Thanks. Thanks. Oh, what? There. Thank you very much. It's a pleasure to be here, and I I do talk a lot. So, definitely, moderate me. I'm someone who needs friction and moderation constantly. So, yeah, thanks thanks, Bobby. Thanks, Cent. It's nice to see some familiar faces, people I haven't seen in a while, but also new ones. So, you know, I'm gonna I'm gonna jump in. I might introduce myself too much. I'm a I'm a law professor who I teach law, but I'm almost always writing in other disciplines, in econ, in in, with computer scientists, like, in other fields, in in, always doing sort of interdisciplinary research. So this is another just another example of that. So I'm gonna share my screen. Use some cruddy slides. They're not great, but they get the point across. Some of the points across. Alright. So you guys can all see that. Right? Yeah. Alright. Cool. So I'm gonna talk about this paper, Friction and Design, but I'm gonna sort of set it up and set up the paper, but also set up a broader research agenda that so Bobby and I have talked a bit about. And it's a collaborative interdisciplinary research agenda, so I think this this audience might really might be appeal to people in the audience, but also sort of might be an opportunity for exploring future collaborations of different kinds. So for the past ten plus ten, maybe fifteen years, I've been working on things, like, sort of trying to explore how what we build shapes who we are and who we can be, what we aspire to for ourselves and for our children. So sort of thinking about world building, through techno social systems. And so what I wanna do is sort of give you a a sense of the big picture. Why am I doing this? Why am I thinking about friction design at all? As a non computer scientist, non I'm not a designer. I'm not trained in design in particular. So how did I get here? But also some of the more broader motivations to give a better sense of this the actual stakes, the complexity of the problems, and thus the complexity of some of those kinds of solutions that we might wanna speculate about and explore more systematically. And then I'm gonna talk about a category of solutions, which I which I call friction and design, and then friction and design regulation. And that'll lead into a series of projects. Alright. So I'm gonna advance the slides. There we go. So to for in the reengineering humanity book published back in 2018, I'm not gonna go through all of that, but it helps to set things up. Sort of I talk about, and explore humanity's techno social dilemma, sort of a wicked global global scale tragedy of common style dilemma, the the stakes of which are who we are and who we can be both individually and collectively across generations. I I can spell that out. If people are interested in q and a later, I won't do it now. It'll it'll take me forever. And it won't take me forever, but it'll take me longer than I wanna take. But the bottom line is that the root problem is not digital network technologies per se, but rather the underlying economic, social, and political logics that have driven and continue to drive the development design and deployment of various, technologies. And you can think one way to, and and, again, that's a complicated story to tell. There's lots of different pieces to a puzzle that when you put them together, you start to reveal, what the dilemma looks like. But one way to simplify it is to think about a series of, design mantras, that are familiar in economics, in engineering, as well as in law. Each of the mantras are listed on the screen, sort of, you know, maximize efficiency, minimize transaction costs, eliminate friction, seamlessly interconnect, increase the speed, scale, and scope of engineered interactions, and democratize whatever, like speech, knowledge, culture, access to blah, blah, blah. Doesn't make a difference. Fill in the blank with whatever you like. And each of the mantras are design mantras are superficially defensible. They all make sense to a degree, but they're deeply flawed upon examination. Human beings are necessarily inefficient. We need friction in our lives. Seams are critical to governance. We need boundaries for ourselves and our communities. Interconnection often should be seem full as the friction at seams provides opportunity or space for governance. Growth in scale and scope are often overrated as they have substantial hidden trade offs, usually involving externalities. Democratize, like innovation, has become a useless buzzword that hides considerations of power, quality, and meaning among many other things. And any of these things, I'm happy to talk about in q and a and go into detail if people are interested. The techno social dilemma that we sort of unpack is not exclusively a social media or even an Internet phenomena. Right? So the deployment of supposedly smart tech, including activity trackers in schools, for example, raises a host of issues, that involve the engineering of a generation of kids and their parents' beliefs and preferences about digital network technology, including twenty four seven bodily surveillance. We can consider even with this example a range of function creep phenomena, boilerplate creep, surveillance creep, social engineering creep, all of which can be associated with the engineering of human human beings. And we can kind of extend the function creep idea and sort of think about Pavlovian nudges. Thank thank goodness this isn't yet been deployed in our school systems, but, you know, the technology certainly exists. But I'm gonna you know, I'll skip over some of the the the ideas. There's lots of different examples of this that that of different social dilemmas that you we we've identified and often analyzed independently. And in the reengineering humanity picture, we try to put it put those things together. So I'm gonna skip over some of the stuff because I know I'm already probably eating in more into the time than I'd like to. So for example, though, we did run an experiment experiment on Facebook using fake happy birthday notifications to sort of identify how people responded in sort of automatic way even when they have reasons to, to stop and think about the veracity of the signals that they're getting. So, again, not a not a uniquely, Internet phenomena. You can think about the GPS. You can think about smart cars. You can think about workplace environments or smart homes. But the basic concern raised throughout the Reengineering Humanities book is this idea that we're engineering, we're increasingly building systems that engineer our behavior and how we think, often leading to automatic mindless style behavior. Okay. So let me let me skip over some of this. And, like, the basic ideas are, okay, so what do we do in the in in the reengineering humanity book among other things, we start to lay out a series of potential strategies for dealing with some of the the problems that we identify. Won't