 Thank you very much for the opportunity and Greg for moderating, and like I said earlier to Greg, the restrictions should apply not only to questions asked, because there's wide range of possible questions that we can entertain, but also to stop me from giving too long answers if it occurs. So I also want to kind of start with a background about how this talk came about, because this is not a usual, this is not my usual work, this is not a model that I, you know, that I sold for equilibrium. Andre and Julian asked me at some point which are blockchains, blockchain businesses replace or disrupt, disrupt platform businesses, platform merchants. And I said that's an interesting question, I actually was thinking a lot about the relations between blockchains and platforms and how they could compete and how blockchains could affect competition in block in platforms market. So let me say, you know, this talk is more about setting up a question and hypothesis and frameworks or how to think about those technologies, because blockchains and blockchains are this new technology that promises decentralization, this intermediation and democratization. And we already had a technology like this, this was internet, and that's led to biggest intermediaries that we have ever seen. So will blockchain really deliver on the promise that the internet only partially delivered on? So this is how I kind of set it up as blockchains versus platforms. And this goes back to the question that Andre and Julian asked, will blockchains this intermediate platforms? I'm actually, you know, flattening the question here. But the background, I'm not going to spend more than half a slide about what the platforms are in general for this crowd. But just to kind of pin the attention is platforms that the defining feature almost on the platforms is that they attract independent users, independent parties, and they connect the participants. Online platforms very often connect them algorithmically. And there are strong network effects, network effects create barriers to entry. But if the platform can get their network effects on the site and overcome the network effects problem, coordination problem, then a successful platform can actually enjoy network effects and harness the benefits of it, either through higher prices or through other ways of extracting value. And this ability to extract value attracted attention of proponents of blockchains who say, oh, we have this decentralized structure, would it be possible, or maybe the proponents of blockchains assert that it would be possible to provide similar services as platforms do. But on blockchain, there will be no entity that would capture this additional value. And therefore the value would be fully distributed to all the users. And this will be a better deal for the users for the economy. So I'm exploring here the question of whether it would really be the case and maybe what dimensions we need to look at. And of course, to start with, I'm assuming that I'm not going to, I will not need to explain a lot about platforms, but we should be a little bit clearer about what blockchains are and what blockchains do. I say blockchains in plural because there are many different designs of blockchains, different blockchains may have different characteristics. But I am going to get into what is the common characteristics of many, many blockchains. But typically when we start, when we talk about blockchains, we start with Bitcoin. And something to be said is Bitcoin was created as this decentralized way of managing digital currencies. So this is a very specific problem and Bitcoin is a solution to a very specific problem. So whenever we have digital goods, digital goods, our digital files are really easy to copy perfectly. And as much as it's great when we want to copy files and send it among the cultures and between our devices, when we are seamlessly copying and costlessly copying music and movies and digital goods that have been created by somebody else, it's called piracy, it's a problem. But if we would be able to seamlessly copy money, then it would be a disaster for this money system. So this is called a double spending problem because you can spend the same digital dollar over and over again because it's just a digital file. This double spending problem typically for digital currencies is solved by having a centralized party that keeps the ledger and basically keeps track of whether somebody already spent this $1 or not. And this is what we have with credit cards, with banks. So digital money has been around for a long time. What Bitcoin brings is the distributed or decentralized digital money where we get away, we can create a system without any central trusted third party to manage the ledger. So by design, by the virtue of being a solution to this particular problem, it is distributed. It means that it's composed of only peer-to-peer notes, none of those notes is more important than others. But it's even more than this. To this extent, there is no, nobody is more important than anybody else, that there is no, not only no trusted third party, but there's also no party that would tell who can belong to the system and who cannot. And this is what makes it a permission less. So this is like the ideal democratic situation. The way that this problem was solved in Bitcoin is that all the transactions in Bitcoin currency are collected in blogs and they are linked by hash pointers and this is what gives it a name, a blockchain, because the blocks are linked in a chain. The validators, anyone can become a validator, validators compete to be able to add this block of transactions to the blockchain and whatever is in the blockchain is kind of commonly agreed on. I'm going to comment on that. But what is even more important in