Decentralizing Media - Balaji Srinivasan keynote - Mainnet 2020

Hey friend. A Musari investor and the former CTO of Coinbase. He is a serial entrepreneur, one of the leading thinkers in Silicon Valley. And as I said before, we owe him debt of gratitude for his early outspoken and contrarian and right coverage of the coronavirus. So Balaji Srinivasan is going to talk about decentralizing media. Balaji, welcome. Thank you.

Can you hear me, Ryan?

Yes, we can.

Okay, great and clear. All right, awesome. Okay. Yeah, so today I'm going to talk about decentralizing media and essentially talk about how we can move from centralized legacy media corporations to where we currently are social media corporations, also centralized to decentralized media. Essentially a network of personal media corporations. And I'm basically going to describe kind of motivation and then a feature list and then current technology is for implementation that implement pieces of that feature list. And then if you want to learn more, if we launch something on this, please follow me@twitter.com Balajs okay, so just jumping in. So just as an overview, you know, there's this famous saying, freedom of the press belongs to those who have one. And you know, you may like or dislike them, but centralized media corporations are not perfectly aligned with your interests. Okay? That is true for both the New York Times company and Facebook, for both legacy media corporations and social media corporations. If you're not the publisher of the New York Times or the CEO of Facebook, you're fundamentally not in control of how you appear in those pages, whether in the page of the New York Times if you're quoted or whatever you publish to Facebook. And now with the Internet, we can actually move beyond this to you know, what I'm, what I'm calling decentralized media. Essentially a network of loosely cooperating personal media corporations, each of which control their own distribution, monetization and brand staffed by citizen journalists equipped with decentralized fact checking. In a sense, you know, you can almost think this is just blogs 2.0, but with everything we've learned over the last 15 years, particularly how valuable user generated content can be, you know, everything we can do in terms of decentralized layers and you know, which, which doesn't try to reinvent social media or legacy assumes they exist and then builds on top of them. So just in terms of some concepts, I'm going to build through this slide, but basically if we say legacy media, social media, decentralized media, what do I mean by the differences here? So first concept in terms of sovereignty, the New York Times has Deploy access@nytimes.com At Twitter, you have kind of Partial access at your username, but@username.com, you have full root access. Second concept, integration. So legacy media just produce content. Social media really just thought of it as pure tech. You know, Zuckerberg used to refer to Facebook as a social utility. Decentralized media combines content and tech. And that is to say it's opinionated in a way that, you know, Facebook or Twitter or what have you, these legacy platforms, or really, I shouldn't say legacy, but let's say existing scale platforms really were not set up to be fact checking. So legacy media does this at the corporate level and will usually not fact check itself. It considers whatever it says itself to be true. Social media until very recently did no fact checking. Now it's done arguably an arbitrary kind of fact checking. Decentralized media, everybody's fact checking each other. I'll describe some tools for this later on. But the crucial idea is that truth can't be centralized by its nature. A truth must be reproducible, independently reproducible. Fourth concept, monetization. Legacy media, you know, simple, you pay them. Social media, you know, it's free to use and you know, you can distribute through it and they pay you a little. And decentralized media, you. You own it all. You can make, you know, however much you want, you're in full control. It's your personal media corporation. Fifth concept. So distribution, well, you know, you're basically a letter to the editor and hope legacy media prints it or give a quote to them. Social media, it's much more free. You know, you can post, but you better abide by their terms of service however they currently interpret them. Decentralized media is fully censorship resistant. The idea is that you own an email list, you own your own domain, you are sovereign from the beginning. Not so much because of censorship per se, because it's not so much that that's going to hit you on a daily basis, but it's more control of distribution. And that's highly related to censorship in the sense that, you know, for example, legacy media and social media are wrestling over, you know, does news appear in the feed, how much does it appear in the feed, and so on and so forth. So censorship is sort of like the political version of it, but distribution is kind of the monetary or corporate version of it. How much do you show up? Six concepts, setup, costs. So legacy media was once high, it's now lower. It used to be you had to buy printing presses and so on. Now it's relatively easy to set up like a, you know, a new Media Corporation. Social media was once low, it's now higher because the standards for a minimum viable product have risen. Decentralized media would be very low initially just a domain name. And I think we can, we can take it further privacy. So mostly named authors and legacy media, social media, some pseudonymity, legacy media, full pseudonymity. And what I mean by this is social media, there's some folks who go by pseudonyms. With decentralized media though, you can have a private domain name, you can publish fully pseudonymously, you can take satoshi level OPSEC should you so choose. Somebody actually did this for their analysis of whether the COVID 19 virus came from a Wuhan lab and they published this on GitHub and they did it fully pseudonymously. That's a good example of decentralized media alignment. So legacy media, you know, they're aligned with the shareholders and employees of media corporations just like social media is aligned with the shareholders of tech companies. And for you, if you're setting up your personal media