focus on the conventional list. Like, there's a bunch of stuff here. None of not not all of which is relevant, but if you look at the second and third bullet points kinda capture some of the ideas. Right? Creating gaps and seams between smart techno social systems and others making things more seemful, both to sort of, mitigate against function creep, surveillance creep, boilerplate creep, things like that. And so we talk about things like network neutrality style rules and different kinds of intelligent, enabled intelligence enabled control systems. So it's like net neutrality is one example, but you can think about the same kind of rules on transport smart transportation or smart electricity or power grid, engineering air gaps and incompatibilities between systems, intentionally, you know, zoning rather than filtering. And I'll come back to the idea of zoning online. Sort of age gating is one of the papers that is a friction design proposal that we're working on or study that we're working on. But then you go to the last part. It's engineering transaction costs and inefficiencies with a prosocial purpose. So prosocial friction in design. And some of the examples we'll talk about in contract law. We'll get to that. We'll get to in a second. Okay. So what the bay one of the ideas that leads from the diagnosis of the problem to some some more systematic study of friction and design is this idea that, wait, perhaps we should be engineering friction into some of our techno social systems so that we can slow down, stop and think, interact more meaningfully with each other, practice and develop the capabilities that are essential to human flourishing. Right? And the counter to that I mean, the other side is sort of how can we resist the temptation of building frictionless eliminating friction or building frictionless interactions. So this leads to a whole series of papers and projects that are all some of the some most of which are still underway. We'll talk about the friction design paper that I wrote with Susan in in primarily in in just a sec. But some of the other things include with Madeline Sanfilippo, smart smart slow governance in smart cities, sort of thinking about how to like, what are the set of questions one should be able to answer before deployment of any smart tech in a community environment. That build that sort of applies both some human rights thinking frameworks, but also sort of Ostrom style knowledge common and knowledge commons frameworks to think about community governance of technology. Some work on governance scenes, a paper with Paul Ohm that should be out any day now in the Harvard Journal of Law and Tech, and then another paper on net neutrality as an example of a governance scene with Blake Reid, both of which sort of entail a kind of friction at an interface, the purpose of which is to enable governance. And so we again, I'm highlighting these because I'm not gonna talk about them, but we can come back to them in the q and a later if it happens to sort of resonate with with, inter the the group. And then I'll probably come back a little bit to the the one in the bottom right corner, which is designing effective privacy preserving age verification systems, which is connected to a I filed an amicus brief in the ninth circuit, in the appeal, of a decision that's was striking down as unconstitutional to California age appropriate design code. And in part, that sort of motivated a project to think about, can we do, privacy, protecting, age verification to enable age gating online as we do often in often in in the offline context? Okay. So the ideas this of this emerging research agenda is it's interdisciplinary. You need to do it needs collaboration across disciplines. It's new research, new policy, new approaches to thinking about governance. It's a sort of a turn away from this a lot of the emphasis in the online world, in the tech space on content moderation and content moderation systems to focus on the design of the systems themselves and the intersection among subcomponents of of those systems. One of the obstacles in at least in The United States to any talk about any kind of government regulation that involves design or, you know, sort of restrictions on design or prescriptions about design is the first amendment. Okay? And so the the paper so Susan, Benash, heads the dangerous speech project. She's also human rights attorney in DC, and she's a Berkman Center, fellow. And so the two of us teamed up really to sort of think about whether we could clear the brush, some of the first amendment brush, the sort of immediate obstacle to any thinking about friction and design, laws that would sort of touch on friction and design, answers is there some space to maneuver within the first amendment where the first amendment wouldn't necessarily just preempt the thinking altogether? Because we've kind of found that a lot of times it seems to just preempt even the conversation. So that led us to start thinking about friction and design regulation as potentially analogous to time, place, and manner restrictions. Okay? That also got me thinking quite a bit about, oddly enough, traffic calming. So I spent a a summer ago, I spent a good two months sort of deep dive into civil engineering literature on traffic calming, sort of figuring out, like, speed bumps and and and traffic design, In part because it was sort of like, how do we slow down sort of some of the designed interactions online so that people have a chance to think about what they're doing? And it turns out that speed bumps and other frictional measures offline that we accept and and often take for granted, maybe we can learn to appreciate them a bit more and sort of think about what the what what those things would look like if we design traffic calming or other kinds of calming measures in the online space. And so if we think about it in the online space, there's things like passwords and CAPTCHAs, which are examples of sort of digital speed bumps that slow us down to achieve a particular kind of moderation or a particular kind of social objective. Before I jump, again, I I forgot that I had the slide, but sort of the the setup in part is so this we've recently filed a or submitted a proposal to the NSF to sort of look at friction to design or the governance of friction of friction in design, in a more systematic way. The as the, the aims at the bottom sort of suggest, we're gonna tax taxonomize prosocial friction design, study its effectiveness in different contexts, and then be able to recommend governance of how how to think about the governance more systematically. And we can, again, we can come back to this later. There's also a workshop that we've organized that's in at the HHII conference in Sweet shortly in Sweden in June that's focusing on friction in often in the context of the design of decision support systems, but also in other other contexts where you want to