 Next up we'll be we'll be a talk by Bruce Calais Kale, that was it Kale, sorry Bruce sir is a digital librarian and he is the founder of the internet archive I'm sure many of you will know that Bruce sir has been working on on internet technology since the mid-80s and now He's here to call upon us to make it better Bruce sir, please Thank you working Can you hear me excellent? So this is this talk is speculative Tomorrow morning at 10 30 if anybody's actually awake We're going to talk about how we're hacking copyright to try to get everything online but this talk is really Dedicated and motivated by Edward Snowden that we basically We've come a long way, but we've been showing that we've got some real faults The last 25 years we've been pouring Amazing amount of effort and great stuff and our personal lives into the worldwide web the worldwide web is flowered There's something completely amazing but We've got some problems some real flaws and this talk is to try to see if we can help address some of these um My friend and hero Larry Lessick said code is law that How we code the web determines a lot of how we live our lives online That if it's not in the code it probably doesn't Exist within how it's working, so we should have freedom of speech embedded in the code We should have universal access to knowledge embedded in the code We should have privacy both writer privacy and reader privacy Embedded in the code, but it turns out That it's not there the actual the web is fragile. It's kind of broken even But it is huge We know this at the internet archive because we collect about a billion web pages every week We're basically snarling up and starting to really understand what it is the worldwide web is and Every web page the average life of a web page is about 100 days Between when it's either changed or it's deleted So there are these pages that are flickering on and off Websites at a constant pace so it's not a very reliable Infrastructure to use but it is massively available. You can get to all almost everybody Except if you live in China So the internet archive is blocked in China It's we're on and off blocked in Russia every so often we blink off in in India based on government mandates So the web is not reliable in terms of the web pages or in terms of how you can get Things out out to people. So we've got some problems there We also have a problem that it's it's not private that Corporations and governments are watching what it is you're reading and they are up a storm so Edwards notes one of his parts of his releases showed that GCH Q was watching the people that were using the WikiLeaks website and Then took those people and then gave them to the NSA and that was enough for them to generate special targeting Even if they're American citizens to basically target the people that have read WikiLeaks the idea of being rounded up and profiled and Targeted based on what you've read is for librarians sort of a major warning flag There's a long and bad history About doing this and it's happening now and it's happening because our technology isn't isn't good enough But it is fun Well, we out of the big three I'd say reliable Private and fun the web is fun and why is it fun? It's because Bonzo people us can go and do all sorts of things with it And we've sort of added to it and manipulated it and made it sort of something pretty pretty interesting So I'd say those are the if I were to have the big three priorities for a next-generation web What I'd want is reliable Private and still fun How are we going to get there if we've only got one out of three I I'm going to suggest and this is a call. This is a an aspirational Talk and I think we within the four Boundaries of this camp we have everything required to be able to pull this off But I think we need to start thinking about building a distributed web a Distributed web that would have these three characteristics and I'm going to try to say a little bit of wow what this might look like and Some of the pieces that might be able to be put together to show a straw man of how this could work Not that it's the right way to do it But that it's at least possible and then I'm hoping that we come together with this as a goal Let's talk about what distributed means so distributed Let's contrast the worldwide web to the internet the internet is a distributed system in the sense that if any particular Piece of hardware goes down or slows down or bogs down gets overloaded things route around it So it's all you basically continue to have the whole thing working even if it gets parts of it gets nuked quite literally The web on the other hand is not decentralized is not Distributed in that way because if you basically control the wires to a particular piece of hardware or a particular piece of hardware You can either watch control or destroy any particular piece of the internet Take another example I'd say Amazon comms cloud is an interesting one It that's runs out of a set of data centers all around the world all owned by Amazon comm And if you configure a thing right they'll move your data closer to where your users are and it'll move it around Hardware that's not working anymore So can we go and build a worldwide web or internet that works kind of like the Amazon cloud does But spread all over the internet and not controlled by a single corporation another piece is Let's build in the user the reader privacy So how can we go and make it so that if you are operating on a distributed web? If your website doesn't exist anywhere in particular Then it's much harder to figure out who are the readers of a particular website So if a website is distributed multiple copies multiple places so that as readers are using them They're using local copies or one's