 Consensus Lab is basically dedicated to building consensus protocols for Web 3. So what's Web 3? If we go from the inception of the internet in the 60s to Web 1.0, that was basically posting read-only content by those people who knew how to post the content to Web 2.0, which we have now with different social networks, clouds, and so on. But when we have data basically siloed in mostly centralized cloud provider, there is a wonderful sitcom paper, this year best sitcom paper in 2021 that says that half of the internet traffic is basically governed by five institutions. And Web 3 is going towards decentralization, right? So decentralization, in a sense, and I will define decentralization in more details later during this talk. But this is basically a set of computer systems, a set of computer programs that basically build decentralized internet in a way such that you don't trust in principle any single component. And this is what is emerging today. And consensus lab is essentially looking into building a layer of strong consistency and a bit weaker consistency is also in the scope of consensus lab that would power and that would act as a backbone of Web 3. So this is the best I could do for a one-slide goals and visions. So consensus lab is supposed to be a go-to place for scalable consensus and decentralized systems R&D. So yes, we are part of protocol labs. And yes, one of our goals is to impact Filecoin and IPFS ecosystems. As you know, one of the protocol labs has several projects. But one of the most well-known ones are Filecoin, which is incentivized content addressable storage and IPFS, which is the same but without incentives. And there are several other protocols. And there is a lot of Web 3 projects that is actually using Filecoin, IPFS, but also DRAN, the libp2p, and many other projects that come from protocol labs. So consensus lab is definitely dedicated for impact in Filecoin and IPFS, but also the entire Web 3 ecosystem. But part of it is actually collaborating with academia. And there are many friends, ex-collaborators. And I hope future collaborators here on this call today and among the registrants who were basically the target audience for this event. And we plan in consensus lab to foster collaboration in academia and other Web 3 projects through different grants. Like these would be more classical funding schemes. But what we are also keen looking into and very much interested in are decentralized funding and incentive schemes of which there will be talks in the following months from PL. But also we might scratch the surface of these discussions during the panel. Consensus lab as protocol labs is working in the open. So there are not necessarily, I wouldn't say not exclusive, but not necessarily patents. So we are not interested in patents. We are interested in open source. We are interested in sharing the knowledge and the code among different projects and different people and basically building the decentralized Web 3 together. In the consensus lab, this is not going to grow to hundreds of people. So as you will see, consensus lab does grow fairly fast. So I joined in the end of June, 2021. So end of this year, July, so maybe three months ago. And we are growing pretty fast, but we are going to stop growing at some point. So the members of consensus lab are essentially going to do also what I'm trying to do. And in discussions with many of you, I also mentioned this. This is trying to leverage our connections and our collaborations with the academia to basically build together and also incentivize this collaboration for building together these consensus for decentralized web. So we are looking for the impact beyond Filecoin ecosystem for the entire Web 3 and maybe reinventing the whole collaboration in the open. So this goes in the direction of the thing I mentioned towards, for example, changing even the peer review systems later on and things like that. We'll discuss this. This is just the beginning. So just to give you a flavor of where we are going. More technically, so this was on a higher level, but more technically, if you look at decentralized systems today, consensus is the bottom right. There are two things about consensus. So if you have basically here, I borrowed the next two slides from Juan Stokes where he describes basically the problems that motivate formation of consensus in protocol labs is that there are distributed application potential today is very, very high and the hopes are high. But basically when you channel this through consensus protocols, and there are more efficient protocols than the one that I mentioned here, but classical ones are Bitcoin with seven transactions per second and Ethereum with 15 to 30 transactions per second. Basically, but these total order services, if you proceed them like that, they are most of the time bottlenecks in the system. And depending on your goals, you can put even the best consensus protocol and it's still going to perform basically how a single machine performs, unless you do something else. So that's one. So one thing is about sequencing the transactions and basically establishing total order and addressing these bottlenecks. But the other is also transactional smart contract execution. So we see in smart contract systems such as Ethereum, smart contract execution is sequential from distributed systems basically decades of research. We know how to, we have an idea how to address parallel execution in centralized system, where you don't have the difficulty of not trusting other nodes in the system. But this is also a problem. So if you split this into ordering and execution, there is a problem here in my previous work, which I did inside here, I kind of, inside