have friction in the design of AI, hybrid human AI decision making systems. Okay. The the the so the paper with Susan again is about clearing the first amendment brush. And it turn well, it it turns out that the friction design is actually quite as I sort of mentioned before, it's quite common in the offline context. You can think about the familiar speed bump. Right? And just, the thing about speed bumps is that they add friction. No one likes driving over speed bumps. It's not it's not pleasurable to anyone. No one likes it. But we've we kind of tolerate the existence of speed bump because we understand to some degree its purpose. Right? And the way that it's interesting. I mean, there's the obvious friction of just any bump in the road, a stick in the road causes some friction. Right? But it's the friction of going over the bump in terms of the of the drivers being forced to sort of internalize the external cost of the speed of their driving speed. Right? That can moderate lead to the moderation of their behavior. I mean, of course, you can still just speed and and go over the bump. It doesn't stop or prohibit driving at a certain speed. But the friction associated with the bump and it can bumps are designed in lots of different ways. By the way, it's kind of interesting, but I won't go go down that road. But the the design of the bump can cause more, like, people to internalize things differently. But it's an architectural feature that creates friction for the instrumental purpose of inducing behavioral change, speed moderation, but also to serve a set a set of different social ends. Right? It's safety, and security in in a particular context. We don't use speed bumps on all roads. We use them in particular context where we care about where there are safe safety concerns, but it's also shared use of the road. Right? There's it's a multi user resource. Right? So there's got pedestrians and bicyclists and kids playing in in a neighborhood as well as the drivers. And if you want shared you wanna manage shared use of the resource, that is a social end in and of its in and of itself. And so Susan and I developed this framework, descriptive framework, largely just to sort of help taxonomize and and understand how we might identify different kinds of friction in frictions in design and how they work. I won't go through all the details. Again, we can come back to this if people are interested. But we started by thinking about it and applying it to the the simple offline examples of traffic calming measures, and then we wanna sort of move into the digital world. And, the first and most straightforward example is is a CAPTCHA, which I'm sure everyone on the call is is familiar with. It's a useful digital speed bump. It's not perfect. There's plenty of you know, we can critique it for a variety of reasons. Right? So it it's not great for for certain communities or worse off or the it's not good for disabled. Like, there's different ways to design CAPTCHAs to overcome some of those runs, but the but the basic idea of a CAPTCHA is that it creates it it, it's an architectural feature of a site. It creates a slight time delay. You perform a particular particular task to show that you're a human, not a bot. That generates reliable evidence, and and enables a an authentication process to work. Okay? And what's interesting an interesting hypothesis I don't know that we've that I've seen anything that tests this hypothesis. So if anyone knows anything, that would let me know. But my hypothesis is that, like, we encounter CAPTCHAs and people don't love them, and they don't like the extra work, but we tolerate them to some degree because we understand why what they're about. Like, we understand that that there is some friction and it's not pleasant, but it's because of security and it's because it because it enables the authentication that that users generally tolerate, CAPTCHAs. I may be wrong, but I think that the the point the reason I raise it is I think that understanding that when we can use friction and design, it's gonna be more palatable and or tolerable to users when in fact we can explain to them what the purpose is, and if that purpose itself is acceptable. So we may come back to that with with some of the other examples. Okay. So we also well, this is the the fast forward what's up fast forwarding restriction, which we talk about. Right? It's a friction that generates delay, requires more effort to get your message to reach a wider audience. WhatsApp did this on its own. It wasn't a government regulation. It reduces virality. It reduces the spread of messages. It's triggered when a mess so, like, the trigger where it's always on or triggered is sort of the scope of application question is kind of interesting, but it's triggered when a message is forwarded a certain number of times. So, again, it's designed as to reduce virality. What's interesting about the WhatsApp example is it's content neutral. Right? It's on an end to end encrypted system. It has nothing to do with what's being said. It just has to do with how often or or it's or the speed or the spread, which will you know, which we may come back to as being a relevant factor. So when what Susan and I tried to do in the paper was, again, to sort of think about when the state was so friction and design is something that private companies could do on their own. Communities can install them on their own too in sort of commons context. But also in some cases, it might be the case that we think that the state, that it makes sense for the state to require some friction and design regulation, where the private market wouldn't otherwise do it on its own. And again, Susan and I were concerned with the first amendment hurdle. The first amendment seems to present a strict scrutiny. Strict scrutiny is the standard of review you would use under the first amendment often when it's content based, a content based law. And often, if people think initially think friction if you're gonna regulate the design of a system, it's it's their speech. It's the it's the it's code after all. And so that's gonna trigger strict scrutiny. But Susan and I wanted to explore friction design as a as a potential content neutral time, place, and manner restrictions. Akin to the WhatsApp example sort of is a one way to easy way to see what that might look like. But to be clear, not all friction and design regulations or friction design measures would be content neutral time, place, and matter restrictions, and not all time, place, and matter sorry. Not all time, place, and matter restrictions are gonna involve friction and design. Right? So we're kind of looking for that sweet spot in the middle where there's where they overlap, where we think that there are some friction design regulations that are if they are designed to be or written to be content neutral time, place, and manner restrictions that they would be more feasible from a