# Blockchain's Problems with Unknown Unknowns - Shermin Voshmgir

- Channel: [Devcon](https://streameth.org/devcon)
- Date: 2017-10-09
- Duration: 55:41
- Topics: blockchain, smart contracts, distributed ledger, governance
- Watch: https://streameth.org/watch/yt-u4lnSx4s7sc
- YouTube: https://www.youtube.com/watch?v=u4lnSx4s7sc

## Description

Blockchain promises to dis-intermediate and disrupt governance by offering an incentive system for decentralised coordination of a disparate group of people who do not know and trust each other. We don’t have to trust people anymore, we only have to trust the code. So the theory goes. In reality, formalised and codified governance rulesets can only depict known knowns and known unknowns, but have very limited capabilities to properly deal with unknown unknowns — aka all those events, that cannot be anticipated, or have not been considered at the time of writing/agreeing on the code. Recent events in the Bitcoin & Ethereum community (ie. Bitcoins Blocksize debate & post TheDAO hard fork) are a testament of these limitations. Based on discussions in this post, Shermin would present and discuss different aspects of these limitations and outline potential solutions.

## Transcript

Okay. Yeah, that's it. Um, so please welcome Sherman and her talk about the unknown unknowns. Okay, my computer did something. Thank you. Um, yeah, thanks also for inviting me to speak. We had some discussions around lunchtime. For those who don't know me, um, I'm Shine. I started the blockchain hub here in Berlin. um and share the co-working space here at the Ethereum office and um so what we do basically is blockchain advocacy and um sometime uh early last summer I got caught up in the DAO um and who was one of the curators together with you and Fabian uh and a few other people and I guess that is the reason why um I started thinking about all the problems around the unknown unknowns and blockchains and this is why we're here today because Well, most of you are aware what happened and we will talk about it uh today. We're also seeing what's happening in Bitcoin right now with the block size debate and um so I'm hoping that we can have a joint discussion around those thoughts that I had and um yeah so it's the first time I'm talking about this. I hope it's not too incoherent. So basically what are the thoughts? Um we all grew up around the promise of the blockchain and words like trustless trust and disintermediation and um so I guess why I was got interested into blockchain in the first place is because what I saw from the very beginning was this this really potential um uh of that technology to disrupt governance and we saw it the first use case is Bitcoin and it really is an impressive use case because it's the first decentralized autonomous organization or cor cooperation. Uh it's a it's a bank without um central banks and bank managers. There is no central entity that controls it and um uh that is impressive. We've never had such a thing uh in this uh shape and form. So why can blockchain disrupt governance? Because of uh the fact that machine consensus on a peer-to-peer network uh can radically reduce the transaction costs of uh of uh corporation. And by that also it eradicates a huge dilemma that we've always had and we still have in this world is what we call the principal agent dilemma. um that you have in companies as well as in government where um for those who don't know uh the term uh the uh the agent for example um the manager is representing the interests of the shareholder or or other stakeholders the principal and uh if there are information or symmetries there is the potential of moral hazard on purpose or uh not on purpose. So um there's time lag, information lag in decision-m in traditionally top- down organized companies and we very often have these principal agent problems and we have the same problems in in in politics and in representative democracy and with bureaucracy. So when blockchain came around that that really uh gave the potential to disrupt governance because everything uh you you have the rules in the code but only until something goes wrong or only until we need a system upgrade. So I guess blockchain is really really good to reduce bureaucracy uh radically but only in the cases of known knowns and known unknowns but never in the case of unknown unknowns. So one of the use cases of course was a a very good example last year with what happened with the DAO however we would like to call call it whether it was a hack or an attack an infiltration or um um called for um something happened and there was no procedure of how to react to it and um so machine consensus is great but it has limitations in three instances and I welcome you to to criticize this or to enhance this list. So um conditions that change over time. We're seeing that for example a very good example is what is happening in the Bitcoin community when it comes to the block size debate for example or what happened also uh with the advent of mining pools. Uh we will talk about this later. Uh human error that was something uh that happened with the DAO. Uh it was human error. Obviously there are tools around that. Uh formal verification uh is a very powerful tool to to make sure that this doesn't happen anymore. Um but formal verification has also its limits within uh what the system can find. So uh we have yeah we discussed about this quite a bit. Um, and uh, AI one day might be a solution around this, but AI might also be the source of the problem because there's always better AI to exploit potential uh, attack vectors. um and