Natalie Mullins 0:25 I've been looking forward to this talk. 0:31 All right, how's everybody been enjoying Vancouver recently, so far on your trip? 0:37 Excellent. 0:38 As like the, you know, the local person, I'm just like really happy to like welcome people to my city. 0:43 If you have any questions or looking for recommendations, you know, hit me up anytime. 0:52 And I think we're ready to go. Speaker B 0:53 Okay. 0:54 Hello? 0:54 Hello? 0:57 Does it work? 0:58 Is it gonna be louder or? 0:59 It's good. 1:03 Okay. 1:03 Okay. 1:04 Hello everyone, I'm Natalie. 1:06 I'm an investor at Bain Capital Crypto where we focus primarily on open protocols and cryptography. 1:13 And we are actually investors in Bluesky as well as a couple other projects in the ecosystem. 1:18 But I also come here as someone who's just internet native and has been publishing original content, music, writing on the internet for a while. 1:25 And I spend most of my time in my investing capacity thinking about new media and new business models. 1:31 And so what I wanna talk about here today is, you know, I think most people here, we share a lot of values when it comes to, you know, why we're here and what we think is important about this. 1:41 But I think that for better or worse, I think when it comes to winning the masses, I do think that we generally have to sort of win on economics and not just values. 1:52 And so I think one of the most important and interesting sort of open questions in this world right now is like, what are the sort of the economic primitives or substrates or questions that we should be thinking about when it comes to how we actually manifest the world that I think a lot of us here are interested in seeing come to fruition? 2:08 So maybe just to start off with something very basic. 2:12 You know, when I talk about sovereign media, when we talk about sovereign media, what exactly are we talking about? 2:17 And I think for me, it's sort of very basically not necessarily a format, but really a paradigm that is about people having the capacity to own the means of their own cultural production, which I think is specifically about, do you have control over how your content is produced? 2:32 Do you have control over how your content is distributed? 2:34 And do you have control over how your content is monetized? 2:37 And I have this caveat here at the end about without sacrificing access to scale, because I think historically, when you look at— go back as far as you want to— I think the kind of Faustian bargain that underlies the media economy has been for a long time that if you're a creator, you will trade some amount of agency, some amount of creative control for access to scale. 3:01 For me personally, if I had to pick somewhere to fall on here, I think I'd probably want to fall on the patronage model, honestly. 3:05 I think that comes with a lot of great benefits. 3:08 But I think generally speaking, this is kind of the bargain that people find themselves in. 3:14 And I don't want to make this sound like it's inherently evil all the time. 3:18 I think there are better and worse places to kind of fall on the spectrum. 3:22 But I think it's important to kind of— I think this is a useful sort of mental model to keep in mind. 3:26 And obviously, there's some good examples of this. 3:29 I think for most people here, like a Google search, where it's like, if you make yourself legible to us, You let us index you, we'll send you viewers, right? 3:37 So obviously, there are some trade-offs that are associated with that. 3:41 Scale requires standardization. 3:43 Standardization requires legibility. 3:44 And legibility really constrains what gets made. 3:47 And I think the thing that I kind of wanted to double-click on here is that I think a lot of times, again, people here who are in these spaces share certain values. 3:56 And we think privacy is important, but I think It's really important to understand or just to keep in mind that a lot of times the reason why people will give up their privacy or will surveil themselves or allow themselves to be surveilled is not just because they don't value those things, but because there's a real economic cost to that a lot of times. 4:15 In our society, in our attention economy, however you want to call it, legibility and economic value are very deeply intertwined. 4:25 Oftentimes when people want to, you know, if people want to put privacy first, that oftentimes is associated with a real cost, a real economic cost. 4:33 And so I think the big point here is like thinking about this in terms of a problem that is systemic and architectural and not something that is merely the consequence of bad choices. 4:44 So, and I think another thing kind of related to that that I think is important to remember is that, you know, I think there's a lot of reasons that this is important. 4:52 Obviously we want creators to be paid. 4:54 Obviously we don't want all of our communications infrastructure to be owned by billionaires. 