particular a setup of Bitcoin, all the transactions are public, so anyone can verify validity of the transaction and anyone can make sure that there is no double spending, because the same Bitcoin has not been spent twice. And while they are public, they are also pseudonymous. We don't really see the identities, we just see the wallet identifiers. And anyone can have as many wallets as they want. And maintaining integrity of the ledger relies on costly mining. So the way that misbehavior is kept at bay, that validators, even though nobody is policing them, they need to really do a lot of costly work. They are rewarded by new Bitcoins and this makes it worthwhile, but this costly mining is one of the main features that assures credibility of the ledger. Now I'm not going to get into the main details of how it works. For that, there is a lot of works and I'm happy to recommend. But something that is related to that there are a lot of economic forces and a lot of game theory involved in the Bitcoin system. But very often Bitcoin is said to be immutable and this is not the case. So you can rewrite the ledger. It is possible. It's just rewriting this ledger on purpose in case of Bitcoin is very costly. And it is costly exactly because mining is costly. Mining is costly exactly because mining is costly in Bitcoin because Bitcoin price is high. And miners are willing to bear this cost in order to get those Bitcoins. And this is why if we have currencies with exactly the same design, but with lower value of currency, they actually have experienced this rewriting or double spending or changes in the blockchain. So here I am going in this talk. I'm going to assume that we have a well-functioning blockchain. And there's a lot of work about what makes a blockchain well-functioning or not under different conditions. I'm not going to focus on that. What I do want to kind of mention is what do we mean by the ledger? So we have a well-functioning blockchain. It's for our practical purposes immutable. And we keep talking about the ledger. And if you read anything about the blockchain is the ledger. The ledger indicates like there would be just one ledger and there would be some reference point that anyone can refer to. But in reality, there really isn't one ledger. There are those independent peer-to-peer nodes. Nobody is more important than others. And each of them keeps their own ledger. And they update this ledger based on the information, the messages that they are getting locally from the nodes that they are connected to. There isn't like all public broadcasting. I'm not sure whether I got the same message that anybody else. And the ingenuity of the consensus system is such that we are going in a well-working consensus system, we are all going to end up with the same copy of the ledger. And this is important, I think, for what the blockchains are offering in terms of interactions and opportunities as platforms. So the ledger is really a short-hand for consensus. It is many, many ledgers. But if everything works well, we have many, many copies of the same ledger. So Bitcoin's blockchain is just one possible design of blockchain. Bitcoin got to the media and there was a lot of excitement and it seemed to have worked. And even though despite initial skepticism, there was no double spending, even though anyone can become a miner and anyone can participate in the system, there were a lot of voices about using the Bitcoin system for moving any other value or for creating other systems that are coordinating and creating consensus in a similar way between independent parties without any trusted third party. Now quickly, it became clear that the Bitcoin, the particular Bitcoin blockchain is not a good solution for the industry and enterprise solutions, for enterprise use. So first of all, public transactions would not be really good for competitive behavior because you would need to reveal all your transactions to your competitors if we would put everything on the same shirt ledger. So anonymity does not really work for a small number of players because you can easily identify them by transactions. There's high cost of mining, which is crucial for Bitcoin to be reliable. And nobody really wants to bear this high cost of mining and it has environmental impact as well. And what is important is that Bitcoin's particular blockchain offers only probabilistic settlement. So at any point, you may learn that you actually don't own the Bitcoins that you thought you have. They just disappear from your wallet because there may also be accidental, so even though it's protected against more or less malicious, but there may be accidental reordering of the blockchain. And it comes from the peer-to-peer nature. So instead, enterprise blockchains are mostly permissioned and there is a number of well-working initiatives there. They are designed based on distributed systems that have been around for a very, very long time. Any internet company is relying on distributed databases. But the new thing and very important thing is that the old-fashioned distributed databases were only within a company. There would be many, many nodes that would need to coordinate and reach consensus within one company. What blockchain consensus mechanisms that were developed since, what they allow us to do is to have this cooperation and bringing together independent companies. So we see the enterprise blockchain where we have a shipping industry like trade lands. This cooperation between MERSC and IBM, where a lot of independent shipping companies are getting on the same distributed system, which is the blockchain system. And it also is possible to have a system where we share this information, but we don't necessarily see the information. So what we have is I can see a footprint