First Amendment perspective. The First Amendment wouldn't get in the way. Okay. So what does that mean? This leads us to, you know, generically to talk about digital friction and design measures of a variety of different kinds. I'm not gonna I don't think it's worth time wise going through this. I probably only have, like, what, sent like four or five minutes, three minute, four minute. Okay. So I won't go through the details of this slide. You saw it in the paper. Some of the examples that we talk about are alerts. Right? So pop ups where you get virtual drag, you've got to clear the pop up. There's the friction. And generating that delay, providing selling information. You could there's a you can design alerts in lots of different ways. This is an area I think that's worth studying. By the way, I should say this. I maybe should have said it earlier with the NSF rules, but I didn't. One of the reasons why we we in the NSF proposal, one of the arguments we made for why we need funding and why this needs to be a re interdisciplinary research done by academics and and done by practitioners, but not necessarily. We can't really rely on the of course, alerts exist on lots of social media platforms. Right? But we don't know what, yeah, we don't know what experiments big tech runs. We don't have access to the data on all the stuff that they're doing. We don't have access to all the experiments that they run that they don't tell us about. And so you need independent independent research on different kinds of friction design measures. You can't just rely on the fact that there are some that exist on certain platforms. I mentioned I should have said that earlier. Okay. So the the friction and design stuff that we talk about in the time, place, manner sort of at the end of the paper, we suggest we suggest the idea that there should be a research agenda. We suggest that there's you could have time based delays, like what kind of friction? It could be time. It could be informational, it could be task based. And what triggers the friction of design could vary from, for example, geographic distance when things are traveling over across jurisdictional boundaries. But we'd also propose the idea of social graphic distance. The trigger could be that when I'm communicating with, known friends or people I have strong ties with, that's different than we would have a different sort of approach to the friction. There might it might be frictionless or less friction than when I'm communicating or sharing information with strangers, non ties. But that's kind of speculative at the end of at at the end of the piece. Some more recent stuff that tries to put friction and design into practice in particular context is a paper with a computer scientist, Moshi Moshi Vardy, in digital contracts, where we propose to abolish altogether, click the contract mechanisms, that produce digital boilerplate and replace it with a, digital bet digital contracting systems that leverage friction design, to enable actual deliberation, to generate to generate reliable evidence that people understand what what it is that they're doing. The simple point about digital contracting systems, which you can generalize to lots of other systems, is that they're designed to eliminate friction, okay, minimize transaction costs at least in one direction. Right? So contracting is frictionless. Thinking is frictionful. Right? There's stacked friction in the direction of deliberating or thinking, but there's no friction in the other direction. I'll give you another example. We could do a live experiment. I'll jump out of the contract example, then we can come back to contracts if you're interested. If you have iPhones and you wanna play around while we're doing this, this is this is a a simple experiment. I'll also just show you slides if you don't wanna pull out your iPhone, what it's gonna look like. But it will take more it'll take probably two minutes sent. So I hope you'll we can do the experiment if people wanna sort of see to experience friction stacking. Right? This is the simplest example. I give it almost every talk, so some of you may have seen this before. But if you go to your settings and you turn off geolocation services tracking, which I'm sure everyone on the call has already done, but just just in case you haven't, go in there, go into your settings, find privacy and security, go to locations. They call it location services, I like to say, tracking. But turn it off. Everybody more or less there? Okay. So now I just want you to turn it on and think about what you have to do. Right? So this is what you know, it's off and it's on. And this is if you've got an iPhone, this is what it would look like. Right? To turn it on takes one click, one step. It's on. And what color do you see? It's green. Safe. Good. Go. Right? Now turn it off again. Right? Turn off location services and note what you have to do and what you see. Right? So you get another screen. You get a bunch of text that you don't really can't it's hard to see. It's you don't really don't wanna read it because, heck, that takes time. That's a little bit of friction. Right? Then you've gotta actually hit it. You don't just tap it once. You've got another tap, which is the turn off. Oh, wait. It's red. Danger. Hot. Don't touch. Go back. Right? I hope people are seeing something along these lines. You might not depending on your phone. But the point is that in the direction of of tracking, geolocation tracking, or in other words, in the direction of no privacy, it's frictionless. In the direction of privacy, don't track me, it's friction full. Multiple steps, more friction, and it's a discouragement. And you'll see this asymmetric friction design or friction stacking all over the place online. You'll see if you pay attention to it. Now this isn't a dark pattern, by the way. This is a not so dark pattern. It's in your face. There's nothing dark about it, but it's still harmful. Right? It's harmful in the sense that it it it is tilted in one direction towards no privacy and tracking and against, prep there's lots of in between designs. Right? So there are a bunch of other design options. You could have symmetric friction, or you could have an asymmetric friction in the other direction, pro privacy, a pro social friction in design. But, of course, absent regulation or absent nudge or absent something else, you're not gonna that's not what you're gonna see. Alright. So I will stop, because I'm probably at time. The last thing I was gonna say was about this other project on pedagogical friction and design, and we have an experiment we're running with with password security interfaces with tech if people wanna go back to these slides, but sort of we're running an experiment now. And then we've got a thing on h gating, which I'm also happy to talk about. I'll I'll get out of my slides and see all of you because you're better looking than my slides.

