The Web We Want by Brewster Kahle (Devcon4)

[Applause] thank you great to be here I I'd like to start with a little story so the Internet Archive started accepting bitcoins in 2011 as donations I kind of wish we hadn't sold those 2700 bitcoins that we got then for mere $5,000 but by 2012 we were wise Ihnen up a bit and we were getting these donations and we were paying our employees optionally in Bitcoin we set up an ATM doing these things and this I was getting interviewed by Bitcoin magazine and the conversation went completely fun and interesting areas I was getting to talk with this young man Vitalik and I it was I now realize I really it would have been a much better article if I had interviewed him as to his and gotten that on paper as to what he saw and how he saw this whole thing evolving but I I think it's actually kind of great that when I wanted to upgrade from Bitcoin - qti it was big it was Vitalik himself that sort of said yeah you might want to use these wallets and and and the like and it's worked out pretty well ever ever ever since what I want is a game with many winners I want a game that evolves I want a game that isn't monopolized isn't structured in such a way that there are no room for survive us or surprises and evolution and coolness um I like games that are fun and inviting and don't do what you expect the only games that I have ever known like this are fundamentally open and by open I mean no centralized points of control I mean decentralized games and I've had the pleasure in my career to be involved in a couple of them one is the Internet itself so during the night late 1970s and 80s helped participate in putting this thing together where if you just played by the protocols you were welcome there wasn't a permission-based you weren't on somebody's platform it just evolved and and mutated along in conferences kind of like this called the IETF and it worked very well there's another game was the World Wide Web where you could go and set up and just be part of it just by playing by the rules but then you could bend the rules and you're welcome to go and transform this thing of the World Wide Web for many many years those who have been great and they're real there were people that were involved in those whether it's Vint Cerf and oak or and I hope there are thousands of us that were involved in the Internet era of trying to get that whole thing on to work I think we felt very successful but we didn't necessarily all get rich right that there was just a way of sort of participating in in this game with many winners but not everybody was a winner in the world wide web Tim berners-lee actually never became completely wealthy but he helped make something very very important happen I think that this area that this conference is based on is all a base around making games with multi winners and let's do what we can to try to keep it that way because the internet and the world wide web have seen some real closures on some real collapsing in the early 1990s there were 40,000 ISPs in the United States but now there are only really a handful that make any difference at all so something's gone wrong now instead of the world wide web where lots of blogs and the like that it's collapsed down to Facebook or Twitter so we have basically a it's not that the web is failing it's the people are setting up castles or closed gardens in the open pastures it's something that I think we really have to watch out for and when we're looking at sort of the structure of the Internet Archive what we were doing we thought that we really needed to be involved in building a new game and that's what got us involved in the whole decentralized Bitcoin world and and beyond and that's what I'd like to talk about today so the Internet Archive I get to work here this cool place we bought a church in San Francisco and converted it to be our headquarters and our mission is universal access to all knowledge the thing that I signed up for in the internet is to try to build the Internet into the Library of Alexandria version two could we make it so that all the published works of Humankind books music video web pages software could be made available to anybody wherever they were forever for free how many people here have used the Internet Archive oh yay fantastic the wayback machine probably did you know that we do books no I did you know we do television so we okay a few people okay so anyway we do a lot of these other things on the at the Internet Archive as well we basically are trying to take all of the existing materials digitize it and put it online permanently we're trying to take what's born digital and make it permanently available because the average life of a webpage is a hundred days and it goes away so we have 300 billion web pages and grows you know anyway a lot and but what I'm going to talk about now is these restructuring of the web itself the web itself is but is being we're being portrayed the Commons is being poisoned it is the people that were supposed to be trustful haven't turned out to be and the real wake-up call for me was Snowden to go and show that a lot of the people that helped build the the web had men turned to use it as a massive surveillance tool then the last election in the United States no matter who you think should be elected or not the actual process of it went very very wrong that we had people trolling and packaging and manipulating fake news structures were rampant and they're still around so we have basically structures that aren't working all that well what I suggest we need is a web that's decentralized a web that has got a peer-to-peer back-end on it that preserves some of the values that we wanted but doesn't go backwards but it's going forwards into the technologies that many of you are really pioneering and putting putting together to try to do this two years ago we put together a call for the decentralized web some it wouldn't it be great if we could build a web that's decentralized and Tim berners-lee and Vint Cerf came but also the ipfs folks all sorts of people were building this when we did this again just last year a lot of these people had startups that were well funded it's on its way but it's still not altogether there