then information as symmetries and I think this is very very important because the assumption in the in the very beginning was that everyone's um interests is well aligned but in in the Bitcoin network is a huge network with many different stakeholders and certain interests are well aligned but other interests are not so well aligned um as we can see for example in the block size debate so in complex multistakeholder environments like the Ethereum network or the Bitcoin network or miners or mining pools have very very different um um interests than for example a company like consensus building on top of it. They have different economic interests. One wants transaction costs um um to go up and another one wants transaction costs to go down. So, oh, I'm sorry. I'm going back. Not quite intuitive. I know. Okay. So, the question I guess the question is always hard fork, soft fork or no fork. So many people say no fork in many qu uh in many cases. So in the case of the DAO and the uh post DAO attack uh Ethereum hard fork, my question to you, how many people did really understand the ins and outs of what was discussed whether to not fork? Well, that was quite easy, but to soft fork or to hard fork and how to hard fork. Was it five people who really understood all the details? Was it maybe 30 people? I don't know how many people would you say. I mean there were two levels to understand it, right? And the the the first level not that complicated was it at least in my view. Which level do you refer to? Yeah, I mean the there were multiple transactions and the the first main transaction or not really transaction but the balance shift. Yeah. Was not that complicated, right? Yes. But the question is like there was this huge discussion around to fork or not to fork and if to fork how to fork. So how many people would you be comfortable to say really understood the ins and outs of the whole debate over this 35 days that we had until the community had to come to a conclusion and to decide whether to be on this side of the chain or the other side of the chain. So that really is a question. So probably we cannot answer it conclusively. My guess was talking to a few people maybe three to five people fully understood it. A group of 30 people understood it very well and uh certain aspects of it and the rest was trusting this group of people. Is that fair to say? I don't know. Is there any opinion on that in the room? Fabian, any opinion? How many people really understood the ins and outs? You didn't pay attention. actually. So the question is how many people really understood the ins and outs of the whole post DAO attack Ethereum hard fork soft fork no no fork discussion maybe 30% just guess 30% of the people understood or maybe even less I mean but yeah so my assumption was five to 30 people five people really fully fully understanding everything and 30 people understanding Well, fully. Yeah, that's that's simple. That's a lot of a lot. Yep. Why did you have 35 days to make decision? Um, okay. That's a very long story. I I'd like to explain it to you maybe later. Um, yeah, that's that's not okay. So, why am I asking the question? Okay, maybe let's go to the next one. Uh, the current one. So the the implications of the software there we can definitely say that too few people understood it because yeah it it was uh wrote back again right yes it was a very important uh part in the decision-m process to decide that the soft fork why a soft fork wouldn't be viable and so too few people you mean how many people would that be who really It depends on on the timing there. Yeah. The beginning grabs two people and after that 30. Okay. So the question here of course is having um the promise of the blockchain being a trustless trust and b decentralized of course is how how decentralized is a system if currently and that might change in the future. uh only 5 to 30 people in this whole network really understand the ins and outs of such a great discussion and uh which is fine and which is fair enough and which is part of the process probably because blockchain is still in its very early stages but the term trustless trust uh we have to be very careful when we use it because it's true until as long as everything uh is okay it's not true anymore as as soon as we need a system upgrade or there is an edge case where we need to take action because in this moment we need to trust the experts. Uh same as for the Bitcoin block size debate. Andreas, you are here. What do you think? How many people really can follow this whole debate? Talking about me. Yeah. I don't know. I mean, you're a wallet developer. So, a lot of people following the debate, but if that's the same question, how many people understand it? I don't know. Maybe really so you would say only 10 or 20 people in the whole Bitcoin Ethereum ecosystem. Okay. But that's of course just an opinion. No, I thought that it would be a bit more than in uh Ethereum because it's an older network. I'm a bit surprised that you're saying but in any case, I mean this is an estimation, right? uh but it's it's really just a handful of people and I mean on Bitcoin there's also not that many developers and I I pay to develop to understand do you think otherwise so in the end what I'm trying to say is in the beginning um we didn't need consensus in the beginning when Satoshi defined the white paper and then it was implemented there was very little need for consensus as the system is growing and the stakeholders are becoming more and when you over time see that you need to uh change the protocol you need consensus. Uh but this consensus