IBM, inside the IBM teams, we tackle this in the context of permission systems, such as high privilege of fabric, but in principle, in permissionless system, the situation is much more complicated. And we are looking at the same problems here again, right? But what are the requirements? Okay, so like what are the use cases and what are we optimizing for if you want? And there, when I had an idea to join protocol labs, I was like, okay, we are going to apply some really, I was recently working on efficient, high throughput protocols. I know other people, other groups were working on highly efficient protocols. We are going to implement really nice scalable consensus and we are mostly done. Well, except that Juan had this nice idea, which I really liked, which is, okay, let's forget about a single blockchain. Let's talk about web three requirements as of porting web to work close to web three. And then suddenly you talk about billions of transactions per second. You talk in some use cases, you talk about data center latency, so milliseconds or tens of milliseconds finality, depending on what is the application doing, then you might be talking about Earth consensus or like the city or the urban area consensus or so on. But you still want to have like unlike in web two, unlike in centralized, internet and centralized cloud computing, you want to have this security against nation states and hackers. For example, the Bitcoin has and so on, right? So if you have this bunch of goals, then you need to look at the problem a bit differently, right? And this bunch of goals motivated basically how I view what consensus labs should be doing and how we structured work in consensus labs. So how do we get there? It's clear kind of if you're looking at the billion transactions of second, that one of the key problems is not going to be the scale only to scale a consensus, a single consensus protocol. So if you look at each of these lines and these circles as a single blockchain name space or the state space, or you can call it a sharp, then you will be talking about multi-sharded system that of course has its own, we know this from distributed systems for a long time, has its problems in the context of decentralized system security related problems, but also consistency related problems. Like what happens to cross-sharp transactions? And for example, if you are sampling the membership of a parent shard into a child's shard, how do you protect against the attacks of the power dilution and so on? But you might not be sampling, you might not be partitioning the state, but you might be looking into some other things and we will have some very interesting talks just after my talk during the sharding session about this, but we are certainly not looking into a single blockchain system. I think we can agree on this. Then depending on a particular shard, you're not looking into a single consensus protocol either. So ideally we don't want thousands of consensus protocols because currently I mean, I can only mention a few that are production ready, right? So they don't halt, will not name particular blockchains that halted recently, but there are many implementation that just halt for some reason that they're not proven. And in the industry, there's just several projects that are really proven when it comes to consensus implementation. So we are not going to expect thousands of those, but we are going to expect a few of those and they're probably going to differ, maybe not, right? But currently as the current design goes, they differ with respect to number of nodes. So you have different design choices. For example, with less than 100 nodes, you would go with a certain family of protocols. If you have tens of thousands of nodes, you might go more towards randomization and so on, right? I'm not saying we should go that way, by the way, for these techniques, but I'm just illustrating what's the case today. But then again, for individual implementations inside the shards, then you have a separate set of challenges. So what is the performance? What are the trade-offs? What is the security? How do you do civil attack protection and so on? How do you prove things correct? So we are talking about formal verification, model checking, are we talking about distributed testing? How is this done? And this is all something that's like the research is there, but when it comes to production and powering, if you imagine powering systems at the scale of web tool, we don't have production protocols that are at the level of similar systems that power web tool, for example, and classical strongly consistent systems that we use in threshold tolerance, for example. And the third thing is this execution, right? That you mentioned earlier. So of course, if you have whatever we build in the first two pillars, I would say, and if you have then sequential execution, the way it's mostly done today, this is not going to go far, right? So we are going to have to look at parallel execution, maybe even computation of our arbitrary data. One thing to think about there is, if you do federated machine learning, where you don't have the machine learning data style or the inside single data center in control of a single, let's say, social network provider, how do you do this, right? Especially if you couple it with encryption and privacy, which I'm not focusing on. By the way, this is what I should mention. So there, consensus lab is one of the four groups that are looking deeply into these bringing cutting edge research to development areas. So we're looking into basically what I'm discussing now, but we have three other groups that are looking into more topics like zero knowledge computing or just applications of