of the information. I don't see the actual data, but if you later tell me that this data was something else, I'm going to be able to verify. So we can have a verifiable or expose verifiable information even without seeing the information. So it solves the problem of the transparency, which was the problem for using Bitcoin's blockchain for enterprise. But for permission blockchains, you need to have some trusted third party, at least to give permissions, but it is possible that this trusted third party actually has no power over any changes in the transaction. So whether permission or permission less and many, many different designs of blockchain, what I'm going to say that the defining feature of blockchain is, and I'm saying that because we don't really have well agreed upon definition of blockchain. So what I'm going to say and defining feature of blockchain is that we have a ledger of information that is reliably shared between independent self-interested parties and reliably shared, meaning that it's verifiable. This information is verifiable even if it's sometimes exposed. And that means that parties that participate in the ledger, they have a symmetric information about what the state of the world. And so it can be verifiable by cryptographic proofs. Whether it's immutable or not, I'm not going to get into it, but even when something changes, we all see this change happening at the same time. And another thing that is important is that the consensus mechanism itself is actually a way of algorithmifying governance. So the government's decision about which of the transactions is valid in their conflicting transactions is decided algorithmically. And therefore it can be much quicker and without involving any management because there is no management in this environment. So what is important is that once we have the shared ledger and we are sharing data, this data may be a code or program that is running with the data. So we can have processes that are shared between different parties. And I don't see your data, but we can have the same program running on your data and my data. And even though we do not see each other's data, the program is going to give us the same outcome automatically. So that would allow us to also algorithmify processes and use of data between independent entities without sharing it, but without anyone having a master copy necessarily or having a power to give us or not give us access to this data. So it's used for many, many applications. It could be in the shipping industry, the automatic replacement of shipment if the conditions of the shipping were not according to the contract or tokens are example of the smart contract. So this is the executable code that is shared between the ledgers or is on the ledger and therefore it's the same for everyone, okay? So on top of the smart contracts, we can actually build applications, the D-Apps. And D-Apps are going to be, to consist of smart, sometimes very often, several smart contracts that are linked together. Some other software on top of that. So if we have a game, then there is some other software on top of smart contracts. And then there's user interface. The very first D-App was a game on Ethereum called CryptoKitties. CryptoKitties were fetching hundreds of thousands of dollars. So it was a, a nice example of how to start this phenomenon. So D-Apps stands for decentralized applications. But something to be noted here is that those applications, even though they are running on blockchains and even if they would be running on permissionless blockchains, they do not need to be decentralized themselves. So smart contracts are going to be need to be decentralized themselves. So smart contracts, even running on permissionless blockchain, may restrict who can use a smart contract. And then if you add to that, that creator of the smart contract, creator of the app, there's another software layer that there's even more decision to be made and a managerial decision, so to say. So the D-Apps may be very much centralized in the sense that somebody is making managerial decisions about pricing, about access, about any other strategic decision as in regular businesses. Okay, so with that, let me spend the second half of the talk on a relation between blockchains and platforms in the sense. So Greg, are there any clarifying questions at this point? Any questions so far? And you're pretty much right on halfway through the time. Yeah, okay, thank you. So the relation is multi-dimension. So first of all, most blockchains are really platforms. And I'm going to elaborate on that. On the top of that, blockchains can host platforms. So those D-Apps are maybe platforms themselves. And the third way in which blockchains and platforms relate is that the traditional already existing platforms can use some blockchain tools like smart contracts and tokens, and there are many examples of this. Here I'm going to focus on the first two bullet points and I'm not going to discuss how traditional platforms are using the tokens or smart contracts already existing platforms. So most blockchains are platforms because just like platforms, they bring together multiple individually optimizing participants. The participants join voluntarily. So this is the difference between blockchain and just any distributed database that we had within companies. And typically the more participants you have, the higher value is joining the blockchain. So if you think about Bitcoin, the more miners there are, the safer is the Bitcoin's blockchain. And the miners are attracted by the price of Bitcoin, but really the price of Bitcoin is coming from having higher security and more trades. So we have this network effects, the two-sided network effects feedback loop running in a very straightforward way. But