yet buildings decentralized systems is fundamentally - harder than MIT building centralized systems and so we're all trying to get there and the way that I wanted to try to help was to be a convener but also by going and putting our own stuff online could the Internet Archive make itself decentralized could the forty five petabytes of the Internet Archive store the three million users a day that are using the Internet Archive could we make that decentralized could we make it so it could get to it from anywhere and nowhere and so that is the journey that I thought I'd go over today and to list a bunch of the problem areas that we're finding that frankly we could really use help and it's not just we use we as the whole web community the Internet community could use help on a laundry list of projects that I think you guys already know the answers to many of them building a decentralized Internet Archive I realize there are there about five central magic properties that were based on that is I think it's time to just take a little notice of just like holy crap the idea of content address ability is wild it's sort of Merlin stuff that if you just have a hash of a file you can just call it from the universe and it can come together and you can have it and you can check that you have the right thing yet the first time I think you never you don't actually need provenance you don't know who you got it from as long as it it checks out with your hash you know you have it right this is a big interesting technological step the magic of encryption public key encryption particularly is in a woven into all our systems that yeah yeah yeah but just take a moment it's pretty amazing in terms of being able to be built in a decentralized environment private communications that are extremely difficult to bug another that's not quite technology it's sociology and politics and economics is a well connected internet but we're all very much based on it and we have to do what we can to try to keep it around and alive yes there of these countrywide firewalls going up and their problems but in general we have a well connected internet to more things the that we have an ability now to have downloadable code that's safe so we're used to be that we had to wait for a Standards Committee to come around to go and invent and then it gets implemented and then it gets finally deployed it's super slow evolution especially in the web browser space can we go and download javascript to make my our web browsers into something else and you guys are doing within the etherium contract structures with web assembly which is excellent but this is magic in the sense that you can evolve much much faster and lastly crypto currencies how do we get beyond the original sin of the web which is advertising how do we do that how do we make it so you can make money by publishing on the Internet right now you have to pretty much sign up to Amazon or iTunes or something like that to go and host your stuff can we go and use crypto currencies and I think that's the big opportunity okay I'm going to do things something that they always suggest not to it actually made people in this in the conference very nervous I'm gonna try to do a live demo of a decentralized Internet Archive if this works you're gonna have to applause okay okay I did try this a month ago in front of a group and too many people went to this website before and things crashed it was kind of from so don't go there yet here I go this is the Internet Archive website so just normal archive.org sort of see what it is what I'm going to try to show is a decentralized version of this that is a lot like this so you kind of say well are you gonna show the anything new has happened I'm going to go and turn on the JavaScript console and show how much mechanism is going on so when I change this to D webbed archive.org it's going to bootstrap all sorts of systems web torrent ipfs gun yjs all in JavaScript not just as clients but as servers of these and then bootstrap all of that and then start getting all of the assets to be the internet archive site okay wish me luck okay here just okay they boot it up alright okay we've gotten out we got them on banners up okay so now we have the basic homepage up in terms of all of these images let's try going to get a movie so here again that C's going and getting a lot of assets of the Prelinger archives which is old archival movies these are old educational films that anyway they're completely dorky and wonderful and popular and so let's let's go for health your posture you know gotta stand up straight they're gonna tell you how and okay this is going to bring in things using ipfs gun and yjs and then the actual video if it works okay we've got one pr5 peers now so it's downloading its five peers coming in [Music] so this is basically trying to weave peer-to-peer into the web itself but there's all sorts of other magic here I know that you're all gonna really want to go see this movie and I'll make sure you're standing up straight and kinda like but web torrent has was doing a lot of this and so it's not only downloading but this web browser is uploading this file already can we go and make this smoothly easy to do and how do we get the bootstrapping system such that this can happen in censorship in a censorship resistant way the Internet Archive is completely blocked in China for instance and sometimes blocked in Russia and occasionally in India and bla bla bla bla bla can we go and keep things up and available through these technologies and I think so but we're not quite there yet so what's the benefits of going through all of this work right we have ability to go into our web sites or home pages or whatever well I'd say there are four big ones one is we can make things more reliable they can be archivable it's like git it's it's you can roll things back you could fork all sorts of nice things censorship resistant and even works in poorly connected zones like say a lot of the United States it offers more privacy one thing that was really spooky to me as the in the stones revelations is the GCHQ of the the NSA of the united kingdom stood in front of the WikiLeaks website and recorded all the IP addresses that came in and it turned