finding process not everyone is an expert. So we still need to trust real people there will be centralization around the experts. So why is it important uh in the case of unknown unknowns? When unknown unknowns occur, when conditions change over time, when unexpected um things uh appear, we still need to trust the design architects of the blockchain networks and the few developers who understand everything and we can try to make an educated decision, but we still need to trust them. Um the next question is I'm so sorry keep going up. Um so the myth of uh trustless trust we will still need to trust experts. Um then there is always uh um the debate whether blockchains or public blockchains they're open source. So based on meritocracy if you put enough time and effort to understand the code you can understand or or learn uh coding or learn um the blockchain pro protocol and uh the languages building on top of it. you can understand it. But there is a limitations to meritocracy because as long as we don't start teaching coding in in kindergarten, there is only a limited amount of people who really have the skills uh or the talent to uh understand what is happening in this field. And uh so yes, in theory the code is open source. Yes, in theory there is such a thing as meritocracy. But but in reality um you need to spend most people would need to spend enormous amount of time and effort and maybe lack the talent to really be able to do that especially based on our current education system. So um all I know is logic. I think this is why we're building blockchains because they're very very powerful. Uh once you define a rule set and because it is autoinforcable um it is very powerful. It reduces a lot of transaction costs but um it reduces those transaction costs because we have the transaction layer. We have the peer-to-peer network and on top of this the transaction layer. This is where uh we define the rules to the game uh in the economic models uh based on game theory. We define which actor gets which incentive to do what in the network. In order to model this uh consensus layer, we need to make certain assumptions about the behavior of the network participants. And these assumptions can be true or not. And mostly they're probably um um some standardized assumption and you need some fall tolerance um tied to it. What if these assumptions are not true? What uh if uh what happens in the case of unknown uh unanticipated or irrational network uh behavior of the network participants and um basically and I had a lot of discussions also with black about like uh the homoeconomicals like um you need to make some standardized assumptions and the assumptions are m mainly based on the idea that network participants will make irrational decisions based on profit maximizing their own uh interests uh as a network participant. For example, the miners um but if we for example look what happened uh with Bitcoin, Bitcoin's consensus layer was originally designed by Satoshi based on the notion of simple game theory. The idea was that every minor running a a computer would pro uh max um profit maximize uh his own interest based on running one or two computers. Um the consensus layer, Bitcoin's consensus layer does not account for um cooperative game theory. What happened after a few years of the existence of the network that we now have in mining pools? So uh the reality or the the uh what Bitcoin was designed to be much more decentralized than it is now because the assumptions that were made on the consensus layer um uh turned out not to be true because now we have mining pools and a handful of them are concentrating almost half of the network. So this is one of the unknown unknowns that evolve over time. Um so the term plutoaucracy and consensus I see like one of the biggest dangers in my opinion around uh blockchains state-of-the-art blockchains as they are designed today is everything is okay as long as the the protocol runs. But the moment we need to modify uh the consensus layer uh the moment we need to upgra upgrade the system the question is who gets to vote and who has a stake uh who has the power to to vote for an upgrade. Now that's very different in in Bitcoin and in Ethereum it's very different. It depends on whether we're talking about the soft fork or a hard fork. And it's also very different in the Ethereum ecosystem when we look at decentralized autonomous or organizations and the token governance rules. Uh they are very independent from each other and particular but in most cases more money in the fiat world buys more voting power in the Bitcoin world. It might be indirectly um well through um um through through hashing power. the more money you have, the more hashing power you can buy and the more vote you have um in a in a in a fork discussion. So in the case of unknown unknowns and given the fact that we come from a world from in extreme wealth inequality, what does it mean for the decision-m process in such cases? Uh it means that certain people have more impact than other people. Is this the kind of how do we deal want to deal with those situations? Um okay, I'm going to skip this. And the last thing that um probably we saw that decentralization or trustless trust um um is a myth to a certain extent. if we don't have necessary decentralized information communication and transparency tools and we saw for example during the DAO hack in a very short period of time uh decisions had to be made we were communicating on social media networks like Twitter like um um uh Reddit etc etc a lot of discussions were happening in parallel on different