maybe just down the hall or something like that It makes it much harder. It's still probably still probably possible but it makes it much much harder to go and make a Big surveillance system that we call the worldwide web at the moment Let's also build in a time axis So that there's a version kind of kind of like get or something like that for web So that you can always reference older versions of websites so that also they can be forked So the idea of websites Let's build in the archiving function at the start the wayback machine, which is fairly widely used It's still not bundled into the web itself Let's go and make it so that the web is self archiving as it's distributed so that there are versions of websites That are moving along and as we're at it if we're going to build a new web Let's see if we can get some people paid other than Google So the idea of a are advertising driven industry. We know how this story ends It's how magazines television or or radio It's usually massive centralization of people that build the critical mass ad networks Let's not do that. Is there a mechanism for us to make it so that we can go and pay for optionally either if it's a tip or demand by the By the writers to be able to make a better web I'm going to suggest this is doable to be able to make this web happen How well we've got some tools that Tim Berners-Lee didn't have 25 years ago JavaScript is actually now an amazingly powerful very weird But amazingly powerful programming paradigm that allows code to move around and to be able to Work we've got the blockchain that's building a distributed database That's at least strong enough to withstand a lot of people trying to get rich by ripping each other off And there's Bitcoin which can be used for a Micropayments system that could be embedded in a distributed way into the system and in the first place and made less Just remember that 25 years ago. It was illegal to distribute a strong crypto system We've won that war But why let's use it some more let's use crypto in and embedding it and building the next web in some Possible ways like it now. There's a lot of systems that are already kind of going. I'm not trying to say oh my god This is the first time I've ever heard of this. No, there is actually a lot of good work That's been going on in these areas in these bits and pieces So I'm going to try to champion bringing things together for a particular end user purpose That then we might be able to rally behind and and get something Seriously done. So these are the systems that I've come across and now more of them I'm finding all the time, but it's people that are building these innately distributed systems and distributed systems are more Difficult to build than centralized systems because you have to think through a lot more protocol aspects and the like So these guys are some real heroes and actually I think most of these programs these projects have people here at camp So what about a bold goal? What can we do? There would be a demonstrable bold bold enough goal that would make it so that people would want to use a new system And and the like what could we do if we could build a wordpress? Type functionality so take the functionality that people get with a wordpress blog And but make it distributed so it makes it reliable makes it private still makes it fun makes it able to be distributed freely so what would it take to be able to pull something like this off What I would say is there we need a couple of components One is it should work with no plugins. It shouldn't require a new download It should just run on the existing web So if you're you know playing along and you just happen to go to somebody's website say my blog and it happens to be Distributed Distributed website it may take a little while to sort of start up because it's got some more JavaScript and mechanism to it But it would be able to come up and run in in the browser. So I'd say that's a Interesting characteristic it needs to have easy to know names this idea of knowing somebody's Bitcoin address and thinking you're actually gonna remember it. I mean I can't even remember a phone number So it has to go have a good naming a naming system It has to be fast as no matter where you are in in the world It has to be able to be fun to basically go and pound out and comment and update There has to be different user roles So we need a distributed identity system that would make it so that you could have Administrators and editors, you know all the different things So I've been using this as that sort of engineering straw man of okay What if we can do a WordPress this way? I think we're not a far step away from making a Twitter a Facebook a Wikipedia and the like but But WordPress is at least constrained enough to be able to do this So what if we take these other top couple steps? Can we make it have payments? So you can at least have tips and wouldn't it be great if we actually had a system that you could publish Things on the web and get paid for it If it's worthy of being paid for that would be terrific. I think that Apple did a great Service by showing us that people will work infinitely hard to make Apple apps for their phones In the promise that you might make money. I can't imagine very many app developers actually make much money But they had this mechanism of making it possible to do it Can we build our next web to be able to survive that and then can we build in versionings and the archive? I'm gonna say yes, and I'm gonna go down each one of these and I'm gonna give sort of a a Handwave of okay. Can we go and do each one of these particular pieces? With things that are going on my mind was blown by seeing in JavaScript when they cross-compiled an Apple to Computer emulator from C