over a large number of those to the NSA for further whatever they wanted to do then the papers we don't know what the NSA did with that but the idea of people being rounded up for what it is they read and had bad things happen to them it had a very long bad tradition in the library world so this is sort of you know puts a big red flag of the idea of just getting in trouble for what your reading is not good enough it's not how the world's supposed to work so let's build a better system - two more things is we can build a business model that works better than and advertising and we can make it fun and malleable I think we could make a system that works a whole lot better and and does better okay here's my laundry list of 13 projects that has been evolving out of the decentralized Web Summit and and the upcoming in the summer next summer the deep web camp in San Francisco these are areas that can really use some help for instance naming in the whole decentralized web world if we're going to move beyond DNS how should that work that's actually a probably a very good place for registries and Ledger's and and and the like how do we do key management on the D Web this is really tricky it's um Dmitry excuse me on Dietrich of Mozilla talked yesterday of wanting to build better key management into the browsers let's help them let's try to figure out what we should do to be able to do that in such a way that it works well for normal people identity reputation how to do mutable data at scale so when BitTorrent we implemented bit torn on all our material so we have 45 petabytes of materials available bit to arm it's kind of cool but it's also static and if any of those files change if you don't get the new torrent it will go and just continuously download thinking that it didn't get a proper version so we have problems by having things be more immutable than the BitTorrent or ipfs world's currently work how do we do archiving in the decentralized web do we need incentive eyes structures I'm hoping that we can get ISPs to go and make their own cached decentralized web nodes even if it's just to serve to their customers so it's basically a content distribution network that's open access so it's not akamai that's only the big boys can Netflix can go and put in that anybody materials just get faster by they're getting out there so the Internet Archive can play a role but can we have lots of different organizations play a role in archiving and caching materials reader privacy is notoriously hard if Zucco says that privacy is sort of not easy to get OPSEC absolute then it's geez well what am you mere mortals supposed to be able to do here but we can make progress we can make it's at least the GCHQ attack on WikiLeaks readers is much more difficult to do takedowns in the era of hashes rather than URLs what does that look like and should they be avoided are they opt-in is it kind of like spam how does it how is it going to work another row of challenges again and what I'm trying to do is is just trying to excite you guys to help out say I think I've got an idea on how to help on say web beyond advertising I've got a puzzle for you it's a crypto puzzle that I've even posed to wit Diffie and he hasn't come back with an answer is can you make a deep urges on the D web what does I what do I mean by that let that my kid is a guitarist he plays songs and so he he hasn't held them what if he wants to make a decentralized website that has his album on it the first ones free you can just play it but the others he wants to sell so he wants to sell for some amount of money what I imagine is we would encrypt those songs with a key and then if then it would go and advertise on the decentralized web page put a coin in this slot and then you'll get the key to go and unlock the song right you could go and rip them off by going and you know unencrypted and it's passing it to your friends or whatever but that's always the case but at least if you wanted to play the game can we make it work without any Amazon or iTunes or even an escrow agent and I can't figure out how to do that because it requires having a secret key that lives say in a public contract and then I don't know enough about zero knowledge proof to be able to figure out whether that works but anyway if we can crack that then you can feel confident putting a coin in the slot and knowing you're gonna be able to get some decentralized content that's for pay can we make the decentralized web work in the browser itself I'd say one of the cool things that these guys did in the area of the decentralized Internet Archive that you just saw is there was no extension there was no new browser it was it works on your phone and actually now go ahead please do go to do a archive.org and just try it out because you're gonna be basically using everybody else that's here to be going and starting to access the internet archive site so can we make it work in browsers completely unchanged and I think the answer is kind of but we're gonna need some help from the browser companies themselves but can tweak you really get going storage and they decentralized web is it gonna be file point it's going to be the Internet Archive how are we going to do a larger scale storage than what's actually on a blockchain governance in the Deep Web I'd say this community is thinking harder about governance than anything I've ever seen in terms of these did the whole Dowell thing is terrific that we need many models to be played with what are we going to do about the D Web and that the standards that are kind of evolved in that space and who can help and lastly can we define our decentralized terms I mean I'm old enough to remember II when II was supposed to be in something or open remember open seem to have some definition at least Stallman had some real definitions of what free software men he had principles and I think we need the same kinds of things around decentralization or it could just become a marketing term that is just wasted and we sort of would have built up a lot of goodwill around a term that'll mean as much as organic and so if we can go and do something in that area I think it would be a very good long-term investment and of course rethinking the UI in u.s.

