Tokenizing the Permaweb - TOKEN2049 Singapore 2022

thank you everyone okay so today I'm going to tell you about awive which is the base layer of the permoweb um so our weave is at a fundamental level uh blockchain redesigned to store large amounts of data inside not just large amounts but like a whole web worth of data inside uh why would you want to do this well since the beginning of recorded history humans have uh seemed to possess this in innate desire to record information for extremely long periods of time uh we build all of these archives we use unbelievable amounts of resources actually to try and store documents for hundreds or thousands of years and we essentially always fail uh why do we fail well because we use centralized repositories of the information and they always fall victim to fire flood or damaged by governments eventually so there's another reason you might be interested in doing this too which is that history is written by The Victors we think Winston Churchill said this not actually sure because we've lost all the records of it um yeah so so if you can build a an archive a data storage system inside a blockchain you can use the democratizing and sort of uh permissionless nature of that system to make sure that every person can have the information that they want to have stored uh forever they have access to that which is is something that is like fundamentally different than the traditional archiving space typically if you want to store data in an archive you have to be like a government yourself or you have to have you know tens of millions of dollars at least to get started if not hundreds of Millions with the permo web what we're doing is essentially making it so that there's one unified protocol for building archives and then anyone can add to that archive for I think it's about half a cent per megabyte typically so if you've got a blog post or something that you want to add to the record of history or a picture or something like this you're gonna be able to do so for fractions of ascent essentially which is a massively democratizing force in in the space of long-term data preservation so another reason you might be interested in doing this is understanding that as George Orwell put it the person that controls the past controls the future and the person that controls the present controls the past so the fundamental observation here is that by changing the way that people think about the future you change the way that they act sorry think about the past you change the way that they act in the future if you control the present time you control information about the past it's a fundamental tactic of authoritarian governments they like to control uh access to information about the past and we saw this most recently You could argue in Russia um before the invasion of Ukraine uh Vladimir Putin spent about a year going around trying to recreate the history of Ukraine he actually wrote this like 7 000 word essay or something you can read it online it's very weird but trying to um yeah essentially redefine the history of Ukraine such that it doesn't exist which was a sort of precursor to the action that we now see by controlling the present try to control access to information about the past or perception of the past in order to change how the future worked so okay that's why we might want to do this uh well how do you actually get it done you you have to solve essentially two simultaneous problems the first is that you need to um I'll actually start with scalability the first is you need to make it so that a blockchain can actually fit large amounts of information inside like normally um a blockchain will fit about a maximum 10 megabytes per minute or something like that throughput rate onto the chain itself so that's obviously not big enough we want to put like an entire web inside this thing that's that's not going to work okay so you need to solve that problem you also then need to solve uh how to make that economically sustainable right so if you can store information uh technically forever where you have to have some way to economically pay for it scalability how do we do it well it turns out that basically all you have to all you have to do um is localize the data so you get all of the data you want to put in the system you split it up into tiny chunks you then make a Merkle tree of all of those chunks and you commit the root of the Merkel tree and then you make the mining process about storing that data so instead of wasting energy uh guessing random numbers you say hey go get me byte number you know seven billion from the network go get me bike number five billion and then you recall them from the hard drives and you test them for a cryptographic um would you say a cryptographic puzzle essentially and and by doing this you can enforce that the miners have to replicate the data in order to get paid okay so that's that's pretty neat and then on the other side now you've done that you have to work out how to pay them the only way that we think this is possible is using an endowment structure um an announcement essentially is you put a large amount of uh dollars one way or another aside you make interest on those dollars and then you use the the yield that you're generating in order to pay for some kind of good um most common example of this outside crypto or outside our we've actually is uh in universities often you'll find that a university has a massive endowment and they uh they use this to pay for tuition fees for students over time they might take you know 10 or 100 students every year and they pay for their tuition by just having a bunch of capital and generating yield on that capital um in aoi's case we did it slightly differently we said okay well we can actually get yield from the declining cost of storage so if you uh look at your endowment contribution which is 200 years worth of storage at present prices up front you see that as the cost of storage declines you get a natural yield on that in fact actually this is very very robust because humans love data and they love making it cheaper to encode data and to store it so you can you get a more reliable yield using this type of system than you would try and bet on the stock market or what have you um in fact over the last 50 years this type of system has yielded a 30.5 return in storage purchasing power um on average which is pretty remarkable but you can't use that yield for anything other than uh purchasing storage so it works great for us but you can't financialize this product okay so what we do with our weave is we say 30 is is pretty robust uh let's let's build a system that will be uh stable as long as that rate stays above 0.5 percent which is approximately a 60x margin of error um and the idea there is that we really want to build a system that is reliably robust of an extremely long period of time in many different economic situations so how we've launched um in 2018 and has been growing more or less exponentially ever since um we see that all of this information inside the system it's about 350 million pieces of data at this point maybe there are 400 million of this time I guess um is available in people's web browsers and that's something that people find really interesting because it means that you can essentially browse this system as if it is a web we call that web the Perma web it's a global permanent web that is a booming noise in the back there geez okay so over time we think that this collection will become the sort of basic yeah knowledge set of humanity we want to make it so that anyone can contribute any knowledge that they think is useful to other people over long periods of time um but it can also store apps which is a little bit tangential you would think but we kind of see that applications are essentially knowledge for computers like it's how do you achieve a task and it turns out once you build applications on top of this system the permanence of the data gives them fundamentally different uh power dynamics so these applications are ownerless fundamentally once the developer has um has published them they cannot unpublish them and that that seems like a kind of trivial thing at the beginning but then when you look into the effects that it has it's actually pretty profound so an example is this this is our Wiki our Wiki is a fully Perma web hosted application um which means that Ricky uh Ricky ah Wiki Ricky as we call him um cannot change this now that it is launched uh that means that the the code if you will is Sovereign inside this system there is no higher authority than uh the code itself and what that code says is we want to entrust this Wiki to a Dao and so you can contribute articles to the wiki and um and you can essentially gain some tokens in response you can use those tokens to elect moderators or to remove moderators and they're essentially the people that get to choose what is actually shown on the wiki and this is really interesting because this is probably one of the first maybe even the first uh real world web applications where the community themselves is Sovereign there is no higher authority if you try and build a system on web 2 you will always have a server somewhere that um defines what happens when you visit the web service when you go to twitter.com Twitter the company owns what happens on that platform if you go to wikipedia.org uh the Wikimedia Foundation owns what happens on their platform just by the nature of the protocol the the information lives in a place and that is a physical location like underneath all of the wiring uh that goes from Wikipedia.