even in other examples of blockchains, if we, for example, take consortium of banks or supply chain tracking system that is like our shipping blockchain, then the more of your counterparties are already on the blockchain, the higher is the benefit because the smoother is tracking and there are no holes in the information. So you want to join the same blockchain that your other suppliers, that your suppliers have joined so that you can have this information moving seamlessly. And just like in the regular platforms, in blockchain platforms, we see a variety of platform types. So we can have blockchain platforms that are differing in governance, control and the role of the platform provider. So here are brought a categorization and typology of platforms that are as well traditional as blockchain. So in the traditional platforms, we mostly focus in our analysis and in economic endeavors on the centrally managed platforms that have a variety of strategic choices on price, access, advertising and the number of items that you see when you are making your choice and so on. And there is some discussion about open platforms when we talk about Wikipedia and Linux, for example. Those discussions about open platforms are often focusing on comparing the open platforms and the centrally managed platforms. There do exist also a co-op and consortium platforms, but we typically don't really explore them much more in our research. So MasterCard and Visa, before really going public, they were a consortium platform owned by the banks. SWIFT is a consortium platform, but there are many others. There are competitors to Uber, there are competitors to Airbnb that are consortia platforms, that are owned by the participants and they are jointly managed. We see the same kind of structure of different platforms in blockchain platforms. So in blockchain platforms, the focus is actually on different, goes in different direction. So we mostly, what is mostly discussed in the public sphere is Bitcoin and Ethereum, which are completely open platforms. There is nobody there, even more open than Wikipedia and Linux because there isn't really, especially for Bitcoin, there isn't anyone making final decisions or vetting or being able to shut down the website, like in the case of Wikipedia. And similarly in Ethereum, despite Vitalik Buterin's engagement, he could know he would not be able to shut down the network. And they do operate very effectively. But we also see a co-op or consortia platforms running on blockchain. So a kind of a well-known example, we'll discuss example would be what was called Libra and now it's called DM, the effort of Facebook to gather other companies and together create this consortium blockchain for managing of stablecoin. There are, so Spunta Banka is a consortium blockchain for banks in Spain that is already operational. And they're also centrally managed platforms that are based on blockchain and they were created solely because the blockchain technology was available. So this would be repolled that is connecting banks for inter, international payments. Lucidity that I'm going to mention a little bit more is a platform for advertisement for mobile advertising. Connecting food is, or kind of IBM food trust as well. Those are blockchain platforms that are tracking or providing tracking and supply chain management for food services. Now Spunta Banka is an example of this co-op blockchain platform. It has, it is owner, it brings together 98 Italian banks. It was promoted and it is promoted but by Italian banking association and it's coordinated. So the key here is it's coordinated by the ABI lab but it's owned and controlled by the banks. And so it is operational for a while now and now it's used only for reconciliation processes. This shared ledger is very useful here to make sure that the banks do not need to go through the multi-steps reconciliation of their own individual ledgers. And there are more D-Apps planned on top of this blockchain for the future. Now, what is important because of the structure, the co-op structure is that nobody can unilaterally decide to increase price for participation in this blockchain, in this platform. It needs to be made jointly by some share of the users. So this is in contrast with Lucidity, which is like I said, a blockchain platform for mobile ads. So the issue, the problem with mobile advertising is that the ads are going through multiple ad exchanges between the advertiser and reaching the final customer. Those ad exchanges have their own metrics. They report in a way that it's not necessarily consistent one with another and advertisers complain that they don't really know whether the ad is displayed to the population that they paid for. So what Lucidity does is it tracks the advertising through participating exchanges. So exchanges that participate in the blockchain can get the ads that Lucidity is contracted to display. And smart metrics are governed by smart contracts which is reconciled, which makes sure that all the metrics are reconciled throughout the supply chain. And there is a transparency provided by the blockchain and the tracking. So we can really track all the ads from the advertiser all the way to whom it was displayed. And this transparency yields higher willingness to pay by the advertiser. So a company like Toyota, for example, said that they are not going to advertise on mobile through any other channels but Lucidity because otherwise they don't know where the money is going. So right now Lucidity is paying the ad exchanges for participating in the blockchain, but you can also imagine that when they become more demanded and they get more of the higher share of demand in the whole market because they are so attractive to advertisers, they may ask the ad exchanges to pay