Thanks. That's that's great. Do you want me to I mean, excuse me. So, yeah, Julie Cohen semantic discontinuity stuff is great. Like, that we need gaps for play is sort of like a motivating Julie was my mentor, by the way, when I was a student. I was I was her student. And she

she

features in, you know, page length quotes and discussions in reengineering humanity. So she's she's certainly been highly influential on my thinking. And sometime probably in many ways I don't fully appreciate. But, yes, so I'm 100% with you on that. And, yeah, so time, place and manner restrictions are things like, you know, you can't drive down the, you know, you can say what you wanna say, but you can't drive down the, residential street or whatever at at two in the morning with your loud speakers blaring. Right? So the time and manner in which you get your message across can be regulated. Those kinds of things are subject to a lower level of scrutiny, than if I if I regulate what it is that you're saying. Okay? So the time, place, and manner restriction gives more leeway to government to be able to regulate speech because they're not regulating speech based on who is saying what, but they're regulating they're regulating how speech is being communicated and when. Does that makes is that what you meant, Sandra, in terms of explaining real quick to folks?

Yeah. And the one other thing so that's right. How how strong an interest does the government need to be? But the other is, is the government employing the least restrictive means, which is the part of the strict scrutiny test, which is very, very one bottom line is if the if strict scrutiny applies, the government almost always loses because there's almost always a less restrictive means for trying to accomplish their stated ends. If intermediate scrutiny applies, you don't have to apply the least restrictive means. It just has to be tailored and suited to those ends. So there's more, again, more leeway. Part the interest is a little less, but also the means employed are a little bit there's a little more room. If that's so I'll stop talking. So the next next that's a helpful listener. Thank you.

Yeah. I mean, that's great. So the zoning based on identity depends upon so the one of the things that, like, I'm struggling with in this current paper with Jan and Noah, I won't get there. Noah Aptorp and Jan Schwarzchneider, there I got it right. They're both two computer scientists and we're thinking about age gating. Right? So, like, age appropriate design codes is one sort of age gating, not necessarily based on content because all of the conversation where you start talking about minors and adults immediately turns to porn on one hand or LGBTQ content. Right? So it's the it's the concern either, like, we're regular we're gonna prohibit access to porn is, like, the case in favor of it, and the case against it is, well, but then authoritarian or misguided states like Texas or something or parents, and I'm quoting someone else. This is not me saying it. It's just like the yeah. There's a there's a concern that, like, misguided or authoritarian states or parents are going to block access to LGBTQ content or educate sex education content. So so therefore, age gating we don't wanna build age gating systems. We don't wanna zone based on the criteria of your age. Right? And Jan, Noah, and I are sort of basically on one hand and then there's so I should say, there's those arguments, and then there's arguments like you can't possibly do it. Like, it's gonna be inaccurate. It's not gonna work very well, or it's necessarily gonna violate privacy. Because in order to do age verification, you're gonna have to give over all kinds of sensitive information to Facebook. Right? Or you're gonna involve the government in building a database of ruin that's gonna know everything that the person has done in order because they've gotta verify that you're of age at each transaction. And a lot of those arguments are just hand wavy superficial arguments. They're not grounded in the latest or current state of technology on, for identity identity or age authentication. And I'm not an expert on those technologies. Noah and and Jan are helping me to sort of figure this out, but it's it's the case for age gating or gating based on other characteristics, which I think is what you were Steve, you were more interested in this or maybe than than just age, depend upon what's feasible in terms of identifying the if it's almost a good thing of it as if then rule. If characteristics apply, then governance rules. In the if part, it's about what does that technology make feasible for authenticating. And the zen part is a political or social or community governance question about how we make and design and apply rules. But we don't in the conversations in law and policy as well as in tech, we don't necessarily separate those two things out. They end up becoming conflated. And it's very hard to talk about the design of a system that would make it feasible to zone so then we can talk about debates about what we ought to zone, like, what the governance rules ought to be. And so the friction design part is is basically Yeah. That was this