4:58 But I think the bigger point, at least for me, that I think is important to keep in mind is that there's a kind of incentive flywheel that's at play here. 5:06 And I think the real scary thing and the real fundamental problem that a lot of things like SLOP or surveillance are really symptoms of is this idea that our incentives kind of dictate what gets produced. 5:18 What gets produced dictates the quality of our information ecosystem, which then dictates our ability to really coordinate and sense-make as a society. 5:25 And so I think the real reason why we want the economic incentives that underlie our cultural production to be aligned is because it's really a societal epistemic problem, even more than just a media problem, if you will. 5:40 So with all that context, I want to talk a little bit about the role of technology, which is the world that I come from. 5:48 And at least in the technology world, we often have this sort of belief or assumption that when we expand access to things, that that in and of itself is sort of the winning recipe, if you will. 6:01 And when I was personally spending time on thinking about this question of like, why do we not live in the sovereign media paradigm, one thing that I really started to ruminate on, well, is, you know, why hasn't democratized access to the means of cultural production led to a sovereign media paradigm? 6:19 And what do I mean by that? 6:20 What I mean by that is if you kind of look at, let's call it, the, you know, modern attention economy that I would date going back to roughly the 1830s with the rise of the penny press, what you see is this sort of secular trend of the main inputs coming down. 6:34 The cost of production, the cost of distribution, has actually been coming down in a fairly secular way for centuries. 6:41 And yet the market structure during that time has become as centralized as ever. 6:48 And so I think that can be a hard thing to see if we don't sort of zoom out. 6:53 But I think there's a real sort of almost narrative violation there around how is it that at the same time that access to what we would call owning your means of production has dropped so much, how is it in that same time that the market structure and the economic value has become so unbelievably centralized. 7:09 And I think a good way to think about this is that Facebook and Google simultaneously represent the most accessible platforms in a lot of ways. 7:18 They are offering us in some ways free distribution and free production. 7:22 But at the same time, they represent a complete consolidation and vertical integration of the market structure. 7:28 And just to kind of give you an example by what I mean by that, if you go back to like the TV era, for example, there was a lot of consolidation then, but there was a real market structure in the sense that you had distributors, you had producers, you had ad exchanges, you had people who did measurement of audience engagement and things like that. 7:45 And then if you think about something like Facebook, they are completely vertically integrated across all of those things. 7:50 They do the measurement, they're the ad exchange, they're the distributor, they're— now with AI, they're the producer as well. 7:55 And so there's this kind of strange bifurcation that is, I think, hard to see. 8:00 And it's often obscured by the fact that these companies are offering us free products and services. 8:05 And they're giving us so much free distribution. 8:07 It's like, how could that possibly be a problem? 8:11 And so I think, at least for me, there's kind of two lessons that I think I've taken away from that. 8:16 One is this idea that access and economic value seem to operate really at different levels. 8:23 Access kind of operates at the layer where the costs are collapsed, right? 8:26 So when the internet brings the cost of distribution down to zero, now we can all have access to that. 8:32 But economic value consolidates at the new scarcity layer. 8:35 So the moment that that happens, actually the most valuable thing to do is become the aggregator and become the curator. 8:41 And so there's this weird bifurcation that happens there where just at the moment that you get access to the thing, the economic value has in some way been stripped out of it. 8:50 And I think the second kind of related point to that is that scale creates new coordination problems. 8:56 And so just because access scales— when access scales but trust does not, that gap is where platform moats and economics live. 9:05 And so I think in some way, at least for me, I think one of the lessons I've taken away from this is that trust is really the thing that we have to scale, I think, in some ways to start to break out of this. 9:15 This sort of Faustian bargain that I think we've kind of been stuck in for a long time. 9:20 So I think, you know, takeaway: technology has democratized access in very important and I think very straightforward ways, but it has not democratized economic value in the same ways or to the same degree. 9:32 And I think that's kind of a, you know, at least for people in my industry generally, a bit of a hard pill to swallow because we like to think that access and value are sort of coupled in ways that they may not be. 9:43 So I want to bring this all to the kind of current moment and talk about AI, which I'm sure is everyone's favorite topic. 