org to you know the data you're actually just resolving that to an address um yeah so in the permo web is very different the data is held in a collective Commons and uh as a consequence you can recall it from any different Gateway there are quite a few different gateways now you could even actually access it directly and so you just don't have that kind of location locking effect uh that you get on the on on web 2 and that means at the fundamental level when a user comes to use an application like this they're playing a completely different game of power this application can't be changed by anyone apart from the community but it doesn't stop there you can build social applications on top of this uh for example you can build a system that is a Public Square that verifiably uh guarantees your right to speech you can speak there today you can speak there tomorrow like no doubt as long as you can access the r weave this will work and you've got to compare this to the to the situation on the centralized web where basically the person that controls the server controls the speech it's a completely different game so another interesting property of these systems is that firmware applications preserve what we would call consumer integrity what we mean by that is when you use a firmware application today you can use it like that tomorrow for sure there is no chance that there will ever change um sounds basic but you've got to compare it again to what happens on web 2. in web 2 you have Gmail for example you sign up for Gmail and Gmail uh says you know here are terms of service you just you just click yes and maybe they're not selling your data today but you don't know what they're going to do tomorrow they might well start doing that in fact actually Gmail did this they didn't mine your data to begin with now they do and I'm locked into using Gmail can't really Escape apart from weavemail and so there's this weavemail client here and I can be confident I can use it in its present State forever if the developer wants to build a new version they absolutely can but like a blockchain like system I'm not forced to update if I don't want to I have rights as a user in this system another interesting feature is the pen web applications require zero maintenance so as a developer when you're building a typical web service you have to make sure that that web service is going to stay online and you have to administer the servers and so on the perm web it's different we've dissociated code from infrastructure so you launch the code that's your job as a developer you pay for it it's like that that we've mail application that we're looking at I think cost something like a hundredth of a cent to store so you pay once you store the application for a hundredth of a cent then the user recalls the application on a Gateway and that Gateway is essentially provisioning the infrastructure for you so as a developer I don't have to worry about infrastructure anymore I just launched the thing it gets recalled by the user uh the user has an economic relationship with the Gateway and that's their job that's their business if you will um and I'm freed from from caring about it so you can build a hackathon project on a weekend or what have you you launch it it just doesn't break it continues being served forever and and this is not like Theory we've been running this thing for I guess four and a half years there's applications from four years ago they work just as well today as they did back then and the developer hasn't paid a single Cent more to host them and the um yeah they've done absolutely no server Administration or anything else like that another thing that we're looking at at the moment is monetization on the web so with the Perma web you can essentially embed smart contracts inside the data itself it's not like uh you know an nft or you say okay I've got my smart contract on Solana ethereum whatever it happens to be that points to a Json blob which points to your data and that's the kind of chain of um of brittle linkages essentially uh where one of those disappears the whole thing breaks well on the Perma web uh we built it to be an archive and that means that you don't just use content identifiers it's a crazy cult of content identifiers in crypto they're a really bad idea um you don't just want the hash of the data because the data itself is almost meaningless you want to know not just the data but who said it and what metadata there was and in in the case of our weave it turns out that that metadata can also include contracts itself so you have an atomic asset we would call it on our weave it's got one address uh you know who put the data there at what time they put it there the title description everything else and also who owns it in terms of a smart contract or really anything else you want to program into that smart contract uh yeah so we think this is a pretty interesting feature of the system people are starting to use it to build a web that is tradable here we have a Reddit like system on our weave it's called now.rweave.dev or actually now and then you put you know dot any R weave Gateway and should return for you in the same way um on now you have assets oh there we go you have assets that are Pages or images in the like on the Perma web and when you like them uh it essentially generates a Profit Stream for the person that created the um the asset itself and then you can trade that so this one has a profit sharing token inside um that means that when people stamp this thing when they get likes there's a a profit flow and you can buy and sell that as a user so that's an interesting thing that the rweave community is is playing about with right now okay thank you so much everyone