because I don't know about you but I I always have these cool things that I want my wife to use and they're just if it's not actually working well she'll just sort of she's my good QA Department it's like not good enough to survive so let's go and make the d web a fun and interesting area to play in to help I would suggest what we what I've been doing is I've been trying out the puzzle if you're driving around or in a shower blank but decentralized so just try it out on your favorite piece of tech I think Google Docs is freaking magic it is an amazing technology but it's creepy so anything that's creepy I want to go and think can I make blank but decentralized Google Docs but decentralized Google Maps but decentralized slack but decentralized Twitter but decentralized Facebook but decentralized we just bought a Tesla which is kind of cool and it is like driving a phone it is really strange it is a connected device the company knows everywhere we are it is it is creepy so how do you make cars that are as cool as a Tesla but decentralized it is absolutely possible to do and one thing you might want to try out is your own thing that you're doing and I know if you're doing things and decentralized tech you say oh well I'm already doing that but maybe your homepage is it maybe your blog isn't so how do we go and make all of this work I wanted to say thank you to Mitra and the Internet Archive engineers that made this demo all work and thank God I did a web tour and ipfs Gunn yjs and others really pertinent a lot over the last year to make this work and they're continuing to make it so working and censorship resistance and for masbath back uploading and replicating petabytes of information let's build the web we want and the only way we're gonna do that is together I'm hoping that this was somewhat useful and I look forward to talking with you any questions [Applause] was I really that unprovoked --iv rooster this isn't like a question per se but I work with ipfs on a like almost daily basis and my company I have I believe solved already some of those 13 challenges oh that you so my question is would we be able to like talk to you after absolutely you can talk to any of a number of people because it's not just me it's now groups of people and we're trying to basically put these together so yes if you have real clues as to how to make progress on these they are this sort of the outstanding less fantastic thank you yeah a great talk thank you so much I'm wondering how you think about incentivization for obscure data stuff that like from the 50s these movies that probably nobody will never really look at how do you keep those around how do we how do we do deal with the long-term preservation of stuff that the longtail right then you know the Hollywood movies are gonna take care of themselves just fine how do they deal with the longtail I think it's by filmmaking compatible systems that are easy enough to use that even libraries know how to use the cool thing about libraries is they're paid to give stuff away it's about a twelve billion dollar a year industry in the United States it's about a 21 billion dollar a year industry in the world they've just haven't modernized so the Internet Archive I think oh it was pretty modern but the library system hasn't modernized but boy are they up for it so when the last election happened I got worried about being in the United States and said gosh let's accelerate this make copies in multiple places so the Internet Archive has got two copies in the United States partial copy and Amsterdam and a partial copy in alexandria egypt for real actually it's kind of cold but we've also then said how about Canada again what can we do there so I went around to our Canadian friends and they said sure sign us up we're up for spending money and our universities to go and help back up hearts if not all the whole thing but I didn't have the tech for them well yeah the idea of having just dead dumb storage someplace is a good but not great you really need continuous usability so they wanted to go and have copies of parts of the collections but also serve them that's I think the opportunity of a decentralized internet archive so if we did this there's plenty of money for keeping oh I don't know mere exabytes alive which basically is everything written plus basically published we're 45 petabytes of data and we're small so it's it's I'd say the long tail we can count on that part of the system as long as it's compatible with the thriving commercial system the library world is difficult to go and drive forward technologies but it can ride other people's technologies it's an idea thank you actually sorry regarding this just disgression I get the system at the theorem San Francisco hackathon it's called laser growth which is a decentralized optical storage system precisely for this long-term archiving sign that man great we need more ideas than this and I think it's a lot of these things are bottom-up if we just wait for the library Congress to do something it's going to be a long fricking time so it's going to be bottom-up and that sounds like there's freedom box there's a bunch of other boxes that people have have tried to make let's make them useful and used yes your answer about moving the archive thank you in the Trump moment was part of my answers part of my question the question is in the Trump boom what projects maybe that you didn't even mention here should we be focused on slash what norms give you pause what projects should we be focused on now there's a lot more excitement around the nonprofit sector than there was even 10 years ago and people are just trying to make some crappy app and you know buy a house for cash or whatever just yeah so there a lot more interest in these the ef-s public knowledge Wikipedia Internet Archive they're our biggest donations from ever other than from from me have been from the crypto world actually the pineapple fund gave us two million dollars a theory a foundation give us a hundred thousand handshake gave us a hundred thousand I'm nice ago gave us 30 thousand we've gotten a lot of support actually so that's that's my Schilling moment here but there is interest in these spaces and I really feel and I you know I was invited here I think it's kind of cool because we're really from a non-commercial space we are for the public good and that that's being involved in the conversations I think is a really good good sign in terms of what to actually do is I've got a list of 13 things that if we solve those I think that we can crack through and make a lawful lot of the websites we use decentralized and