different cryptographic primitives to blockchain. And one group looks at crypto economics or more from a game theory perspective. Of course, we work together. Like I cannot look at these problems that I'm showing here without considering those, but I'm focusing on these now, right? So just do that. You don't think that I forgot other things to mention here. There is also the need to support, I don't know, web assembly, virtual machines, maybe Ethereum virtual machines, maybe some other virtual machines, right? So if we write a so-called smart contract instead of simple cryptocurrency transfers, the situation gets complicated, right? And we got that far basically as things stand today, but there is a lot of work to be done. And essentially consensus lab is focusing on these three things. So we are dividing our work. We are building single systems and we are interested in basically building these systems together with, again, with the other partners, as I mentioned before, but focusing on these things, right? And these are the three vertical pillars of consensus lab. But then you see basically in red, some challenges define horizontally relevant topics that basically are relevant to all three pillars. These include security, which is pervasive. Also, how do we implement things correctly? How do we prove them correctly? This is also pervasive if you're interested for consensus lab and so on. So essentially this also covers many of the sessions that are consensus day sessions today and tomorrow. And I'm very happy that basically we got such a response from the community. So all the papers that we got were extremely relevant. I don't know if that was one or two exceptions that was not relevant to what we as a consensus lab are looking at. So that wasn't really great. Consensus lab takes this perspective of connecting the dots approach. So again, because it's about web bringing web to work across to web three, the idea is not to look at these problems from the perspective of a single blockchain project. So it's not about here scaling file coin consensus, but it's really like, okay, can we think about the whole space, the whole blockchain, whole decentralized system space and build an engineer system such that for example, you reuse things that are built by separate, by different teams. And my impression of the current situation when I was moving from permission world to permissionless world is that web three projects, largely not exclusively, right? Protocol labs is working a lot on collaboration across different teams, but largely they are, they appear as islands of research and development among maybe competing teams, whose blockchain has more transactions per second and so on. So I think we should need to step, take a step back and basically look at, like imagine that you're a systems engineer, each of us is a system engineer that's supposed to build decentralized internet himself. Like how would you think about the problem? And of course then later on, you wouldn't build a decentralized internet yourself, but we will collaborate. That's by the way the challenge that we're going to talk about, how do we do it? But this is the thing, right? So you're not building a single blockchain here, like how would you change the world? How would you put all things together? How would you connect the dots? That's the approach that we would like to emphasize and that we'd like to work towards. So in this context, I would like to also take a step back to look what we have done so far and to have a look, especially from the consensus perspective, like what's there today and what do we have and what can we use in a which way? So basically this needs to start from proof of work, right? So we cannot avoid elephant in the room. And as we were kind of trying to improve certain things that you see in red here, that are normally seen as drawbacks, then we were going more towards Byzantine fault tolerance and proof of stake protocols. So Byzantine fault tolerance was something that exists for 40 years. So I did my PhD actually, I started my PhD 18 years ago on Byzantine fault tolerance consensus. I defended my thesis one month before launching the Bitcoin, so this was there. I mean, these protocols were there, they were not as advanced as today, but they were there, but they were for some reason, they were not deployed. And as we will see, there are certain reasons why this is the case and why this is still the case. But proof of work, I mean, it has very high energy consumption. This is undebatable, right? So currently it's larger than Poland. It's 0.1% of world's energy production. And the consensus itself, if you look it from the computer science perspective, I mean the guarantees are not much, right? So the throughput is not much, the latency is not much. I mean, it's much, it's a lot, but it's not very good. And there is no finality, right? So depending on your perspective, you can, certain things are actually good, but we need to see like in this progress of replacing proof of work, like how far did we get? And we got some new applications, right? So we got smart contracts, we got decentralized finance, but what we also got is probably a huge confusion on the market. So I did a snapshot of CoinMarketCap today and this lists 12,000 cryptocurrencies. So the question is, are these all decentralized? Are these all useful? Can we, what do we do with this? Like what are we building here? And here in a recent paper that I had, so I'm really, really thankful during this call, I would like to thank Stefan Schmidt and Seth Gilbert for inviting this position paper that I wrote for bulleting of the European Association for Theoretical Computer Science Disability Computing Column. It