The Wolf Of All Streets - The Endless Opportunities of Decentralized Data | Sam Williams, Arweave

Episode Date: March 15, 2022

The Russia/Ukraine crisis has prompted floods of misinformation across the web. With social media platforms like Instagram and TikTok, it’s become more and more difficult to break through the cloud of misinformation and fake news. Sam Williams is tackling this head on through the creation of the permaweb. The Founder and CEO of Arweave (https://www.arweave.org/) has built a collectively owned, permanent hard drive to prevent anyone from rewriting history ever again. On this episode, Sam joins host Scott Melker to discuss the importance of archiving fake news, the endless opportunities of decentralized data, and the future of decentralized social media. ••• FIND US ON SOCIAL ••• Scott Melker: https://twitter.com/scottmelker Sam Williams: https://twitter.com/samecwilliams Production & Marketing Team: https://penname.co/ ••• JOIN THE WOLF DEN NEWSLETTER ••• 📩 https://www.getrevue.co/profile/TheWolfDen/members

What's up, everybody? I'm Scott Melker, and this is the Wolf of Wall Street's podcast, where two times every week I talk to your favorite personalities from the worlds of Bitcoin, finance, music, trading, sport, art, politics, basically anyone with a good story to tell. Now, we've seen cryptocurrency talk about Web3. We've seen all the innovation happening in this space. And one of the most exciting use cases is fixing the data storage problem that has seemingly existed for a very long time, especially because all of that data is largely stored on centralized servers and by centralized authorities. Now, today's guest, Sam Williams, is CEO of Arweave, and they are really, really breaking ground in how we can completely change that. And not only that,

it's having a major, major impact on what's happening in Ukraine at the moment. So it's hyper relevant right now. Sam, thank you so much. I can't wait to discuss this. Thank you for taking the time. Thanks so much for having me on. So we should probably get the background on what Arweave is and what problem it's solving. So can you talk about that a bit? Yeah, put simply, Arweave is a permanent information storage system. It's the first one that has existed in history, we think. The way it works is we've scaled a blockchain so you can fit very large amounts of data inside the chain itself and then paired that with an endowment essentially when you put a piece of data into our weave you put like a reasonable amount of capital up front uh to pay for 200 years of storage and then as the cost of storage declines over time you you generate essentially interest in the form of storage

purchasing power and we just use that to pay for the storage. And that means that you're essentially from the single payment, your interest is covering the payment storage essentially perpetually. Yeah, so a blockchain is the perfect way to do this because it is a ledger that no single person controls. And the consequence when you put all this stuff together is that you have a ledger of history

which is out of the hands of any dictator to change over time, or any individual, frankly, to come along and to try and put their thumb on the scale of how people see the past. And we think this is so important because as George Orwell put it in 1984, he who controls the present controls the past,

and he who controls the past controls the future. So by changing the way people, and you see this with dictatorships and authoritarian regimes in general all the time, controls the future. So by changing the way people, and you see this with dictatorships and authoritarian regimes in general all the time, by changing the way people think about the past, you change the actions that they take in the future and the way that the world works. And so, you know, we think it's a fundamental necessity to put archives out of the control of individuals. Essentially, we've built the infrastructure to do that. Sort of the Churchill statement, history is written by

the victors, right? We never get an accurate appraisal of history because whoever wins gets to tell you what happened. Yeah. And history is written by the victors on day one, you know, or after victory day, whenever that is. But then it's rewritten endlessly by people that want to encourage their population to support whatever it is that's the next initiative. And we saw this in Russia with Ukraine, right? Like the line that President Putin is pushing is essentially Ukraine was never a state. It's all rewriting history as a pretext for action in the future. But we want to make sure that all of the information about the present crisis and frankly all of the other historical events around the world are stored in such a fashion

that no single person can come along and change it later. Just in the same way that in Bitcoin, nobody can come and alter the ledger and say actually I have all the money. It's exactly the same technique, just applied to massive amounts of data. across social media for years now having an effect on international relations, on elections, things like that. What prevents people from storing fake news on our weave? Absolutely nothing. And I think that's absolutely right. Here's why. If you look back to the last, I'd say, major theater war of this scale, it's probably Iraq 2003. Why did we go into Iraq in 2003? Well, the stated reason was weapons of mass destruction. How many of those stories do we now actually believe? Okay, let's go back to the last major theater war

before that, it was Vietnam. Why did we go into Vietnam, the Gulf of Tonkin incident? Did it actually happen? Well, what we know for sure is it wasn't. It didn't happen as reported in the newspapers at the time. That much is clear. What actually happens is a bit murky. 1939, Germany invades Poland. What does the New York Times write in the paper the next day? They say that they essentially write the Nazi propaganda line, which was that Polish forces took over a German, I think it was a radio outlet. I'm going to push. I believe it was. Yeah, right.