for participating in their blockchain because the ad exchanges are going to have very few ads outside of the blockchain. And they are making unilateral decisions on the prices. So with that, out of those examples, I actually want to go back to emphasizing what blockchain technology is bringing to platforms and how does it differ from internet? So internet already was providing us with more content because it was easier to transfer information between individuals, but the content is subjective. On blockchain, you are getting with shared ledger verifiable and symmetric information. And this allows us to algorithmify processes and use of data between independent entities. So like unifying the metrics. And just as internet was allowing us for algorithmification within a company. So now we can do it between the companies. And this may allow us to also algorithmify governance like in the case of Ethereum or Bitcoin. And permission less access may also open access to services that before were restricted like for example, financial services. So we see it in the explosion of DeFi applications. So this algorithmification of processes and of governance may be especially beneficial for open and core platforms in order to improve governance and monetization because now we can have channels to move money around in a completely open system. And before, if you wanted to move money around you needed to set up some kind of monetary like either connection to a credit card or to, or maybe your own bond but somebody would need to keep track of money. Now we don't need to do that, okay? So going back to those different types of platforms, pre-blockchain and blockchain we can actually see different or I like to divide them into different categories. So I would say that centrally managed platforms whether they are blockchain or not blockchain they have very common feature. They make a lot of those strategic decisions unilaterally and the centrally managed blockchain platforms they're using blockchain as technology but the fundamental economic forces are very similar. And this is why I called them, well, I initially wanted to call them Tirol or Shed Coyote Julian platforms but it's quite a mouthful. So I went for Toulouse platforms. So Toulouse platforms are centrally managed and they do have a wide variety of strategic decisions that they make unilaterally. And I contrast them with Nakamoto platforms which use blockchain and attributes of blockchain to improve the centralization of decision-making. And this may allow for those consortia and open platforms to operate much more efficiently and allows for also monetization. Okay, on top of the platforms that the blockchains being platforms, like I said, the blockchains can host other platforms I like the apps. And what is important is that blockchain can give rise for example, blockchain that is completely decentralized like Ethereum or Bitcoin can give rise to new platforms and intermediaries that are centralized. So it doesn't, so we can have examples of Nakamoto platforms and Toulouse platforms that are coming from Ethereum, for example, or that can be built on even with Bitcoin. So Uniswap is an example of this open and decentralized trading up. Many millions of dollars are going, are being traded through this exchange but it is operating without any trusted party that would be setting prices and collecting the surplus and extracting value. On the other hand, the cryptokitties that I already mentioned, the game, the people who set up the game, the company that set up the game has full power and they make a lot of decisions including pricing and how large a cut of the trading they are taking and they can change the decision. So this is an example of a typical Toulouse platform and they even made decision to move the D-up from one blockchain to another blockchain. So strategic decisions are even more than just pricing and access. And OpenSea is an intermediary facilitating creation of NFTs, I didn't talk about NFTs but it's basically a trading platform. So it helps you set up your NFT and it is also an auction platform. And it is much more similar to very traditional platforms and intermediaries and any other auction platform like eBay rather than similar to Ethereum being decentralized. So it is again a Toulouse platform, okay? So I think that instead of asking whether platforms will replace, whether blockchains will replace platforms, we should be asking whether Nakamoto platforms would replace Toulouse platforms. So let's suppose that we have a completely open and decentralized right sharing app like similar to Uber but it would be a Nakamoto platform. So Uber is not taking the fees, it's not taking the cut, it's distributing all the value to the users. So would it replace Uber? Would it be more attractive than Uber? Would it be cheaper and better? So in fact, we can inform our thinking from already existing studies about competition between open and proprietary platforms. So what we're learning from those studies is that while open platforms often offer lower price for a given size of the network, they face challenge in actually coordinating and growing the network. And this is because the proprietary platforms can offer subsidies, can actually go and advertise much more. And they can quicker improve quality. There's also an important element that I don't think has been explored very much in the literature is if you have, for example, consortium platform or open platform, there is a big difference in incentives, in objective function. So a centrally managed platform is going to have objective to increase profit of the platform. Whereas if you have an open platform or a core platform, it's going to have the individual members have different objective than maximizing the total value of the total