you wouldn't even know it that would be the best case it's like oh yeah of course it's decentralized it's like of course it's HTTP all right of course it's and just how do we go and make it all just go decentralized my question is kind of two-part you mentioned earlier about China and not being able to access I wondered if the D Web archive.org is accessible in any way in a place like China that has not yet China is actually really difficult to crack it'll probably be easier in Turkey or Iran or some of the other places that were blocked I think no more often in Iran now so there's a but because you have to go and get rid of every access to the DNS for instance and that is challenging so China blocks us largely by messing with our DNS and by blocking things going to our IP addresses so we don't have that yeah and it's a lot of the bootstrapping problem and that I think is going to require either browser manufacturer cooperation or extensions okay that was part of my second question when you mentioned going beyond DNS I just wonder if you could expound more on um Mitra who's been working on this has been making a cool system to go and make it so that if you have D web in your domain name D web archive.org it mutates that into AD web me slash arc-flash archive org or something like that and it'll do it for any prepend like that so if you have d web in any domain name it'll swap these around and it then looks it up in us in a sequence of databases to find who has the authority for that part of the subtree is there something we can read more yeah it's all open source but I'm not sure it's the perfect thing but people are trying and to basically build a sort of DNS that's open and resilient and doesn't have some of the problems we saw the last time around I think hello and whenever I should get the yank please yank me but yes yes sir Sam for the dog and who this who can upload content to the Internet Archive you can Holly the content go burn or curated it can't get the Layton's on how who is okay in the platform how do you upload to the Internet Archive and can it get deleted somehow is that the question so on archive.org there's a button upload please hit it please hit it now just hit hit hit hit upload and just try uploading something it's a little clunky it's worse than what you but it supports YouTube its videos and books and all sorts of things and makes it into lots of different formats and keep tries to keep them alive forever but we do take things down and we do under US law we've tried to push back but we are under US law so if people make a valid copyright complaint and prove it then we will make it not available publicly if you want to go and use the archive at scale use s3 archive data we're trying to be bug compatible with Amazon comms s3 cloud s3 amazon.com so you can use your bucket and upload things at scale please let us know if you're gonna do many terabytes or many petabytes but we're very interested so we you that's how all of the materials we get 23 terabytes a day coming into the internet archive all of it comes through our s3 implementation so you can use s3 really easy or just go to the website and hit upload and and please do okay thanks yeah thank you for the talk I thought there were a lot of really great points that you made I guess it's this last question so in terms of you know just trying to get more people can it be the second to last question okay sorry yeah so more people from around the world accessing the same internet trying to not move in this to internet parallel that's emerging with nations like China kind of having a separate internet yes so how can we prevent that with better interfaces / what could we do to it like you said earlier you have to take some things down based upon US law that seems to be you know both a good thing in terms of preventing malicious content but then also a bad thing in terms of potentially having a burning down of the Alexandria library yeah oh man like how do we prevent that I I'm not sure I've talked about this in public why we're blocked in China as best I can tell so in general company countries will come to us and go and say this particular thing is not legal in Germany or whatever under our expansive copyright terms and so if if they basically make a good case that it's illegal in their country then we'll block it for users in that country and that has allowed us survive very well the u.

s. is actually pretty liberal unless you talk about child porn or almost anything dealing with Muslims but otherwise the United States allows a lot to be put out there in China there were some uploaded videos to the Internet Archive and they came and said you have to take them down and we said are they illegal in China and they said yes those are illegal in China we said fine we'll block them in China and they said that's not good enough you have to block them for the whole world and because they said that we said we can't do that that's not how we operate and they said okay chunk so 45 petabytes went offline including millions of books digitized by the Department of Education in China that have been done as a collaborative project so I think we have to um for those of us there work within the governments and the like is we just have to make sure that we're not overstepping our bounds and allow some of this experiment to continue that's I bet that's the only thing I can think of and make multiple copies all right Brewster I know he's gonna hate me so go fast okay I'll be really fast this is the same question I asked you the last time you gave this speech which is if advertising is the root of all evil and getting people to pay for it will fix it then what do we do about the fact that we have a hugely unequal world and the ability to pay is very very widely dispersed and that all the stuff you pay for right now is just as abusive as the advertising stuff Netflix buys and you Apple locked you down and so on damn IIIi think I've got to go now so yeah how do you even have payments work in sliding scale structures and and having an equal world when there's vast inequality in wealth she I don't know I think we need a lot more experiments going on than the few that we've got going and I just want to see more things happen out there it monocultures don't work very well advertising based systems end up with newspapers television Facebook Google so that's not good enough I'm I'm I'm Oh generally a free-for-all kind of guy there's ways of making that work I just like to see some more tools in our in our tool box sorry not be more satisfying thank you very much we're around for the rest of the day you