into JavaScript, and I could click and run old style games in my browser It was just I I don't know quite why it surprised me so much that you could take a 20 year old computer whether it's Macintosh or IBM PC and run it at speed in JavaScript, but it meant that something has changed for me that we could actually implement Most of the server capabilities that we're using in our websites in the browsers themselves not only could we do that we could do the whole peer-to-peer levels of going and sharing and moving the data around in JavaScript itself Oh, and I actually there's Marcel Who's whose food hacking camp? He basically took my blog and moved it into a peer-to-peer system So I'm gonna do what you should never do a chaos, or I'm gonna try to do a live demo All right, so okay if this works you have to applaud if not Then don't say anything and you just sort of say yeah, you shouldn't have tried that okay, all right Okay, so so this is My normal Firefox and okay here it goes Okay, what this is is my blog running in a peer-to-peer file system called IPFS And it's running by asking my own machine and it's calling out around the net and going and pulling in the pieces To be able to put together the files that are my blog and the cool thing for me about this is it's not just web pages Can you say hey? Well, why don't you just put it all in a torrent? Which it more or less is but we also have this extra cool thing that It's got a search engine built in on the client side So you say well that's kind of neat that the idea of you can take a search engine or database features and be able to then Be able to Oh 404 document I found yeah, that's a good sign Yeah, okay, here's where the demo is falling down here Yay, okay, so basically I'm able to search for stallman and find other Blog post so what I what? Marcel did is he took my blog and he took the database functions in the search engine made it into JavaScript So it's moved around as files. It's now in my browser And I'm sort of having a WordPress like experience even though it's in my browser and he says say well Yeah, gosh, that's all it's easy to do but if you take that step further you can really build whole websites applications and move them around pretty efficiently as best we can tell in terms of of Using the peer-to-peer technology to avoid needing a server infrastructure We need an easy naming system for this So you can just type in say Brewster's blog or something like that And have that work And there are a couple of systems that are starting to work on this using some of the distributed registration systems in the blockchain for instance Namecoin has explicitly been doing this using the blockchain Underneath Bitcoin and Ethereum has been using is building a new Block block chain to be able to support other types of contracts So the idea of having a registration system that doesn't have the problems of the current Registration system which might be a first first come first serve style naming or other types of naming structures Is completely doable as if we use the block change type structures We really want performance and we want versions and here's something that I think that things like the Internet Archive and other Organizations like us can play a role by playing big honking caching servers If you think of the wayback machine, we've got oh, I don't know a total of the Internet Archive We've got about 25 petabytes of data spinning and we put out a lot of Materials we've got some replications around the world the idea of Making that such that it could be the actual websites themselves not just snapshots of the websites I think we could get to be caught to catch on not with not just cultural institutions But ISPs because ISPs want to make sure that their users have a really snappy environment to be able to be served So if we can make it so that other people can cache things then it's not just relying on other people's PCs That have read that website recently because if we don't have that then we we do have just a peer-to-peer system And at least my use of BitTorrent just means it's slow come, you know starting up So we really need snappy performance and having these large caching organizations. I think is completely completely doable another couple check marks here up updates and distribution so one of the problems with WordPress excuse me with With BitTorrent and actually we've been working with BitTorrent incorporated on this in their Maelstrom System is updates are a little bit clunky and weird That basically it's around hashes where everything has to be unique and it has to be exactly the way it was But can you build another layer on top of it in such a way that you can have okay? That's not the the current one. We have got a new one So use the new version of the blog unless you want the old one and both BitTorrent has them their mutable torrents and IPFS has a partly working system For doing this naming so I'd say that's that's kind of a hand wave check mark that we're doing pretty well on that particular point So as I said, I'm just going down this list trying to say it's doable to go and build a completely serverless web infrastructure Okay, here's a cool idea I think one is to have a distributed identity system that actually uses Bitcoin addresses as your identities. So with strong crypto we can do now Distributed identities We haven't done it very well yet because mostly you know, we use our Facebook or Google IDs on all sorts of things all over the net and that's got its own problems So is there a mechanism for us to use the blockchain and Bitcoin to be able to pull this off? I think so because it's got a very active community to try to support it But they've got another extra feature if we actually use an address as