False pretense created over and over again. And so what we've got to understand is that fundamentally false pretexts are unfortunately a part of history and particularly when it comes to instigation of wars half the time don't quote me on that but a large proportion of the time that is why we do it in the first place and so if we didn't archive that record of history that this this misinformation was out there you look back and you say why did we go to iraq in 2003 doesn't make any sense. But you have to store this stuff because it is the information that the leaders and the populace had on hand at the time and it affected the way that they thought about things. So we try and make, are we the neutral ledger of all of the information that people have when they're assessing the situation in the moment? And then we allow historians over time to come back and look at that and say, well, was that actually true?

And, you know, it's pessimistic in some sense, because, of course, you can't go back and say, oh, that wasn't true, please change history. It doesn't work like that. But we can at least get a clear picture of it. And then hopefully, you know, maybe we can be more aware of misinformation in all the forms it comes in. And perhaps in the case of weapons of mass destruction in Iraq, you know,

maybe that was just mischaracterization, or we just jumped to the wrong conclusion. So a dozen different opinions you can put on that, but one way or another, it wasn't true, right? It wasn't true to the extent that was portrayed. So maybe if we have this sort of accurate ledger of history that doesn't get rewritten all the time, we can have a stronger appreciation for the importance of being careful about what the information that we trust is. And interestingly, something I've never really thought about, but putting the fake news on the ledger is actually a very important part of the current historical context for what's happening. So the fake news itself, even though it's not true, when you look back is a very good sample of what was actually

happening today. Fake news is a huge part of the world right now. So even eliminating the fake news would never give you an accurate appraisal of what it was like to live in 2022. That is precisely correct. I mean, it's been really interesting to see around this Ukraine stuff. Obviously, we've spoken to lots of reports about it

and always this misinformation question comes up, but i see it from essentially the opposite point of view when we first made our weave which to get into that you know the world was i would say more or less destabilizing back in late 2016 that sort of time and we i was doing a phd and i i took some time when I should have been building this operating system and my computer science studies. Instead, I looked at the history of authoritarianism around the world and how that actually developed. And we saw that one of the things that happens is this modification of records of the past, or at least modification of the perception of the past.

And part of that is typically removing access to old records um and we thought well okay maybe the one small thing that we can throw into this colossal mix that is happening but we could actually like you know we could actually do we have the skills to to address is well let's just make it so people can't edit records in the past that seems like a good idea and so so we we started doing that and when it got into production a year later um we the first things that we wanted to archive into it were fake news deliberately and and so the example of this was and this was the first major you could say kind of success in some form for the network obviously it's a horrific event, but from the point of view of recording history,

it was, this is what we set out to do and it worked. Someone from the community in, I think it was December 2018, yeah, December 2018, recorded the first English language, Russian propaganda outlet piece about what was called the Kerch Strait incident. Essentially what happened was a bunch of Russian sailors captured a Ukrainian vessel in the Sea of Azov. I do that every day. Yeah. But they captured this vessel. Okay, and this was quite a serious international incident at the time.

They managed to get the first English language piece from a Russian propaganda outlet, and they stored it. And then on reflection, they noticed, wait, this is really pro-Ukraine. And it was only online for 14 minutes. And then whoever was the ram, Sputnik at the time, memory-Ukraine. And it was only online for 14 minutes. And then whoever was the ran Sputnik at the time, memory hold it. They said, no, no, don't put that out into the information space.

And then they replaced it with a much more pro-Russian article. And so by recording what it is that governments want people to believe, they actually get quite an important view on history, if you will, regardless of whether that's true or not. That's a whole different thing.

But to at least be able to assert with authority, well, at this particular point in time, the government of Russia, their outlets, would you say, are saying this line. And then if they change it later, well, we can detect that change and say, well, look, you're not being internally coherent. What happened in the meantime? So this got picked up by and used in a pretty major piece for a British newspaper, or say high circulation British newspaper. And that was the first time the network was used sort of in practice like that. And ever since then, it's just been growing and growing.

So the idea then is, I mean, you're focused on capturing as much information as possible and being able to store it indefinitely and let the oracles and historians sort it out in the future. Yeah, I think determining truth is like a profoundly difficult thing that frankly, nobody knows how to do. Like, I think arguably, the way that humans have evolved over hundreds of millions of years, life has evolved around hundreds of millions of years, is to address the fundamental question of how do we work out what truth is?