corporation, it's aligned with your shareholders programability. So legacy media closed API and scraping, you know, it's, it's really not meant to be programmed against NY Times for example, has an API, but you know, it's, it's really not something that you have some access to it. But to build a full company on top of it, you probably wouldn't be that easy. Social media is a much more open API. Lots of people write programs with Facebook login or Twitter login, but famously they've also used their platform in aggressive ways. Decentralized media would be a culture of fully open source and using technologies like DNS, things like GitHub pages are examples of this. I'll come back to more later. Finally, in terms of the general orientation as how I think folks should approach this, legacy media is, you know, writing as a job, social media is writing as a hobby. And decentralized media is, you know, I think of it as writing as a duty duty as a citizen to get it out there. You, you may want to make some money. It's like a serious hobby. But the idea would be a thousand, you know, citizen journalists making $100 or 100 citizen journalists making $1,000 versus one corporate journalist making $100,000 a year. You basically decentralize it. Rather than have one generalist who doesn't really know anything in depth. You have, you know, ideally a plumber who's writing, you have a delivery worker who's writing. You have folks from many different professions who all have their own specific expertise. Okay, so that was a desired feature list. Implementation will be discussed at the end. I want to go into a little more depth in each of these features that we discussed. So sovereignty. So as I mentioned, with decentralized media you own the domain name. You know, with Twitter you can have your account suspended. You can be, you know, like, like deprioritizing the feed you have to pay them for. Here you can just get your own domain name. That's kind of one of the things that I would consider principle of decentralized media. You own your domain name. Eventually we'll want ENS, IPFs, stuff like that, integration. So one of the concepts I think about when I think about decentralized media is full stack narrative. There's a great site called wistia.com that gives an example of this, which is better shown than told about. But if you Google 110100 and watch this, it's essentially three videos that, that go beyond content marketing to essentially make Wistia.com into something that's developing and shipping its own TV show. So that's, that's different than isolated blog posts, that's different than isolated tweets. That's a coherent story arc. And the idea is that with decentralized media, every company a media company. If over the last 10 years every company became a tech company, you know, whether you're in taxis or hotels or what have you, you had to become a tech company over the next 10 years. I think every company must become a media company because otherwise they're not going to get their, their own NAR out there. They're too disaligned with legacy media corporations. That means not just tweets and podcasts, but TV shows, interviews, memes, the fulcrum. It you're a media corporation, that's actually a big part of your advertising budget in a sense. So next concept, fact checking. You know, the, the old school way is something like PolitiFact, you know, or it's relatively new. But the concept is you just trust a single entity. You know, social media. Now Twitter's been rolling this stuff out like, you know, get the facts about COVID 19 tagging, misinform. But decentralized media, there's a number of different approaches here and that's kind of a whole talk in its own right. But there's a great book called the Truth Machine by Michael Casey and Paul Vigna, formerly of the Wall Street Journal that talks about the fact that Bitcoin is actually decentralized access to truth in the sense of there's folks who share nothing in common. They're in Japan, they're in Brazil, they're in France or in the US they don't share common legal systems, they don't know each other, they don't share a common language, anything like that. But they do all agree on the, of the bitcoin blockchain. And that's a very important concept of decentralized truth. And it potentially can be extended with things like oracles, for example. So that's something I'm very interested in, is using computer science and explicit references to potentially allow people to fact check in a decentralized way. So you don't just let it be a complete free for all with no objective truth. But you start to think, okay, how far can we push decentralized consensus? Maybe it's only within a group, but how big can we make that group? Okay, monetization, so, you know, substacks a major part of this. You know, like basically you start to think of your writing as something that's revenue generating and it's not even really, you know, for some people. They would try to do this as a profession of course, and that's fine, more power to them. But you know, just allowing folks to make, you know, a few hundred or a few thousand dollars a year to spread it out more. You know, it may be that not everybody has, you know, 50 or 100 essays in them a year, but if you have two or three based on your own unique experience, if you could get paid for that, that'd be really great for a lot of people. That'd be, you know, a nice, maybe it's a, it's a, it's a computer or it's an iPad or something like that. It's the encouragement to get citizens to do something more than it is, you know, like monetization, like, like just for, for a big corporation. And one way of thinking about this is a lot of the money has gone out of media. And so a lot of folks in media have to either have trust funds or they have to have, you know, wealthy spouses or you know, mom and dad are paying for their, their apartment or supporting them. That's kind of like an open secret within legacy media. And you know, maybe the news trust fund is a full time job in the sense of you're a chemical engineer, okay, you're, you're somebody who works in design or architecture or something like that, and you know something about a topic area and you write about it and you're doing it because it's your Duty to inform the public. You're not doing it because you're trying