profit of the platform. And this fundamental economic, these fundamental economic forces are not changing if we have a technology like blockchain that is even if it's allowing for open and core platforms to operate more efficiently. So Nakamoto platforms may or may not drive to lose platforms out. I'm not, I'm just making hypothesis here and it is actually a subject that could be a good subject for future research. And I think that in this crowd, I want to convince you that it's a good topic to take on and explore. I suspect that Nakamoto platforms can increase competitive pressure but mostly for open and core platforms. And for, and as core platforms are more efficient, maybe it will increase pressure on the to lose platforms. They may decrease barriers to entry by removing this participation restrictions, but this probably applies to small set of industries where those restrictions are already existing, okay? So in conclusion, what I am arguing here is that blockchains and when working property, again, I'm not, I did not get into those conditions, they offer a short ledger of information and this facilitates linking processes and automation between independent entities. And this is a new aspect. And it may facilitate this centralized governance which we make open platforms and cooperation core platforms more efficient. But blockchain platforms can be both centralized or not. So there could be to lose platforms, what I call to lose platforms and Nakamoto platforms for a core and consortium, core consortium and open blockchain platforms, they may be more efficient than non-blockchain open platforms. And that may provide this competitive pressure on the centralized platforms. The open blockchain platforms may provide access to services not available to individuals otherwise for at least some set of services. And like I said, Nakamoto platforms may increase competitive pressure on to lose platforms as well. But in the end, I would say that all types of platforms may benefit from blockchain and what is still an open research question is which type of the platforms is going to benefit most. And just to kind of finish, if you are interested if this talk succeeded in interesting you in blockchains and you want to explore blockchains in relation to platforms, you may also like that. So together with co-authors I wrote a summary or overview of literature on economic forces that are playing out in the cryptocurrencies which is one of the open platforms. So this is a clear link between economics and blockchains. And also the book that I have with Mikhail Sarvar and Guillaume Herringer now is going into the second edition and both the book, the chapters of the book and the gel paper are available on SSRN. So for the book, those are preliminary chapters and in fact, any feedback would be very useful at this point if you have. Okay, so let me finish here and thank you very much. I hope I'm going to get more co-authors working on and competitors working on blockchain and platforms. Thank you. Thank you, Hannah for the expert timekeeping. Now we go to Andrei for five minute discussion. Sure, so again, this is a pretty different talk from what we normally do in keeping with that. So I'm not going to discuss this, we normally discuss a paper, but it's more like I want to drill a bit deeper into some of the points that Hannah made. Anyway, Hannah, this was great. I think it was a great setup for the, and I guess rephrasing of the question that we initially had in mind when we talked about you presenting this. So just to sort of push the discussion forward based on what you talked about, I think what would be really interesting is to see if we can nail down a little bit more clearly what the main trade-offs are between the Toulouse platforms and the Nakamoto platforms. So maybe say, let's say I'm thinking of basing a platform on blockchain. So I can be somewhere in between a pure Nakamoto platform and a pure Toulouse platform, but I'm considering using the blockchain. So we'll be the main trade-offs. What are the main benefits and the main issues? So I think there's some interesting economic issues here. I think you hinted at some. So I'll just say a couple that I think are pretty obvious. So one of the great things about basing, so if you're basing your platform, like we're using some elements of the blockchain, I think a key thing, which you hinted at, but I think we can discuss in more detail, is there's some aspects of governance that instead of being unilaterally decided by the platform sponsor, I mean, I think of this being like a dictator, right? I mean, I can choose to do whatever I want in terms of governance. There, sometimes there might be some benefits in having those algorithmically done. It's almost like a credible commitment to have a democracy in some sense. So that could be one benefit. On the cost side, I think you mentioned some, right? So I mean, there's slower decision-making. It's a little bit harder to solve the chicken and egg problem. I can also think about issues around with the switching. If we're basing the platform on the blockchain, presumably this, I mean, I think my intuition is the switching cost would be lower. So it's maybe a little bit harder to have winner take all than with traditional platforms. So I mean, I'm just curious what your thoughts are. So can you elaborate why you think that the switching cost would be lower if we base our platform on it? So that's a good one. So actually, I think maybe that's one of the key ones because that's one of the key questions, right? That we always do with network effects. Do we think it's gonna lead to tipping or not? So I haven't thought this entirely through, but my idea was something like, if I'm basing my platforms on a blockchain, I mean, it's a bit like