused on the collective good and the public good rather than, say, individual rights and things like that, right? And so obviously, as I said earlier, I am certainly not an expert on this topic. And so I was wondering, you know, is there anything that we haven't really touched on in this conversation that you think is important for people to understand or

to know when we talk about data governance or data management? No, I mean, I think the conversation's been great. I guess what I would just add to it is that, you know, I think at times it can seem quite conceptual or theoretical to say like, well, what is it that exactly that makes datafication wrong? And what exactly would the right solution be? And this seems pretty good, but not totally good enough. But I think it's really important, particularly at a moment where I think there's appetite and possibility to reform data production and the ways that we kind of regulate it to get it conceptually right. Because there's this amazing quote from this amazing race scholar, Gary Peller, who talks about the responses to racial injustice in the U.S. in the 1960s, kind of going down the route of formal equality and in doing that, marking the limits of their own reform.

And I think it is important to get these theories conceptually correct so that when we implement responses to them, we don't end up marking the limits to our own reform and thwarting things that we want to achieve or thwarting attempts to respond to injustice one or two steps down the road. So, you know, I think I would just sort of emphasize that in these moments of regulatory possibility and when we're really doing lively theorizing, it's actually, I think, really important to conceptually get at why we care about certain forms of injustice and let that really guide our institutional responses to that injustice. I think that's such a fantastic point and a great way to end this great conversation. Salome, thank you so much for

sharing your knowledge and your insight with us. I really appreciate it. Yeah, thanks so much. I really enjoyed the conversation. It was great to be able to have the chat with you today. Salome Filiun is an affiliate at the Berkman Klein Center and a joint postdoctoral fellow at the NYU School of Law and the Cornell Tech Digital Life Initiative. You can follow Salome on Twitter at Salome underscore Filiun underscore. You can also follow me at Paris Marks and you can follow the show at Tech Won't Save Us. Tech Won't Save Us is part of the Harbinger Media Network, a group of left-wing podcasts that are made in Canada, and you can find out more about that at harbingermedianetwork.com. And if you want to support the work that I put into making this show every week, you can go to patreon.com slash tech won't save us and become

a supporter. Thanks for listening. Thank you.