And we do that in groups. And even then we get it wrong all the time. See weapons of mass destruction in Iraq in 2003, which is completely false. There's only a tiny nugget of truth in it. But we all believed it at the time. There was a very high perception of its truthiness. So that isn't part of the scope of the system that we are building directly. There are people building stuff on top of our belief, which is like, well, can we use

markets to sort of, if you will, do like a perpetual future on the truth of a statement, something like that to reward people for saying things that are true. So it's an interesting spin on prediction markets. There's lots going on in that area, but at the base layer, we just want to make sure that the data isn't changed over time and can't be removed from the data set and all people have access to it to assess it at different points in history. Right. So I guess then it begs the question, what are the limits of the existing systems, right? And what are the problems with having centralized data systems because i mean there are systems that can store a whole lot of data right i mean that exists yeah for sure i mean the remarkable thing is that archives which is essentially what album is like a sort of archive of steroids in some sense they haven't changed in like a thousand

years which is really quite profound like there's no real disruption as these things have aged they just they run in the same way and they're all vulnerable to the same problem which is a centralized point of failure right so there is a place that if fire flood or the government come and destroy the data it is gone that is the end of it i'll be was like totally different. He says, well, why don't we, first off, let's not work as a charity. Let's be a business, decentralized business that anyone can take part in, in the model of.

So we don't have to rely on donations. That removes one big area for corruption, right? Like who pays the donations to store the data gets a good say in what data is stored you're trying to sell store something for 500 years okay well then what's the likelihood in the meantime some donor comes along it's like a serious investor one investor but serious backer of this project and says actually like this information they're going to say something like isn't that useful right and then well there it goes gone um so that's uh that's one of the things you can get rid of but fundamentally you store the data and it's replicated around the world in like hundreds if not thousands of places

impossible to remove impervious to fire flood nuclear war you know almost all doomsday scenarios. There's deep irony when the pandemic came along, hadn't thought so much about that, that failure mode, if you will. That's actually the one that scares me the most in the long term of the archive, because people have to come along and manage the infrastructure of the internet. And if they stop doing that, because there's a pandemic that's so extreme, fortunately, of course, while horrific, COVID-19 was not above it, right? People turned up every day to manage the wires

that keep the internet running. But, you know, if they stop doing that, that actually could be a potential problem for the network. But, you know, we've built sort of, what do you say, doomsday scenario planning situations for even that. But my point is more general that, yeah, this radically increases the veracity of archives on so many different levels.

Is it infinite? Is there any end to the amount of data that you can store? At a protocol level, the maximum data stored inside a transaction is two to the power 256 bytes. So that's enormous. Obviously there needs to be hard drives to back it and that's why there are the incentives in the system.

The cool things about the incentives are that if the cost of storage declines to the rate above 0.5% per year, then at the end of any given year, you end up with more storage purchasing power than you had at the start. And so if you look at the history of data storage prices, you see that the normal decline rate is about 30%, three zero percent. So we're like a 68 safety margin there um and obviously for every year that passes and you don't use that full um or there's an excess on top of that 0.5 percent

you just store that for the future so right now there's tokens in the endowment to pay for storage of the data for around a thousand years so it's pretty it's pretty robust it's pretty robust. But of course, at some point, like, you know, hundreds or thousands of years from now, someone comes along with an even better protocol for storing data permanently. And that's great. And then likely what will happen is someone just puts all of you inside that archive in the same way that,

you know, anthologies of poems. It becomes a block. Right, right, right. Exactly, exactly. Yeah. And so you find this pattern all the time that the archives get nested inside each other as things go through history.

So I would say that successfully, our project is that the data set is around until after the last human-like thing that might be interested in reading that data is here on earth or who knows wherever. So is that data, is the data, what you're storing primarily user-generated or is it partially user-generated, but is it also partially like just consuming everything that's on the internet? How do you actually, right? Yeah. So what happens is it's a neutral thing anyone can come along and

store any piece of data uh given that they're willing to pay into the endowment and the endowment like you'd imagine this is expensive the thing is that storage is so cheap that like actually it's less than a cent a megabyte so if you want to store it with pages and like that it's like a cent or two um yeah which is really nice because it means that the addressable amount of data is huge because there's so much data that it's worth storing at that price permanently. Yeah, so people come along, they, like

in the case of Ukraine, where we've now stored as a community 17 million records, I think, today, maybe even 18 now, from the conflict, which is just an enormous archive. And it's one of the strange things about the Ukraine crisis is that, and it speaks to some extent to the horror of the whole thing. It's a highly developed state

where one could not previously have imagined that these things would happen. Yet, here we are. But one of the effects of that is that there are citizens generating signals about this war that they are posting on social media all the time. And whereas previously I saw someone describe this eloquently as the TikTok war.