to maximize clicks alone. You're doing it because it's your duty. Next concept distribution. So, you know, as I mentioned, you know, substack is good for monetization. It's also good for distribution. Those go hand in hand. Actually. You basically get your own email list. You're not subject to just having followers that you can't directly reach without the consent of Twitter or Facebook people. Actually, they often mistake this. They think, oh, what's important at Twitter or Facebook is your data. Actually, it's pretty easy to export your data. The hard. The important thing that Twitter or Facebook have is your access to their social graph. That is to say, you know, whatever followers you had, you can't, you cannot contact them without the consent of Twitter or Facebook. And that's actually the important part which getting your own email list gives you. So next concept setup costs again. Domain names have become very cheap. And when we think about this, in a sense, I'm just talking about setting up a blog for each person. But I think we can do way better than the old age of blogs now that it's like 15, 20 years later. There's things we can do with DNS that are in between a full index like Google of every site and a full hosting thing like Facebook. If you had a network of sites that all had the same technology, they could have sovereignty with some degree of cooperation. And I think there's something there. So privacy, you know, so this basically is, you know, if you want to be satoshi, if you want to publish something that's really important, you know, the traditional way of doing it is if you're, you know, like Deep Throat going and talking to Woodward and Bernstein or you're Snowden, you need to go through and, and existing legacy media outlet. Now increasingly things like the analysis of the Wuhan virus that I mentioned before that just got posted to social media. But the ultimate is decentralized media and the COVID virus analysis. Also another example of this where you post it to your own site and now the source can just speak for themselves. They don't need to be mediated. They can say things in their own words. They don't need to trust a reporter to protect their privacy. There's an example a few years ago where there, there was a source who spoke to the Intercept and it got leaked somehow and then that source was, was jailed. And so it doesn't even need to be that, you know, the journalist would, would mess it up on purpose. It would just be something where you know, you've just got two people who can screw up rather than one person. So if you want satoshi level opsec, it's really sensitive and important information, decentralized media may be superior there alignment. So again an obvious one. You know the New York Times is a company, it's a corporation, it's got proxy statements, it's, it's worth reading their financials just like it's worth reading Facebook's financials. Ult ultimately they have a responsibility first and foremost to their employees and shareholders, which is completely reasonable. If you have your own personal media corporation, you're responsible to your shareholders who can be whoever you want. It can be your friends, it could be your co workers, it can be your colleagues, can be other people in your industry. You can slice and dice that that cap table if you decide to take outside investment or do whatever you want with it. Programmability. So you know, I think GitHub pages and creative Commons are sort of in the spirit of what I'm thinking here. GitHub pages, your back end is open source. That is to say your backend is a git repository. It's not a CMS in the traditional way. It's something that's on GitHub and you can export it anytime. And that combined with Creative Commons. Well, what if we actually had a new, not media company, but a new media community where all content by default was open source, that would allow for natural language processing, that would allow for scraping, for integration, for named NT recognition, for lots of interesting things. And I think we're really just at the beginning of something where all content is, would be open source and you might monetize another aspect of it. And finally duty, you know, citizen journalism, if you, if you go and Google it, it is, it is a term, it's in the dictionary, the collection dissemination, analysis of news information by the general public, especially by means of the Internet. So, so it's out there, it's, it's, it's an important concept. And I think, you know, more and more we're finding that, you know, unless you yourself go and participate in the news gathering process, you don't really have the ability to discern what is true and what is not for decisions that could significantly impact your health. Like something like COVID 19 or something like, you know, civil unrest. So that's writing as duty. So in terms of implementation, you know, these are concepts, this is like a vector for, you know, for what decentralized media is and could be parts of These pieces of this have been implemented, but not all of it in kind of one package. So what do I mean by pieces? For example, blogs of course had independent domain names. You know, they've been around for a long time, 20 plus years. Substack has, has the email list, monetization, very important components. Stripe Atlas has incorporation, YC seed funding, GitHub pages, open source content. All existing legacy media is content for critique. All existing social media, you don't have to give it up. Those would be tools that you use to generate content, to distribute, to find influencers for your personal media corporation, which of course you can use to employ others. And then finally crypto I think will be an important component of this and the one that really hasn't been rolled out yet. Crypto first for payments simply to pay people in any, in any country or take them to any country, and then eventually hopefully for truth in the sense of Vigna and Casey's truth machine. So I know this is a list of technologies. This is, it's something where all the pieces haven't come together yet. But this is directionally how I've been thinking about decentralized media. So thank you very much and follow me on Twitter or DM me if you have any questions or comments or critiques. Thank you. Thank you.