our online identity Then you know how to tip somebody so if every web page is signed with a Bitcoin address Then you'd actually know how to pay somebody money for what it is It's going on there and I'm hand waving But I believe we could also make it so that people could stipulate that you have to pay Something to be able to get something to unencrypt say a movie that you'd be able to buy Over the net in a distributed way without ever knowing who it is. You're you're paying So I would say through this that we have the possibility That we have all the pieces. They're not woven together to be able to build a word press like Content management system which according to wordpress.com is about 25% of all the websites out there Maybe not the most popular but of but there's a lot millions and millions of them if we can build something with that level of functionality to it and plug in ability and distributability That's completely distributed that has all of these characteristics Then I think we've done something pretty Amazing the idea is to have a web this time around That is reliable no matter where you are who you are um If it's gone away or not or if the server is going as up or down or blocked That it's private for both the readers of the website, which is extremely difficult to pull off But also the writers so the idea that you can go and publish anonymously But you can also read without being spied on or make that much harder and still make it really fun and Malleable enough so that millions of people will go and play with it that if we make our websites Basically apps that move around we could version our underlying infrastructure more easily than we currently have to wait for the browser manufacturers to go and and put in new features if if a lot of the features are in the JavaScript layers that are moved around we could have a lot more fun in the whole thing sort of evolving in a really fast and interesting way So including Locking the web open. I think it could be something we can do it's something I suggest we have to do we can bake the Freedom of the press into the code itself if we do this right we can make the web openness irrevocable That we could make it so you couldn't take it back away from us That we basically be locking the web open Building a distributed web. I think we can build this. I think we have the right community to do this and based on Edward Snowden's Revelations, we should do this. Thank you very much Okay Very interesting talk We have some time for question and answers so There's Mike's left and right if anybody has questions Please form a row We'll start with the mic here on the left. I Think there's someone there. Yes Please okay It's a pity that I can't see the speaker from this position And he can't see me. Well, maybe we can move there that we can at least see each other So I think it's really great idea, but The problem I have is this sounds like web 1.0 Because if I look at my table All of the data I use and care about come from databases and do systems and whatnot and There's no one nobody can replicate it except we can do that So I think therefore for a lot of say political documents and stuff like that It sounds like a really good idea, but for a lot of practical stuff Like well, how do you do Facebook or Amazon or whatnot? It sounds like Well, not really that easy to solve Um Does it solve all problems? No I think it might solve for a particular set of communities that we find a lot of at the internet archive That there are a lot of communities that don't want to be tied to a particular piece of hardware or particular company Or the first original person that started the website and they've been pouring themselves into some Website there, you know and who knows what's gonna happen to it because they've got sort of an arbitrary person in charge This sort of makes a more distributed structure They can at least support those sorts of communities So the internet archive often plays a role for going and hosting those Is that web 1.0? I I guess so I think If the web 2.0 is the being basically being able to interlink into other people's websites I think that that is up to us to do well in building this thing again So I don't think we have to lose that Going forward 2.0. I mean more like Contact contributed by the users of the website instead of the site owner So so everything that has common sections web or whatever all user generated Content, which is typically stored in databases and is not Available as as version files that you can put into a distributed File system or what not or first it's I'm not quite sure I understand But I think that this system is like when we implemented this on top of IPF IPFS and when we implemented on top of BitTorrent you can go and take a somebody else the website that you're using And keep a full copy of it and use it for you can take it apart you can use it So it's not all hidden from you So it's actually moves around that the website is encapsulated as a set of files That moves and is replicated and updated is this actually I'm not quite sure I'm okay. Thank you Okay, we have on the right hand there. I think we had someone at the speaker Yes Is it's gonna be free software? Of course Yes, I I think it I think it has to be I think it has to be Something that you're free to build on and change and mutate and and I happen to be a big GPL fan But I don't think it requires being GPL style free All of the components at these layers But I think they'll be the ones that are successful. Yes On the left hand over there, please. Yeah. Hi First of all, thanks for your talk a lot of us I think at the camp have been thinking about pieces of this for a long time And I was a little surprised. I mean disclosure. I work on tour I was a