So back in 2003 in Iraq, you got your information from three different sources. You got it from the Iraqi government, you got it from the US government, and you got it from journalists that were embedded with the US government, you got it from journalists that were embedded with the US military, which were publishing only at the whim of the US government. Those were how you heard about the war. You didn't hear about anything that those three different sources didn't want you to

hear about. Whereas in Ukraine, it's completely different. People just saw a photo earlier from a guy called, I forget his name, but it's just like a guy who went for a walk in a forest and found an abandoned surface-to-air missile, like a mobile system. Yeah. Yeah, and it's just like, wow, that's crazy. But he took a selfie of himself with this thing. It's got geolocation data. So it's like, okay, well, now we have this record of history, which says that at this particular place,

at this particular time, and it's timestamped just like a transaction of blockchain, like Bitcoin can never be changed. It's really robust. At this particular point in time, this guy discovered this abandoned vehicle and that tells us something about the battle that happened in the

associated area and we saw it if you will firsthand and historians in the future will be able to see all that data firsthand another example be you know in the history books people write about well before world war ii broke out there were there were troop concentrations on the border with Poland, or whatever it happens to be. It's like, okay, well, how did that happen? What were the concentrations like? We don't really know because it's been summarized so many times over the course of history. And now we're just reading the summary of the summary of the summary of the original. And maybe we've got like one or two little fragments from the time. It's nothing like what we have today.

Today we have recordings that are verified. Satellite imagery. Satellite imagery, like people just in the cars, they stop the car next to a train track and they're like, oh, that train heading towards the border with Ukraine has a massive number of tanks on it. They record it. It goes on the internet. It gets sucked up under our weave.

And then in the future, historians can look back and say, on this particular date, on this particular time, a battalion of this size moved into this particular position. And this is just like a level of fidelity of the archive. And I would hope a level of understanding of the events that come as a result in the future that just hasn't existed previously. It's a whole different animal. I talk about this all the time when it

comes to sort of, well, information overload or what's accessible to us now that wasn't. I'm 45 years old. I vividly remember the 1990 Gulf War, right, when Iraq invaded Kuwait and we got our news once a day from Dan Rather or Peter Jennings, right? You got your one hour fix on what was happening, but you didn't have the internet to see what was going on 24 seven. Now, if that news was accurate, some would maybe say that was a superior way, right? You weren't trying to consume it 24 7 you weren't but as you said that was news funneled down through multiple sources for a specific reason that's all you had all your specific reason is precisely the point here yeah well and i think the truth of the matter is and you see this happening in the background of this whole crisis in Ukraine, the propaganda apparatus of each of the countries is getting barred from the other countries. Now it's very hard to access RT

in the West, and it's also very hard to access Voice of America or BBC inside Russia. And so you see this bifurcation happening. My personal opinion is that fog of war is so strong that the likelihood that we get any... That we get an understanding of the present situation that will be the dominant understanding of the situation 200 years from now, today, is very, very low. I think that social media, frankly, makes it a lot better. But nonetheless, I think that our understanding will shift over time

as we have more time to reflect and hopefully look at the archives that are being built, like study the signals from history. I mean, they're coming in at a rate of like, you know, 17 million things just stored in our weave in the last two weeks. But that's going to take a long time to properly comprehend. Well, I was just going to say, if it's 20, let's call it 2050, and I'm a researcher and I'm looking back and I want to know what happened in the Russia-Ukraine conflict

and I access the archive in Arweave, how do I even search that and start to make sense of it? Because now there's millions and millions of pieces of information for me to sort. I'm assuming I have some sort of like superior AI that does it for me, or I'm a robot by 2050 anyways, but you know, in my 2022 mind, how do I access and use that data? Well, I think you're absolutely right. Essentially, it's AI assisted, almost certainly. And that is essentially Google search, right? There's this really interesting thing that's happening.

That is AI. Right, that is AI. It's becoming more like the web. The web started as this concrete set of relational knowledge, basically. So you'd have a piece of knowledge on a web page webpage and it might link out to other pieces of knowledge. That was basically the idea. They called it the distributed knowledge graph right then. Over time, it's turned more into this sort of

what we would call like an open data lake. And now we have machines to help us access the right pieces of that data lake over time. And more of more like this is happening at an increasing rate you see the google well they're just pulling out the information you would want to see from the page into the google search results itself they have like these little fact boxes they pull out like the definitions of stuff from wikipedia they even have like frequently asked questions that they've synthesized from the data and i think that's just going to get

more and more extreme so one of the things that we expect to see from Aave at that point in 2050 is it's just a big open data lake that never forgets. And then you have systems to help you explore it more or less. I would really hope and I'm excited to see that in the Aave ecosystem, there are already people starting to address this, two projects called Asearch and Open Index Protocol, where they're basically looking at how do we build search engines

that are provably running a piece of code which is open source? Because I'm like 95% sure, 85% sure, that five years from now, the big thing is not going to be censorship on the web. It's going to be, there's no good term for it even yet, but basically the manipulation. Extracting the truth. Right. Extracting the truth. Well, I think it's what truth is precisely the question, right?