open source, right? I mean, there's a mechanism in there that is outside my control. And I'm thinking others can basically build very similar platforms on the same base, right? So like for instance, so we'll be a good example. Let's say the right sharing part because there's a bunch of articles online saying, Uber's the bloodsucking intermediary and therefore we're going to create, there's going to be some blockchain initiatives that say we're going to create like a million competing Uber's, right? So like there's no more power to the intermediate. Or all the benefits accrue to the drivers and the riders. Well, I mean, again, I don't really believe it's that extreme, but certainly if, for example, if the identity and the ratings and the information about the drivers and the riders is on a blockchain, then that takes away a lot of the power, a lot of the strength of the network effects that you could accumulate around the proprietary platform. Now you still need some proprietary elements but wouldn't that take some of those away and therefore make switching costs? Like if I'm the driver, I can basically take my reputation, if my reputation, my identity is on a blockchain, I can, I should be able to more easily switch between different competing platforms, shouldn't I? So yeah, so I guess the question here is whether the information is public and portable, right? Because why wouldn't you be able to take your Uber ratings out? Why aren't you? Well, because it's controlled. So in the case of proprietary platforms, I think one of the things they do is like a lot of it. So a lot of it is completely controlled by them. Yeah, but you can have a screenshot of your rating and this is your rating and why can't you take this rating to another platform? And so if we are going to, so I think that what you are assuming is that we are setting up a completely open, decentralized platform on blockchain. So like I said, on blockchain, we can set up a platform that is centralized and limited so you can see the information but you may not be necessarily be able to take it out. I mean, you can, to the same extent that you can take a screenshot of your Uber rating and you can take it out. But the, just because it's on blockchain does not mean that it's interconnected. By the way, so I think screenshot, so I think screenshot is a bit like, I don't think it would be credible enough, right? I mean, I think screenshot would not be enough. So you need to take a little bit more than that. But I think I see what you mean. So it would be, so I guess you can play around with how much, so like we can play around with how much control you can give over the information and how portable the information can be, right? For the participants. So for example, are you going to, you can set, you absolutely can set up a platform on a blockchain where all the information about the drivers and the ratings will be publicly available. So then it would be easy to verify that indeed this platform, this platform, this driver has five stars. You can do that. Would the platform want to do it? Because why wouldn't Uber make it public? Instead, what you are seeing, when do you see the ratings of the driver and you see it only when you call and you are much with the driver? And the blockchain app can also show you the rating only when you have called the, when you have called the cap. So just because it's blockchain does not mean that everything is transparent. The blockchain app, the D app can make decision about what is transparent and what is not. So for example, the very first D app, which is CryptoKitties that composed of four smart contracts. And we say, so if we have a smart contract on Ethereum, we can actually see the code of the smart contract. And we can see even the ledger of the smart contract. So whatever is held in the state of the smart contract. But with CryptoKitties, the genome of the CryptoKitties is actually held. The algorithm that is assigning this genome is held outside. So just like I said, we have smart contracts and then we have a layer of other software and then we have the interface for the app. So where does this, where do the ratings reside? Do the ratings reside in the smart contract and therefore they are visible or do they reside in the layer of the software? So you can have like Uniswap has everything only in the smart contract. So everything is smart contracts. Everything is visible on the blockchain. So now the question is, if you are setting up a DAB, you are making a decision about making it completely transparent and committing to this transparency or retaining some sort of power. And so... By the way, let me, so I just want to bring it back. So I think this, I want to bring it back to the broader question because I think that's kind of what I'm asking, right? So like if you think about this continuum between pure Nakamoto platform, which would be like, I guess, as close to, everything's on the blockchain, like maybe not necessarily transparent, but everything's on the blockchain versus on the other side, pure traditional platform where everything's completely controlled by the platform. You can be somewhere on this continuum. So I think then the question is like, what are the trade-offs of moving between one end of the continuum to the other? So I just want to sort of bring it back one line. So switching cost is one. So it's interesting to think about my intuition, my intuition was like, if I move from proprietary towards Nakamoto, switching costs should go lower. The likelihood of tipping the market might go lower. But you're right, it may not be entirely obvious. But let me also ask about, I just want to make sure like we also cover like the other, there's a bunch of other aspects, which I think are important. I want to let you give