little bit sad that you didn't include tour in your presentation because reader privacy is achievable today with free software with the tour network and end-to-end anonymous communication channels with an address that is as complicated as a Bitcoin key Are available so that you can host any TCP service and that's free software run in a distributed peer-to-peer network right now But you're you're you're right tour has many things to offer that could be built into something like this And I'm just not I don't know enough about how What I think of is tour is mostly trying to obscure the routing through the network And then you're still landing on a single place And I think that we could take for certain types of applications web applications a lot of websites most websites We don't actually have to have it in one place if we if we allow things to go And be replicated in many places. I think we get a lot of value That we wouldn't necessarily just by going and solving it at the routing level, but even this this weekend I'm kind of in awe of the people that are here There are people that are trying to solve this this type of problem at many different levels And so and we needed it every different level because for instance this approach doesn't handle Telecommunications worth a damn. It doesn't have any of the real-time aspects That we need to solve in other ways And it has to be where you're actually landing someplace particular. Let's obscure how it gets there Well, the question that I have is actually not about that. There was more comments It's a German leftist tradition, I guess to give a five-minute introduction and then ask a self-answering question but Effectively, I think you said something in your talk that really shocked me, which is we're gonna get to an end state Where this is it you're we've done it They can't censor the internet anymore and I think from a political strategy perspective That's like declaring suicide and I wanted to humbly suggest that we need to think 20 steps ahead not to the end of this step Which is one more step even though it's a big step and it takes a lot of work I think that if you are successful in what you were doing you will have a lot of trouble from very powerful people and We have to plan for that it can't be the case that what you propose is the end step I mean you're you're going to create an un-revocable Huge history of a lot of things and there will be powerful people who want to erase those things And they will attack those systems and they'll attack the people that run those systems And we need to think about how that works and what the internet infrastructure in general looks like in the future And also what society looks like with these kinds of systems And we should also talk about that in addition to this and I just hope that it's not the end state because No sounds like Trotsky's Revolution theory, but without the part where you acknowledge the ongoing work. No, I I guess I I Misspoke No, I think this is just a step that we can take now because of the maturity of the technologies and some of the Motivations we've seen I'd love to see all sorts of mutated weird-ass things happening on the internet I want to see it so that we don't have to wait for these large software vendors to go and do new types of Experiences on all sorts of devices. We just have some really interesting tools available to us the 20 years ago We didn't and now it's time to move at least another step forward. So thank you for bringing that up Okay, I think we have Maybe time for one more question On the right hand my right hand. Yeah, this this ties in somewhat Some of the comments there So some of what you're saying sounds very familiar and the reason it sounds very familiar is Two three years ago. I think it was when the Congress was last in Berlin. There was a session held in Seabase Which was all about People trying to do distributed Privacy preserving ways of doing social networking all the stuff that people are doing on Twitter and Facebook and so on but in in ways that have the kinds of properties that you're talking about here and I didn't follow a lot of that in detail. There was a follow-up session that we couldn't next to Congress There was a mailing list. There was all sorts of projects involved a couple of which I saw up there But I want to give you the impression that there was a whole ecosystem of projects that were working in that space and Very similar goals, but you're sort of looking at a WordPress set of functionality Which is a pretty limited subset. Yes It'd be a shame to miss out on the opportunity to do something that covers both use cases Yes, you're absolutely right I'm aware of some of these systems and I know that I'm not using them So I was like, okay, where did what happened to them and some of it? I'm Imagining is some of the critical mass problem that you have to get to a critical mass and some of these social networks And so by particularly picking a blog or something like that that might have some Content that you might want to be able to get to you that other people don't want you to get to that that is a small enough piece that could Independently exist and start to expand if we get some of these underlying pieces right of Whether it's a distributed file system or some of the distributed naming systems distributed payment systems We can weave them together in new and different pieces I'm quite conscious of the IPv6 problem of just sort of everybody has to seem to move or it doesn't happen And so I think we should need to pick our goals carefully So that incremental wins will be still a win enough to gather next momentum for the next step But yes, I'd love to not be using Twitter anymore So thank you very much Okay Thank you, Bruce sir