9:50 And I think the useful way for me that I like to think about AI is that it sort of breaks the existing covenant, right? 9:56 So in a world of omnipresent AI, we fall into a situation where your agency can no longer reliably be traded for access to the economic benefits of scale. 10:06 And I think, you know, Going back to the Google example is kind of a perfect one because, you know, you used to, you know, make Google— make your content legible to Google, let them index it, and you would have faith that they would send you some amount of viewership. 10:19 Now with AI, you think of ChatGPT as a kind of layer that sits on top of that. 10:23 And so you're seeing these crazy numbers about how the impression-to-viewership ratios are changing because, you know, AIs are pinging people's websites thousands and thousands of times and sending fewer and fewer viewers. 10:36 And so if you are someone who is taking the time to really produce high-quality content, high-quality information, high-quality cultural production, you find yourself in this weird space where in a lot of ways you're no longer economically incentivized to actually put that out on the open web. 10:52 It's actually becoming economically rational for you to paywall that, for you to put it in private silos. 10:58 And for many cases, people are just actually taking it offline all in entirely. 11:02 And I mentioned at the end of this presentation that there's— I'm noticing, at least in my world, a rise in physical media publications where people are almost going back to not just physical books, but physical magazines where they actually are skipping the whole— they want their distribution to basically be offline in many ways so that the AI can't scrape it or train on it or whatever the case may be. 11:22 And I think Cloudflare, who's a sponsor of this conference, their CEO has been really loud over the last year about this Dynamic and has actually been trying to offer some services to creators that allow them to paywall their content to AI agents. 11:34 But I think there are also challenges with that setup for many reasons. 11:41 So I think the challenge with this is that it's creating this kind of illegible web, if you will, which is what I was just speaking about, where people are incentivized— if you are— the more kind of care and quality that you're putting into your work in a certain way, the more incentivized you are to sort of keep it closed off, keep it paywalled, keep it siloed. 12:03 And so I think the outcome of that, if you play that out long enough, is that all the good stuff is basically hidden, paywalled, unavailable. 12:11 And the open, free web is largely just kind of various forms and flavors of AI slop, which I think to some degree we are already feeling that and living in that a little bit. 12:22 So these are some kind of somewhat funny tweets that I saw that I think kind of capture this sentiment in the zeitgeist in different ways. 12:30 But I think this is kind of where we are, quite frankly, in the current moment. 12:36 So the opportunity for @protocol— where does @protocol fit into something like this? 12:41 And I think for me, using this framing, I think the most interesting and important thing about open web protocols and cryptography in a lot of ways is that they ease this tension between scale and agency, where you don't in the same way have to— getting access to scale does not require the same sacrificing of agency in the same ways. 13:06 And so this is kind of a funny thing, but I do mean it seriously, where I think that part of what we're seeing here is a— I think part of what we're seeing here is a return to gatekeeping in a certain sense where people are wanting to not be forced into one sort of mode or the other. 13:28 And they want to be able to play around in the design space that allows you to have some amount of scale and some amount of agency. 13:34 I think there are some examples I talk about towards the end of this presentation. 13:38 But I think even just as a basic one that people are here —think about something like a Black Sky, where you have a place for your community that is for you and specialized, but that doesn't in any way preclude you from being very deeply tapped into and integrated with a wider, more scaled web and more scaled experience. 13:59 And so I think open enough to scale, closed enough to actually mean something is a good way to think about where I think a lot of people increasingly want to fall here. 14:07 And I think some of this is not even just an economics thing. 14:09 I think increasingly— I was joking with a friend the other day that being famous is not really fun anymore, because it intrinsically comes with so much. 14:17 To be exposed like that in the open web, you expose yourself to so much harassment and so much negativity that I think people are finding— it's getting to a point where some people wonder, is it worth me even being online? 