And right now, Google is totally, totally like black box. It recommends things from this open data lake to you. You have no idea how it got to that conclusion. You have no idea what bias is putting into it. And we see this, you know, China is unfortunately the, how do you say, they're leading the way in a negative sense on this. And Weibo, back at the beginning of the coronavirus we saw that there was this huge outcry for freedom of speech which was amazing

very very very rare to happen in china it's on social media um so people in our community like immediately started trying to build archivists to extract all of those conversations and install them before they were censored and saw them in an anonymized way so the person stored it isn't isn't a risk but we can still get like accurate data about it and um but they what they found after they built this archive because of course that only lasted like three days they went back and they're like well how do i detect that something has been censored and so they saw well okay i can go back to the post. Is the post available?

No, it's 404. There we go. Censored. Okay. Interesting. The second thing that you can do is, well, just search for the exact text of that post. And what do you know? The ones that censored, like 2%, the ones that just didn't show up in search results

have been sort of, there's no word for this was like shadow banning on an individual post basis that was like 60 percent of the of the whole data set and then they built a tool on top of this called um wayblocked which was like okay well let's just highlight all of that data yeah right very transparent yeah well you know tell people transparently what you're doing and i think that's the that's the right way to act in the world and in this case it was well let's just make a website that highlights everything the chinese government tries to and tries to censor and um that was a really interesting um you say episode ended up and of course it's a permaweb app so an application that lives

on top of this archive. This is where things get a bit weird and, to some extent, more interesting even. What does the archive really do? Well, it really puts data outside of the control of any individual. Data, of course, can be applications

itself, and the network is queryable. It works like a database. So you can build full web applications on top of the system. They just run forever, and they don't have any node net. And so they built this web application that just highlighted all of the stuff that was censored in China, and it's available from many, many nodes in the Arweave network, properly decentralized in a distributed way, which are very, very hard to block.

And so now you have this web application that's openly available to people in the Chinese mainland, very, very difficult to block. And it's highlighting only the stuff that the Chinese government censored, kind of like the Streisand effect machine. Yeah.

I love that. Well, what about, so what about the interplay here with our personal data? I mean, I think that the, one of the biggest, certainly narratives at the moment when we talk about big tech and talk about Google, I mean, Google obviously is showing you results because they know you better than you know yourself, right? Based on all the data they've collected. Certainly Facebook, the same way. all of this private data that we wouldn't necessarily want shared until the end of time end up on our weave as well as part of sort of this historical record yeah so there's there's

ways of managing this in the system essentially all of the miners in the network can see the data that is being stored there is no there is no encryption of rest which is important because that also makes them responsible for the data they store. So you as a user can come along and say, hey, please kindly remove my data. And then, yes, it's up to the miner to decide whether they're going to do that or not. But if they are, for example, in the EU, where there's GDPR or in an economic area that is GDPR, would you say, exposed, then they have to, by law, remove your data. And I would expect that the army of miners don't really care about storing people's personal

data. That's not the point. So they'd be happy to remove it. They're there for storing history in some sense, and also uncensorable web applications, which is a whole... Yeah, I mean, it plays on this component which is well what happens when you make a piece of data permanent and not controlled by any single party where you can have web applications that run inside the system that have no centralized controller we

think that's just like a revolutionary change in the architecture of the web because it used to be like if you wanted to use a web service one way or another um you had to trust the party that built it the developer so your relationship is the example i always give with gmail right like you got your gmail address like 15 years ago at this point and you signed on you said yes google i shall basically follow on with whatever privacy policy you use in the future. The application can change at will. And I can't do anything about it. I signed, tick the box, let's go. As if anyone's ever read that privacy policy in the history of time.

Right. And then you had no idea that they were going to sell access to your data, put intrusive adverts on it over time. That was just not part of the, not part of your comprehension of the deal you were making. But with a PromoWeb app, an application stored inside our web running inside the system, querying the network as a database, you get totally different realities. You have an application that the developer can issue a new copy of, but they can't force you to use it. So if you like Gmail 2008 version, no one can force you to upgrade. So the power relationship between you and the developer is different.

If you want to use Twitter and you want to build up like a following, that takes time. That is a real investment of your scarce attention attention and it can be taken away from you at any time when twitter the company decides that what you say sure yeah right well you can build applications where that can't happen either but actually your access to the data sorry to the application to the platform is maintained with a guarantee and And the only caveat to that is a simple part that says someone in the network and in the world must be willing to store it. So if it is illegal globally, then it won't be stored.

Or if it's personally identifiable information or something like that that you want to remove later, then probably that won't be stored either because people just don't want to store it. But there's that fundamental guarantee or integrity guarantees that are offered by the system that you really can't get anywhere else on the web. That's something we're super excited about. So you talked about a few of the applications that are being built upon it, which were primarily around search. But while this dreams

or even what's happening now, what else can you build once you have this archive and it's decentralized and people want to access it? Because it seems like once the wheels get turning, it's probably pretty endless.