Wow. Incredibly insightful. And I think, you know, one of the things Balaji that you, you touched on that I felt instinctively for a while and I'm almost embarrassed that I've not been able to put words to it given I used to run Coindesk and now, you know, running another information business. But the concept of every company being a media company, much like every company became a tech company.

Yes.

And I guess what do you think the biggest challenge you laid out a lot of variables to what this could look like. What do you think the biggest zero to one challenge is in taking the next step in terms of evolving everybody to citizen journalists and decentralized media. Because there are some structural impediments I think that are larger than others, although a lot of them have been eroding over the last few years and this year at a hyper accelerated pace.

I actually don't think there's any major structural impediments. I think it's just a function of execution. What I'd like to do to start is basically just get a community of 1 10, 100 folks who basically just set up their own blogs. The technology stack is important. That's something I've been thinking a lot about. Essentially the ideal would be I'd want something just get a little technical that has the usability of something like ghost.org, but that has an open source back end like GitHub pages. You might decide to just go with GitHub pages for now to start, but it's not quite as full featured as Ghost. That's something just from a technical standpoint I've been thinking about. But then the idea would be you get 110, 100 folks setting up blogs, then something that really hasn't been done is to think about whether you can network between them in a more sophisticated way than simply links. For example, in DNS you can write TXT records. You can do things with DNS, you can also use other layers obviously, like blockchain or crypto layers. What I don't think people have really thought about is how to take websites and organize them in any way other than an index. Maybe you can make hierarchies or rings or topologies of websites where each individual proprietor is is sovereign, but they decide to cooperate. For example, you could structure them in a hierarchy where there's a top level index and then it's got three categories underneath for bio and crypto. And then each one is actually pulling from an RSS feed of a bio site, a crypto site, and so on that are dedicated and it gets kind of mashed up like that. So I think that's kind of an interesting concept. Like a topology of websites hasn't really been done yet to my knowledge, so that's kind of something I've been thinking about.

Perfect. Well, Balaji, thank you so much for joining for this presentation. Hope you're able to hang out. I'm sure some people might have questions in the chat, but appreciate you taking the time and we're excited to see what comes next.

Great. Follow me on Twitter and DM me if you have any questions. Thanks everybody.

Thank you.