14:29 Is it worth me even trying to maintain this public presence in the same way? 14:33 It feels like it really takes a toll. 14:34 Toll on people in a lot of ways. 14:37 And so I think in a lot of ways, at least in my personal opinion, the successful outcomes and what types of things will be interesting to build here will look a lot more like guilds and collectives than traditional startups. 14:49 And so I love the previous speaker who was just here because what he was speaking about in a lot of ways is exactly this, just kind of for the journalism era, where you have all these different sorts of new ways of coordinating with people and forming new types of— basically new types of institutions in a lot of ways to serve your community. 15:10 And yeah, I think thinking about it in terms of new institutions is really a useful way to think about it. 15:15 But I think it's funny for me as a VC because I think a lot of this stuff will, and even I'd go as far as to say should be eligible to venture capital in a lot of ways. 15:25 These sorts of entities will be— the economic substrates that they will have to exist in will be different. 15:33 And I think the funding structures, it goes all the way down to that. 15:38 And so I think just to kind of pose to the audience, I think a useful question to ask is kind of like, where are people already experimenting with high-quality cultural production? 15:48 And where is infrastructure for that missing? 15:50 And at least for me, I think I have a couple examples that I'll share with you guys that I think are interesting experiments that are also being run. 15:57 Obviously, I think there's been a lot of talk here at Proto already about the research and science communities. 16:02 I think that's a great and really, really, really interesting space to spend time in. 16:07 And I don't know how many people here are familiar specifically with the biohacking peptide community, but there are communities of people who are now sort of doing scientific experiments on themselves. 16:17 They're sharing health protocols, and it's this It's this weird sort of thing where it's almost like this new kind of scientific or research institution. 16:27 And you can imagine that there's a certain level of infrastructure that will be needed for that, whether it's verifying that or testing to certain protocols that you've taken or that you haven't taken, or people being able to trust, obviously, in a context where you're essentially giving people medical advice in a lot of cases. 16:44 But you can imagine that that will need more infrastructure, especially in the case of verification, than just your run-of-the-mill social media platform. 16:52 I think a lot of the work that's being done by the Blue Sky team with permission data is incredible because that's exactly the kind of architectural primitives that we will need to open up this design space a lot more. 17:03 And I'm incredibly, incredibly excited about that. 17:06 There's also this idea of community computing. 17:08 I don't know how many people here are familiar with the Dark Forest operating system. 17:12 I honestly wish this project was built on top of that protocol. 17:15 I'm kind of salty that it's not, but this is essentially like community operating spaces. 17:19 And so you can sort of create a— a lot of them have sort of like mini wallets and banks inside. 17:25 They have sort of community apps, and you can really think about it as a kind of like community computer. 17:31 And these are sort of meant to be, again, like they're not closed off in the sense of like they're super, super private, but they are closed. 17:39 Excluded from the wider web in sort of important ways. 17:42 And then obviously, you know, cultural collectives like Black Sky and many of the other groups like that that people will be familiar with. 17:50 So, kind of close out here, I think just maybe the short way of putting this whole talk is that, you know, the economic medium is the message. 17:59 And so I think, you know, as we kind of continue to grapple with what is this new paradigm going to look like? 18:08 I think it's important for us to keep in mind that as the primitives and the architecture changes, the economics will have to change as much. 18:15 And so even though there's always going to be a bit of a tendency to think back on how can we do ads, or how can we do these things or that things, we will have to suspend our imagination here and understand that the economic primitives that make sense are going to be really native to the form factors in a lot of ways. 18:35 And I think that's probably the main point that I wanted to make here. 18:40 So thank you guys so much. 18:41 That's all I have for you. 18:42 And I'm happy to take any questions if anyone has any. Natalie Mullins 18:51 Excellent. 18:52 Thank you very much. 18:52 There's a lot of chat happening on the stream. 18:54 So I just want to acknowledge everybody that's tuning in right now. Speaker B 18:57 So we have a few minutes for questions here. Natalie Mullins 19:00 Anybody have anything they want to ask Natalie? 19:07 Oh, oh, oh, oh, great. 19:08 Here we go. 19:12 Thank you for the talk. 19:13 Love those ideas you're presenting. 