appears this month. And I have basically looked at these things that also I needed to look at to position, basically my view of how consensus labs should continue research. So this includes like, what is Bitcoin actually doing? And is it sustainable in what it is doing? How do we define the centralizations and are we trading in some key properties in this quest for green, very efficient performance? Are we doing something that we shouldn't be doing? So the rest like 15 minutes are about these topics, right? So in that position paper, I went briefly to see how, because there was no, I had an impression that there was no consensus in how do we define decentralization. Then I found a nice text paper by Carmela Trogozzo and others, there was George and Danez Hary-Halfin and I'm skipping one other author, I'm sorry. So, which defines very broadly, decentralized systems as a subset of distributed systems and I like this definition, I think it's very good. Where multiple authorities control different components, but no one is fully trusted by all. So I think this is a basic definition. I think if a system, if a distributed system doesn't satisfy this definition, we shouldn't call it decentralized. This is what I firmly believe. But then again, it doesn't distinguish like global scale permissionless systems such as Bitcoin from a four node, you know, four organizations which run a BFT protocol. So we need other facets of decentralization. And then in the literature, there are several of them, right? So there is resilience, which we know from photo and computing. This is our typical F or T. In blockchain world, it's typically called Nakamoto coefficient. It's basically the same thing. Like how many authorities the adversary needs to compromise to subvert safety of liars or liars. This is our again, F or T in the classical system. Other decentralization facets include like governments, you know, is somebody who controls the updates to the software, how easy is to submit proposals and so on. So they should be like there are certain parameters that I won't have time to go in. But in my paper, there are some pointers, including to certain other papers that go in even more depth. Whether protocol skills in protocol incentives, like is it incentivizing participation from inside the protocol, by giving, for example, block rewards or transaction fees and so on. And there are certain aspects that I'm not focusing on, but which is very important. So we saw this with the outage of a certain famous social network yesterday. So I think we can network level is always important, right? So we cannot forget that. But since we are just to focus on more distributed systems aspect, then we are focusing on these facets that are in black, right? So one very known one is openness. Like whether the system is permissionless or permission. And here we have a classical definition that permissionless systems are roughly speaking in English, those in which anyone can participate. So participants essentially self-elect themselves into the system. And in permission systems, there is an authority, you know, centralized authority or maybe decentralized, maybe existing members, let new members come into the system. This is also possible, which elects participants. Normally we would say that permissionless systems are more decentralized than permission, right? But is this classification expressive enough? So for example, our old permissionless systems equally decentralized. And then basically what I proposed in that paper is a refinement of notion of permissionlessness. Permissionlessness, if I may put it like that. So this is called equal opportunity of inclusiveness. So I was lazy enough to copy, paste the definition here, but there are essentially two properties that a system which satisfies inclusiveness or provides equal opportunities needs to satisfy. Needs to satisfy resource symmetry and genuine openness. So resource symmetry says just if me and George bring same resources to the system, we'll actually have the same power roughly speaking, right? So our maybe expected rewards will be the same. So the system will not distinguish ourselves, the two of us based on our identities, right? And it's easy to see that, you know, this applies to permission systems in a sense that no permission system will be resource symmetric because it's permission, right? So if I'm out of the system, then regardless of how much resources I have, I'm not equal participant as one who is in the permission system. But it doesn't really apply to tell us the difference between for example, proof of stake and proof of work. To both proof of stake and proof of work, you would get resource symmetry. What you don't get, as I will briefly argue later on, is this genuine openness. In certain permissionless systems. So basically the property says that system cannot reach a state in which it is not open, which is not permissionless. So specifically a new participant investment that would match an investment of existing participant must never depend on the action of the existing participants. And I think this is the property that captures the very difference in decentralization between proof of stake and proof of work, which is important basically depending on the use case. So we always need to watch the use case, for some use cases it's not important, for some use cases it might be important. But basically proof of work systems are inclusive and this is what the paper continues to argue about. So basically if me and George bring the same mining power into the system, we're going to have the same power and same expected rewards, right? And it's genuinely open. Like even if someone, imagine someone controlling the entire