Can you send it to me? That would be that would be fantastic. And I'll send you the draft we're working on, so maybe we can kind of yeah.

Thanks so much, Brett, and Sanh. So I've been thinking a lot about the EUX provision that prohibits the use of an AI system that deploys subliminal techniques beyond a person's consciousness, which was literally the text that they use. And there's been a lot of critique right that the youth doesn't give a practical definition of what they mean by subliminal techniques and manipulation and so on and so on. But I wonder what you think regulators are worried about, and you've I'm sure you've had conversations about friction in design with, like, policy makers and regulators, and I'm, like, curious what you're hearing and the potential use of the framework.

Yeah. I mean, I I wish I was having conversations with regular well, I I do and I don't. So and, anyway, said mindful of time, I should stop because I'm I'm such a talker. I could already see Scent, like like, grabbing his chin being like, oh, boy. I wanna do the research first. I wanna know what kinds of friction and designs, how they work, whether they work I'm sort of wanna do the social and the academic and the technical science work first before I talk to regulators and tell them what to do. Because I feel like we don't do that. If we don't do that first, you end up getting ham fisted regulatory efforts that don't work. So I'm not actually talking a lot with regulators about stuff. On the subliminal stuff, it's subliminal advertising. There's a history of subliminal advertising and stuff. The idea is I think there's lots of different techniques where so so this touches on, like, dark I think what they're concerned with in Europe and stuff is dark patterns. Things that are basically taking advantage of predictable biases, cognitive biases that people have, and you're sort of leveraging them asymmetric friction, to take advantage of those known predictable biases that people have. But there are a variety of other ways, like engineering of people's beliefs, for example, that don't necessarily have to do with taking advantage of a cognitive bias, but nonetheless might be a kind of subliminal, you know, be a concern about subliminal advertising or subliminal techniques, right, sort of shaping what people think without them knowing that you're shaping what they think because you've presented certain information a certain way. But that that's the the techno social engineering stuff is all about sort of different techniques for for doing that.

think there's other people, so I wouldn't

Okay. So we could this is like a twenty five minute answer, but I'm gonna give you, like, the the gist of it. And I again, I could share stuff. Like, so we're running experiment. We're thinking about pedagogical friction design right now is largely focused on teaching. It's a kinda new I think it's sort of a newer novel kind of digital literacy. And what I mean by that, it's teaching people about what they're doing while they're doing it. So if you engage with the password security meter right now and I'm gonna use password security only because it's an easy example where we know what people don't know and we know how they work even though we should need none of us should be using passwords anyway. Like, we should move on. But putting that aside, we're gonna keep using passwords. If you engage with the password security me right now and you put in Jeremy, it'll tell you it's weak. You put Jeremy you put a one at the end, it'll move the bar a little bit. You put exclamation point, it'll go green and strong. You've just been misinformed. Like, password security meters misinform people, and they teach them the wrong things, and they teach poorly about password security while you're doing it. So we're running an experiment that would would would wanna teach people the right things about password security while they're doing it. So they're like, while they're learning, they'll learn something about password security so they actually are properly informed, better informed, when they create and that's just an example. So then if we can build the frictions that while you're doing whether it's pop ups or frictions or it's requiring you to answer a question or or, you know, the methods used to teach might vary. We were trying to we're testing a few. But if we can get some of that right, then when we go to contracting or we go to privacy consent mechanisms, if we can design those systems to actually teach people about what they're doing while they're doing it, they might actually be able to act upon their own preferences and not just be reacting to the stimuli that are designed to get them to respond a certain way. Does that make sense? So it's less about learning environments. Like, I wanna teach you about history of something.

Mhmm.

And how do I design a learning environment so that you can answer complex I understand that. That's a cool question, but it's not exactly what we're not that broad in scope yet. We're kind of a little bit more narrow.

I'm muted and blurred. As I pull this off my shelf, thanks so much, Brent. Freedom is an endless meeting, which is about democratic organizing. You know? It comes to mind that that's the the usual complaint about democracy is too much fiction, not too little. And and the argument for fiction, amounts to an argument for bureaucracy, which I love to hear, good argument for bureaucracy. Is that is that, but it's also the demon, especially for a lot of people interested in online governance. They're here to reduce friction. They came to the Internet from real world to have less friction. So so what's the what's your romantic spiel for bureaucracy?