19:15 So one, I was interested in how you thought about the role of curation because I feel like that sort of overlaps the scale agency economics gatekeeping a word you use? 19:29 Like, do you see gatekeeping as similar to curation? 19:32 Is there overlap there? 19:33 I think so. Speaker B 19:34 I think it's more like a localized curation is the way I would think about it, where, you know, a lot of these platforms right now are kind of built on a more global curation where it's, you know, it's the algorithm that's amassing, you know, all this, all these sorts of things. 19:46 And it's not the way in which content is being curated is, I guess it's personalized, but it isn't attuned to the local context of a particular community. 19:54 You know, when you go on Substack, there isn't like an algorithm for science content. 19:58 That's curated by people who are knowledgeable about that particular topic. 20:02 It's just kind of this flattened global experience. 20:05 And so I think part of what I'm trying to get at there is that, especially with the gatekeeping and like the guilds, is that that role, which is itself economically valuable, should in some ways be returned to the context in which they are economically valuable. 20:19 And in which people have, you know, real understanding of what that should be and can, you know, be in conversation and governance and all those sorts of things about how to actually do that in a way that makes sense. Natalie Mullins 20:29 Just to tap in on that, in your, you know, you're looking at a lot of different startups and all, and, you know, across the space, what are you seeing in terms of how AI is like a new meta layer for curation, you know, between all the content and then the humans that are, that have the taste, that have the capability to curate for their communities? Speaker B 20:51 Yeah, I think truthfully I've been a little bit, um, I've been a little bit disappointed, or I feel like I haven't seen, I haven't seen, I think, I think that space is promising because the AI is, is interesting in that you can wield that scale in terms of personal agency. 21:07 I guess it's kind of the way you're, you're framing it, but I, I haven't I haven't seen anything yet that I feel like really hits the nail there. 21:14 I think part of what I'm sort of gesturing towards with this talk is that curation is in a lot of ways a very human and context-dependent job. 21:23 It is very— it's economically valuable in that sense, in that I think part of what will restore our sense of community and ability to engage in meaning-making online is that that function gets returned to human beings to some degree. 21:37 I think, you know, obviously we can't sift through all the millions and millions of content, but I do think that before we can talk more about kind of how the role that AI plays in that, I do think there's a— the more pressing conversation is restoring that functionality to humans a little bit is the way I would think about it. Natalie Mullins 21:54 Thank you. 21:55 Right in the back here. Speaker C 21:59 Yeah, so on the topic of guilds and collectives as like kind of the new form formation. 22:04 How do you think about those kinds of structures getting the right capital formation together? 22:12 Because one of the things that I've observed on my startup journey is that a lot of times VCs are pattern matching and looking for more traditional two founders owning the vast majority of the company, when a guild and collective has like all this, you know, kind of governance complexity? 22:32 Yeah, great question. Speaker B 22:33 I mean, the first part of that question is I think that's, that's a good, that's a good example of why I think VC is a hard and not ideal funding route for a lot of this stuff, because venture, venture math just simply requires scale at a certain point. 22:46 And so if you are not angling for that, it will distort your incentives in exactly the kind of the ways that I was getting towards at the beginning of the presentation. 22:55 So maybe that's one point. 22:56 I think And then maybe a secondary point is I think we're kind of back to maybe the person who was presenting before me. 23:02 We're kind of in the midst of that experiment now where people are trying to make new formations and explore philanthropy or having a more direct relationship with those who consume your content. 23:17 But I do think that remains a little bit— I think that remains a little bit of an open question in the sense that I don't have a specific specific primitive to give you, but I do think that for most things, what ends up being the kind of economic mechanism that sustains you will be very native and specific to how you actually generate value in your community. 23:38 Journalism is monetized and is valued differently than visual media, for example. 23:42 And so I wouldn't expect a guild that is built around the production of journalism or to have the same business model or funding structure as a guild that's built around the production of paintings, for example. 23:55 I don't know, just to take something else. 23:57 And maybe that's part of my point, is that I think the business models have to be more local and more form-specific, I think, is part of what's happening here. Natalie Mullins 24:10 Follow-up? 24:15 What are your thoughts on micropayments? 