power in the Bitcoin mining, as computing power essentially grows outside the system, as we as humanity invent, just have breakthroughs, like this beat moves low, this, just the computing power, historically I cannot prove that it will continue like that, but I think it will be difficult to find the computer scientists that don't agree that our computing power will actually continue to grow. Be it like parallel computing with quantum computing you name, right? And that grows and that computing power is free to join the system. And there is nothing the existing participants can do to prevent new participants from joining the system. We can assume free like, for semiconductors, if we need to build and wait from China semiconductors, there is nothing that theoretically prevents you to build your own semiconductor facilities and so on, right? For the same investment. So just the rules of the gamer are kind of different from proof of state. In proof of state, there is resource symmetry but there are genuine openness not so much. And I think this captures this main difference. So if existing participants have 50% of the state assuming non-inflationary state and they are unwilling to sell, new participants can just not ever reach the existing participants power. And in my opinion, this captures the difference, right? So it captures the difference that in the end proof of state systems are kind of more similar to permission systems in a sense that they can reach that state. Arguably some are not yet in that state but if you're looking at a very long-lead systems like we are facing with decentralized systems and options, this is a pretty fundamental difference. So okay, I mean, again, we should evaluate the value of inclusiveness depending on the use case. And I would also argue that if Bitcoin use case is a payment system, then it's not worth the energy that it's spending, right? So yeah, it's nice to have an inclusive payment system. And maybe we can scale Bitcoin on layer two and other layers, but maybe inclusiveness is really not worth this energy that we are spending on Bitcoin. But what I'm proposing is to consider a different use case for Bitcoin because it's actually not the payment system. It was marketed as a, so it was marketed, it was basically in a published, you know, white Satoshi's white paper, it was called peer-to-peer cash. And what happens if you think of Bitcoin as a decentralized money for humankind? So what happens if you imagine the world in which this becomes actually dominant money? So this is the thought experiment that I was talking to myself for few months ago, like few months ago, and basically trying to imagine the world in which Bitcoin is the money. You just try to, you know, make this thought experiment. And actually the results for me at least profound that I decided to share it with the community in this paper, right? So basically it needs to do with the current fiat monetary system. And I'm not going to talk about preferred players. So we know that, you know, banks due to continue on effect and so on are close to the source of money and these things. I think this is very well known. I want to talk about different incentives under the fiat monetary system and under the Bitcoin monetary system that actually are going to justify the energy expenditure of Bitcoin. So in Keynesian economics, these are basically I have, this is the economics that basically the economic theory and one of the most famous economics that drives the monetary system that we have today and the economy research that we have today. And basically there are a few interesting quotes in the long run, we are all dead, which kind of suggests that nobody here is focusing on the long run. It also talks about silent inflation. This is the cetacean in the lower left corner and so on. So economic growth is the key thing. So your business in order to be recognized, it needs to grow. You're measuring this growth in the fiat currency and you need to spend. So these are the characteristics that shouldn't be too strange for all of us in this current monetary system. So I try to summarize what happens in like one slide in the current monetary system and definitely we have an inflationary system. So governments and central banks, they create money and commercial banks, they also create money because they don't have fractional reserve bank. Let's not go into reasons like is this justified or not? It might be justified. For example, we had COVID crisis, many businesses fail, governments maybe needed to print money, let's not go into good guys and bad guys, right? So it's just what happens in the system, right? So if you have an inflationary system and you have somebody who can print money, it turns out historically that they actually print money for one reason or another. What happens to incentives of people and businesses in such a system is because you have silent tax by inflation, people are incentivized to do two things with the savings. So either to spend or invest. And if you invest into businesses, they basically need to grow again, measuring this fiat, which means that they get, need to get more revenues and more profits which basically promotes consumption and spending. So this is very like almost like first order logic and it goes with Keynesian economy well. So if you read basically the goal of Keynesian economy is to promote consumption as well. And what we have is the economy is what we have today, right? And this is orientation to short-term profits and no long-term outlook. So basically, then this is my claim. I don't have a proof for this, right? But my claim is if eight billion people live in this economy, then basically this is what caused over pollution and