Right, exactly. I mean, you can take any Web2 company or service and you can make that service again without the company. And depending on what industry you think, that is worth slightly more or slightly less but it is pretty objectively a step up for the user to have those guarantees the platform won't get worse um and to have a guarantee that you can't just get arbitrarily removed from it uh outside of would you say the

the rules of engagement that are embedded in the code itself. Those are almost universally beneficial properties. There's areas where they matter more or matter less. I'd say like email. So there's a system called Permamail, for example, which does exactly this. It's email on top of Arweave. But no one can de-platform you and no one can add interest to your adverts.

There are lots of user interfaces to it that don't sell access to your data. Well, great. They're never going to sell access to your data. They can't. That's pretty cool. And things like social media is a big one. I think that Twitter has basically become a global public square, or at least was. But because they act as an obviously American company that is not neutral, they're losing that. So there's this opportunity to build a real public, you know, global square, which doesn't, where your rights of access are maintained, guaranteed, essentially.

I think that's super, super powerful. And I certainly feel, and I know that lots of others do, because you've invested the time to reach an audience that, you know, it's an interesting thing because like Elon Musk does it, for example. Elon Musk's time is extraordinarily valuable, right? But he's invested that time in building a way to speak to people on Twitter.

Well, if Twitter comes along and just says, sorry, time's up, get lost, or, you know. That's billions of dollars for him. That's literally billions and billions of dollars if you quantify his time in dollars, certainly. How much does he make a second? Yeah. So sum that up over all of the users and multiply it also by the risk, because it doesn't matter about the people that are, of course, it matters about the people that are the platform, but it also matters about the people that are worried they will be deplatformed because there's a cost to that as well in the same way the bitcoin right

we love bitcoin because there's not going to be more than 21 million tokens and who knows there might be a fiat system where they just stop printing tokens like uh the swiss i was gonna say token the swiss currency they don't print so much. They're just tokens anyways. One day, one day they will be, so it's fine. One day, yeah, not too long in the future, I think. But, you know, the Swiss token, they don't print so much of, but there is a cost to the risk that in the future they might print more of it.

So if you sum that up for a social network, we're talking about extraordinary amounts of money. But I'll leave you just to get rid of that and say, okay, well, you are a sovereign person in this network. You can't be denied access outside of the rules of the game that you agree to play at the beginning. And if someone comes up with a new version of the game, you can play it if you want.

You're not forced to. So interesting. And so true, even anecdotally from personal experience, you know, I was a part of the Twitter hack when Kanye West and Bill Gates were tweeting out, you know, send me one Bitcoin and I'll send you two. My account was one of the ones that was compromised and obviously made me realize that I needed to build on other platforms, get people's email addresses, build a YouTube channel, all of which, by the way, can individually still be deplatformed. But you're right. The amount of time and effort spent making sure that I was safe from another Twitter takedown is like a full-time job.

I never really thought about quantifying that in terms of my time or value, but it's incredibly powerful when you think about that. Yeah. And there's the value that doesn't accrue to the platform by people being afraid of this. You see this a lot in the developer space, right? So you've got an API for your cool new service. Rewind 10 years, imagine that's Twitter. Okay. People want to come along. They want to build, they see a niche where they can build a business. Okay, but maybe they don't do it, because they know that actually, you can just turn off access to that API arbitrarily, and the business is over, all of that time invested is

gone. And so the value that would have accrued to that platform, now just, it's not even like there was a risk, it's just never got created in the first place. So if you provide, develop the level guarantees of kind of integrity of rights, if you will, essentially, then you can also allow more value to be created inside the ecosystem you're building for whatever platform it is. And what do you think the future of decentralized social media looks like?

Is it just, I've been saying, I don't know what it's going to look like, but I think that's going to be one of the major stories of the next year or two. I mean, that to me is the most obvious usage of decentralized technology is to decentralize social media. But what will that look like? Can we have something as robust as a Facebook, an Instagram, a Snapchat, a Twitter that is fully decentralized or are there limits to how that could be built? Well, the cool thing about Arwe is that as described at the beginning, the limit for even a single transaction put in the network is far beyond the amount of data that exists by someone who's going to work this out while doing

it. But it's going to be like many, many orders of magnitude, maybe even more atoms than there are in the universe, that kind of thing. So the answer is on the scalability side, no. From a protocol point, there are no limits. I think on the infrastructure side to build this stuff, we've just about turned the corner on like, hey, actually, anyone that can build JavaScript,

CSS, HTML applications, they can build these applications in a fairly easy way without having to learn a lot of specialized stuff. That really only just happened two or three months ago. So at Arweave, we're trying to sort of YOLO scale this. We realized like now is the time. We're frankly the whole- That's the crypto way, man. Yeah. But I think we're late. I think we needed

decentralized social media to be growing, experiencing the network effect of hypergrowth like a year ago. That's where we needed to be. We weren't too much focused on the tokens, not enough building. Now is the moment. And so what we're doing is we're focusing on, well, how do we get the next thousand founders into the Arweave ecosystem in the next like 18 months to two years? And we figured, well, there's all these developers out there that are basically junior level. They can just about build web applications. They're looking for their first jobs in the world. Well, screw that.