I think we have good friction and not bad friction. We want friction that's pointed in the right direction, not the wrong direction. I've actually come around to thinking it's not about it's not about frictionless versus friction full. It's about prosocial friction, like friction that's the you know the instrumental function of the friction. Like, the friction that's designed with a purpose as opposed to friction that's either purposeless, and so it's just an obstacle that you're tripping over, or it's friction that's direct directing you in someone else in some other direction that's against your own or against the groups. So I it's not you know, like, Mike, I'm not gonna defend bureaucracy for the sake of bureaucracy other than, like, you know, there's reasons to have, you know, community governance mechanisms require different degrees of participation, different levels of expertise, delegation of authority in different ways. And so sometimes you need there's the defense of bureaucracy is sort of because governance sometimes requires it. But, you know, my I I guess my you know, I I don't know that, like, it's it's a all or nothing, turn it off or, you know, like, let's maximize friction or let's eliminate it. I think it's about, like, purposely using it in pro social ways. And that requires, like, more consideration of what it can do, and then it also requires considerate more debate about what the values you're trying to achieve are when you design things. Does that make is that responsive to your

Well, I mean, it only it only causes more questions, but but, yeah, that that gives me a sense of where you go. Thanks.

got nothing all day.

Thank you all.

Yeah. So thanks, Brett. That was really great and relevant. Seth and I talked about this a lot after talking about this yesterday. So I I think that, you know, Seth and I often fall into, like, professor x magneto type relationship on these sorts of things. I've I mean, a lot of my relationship working relationships fall into that, and I actually really like it, and I'm always magneto, so it's fine for me. But, I I guess, like, the the suggestion I would make is okay. Let's say we develop some efficient bureaucracy such that no one has to really work that hard, and then we end up with, like, a skills decline because people aren't cultivating their skills. Could we not say, okay. Let's take all that time we're spending we're not spending in meetings, and instead spend that time playing something like Catan or, like, have a a more specific intentional games night where we can cultivate related or overlapping skills in a way that's not bureaucratic, but still cultivates it's still skill building. It's still friction sort of experiencing, but it's more intentional. And it's no longer as coupled to the bureaucratic process. That's kinda where I would go with this, and I'm curious how folks and you, Brent, think about that idea.

Yeah. I mean, I don't know about the the the premise, but to start with, right, that, like, we let's suppose we develop an efficient you know, like, we eliminate the need to, you know, have meetings. I like meetings sometimes. I just don't want too many of them. I like this meeting. This meeting's been fun. It it hasn't been so bureaucratic, has it? There's been some friction, but, you know, Sense done a pretty good job of, like, managing things. Anyway and it's been it's been not gamey, but it's been fun. I I don't know. Like, I but, yeah, I think what you're saying makes sense. Like, how like, like, this we make this point in the contracts paper. It's a little different, but I'll just make it to sort of highlight I think what you're saying. Like, people who descend frictionless e digital contracting will often say I say it's autonomy destroying. It destroys the autonomy of people. People are no longer acting on their own interests. They're act they're just clicking away and they're not participating in the and the response is, yeah, whatever, Frish. Like, don't worry about that because all the time they save from actually having to deliberate and think about the contracts they enter into is now repurposed. They get to use it to play Catan and and do different things. But the question that the empirical question that never gets proven or or get gets asked much less answered, is this in fact that what peep what is in fact that's what people are what what people are doing with the supposed freedom that they've been given from having to contract, or they just drawn into other mindless, nonproductive, non Katan like activities where they're just still clicking and swiping away because, hey, they've been trained to click and swipe away to enter the contracts, but they're trained to click and swipe away and lots of other things. In other words, like, the the idea that a like, you've eliminated, you know, the need to go through contract. You know, contracting has become so easy and frictionless, so it's autonomy creating because it gives you more freedom, all depends on whether or not you've actually have time in the day to spend to do things of your choosing. And so I guess my answer, Daniel, would be, like, it all depends on whether we collectively all choose to do Catan and build Catan plane quarters or, you know, systems as opposed to other kinds of activities that continue to just sap our attention and drain, you know, drain our ability to to to otherwise flourish or, you know, play Catan or do things we'd like to do. I don't know if that's really respond if you're smiling. What are you thinking, Daniel? Like, hey.

You sidestep my question or something.

No. No. No.