basically damaging the planet. And I would like just to ask you to look at the different thing. So let's look at the Bitcoin monetary system and its incentives. So in the Genesis book, there is an interesting motivation. I think all of you know this, but let's repeat it briefly. So in basically when country on effect showed last time, this was in 2008, basically there was a failure of economic system and certain players will bailed out. These were the banks, right? There is the cover page of the times that Satoshi put into the Genesis block. And this is basically because governments were first bailing out the banks. This is the effect of country on effects for people organizations who are close to money. They benefit from the money first and money is not evenly distributed, right? To all the people. Bitcoin monetary policy, we have something else. And this is, this didn't change in the inception of Bitcoin. So we have 21 million Bitcoin, which are divided in 100 million smaller units. And we have halving basically 50% of all Bitcoin were mined in the first four years, then another 25% until 2016, then another 12.5% and so on. And we are switching to Satoshi. So basically block reward would be less than Bitcoin in 2032. And last Satoshi will be mine in 2140. By the way, it's really interesting how someone comes up with these incentives until 2140, but that's a separate discussion that we might, if we were at the conference, I would be happy to discuss during the coffee break. But very, very interesting. Supply over time, we are here, right? So we are like roughly 90% of the printed Bitcoins. So there will be no more. So what are the incentives? Like we saw that incentives in the current system are to spend. So what are the incentives in this system? So it may be seen as a volatile investment. If you look at comparison to Fiat money, yes, it's volatile. But if you just take the opposite look, of course, then Fiat is volatile to Bitcoin. There's just a matter of perspective. We know this from Einstein, right? So just what's your system of reference in these things? So, but what does it do on long-term? And you can see this when you zoom out from the price of Bitcoin compared to Fiat currencies. When you zoom out, you see that encourages long-term savings, but this argument is not necessary. So it's just because it's a closed system. So in Bitcoin, you trade, you basically just transfer Bitcoins among the people who accept the game. And it's a closed system without any relation to the outside world, in a sense. Then as the network adoption grows, it's more and more people adopt Bitcoin and you expect this to happen because people recognize the value, then the value of these Bitcoin growth. So it encourages basically saving and holding. This is known as hodling. So some people don't, even if they don't think like that, usually, because anyone of us can trade and basically the hodling term came up as this is actually a hilarious post on Bitcoin talk forum where kind of a bad trader calls himself a bad trader and immense the term hodling. Very interesting. So if you hodl and save and your savings are not devalued over time, then as a human, you incentivize to spend on things you need as opposed to things you think you want in the fiat system. And because the savings are preserved in time, there we get personal freedoms and time back. So essentially, you're not scared that if you work and you saved in Bitcoin over time, you're not scared that somebody can devalue this money. And basically you're allowed to do with this, whatever you want. You're not forced, for example, to invest in other economical player or in other businesses. For businesses growth is practically meaningless. Like, can you grow your revenues infinitely in car debt money? No, you cannot. So what you need to focus, especially since these customers now go basically for them, this money is scarce. And if this is money for the whole humanity, they might not give it that easily as they do it now. So what businesses need to focus on is providing customers true value. And the success, arguably, is not going to be measured or businesses in future is not going to be measured in revenue or profits because that doesn't make sense actually. It's going to be measured in the network effect much like it is today for networks like Bitcoin, Ethereum, I would say Filecoin and others. So those that actually change and impact people lives they're more valued than others. So society changes completely. So if a billion people are more incentivized towards savings as opposed to spending, you have basically the whole behavior of the whole species differ. And basically decided to bother you with this argument because I didn't hear it repeating much. And for me, this was the turning moment where I felt like like in matrix basically waking up when I understood this. After I need to say, I was not very smart. I looked at Bitcoin for 11 years basically since it was invented. But this was the turning point for me. So equal opportunities of Bitcoin does not mean economic equality. We should be aware of that, right? So Bitcoin provides and proof of work provides economic equal opportunities but not necessarily economic equality. However, these economic inequalities are much easier to address in Bitcoin system than in current monetary system. So we have remittances as an example. So you don't need to go through intermediaries to transfer money from wealthy and members of the family to poor ones. And this was one of the motivated use cases for El Salvador adoption of Bitcoin as little tender just a month ago. If you have enough Bitcoin like more than others you can easily figure this out. And I challenge you to do this