Instead, if you just work your way through this funnel that we've set up, which starts with like a hello world application on our business, followed by like building an MVP, establishing some kind of entity, whether that's a DAO or a company, whatever it is, and then getting your first users and so on, we'll just progressively invest eventually $100,000 in any person that does this in a reasonable way.

And it's open to anyone to start at any time. Like we really want to hyperscale this to all the people in the web 2 space or they were just getting started in their careers and bring them on board to do this. Oh, and if you are interested in that, sorry to plug, our Weave.build, go for it. I mean, I think that that that to

me is one of my favorite and i think the most impactful parts of the crypto space in general is that most serious protocols layer ones whatever have created these foundations right to to fund development and building so that what you just described can happen. Forget college, you're talented. Come get paid to build something that could literally make you a billionaire, because it not only doesn't align like the incentives of the protocol, obviously, to get more users and to scale, but it also aligns well with the incentives of the individual who just wants to support their family or get rich or whatever it is. That's always been the beauty of blockchain. I mean, Bitcoin is the same

way, right? You wouldn't attack the network because you wouldn't be incentivized to do so. But it's just amazing to me on blockchain, it seems like the incentives for the group and the individual seems to always align. Yeah, it's really cool. I mean, on a personal level, I find that this is some of the most rewarding stuff that I get to spend time on. I remember the first, back in 2017, we were just getting started. I had three months of my PhD left to go. And so I started thinking about, you know, what is it I can do next? And I realized there was this opportunity to build an archive inside of blockchain, basically, and to make that scale and make it economically sustainable. And I remember the first guy that gave us 5,000 bucks.

And I remember the emails with him. And it was a profound thing to be given that opportunity to to quit what i was working on and just really go after this thing i was so passionate about and now you know five years later looking back one of the coolest things we get to forward that opportunity on to so many more people that is amazing and um and the cool thing about it is you know if you are in college or whatever like you just stop for a little while, try it out. Okay, it doesn't work. What have you lost?

Yeah. And to be honest, like, college applications, they love when people have, you know. Entrepreneurial ventures. Exactly. You're padding your college application by programming. So what's the worst that can happen? So yeah, I feel there's profound meaning in that. And it's an exciting time, for sure.

So where do you see this in 10 years, right? I mean, because you're clearly an exponential thinker, right? You're speaking in numbers that, and I think that that's a rare talent and probably what separates the people who are extremely successful in blockchain from others is that they just have this ability to view that sort of exponential growth and hockey stick. I mean, you literally just said casually that you're talking about numbers like atoms in the universe, right? For how much data you can store. Most people I don't think can think that way. So if you gain exponential traction this year, I mean, do we have an equivalent internet to what we have now built on our weave 10 years from now? I mean, is that what we're talking about here? In theory, I'm not saying that will happen,

but it's within the realm of non-zero chance. Yeah, so I've been tracking the growth of our weave against the growth of the early web and we are way ahead of schedule. So that's pretty cool. I like the think globally, act locally thing. So I can see we're

on the right track to build something that is enormous. But there's no point spending time thinking about that. I was up till 4 a.m. last trying to help the team uh tune the databases that run uh one of the major rv gameways and like that that is my focus i don't care about the long term will happen time time has a way of elapsing with you like you don't know but what needs to be done like today is always my focus and and for example we just launched the rv build rv.build um program this week and so there's a hyper focus on like how do we get us

the documentation is all in place so that a newbie to the network can come along and learn everything they need to know from the start in an as easy a fashion as possible let's make sure that the the infrastructure is of as high quality as possible so that it works nicely when they use it. This is what my focus is on every day. Are there any places that centralized systems are superior? That's an interesting question. I think that the notions of centralized and decentralized are kind of misnomerless, very beast. The reason that we care about decentralization is not, you know, that's the tool. Like if you could have Bitcoin that had all the same properties, but it wasn't decentralized, then I wouldn't care.

That's great. What I care about with Bitcoin is that it protects the rights of the user of the service, which in this case is, let me transact and make sure that there are only 21 million tokens. In our view, it's, let me store data and make sure that data is around forever. And no one, not even a DAO, not a company, not like a crazy ruler that gets 51% of the hash rate can remove it from the data set later. So sorry, I focus less on decentralization more on preserving the rights of users. I think that if it's okay to rephrase your question, are this services that are better if they don't have guarantees for preserving the rights of the users? The only thing that comes

to mind is a well, maybe there's a case where what if you get it wrong? Is the developer in the first place, like it's hyper thing that comes to mind is a well maybe there's a case where what if you get it wrong as the developer in the first place like it's a hyper early stage like trying to you know you can imagine a startup that's pivoting 10 times a minute and and they're trying to put something out that does something really important like i don't know in the biology space, like gene editing, something like this, maybe, maybe there's like a very small niche there where we're being able to revert on, on those user guarantees is a pro social thing. But I think largely, no, it seems to make sense that if we enter into deals in the world,