I mean, it's it's a great answer. I I think that I mean, I I like I like I think this is a thread worth developing. Because, like, when I say katana, I don't mean, like, literally just a katana. I mean, like, skill building through play in a way that the skill building is actually more effective because it's more designed versus, well, maybe, like, an inefficient skill building. Like, if I'm in a meeting and, like, you know, I've run dozens if not hundreds of meetings in my life for hundreds of hours, and most people spend most of the time zoning out. Right? So I would argue that that's actually not a great format for skill building. Could we not take that time and use it more effectively and more intentionally, through other skill building activities? You know? People say, well, a meeting is not really a meeting. It's a ritual. It's a chance to be together. I'd say, well, why not use that time to have, like, a a a an actual dinner with a discussion topic? And that way, more people can spend more time meaningfully engaged. That's kind of, like, the best case scenario to your point, Brett. In practice, is that gonna happen? Like, maybe not slash probably not. And so having built in friction, even if it's inefficient in some on along some axes is at least more consistent because it's it's sort of more of a speed bump. I think I think that's kind of the other way to look at it. You know, you could say maybe we're exchanging one problem, which is friction for another problem, which is how do we prop up the kind of culture we need to to really commit to games nights and dinners nights if they're not necessary. So it could just be we're switching one problem for another. You could say, well, maybe that second problem is the better problem than the first one, and so we should solve that problem instead of the first one. I think those are the ways I could approach it. But I think that, like, if we're I think on some level, if we're saying, okay. Well, we can only have coupled, governance and seal building. If we decouple them, then that increases the amount of degrees of freedom we have to solve the issues, which seems like on some level is a good thing even if it introduces secondary challenges to solve. It's kinda how I think about it, but I could be way off.

Yeah. I don't think you're way off. I I get it's useful to sort of think it through. And now I guess one other thing would be to what extent does what does how far off you are or how spot on you are depend on the context, like the specific context we're talking about or the community that we're talking about. Do you know what I mean? Like so I think it's gonna probably varies considerably. There's not, like, a universal answer to the, you know, are we capable of, you know, decoupling and then sort of shifting to these more pro social, you know, activities within which we know we build skills and flourish as opposed to not. Part of it is like people's preexisting or already determined preferences about engaging in those activities will shape whether they'll shift to that behavior or not in the first place. And so then you gotta think you gotta think about, like, over time, where do people's preferences come? This that this is the project I this is the book I haven't written that I always wanted to write, but I can't get myself to do it. But, like, where do our preferences come from? Where do our beliefs come from? Like, how how over time do you shape them in prosocial ways without, you know, running into the obvious problems? Anyway.

Yeah. So before I tell you that, can is the chat gonna be saved somehow? Because there's so many good things in the chat that I have not been able to even to be honest, when I do these things, I'm such a talker. I listen to talk. I listen and talk, but I don't read the chat. So, yes, is that like, it will be

I

Well so I think I've I mean, I I like to hear how Seth is because Seth I've I've you impress the crap out of me when I see what all the stuff that you're doing. I I always have lots of things ongoing. I never short on ideas. I like to work with other people. I think I'm friendly and, like, you know, work well with other people. So I always am sort of able to do that. I don't know. I never work on weekends. I don't work after 6PM. Like, I play I have three kids. I play soccer four times a week. I play basketball. I, you know, I do I put sometimes during the middle of the day with students. So I've I've always it's always about balance. It's always about doing, you know, getting the right thing. But when I get into something, I get into it intensely. But, yeah, managing time, I don't know. I just sort of again, if a lot of things ongoing all at once and then just sort of I go with the things that are, like, the most pressing in my head or for which there are deadlines. For yeah. For getting in contact and stuff like that, again, like, the easy email works. I'm learning to like, I was on Slack. We had a governing knowledge commons Slack thing that I created years ago and then it died, and, like, no one used it. But then in the last six months, I've gotten back on Slack for this and for two other, like, contextual integrity stuff. And so Slack could work, but I think just email me if you're interested in stuff, any any of the things that I mentioned. Or if you're working on good things too that are related, send me yours. Don't be bashful. Like, we live in a like, it's very hard to find other people's stuff. But if you're working on cool stuff that it all relates to stuff that I've, like, hinted at or whatever, just send an email and don't be bashful at saying, like, you could use this or, like, or cite me or whatever. Like, I I I don't care. Like, I'd rather I'd rather get the pointers than, you know, not. So yeah. And if you know I mean, the the like like, for collaborative research stuff, it's it's just for the knowledge comes stuff we do, we run conferences, and we invite people and collaborate that way. Like, the NSF thing, like, if if it takes off, we will need to find other like, we'll branch it out because it's a planning grant, the first part, then it'll be a project grant. So it'll be sort of building out from there. If anyone on here is a knowledge commons person, we'd love to do GKC in Europe, by the way, since since you're in Berlin. Like, the the idea of building a a governing knowledge commons community in Europe to sort of fund and then run conferences and then kind of keep doing the social science. I think that would be kinda cool. Anyway, I I'm a rambler. They told you something. You gotta shut me up.

Thanks.