in fiat. And if you think it's easy, if you have billion dollars there are less fortunate countries in which billion of the local currency is not much. I was, I had a unlock which I don't want to talk about much of holding actually a bill of 500 billion local dinners in my head. So that depends on like how lucky you are where you are. But if you're in the non-inflationary monetary system it's easy to see if you have more than others. If you have more than others maybe you can give and donate and help other people. And also you can do Bitcoin proof of work mining anywhere on earth. Economies of scale apply. It's not easy to mine on laptop anymore. It was by the way, it's not anymore but now, you know, if El Salvador mines on a volcano like this car started on geothermal energy or if in Africa very basically I have a list here of renewable resources they're actually spread across the planet very equally unlike some other resources that we are relying on today such as oil. So I would stop here and invite you to read the paper and just to say that we need to think about things in a context and Bitcoin is not a payment system. So Bitcoin is just the money if it succeeds it's money for the human kind of the future or otherwise it goes to zero that there are just two numbers for Bitcoin. And the world in which Bitcoin becomes dominant money is actually very nice work. How do we go about economy in that world is might be a kind of a challenge because we don't know if we are afraid of deflation of the economy we might want to give governments the power to change monetary policy and so on. The thing is, there are economic theories about it. So in 1936 there was a paper by Henry Seymons from University of Chicago that was called rules versus authorities in monetary policy. And it was that's an amazing title, right? So it's definitely what Bitcoin does. And I find this reading Friedrich Hayek. So Friedrich Hayek is a Austrian school economist Nobel Memorial Prize like the fake Nobel Prize in Economy winner from 1974 for his works and the Constitution of Liberty from 16th actually talks a lot about welfare state in the deflation of the economy. So I recommend this reading. And basically he claims that arguments advanced by Seymons in 1936 are so strict. This is a quotation that I'm using in my paper are so strong the arguments. The issue is now largely on how far it is practically possible to tie down monetary authority by appropriate rules. But noticed like nobody as I get nobody's bad nobody's good. Let's take this position. It was not possible before. If we have seashells as money somebody would bring more seashells. If we have stones somebody wouldn't bring more stone. If we have gold somebody would bring more gold. If we have fiat money as paper money somebody would bring more paper money. If you didn't have technological solution to prevent somebody bringing more money until Bitcoin. So basically now you have. So there is a technological solution in which you cannot possibly attack and you can double spend if now comparing to the power of nation states, right? So you need somebody who controls more energy than Poland to double spend the point. But as we know, you cannot change the supply even if you do that. So even against the most powerful attackers you cannot change the, in the presence of most powerful attackers you cannot change the monetary policy. And I think as a humanity we should use this because of the technology we have a solution to the problem, which was discussed in thirties. We should think about this in that perspective from my view. So what I propose next is basically to embrace Bitcoin as the money of our future because it's good for humankind and planet Earth. And I try to argue there are more arguments in the paper. The question going back to decentralized computing is what do we do now? So I'm definitely supporting research and reducing energy expenditure, but without trade-offs. So okay, if we have proof of stake that might be useful, but not for layer one. Might be useful for something else. Let's use it where it's useful. For layer one, if we want to improve Bitcoin let's not focus on trade-offs. Let's not sacrifice security in decentralization. Let's find something that's not weaker if not stronger than Bitcoin or prove some impossibility results here. For other things, until we do that let's leverage Bitcoin slow but very secure proof of work on census to secure the rest of decentralized internet. Because if there is no nation state that can attack effectively, maybe except few but not maybe not for a long time, like in future. Let's leverage that security to help with security of other parts of this big decentralized system that we're building. And in that context, I'm inviting you, I'm almost out of time but I will invite you to essentially continuation of the talk in some sense, where Sarah will zoom it tomorrow in the checkpointing session. We'll talk about securing membership and stake checkpoints of business in full torrent and proof of stake blockchains by anchoring them into Bitcoin. So basically what Sarah will describe is the idea that we have and that we are building and I'm inviting anyone interested to join in the context of our open collaboration, of course, basically anchoring membership and hash of the, like checkpoints of the state of a proof of stake network in the top route enabled top route is important when it kicks in in November in a top route enabled Bitcoin. And we, for example, if we need to store the actual state somewhere proof of stake network could build its own content available storage or use file coin and IPFS. And by this you can easily prevent easily as the quotation marks prevent long range of