24:17 Do you see any viable future for that as a means of supporting creative output? Speaker B 24:21 I'm— the verdict is still out for me on that. 24:25 You know, for example, like Cloudflare is— Cloudflare's model, as I understand it, involves some sort of micropayments where the idea is that you will paywall your content to AI agents and they will do a micropayment. 24:36 Quite frankly, in the crypto space, there's a lot of hype around that right now. 24:39 And I'm not closed off to it, but I guess I just, I haven't seen it play out yet. 24:44 And I'm also sensitive to the fact that, you know, creators have bottom lines. 24:49 And so, you know, I think for a lot of people, at least in my personal experience, a lot of people, even though it's not perfect, would prefer a kind of a Substack model to a micropayment. 25:00 You know, I think recurring revenue is something that creators are very interested in. 25:03 And so, you know, I'm not closed off to it, but I haven't I haven't seen a formulation of it yet that feels like it's really deeply sustainable, I guess. Natalie Mullins 25:11 Yeah, just to follow up on that. 25:13 So in London at Protocols for Publishers just last month, micropayments was a topic that came up, as it comes up every time we gather together people. 25:21 You know, and there's a lot of research that has gone back and looked at micropayments over the last 15, 20 years. 25:28 Knight Media Forum also hosted a bunch of people talking about that just recently as well. 25:32 But there's like two things I want to raise at that point. 25:35 One is speed. 25:38 The very simple way to be able to like tap to pay. 25:41 Like one of the challenges that we have right now on the open web is that it is not easy without installing an extra, you know, a plugin or something like a wallet and all of this type of stuff. 25:52 And so that's a challenge. 25:54 It has to be much smoother, just like going on the subway and just like tapping to go through. 25:57 Like tap to pay. 25:58 We have to get to that smooth UX. 26:01 Another thing is, is in the media world, a lot of it isn't just a matter of how much a person pays. 26:12 So like when an article comes up and a paywall comes up and that article has like a certain value attached to it for advertising revenue, and the idea is like, oh great, if you just like, for people that are not subscribing and they just micropayment you just collect enough micropayments, very tiny little micropayments, to be able to get that advertising equivalent to that advertising value, you're fine, right? 26:32 But that's not actually how the economics of a lot of these large newspapers work. 26:36 They package their audiences and sell that information and data and like open that up on a different market. 26:41 There's a lot going on behind the scenes that that is not accounting for. 26:45 And that's also a challenge. 26:46 So even if you'd have to really— it's not just a sense of like, can we hit like the 30 cents? 26:52 You know, it's actually a lot more than that. 26:54 So that's another challenge. 26:55 Yeah. Speaker C 26:58 Hi. 26:59 There was mention of bifurcations in reference to what I was considering more. 27:09 Maybe I was thinking more of the term partition, but like with bifurcation, I imagine like a splitting into two modes of behavior. 27:22 Also, with regard to the partition, I guess I also noticed it in the political sphere, as most evident in Epstein's email with Peter Thiel, where he said Brexit was just the beginning, and that there was return to tribalism and a shift away from globalization, which I guess spans more than just the media ecosystem, the political and cultural aspects as well. 27:56 But I was thinking of partitions because like people are, yeah, separating from each other. 28:04 So was there by multiple modes of behavior that you were referencing with bifurcation? Speaker B 28:11 [Speaker:JESSICA] Yeah, so what I really meant by the bifurcation was that I think that it's kind of like the dynamics of the open web sort of increasingly being incentivized to kind of be the freely accessible, like lower quality, lower brow, kind of more sloppish stuff. 28:28 And then on the other side of the partition, if you will, is all of the more higher quality, more private, more intimate, if you will, more local stuff where people feel, at least right now, that people are scared, I think, to put the stuff that they really deeply labor over in the open web sometimes because will it get trained on? 28:50 We don't know. 28:52 It's more nebulous how you can capture the economic value once you've done that, especially in the world where everyone's Claude bot is crawling the internet at all times. 29:01 And so that's kind of what I was speaking of. 29:03 And this kind of like— that's sort of a bifurcation between the illegible web, as I call it, and the more legible web. Natalie Mullins 29:10 Excellent. 29:11 All right. 29:11 Well, I think we should close off there and give us a minute or two to transition to the next session. 29:15 But thank you very much, Natalie, for your analysis.