which is what we should be doing when we use a web service, when we use something on the internet, you are entering into an agreement that you're going to use Twitter and it's going to make money for them by ad revenue, and you're going to get to use the service. I don't see why that shouldn't be said in stuff. Or the same with your email account and so on. It strikes me that being transparent and clear about the deals

people are getting into upfront is almost universally better. I agree. Yeah. I mean, but I think in some places in the blockchain space, I mean, I've had conversations with quite a few CEOs of layer ones and layer twos, and almost unanimously they say,

well, if we had a billion users right now, we would not be able to scale, right? Our blockchains are slow. They get expensive. They get congested. I mean, the Visa network functions around the entire world. It's centralized, but it's fast, right? So of course you, you sacrifice the decentralization, all the things you talked about, but I guess at this moment in time, centralized, centralized companies often do have some advantages now so i think the yeah okay so one of the things we spent a lot of time in the bear market working on

uh was how do we make it so that are we scales arbitrary at the protocol level like the implementations can need work in the same way that like a web 2 company is they they scale the software itself needs to change. But how do you make sure that you don't end up in an ETH2-like situation? And the cool thing is we've solved that problem. Like I'm not being objective about this. I know the network in and out. And essentially what we've done is by having just one big,

it's going to be technical for just a moment, but one big Merkle tree, so that there's just one big data set and then one contract, which is the endowment to back it. We remove the scalability questions because you're only adjudicating one thing at once. And then you encourage users, because of course there is limited, you could say like block negotiation space. So in Bitcoin, that would be one megabyte.

Yeah, because of that, you encourage users to pool their data together. So instead of bidding higher to store the same number of bytes, you just say, okay, well, you and I both want to use the network. Let's send out data entries, let's sign them, send them to someone or a decentralized network,

even like this thing called, you know, we've never called bundle a dot network without the E obviously. Is it like a roll up or a zero knowledge? It's kind of like a similar equivalent. Like an arbitrarily scalable roll up for. Right. And, and it really is pretty neat because it lets you sign your data and pay for your data with like MetaMask and all the different wallets and tokens from all

these different blockchain ecosystems so you can have like a solana you know plumber web app that kind of thing it's pretty neat but but if you do this you're incentivized to batch transactions together and then submit them all at once and you don't get the roll-up scalability issues either because it's it's just data like there's no contention between... Where all of that stems from the smart contract world? I feel really... I feel sorry for the people that have to work on this frankly

because it's tough, really tough. Well, if you and I have a transaction, it's going to change the same piece of state. One way or another, that transaction has to be ordered. One of our transactions comes before the other. And we can't calculate the other transaction before we calculate

the first one. Like that is complicated. You can't really solve that at scale. And that's why you see smart contracts, like as they scale, scare quotes, I'm not sure if your users

view your videos or this. They do. Okay, cool. Yeah. As they scale, they become more centralized. Well, with Arweave, you just don't have that problem because we're solving an entirely different set of things. So you can fit the whole web

in a single Arweave transaction trivially. It's just about having enough hard drives online. So yeah, that's one of the most exciting things about looking forward to the next 10 years. We're just not going to get into an ETH2 style. Everyone wants to use the service. Now it's broken. It's a super, super cool, very, very exciting.

Absolutely mind-blowing. Yeah, it's mind-blowing and incredibly impressive. And I hope in 10 years, we're having this conversation again and the entire internet is built on Arweave. Why not? Or the next iteration. So where can people follow you? built on Arweave. Why not? Or the next iteration. So where can people follow you? Obviously, arweave.build, if anybody's interested there.

You talked about that. I want people to definitely check that out. Where can people follow you and everything else that's happening with Arweave after this conversation? Yeah, so I post on MetaWeave, which is, of course, the Arweave-based Twitter system. But you can also find me on Twitter at, oh God, the worst Twitter handle in the world. Sam E.C. Williams is the Twitter handle. Could be worse. Yeah, I don't know.

Well, anyway, that's where you can find me on Twitter and my stuff automatically reposts from MetaWeave. So you can see it there and you can also check it out. Very cool. Well, thank you so much. I really appreciate it. And thank you for giving everyone this, I think, very inspired vision of what the future can look like. Because it's scary out there right now.

It's scary out there right now. These are hard questions that people are very worried about with their data and the news. And it's wonderful to know it can. are very worried about with their data and the news. The world is reforming. It's wonderful to know it can... The sad truth of the world is crisis is opportunity. The world is reforming right now. We have the opportunity to push it in a direction

that is pro-social over time, I think. That's what's at stake and you see it unfolding everywhere. I believe you can do it, so thank you. Thanks.