tabs that are also some one other thing that's basically plaguing proof of stake. So this is tomorrow, 5 p.m. UTC presented by Sarah. I really invite you to join this talk. So basically now we're sharding in our back to consensus law goals in our sharding scheme. Maybe we do checkpoints of, you know, the stop level consensus into Bitcoin but we continue scaling, like leveraging security for Bitcoin with continuous scaling. All the other again, coming back and finishing all the other talks and sessions that we have in consensus days are very, very relevant, right? So essentially we have today sharding then we have scaling and performance we have stronger security and the final session will be on peer to peer communication and efficient mem pools. Tomorrow we have the session on checkpoint that I already discussed and basically on eventual consistency and so on. So I'm again thanking you so much for these contributions. I think this is really important. I think these are really interesting as individual contributions and I would invite you whenever you hear the talk to think how we can use it in the context of a global decentralized system, right? And there will be no mark or no protocol labs no, no, anyone to tell but like we need to work and collaborate as a community towards this but we need to have high level goals in mind like really important goals and basically these goals. So let's build a future that decentralized collaborative and good for humanity. So with this advertisement of our group and basically the announcement that we are also growing and hiring I can take some questions. Thank you so much for your attention. I hope this was interesting. If controversial, that was also the goal. So in the definition of genuine openness for proof of stake, the antecedent was that that the person who is holding the non-inflationary token they might not sell it to the other person, the newcomer but by not selling it isn't that person depreciating the value of the token and harming itself? Like should there be some crypto economics flavor in the definition of genuine openness? That's my question. Maybe as you will see the definitions in English. I think not. I think there is a fundamental thing still. I am encouraging also in the paper formalization of what's going on here. The main difference that I want to highlight if I'm holding 50% of the stake and I'm not willing to sell it that's whatever is my motivation even if this take depreciates that's different from me holding 50% of the mining power and not be able to do anything to prevent you from bringing the same amount of mining power in proof of work. If you see the difference. Yeah, that difference is clear but the consequences are not accounted. If you are not like if you don't allow someone to take part in the staking process by giving them the stake in the market then that stake will devalue and there is no incentive from Alice's perspective to do that action itself, right? It's open. I wouldn't answer this it requires deeper economic analysis. For me it was sufficient to see that there was a difference. Fundamental one. So yes, I agree that proof of work is definitely more inclusive than proof of stake but what I wonder about is about the symmetrical because we can already see that there's a tendency of there's energy prices are very different in certain regions in the world. So that to me in a certain place I me investing the same money into it won't have the same result as someone else where the energy prices are significantly lower and obviously someone already holding a lot of hardware when the GPU prices were not as insanely high as currently also had to have invest significantly less than I might have to invest now as the price has increased. Absolutely right. I did have a disclaimer that I sneaked in which I was assuming for proof of work a fair market for energy and computing power to just to make the argument. It's a bigger thing. So of course you can also say, yeah but it takes me time to produce the semiconductors and so on that they're different. They're different ingredients to the game. I completely agree. Which still doesn't so at the fundamental level if you imagine like the ideal world like if you see me like this if you're building a game in which you see me like this right. So basically you would have clear difference between proof of stake and proof of work as games. And of course in our world there are different economic parameters that play into and that essentially impact the calculation a bit whether what we want to calculate but in my opinion they don't impact the fundamental difference between the two games. Yeah, fundamentally there's a big difference between the two. Yes. Yeah. I had a second small question which was to a later point in the terms of the money spending in compared fiat to cryptocurrency where from my experience in different crypto communities I've seen like people being more likely to spend cryptocurrency on all kinds of things being at pyramid schemes, NFTs, games, gambling completely random stuff compared to fiat. That's the thing. So I'm not saying everyone is incentivized to save. It's just that if each of one of us is essentially is a bit so I mean these people in order to make that argument you need to say that these people who are gamblers in the cryptocurrency world wouldn't be gamblers in the fiat world. And I don't think this is the case. I think that like if you look at individuals if they actually save a bit more resources in one world compared to other they don't need to do it relatively to each other but just compared to themselves to their copy in another world. And you multiply that by eight billion you're going towards this conclusion.