us some time to articulate. So what are the other trade, like if I move from proprietary towards Nakamoto along this continuum, what are the other, I mean, there's some benefits and there's some downsides. So can we try to sort of like have, at least at a high level, phrase those? Well, so first of all, you don't need to set up a company to set up at the app, right? You can set up a smart contract and you didn't need to get anyone's approval on Ethereum. You don't need to go through the process of setting up a company and hiring people officially and paying insurance and everything else that is related to cost of running business. You can have it at the more shoestring, so to say. So it is far more flexible. It is kind of similar to guys in a garage setting up Apple, but in the new world. So it is easier. And yes, you can make it more attractive by setting it up in an open way and at least promising that it will stay set up open way. The downside is that a lot of very few people can actually scrutinize their smart contract that you are bringing on. And unless we have like companies that are doing the auditing, then there may not be enough trust. But it's probably cheaper to set up on the blockchain. Now, you also need to worry about all the errors. If it's completely automated, then the errors are difficult to correct. Nobody can overwrite. If you buy a mistake, you sent your transaction fee to be 300 Bitcoins, it's a little chance you are going to get it back. Right. And so the more decentralized it is, the more cost upfront you need to push to make sure that it's working the way you want it to work. And Ethereum is full of smart contracts that are so malfunctioning that nobody wants to call them. Yeah. So is it correct to think of this as some sort of like, I mean, you need to get the mechanism, if you're going to go towards nothing with the platform, you better get the mechanism designed from the get go because once you let it go, there's like, that's the whole point, right? I mean, I can't really change it afterwards. It's a lot harder to change afterwards. So it can be changed afterwards, but then the question is who can change it? If you are doing it in a completely democratic way, that it's then basically decisions are not made. I mean, even in democratic countries, we have the executive branch that is making sure that things are working and the oversight and everything. So if you have, so Bitcoin, Bitcoin itself cannot fix a lot of problems and whenever they try to fix it, they need to fork because of the difficulties in making decisions and making changes in a purely democratic system. So then if you give some people more power to make decisions, the more people, the more power the elites have, so to say, the more problematic it becomes in terms of extracting value. So there's also, I'd like to sneak in one more question if we could and there's an interesting question in the chat from Andy Hubd. Andy, would you be willing to unmute and ask the question yourself? Yes, not a good background. My question is information design, like, so this is a branch of Micro Theory, so you commit to certain statistical experiments that you give to some strategic interaction. In the ability app, we could think, for example, of giving information where there is high demand in the next place. Their optimal information design would be to not give it to all agents in the system. And I'm wondering, is it possible to, for an acar-motor platform to commit to information from this form, which is only some of the agents get information to some parts of the information and they can trust the system, given to us. Well, so in a way, we do have systems like this that give you information on us-need basis. So you can have an even open blockchain and open Nakamoto platform that gives you information only us-need basis, but it's going to be the same information for everyone. So if somebody says, so this is kind of a zero, so what is important is that we have the same information that is verifiable. So you can't, so for example, what Waze very often do or other maps, directions apps, is that you want to go from point A to point B and they are not going to give you the same route as everybody else because then it's going to just create congestion. They're giving different people different routes, right? So I don't know whether you can, well, you can do it if you have additional layers of software. But if you have information directly on the blockchain, that I don't think that it can be done, that shared information is different because if you're making query, it should be the same information. So you can have it in a D-app that is somewhere between the Nakamoto and Toulouse. The question is where this has been formalized. I find if this impossibility is stable, I would find that very interesting. Do you have any? No, so I don't, so actually, so I haven't, and this is an interesting question. There may be results in, so the definition of just pure stable, pure smart contracts that are directly on Ethereum, for example. So this is what a lot of blockchain, smart contracts on blockchain say, that you cannot have non-deterministic calls. So all the calls need to be deterministic, which means that if you call a function, it needs to give you the same, with the same conditions, it's going to give you the same return. So this would be probably equivalent to this impossibility statement if we turn it right. So we reached the end of our hour together. What I would suggest is that I'll stop the recording now and as I said, there's the opportunity for people to stay on the call if they'd like to continue the discussion on a more informal basis. Otherwise, I hope everybody would join me in thanking Hannah for an interesting and very informative talk. Thank you.