networks very often it was hard to distinguish the troll from the expert and we were using social media tools from the web two to make decisions in the web 3. So it would so one of the biggest problems was to decide who is an expert who's a troll and the next so for for many of us for insiders it was a bit easier if you know certain people who you might call Bitcoin maximalist or who have certain stakes in other systems you might understand that but an average token holder may might not know that so it's really hard to distinguish um in the decision m making process in the discussion questions happening on the different social media channels, who's an expert, who's a troll. Information aggregation is a huge problem. Very often, uh I remember we had our um um DAO curator Skype channel and were two 12 people there. Even among those 12 people, a lot of the curators didn't know what another curator had posted before because there was like the information and the opinions were all over the different networks and it was really hard to keep up to date with the discussions even among uh the curators. And uh so if we want to make uh blockchains happen and and work well for for situations of unknown unknowns, we need to build um decentralized uh information aggregation and visualization tools where you might get like uh pie charts where you get information um of okay expert opinion is more distributed. Uh but you know like on with one click you would know that 50% of the experts said no fork 30% said soft fork and 20% said hard fork and then you could deep dive maybe into their blog posts and read up on why they said what they said. We don't have those tools yet and it's the wild west of information and it's hard to make an educated decision especially when you're not uh an expert. Uh so why am I doing all uh saying all these things? I think um Vitalik wrote this excellent blog post on the meaning of decentralization and what he wrote was uh he tried to break down the different levels of decentralization. So um the question is how politically decentralized is a network uh meaning that no single entity controls it. Um how is it architecturally decentralized? Meaning that no uh in there is no infrastructural central point of failure. And uh basically blockchains are politically decentralized more or less um depending and and architecturally decentralized. But blockchains and this is why they are so powerful in reality are logically very centralized um because code is um uh is centralized and it's deterministic and um and uh has really um no it is really hard to um to write code for unknown unknowns. So currently we're in the situation where we deal with these unknown unknowns with on the-fly human intervention as we saw with the uh DAO um and the post DAO hard fork as we're seeing also now with the block size debate uh in Bitcoin and hopefully in the future I think in order to really be able to say that we live in a decentralized world where we don't have uh centralization around expert opinion. Um we need to make sure that we visualize power structures in these distributed networks. Uh that we have mechanisms for governance beyond code and uh possibly also um governance mechanisms on the blockchain level. Um and uh we need better information and communication tools and we need decentralized arbitration systems. And I think this is very important because if we don't have these things, the coders of today will be kind of the trusted third parties in the decentralized network that we have to trust. So we need to make sure that we make these decision making processes much more visible and code better understandable. Um and yeah I would like to conclude that if we assume and this is a bit what I'm saying that seeing that uh blockchains are might be the future constitutions of our future societies as citizens and as as um consumers alike. the more services we start building on blockchains, the more the blockchain protocols will become our constitutional foundation. So we have to make sure that they're flexible enough and um and stable enough but also inclusive enough um so uh we can really talk about the decentralized and trustless yeah decentralized world. I wouldn't talk about trustless trust because I would assume that doesn't exist really. Okay, thanks. [Applause] Any comments or questions? What do you mean by making code understandable? It's uh it's a term of uh in term of make it uh like the code self or to make it localization out of that. Um I guess both. So having better human readable code would be great. We were discussing this uh at lunchtime. I don't know how possible. It would be great to have human readable code but the problem also with these huge networks is that uh the code is very complex. Even if you could read it uh there is a limitation to fully uh understanding all the implications of the all the different modules and aspects of the code. Um, how hard is it to from your point of view to um to fully understand the interaction of a smart contract with the Ethereum virtual machine or how many people could really understand it. The more the more complex the smart contract gets, it depends on the on the complexity of a smart contract. I mean, I think it's not not just about being able to understand, but I mean, even if you're able to understand, you still have to invest the time and read it. And so, it's it's so the problem is also at at another level. So, what you're talking about is delegation. So, delegation of time. So even if I would be interested or if even if I could understand the code, even if if I were interested to learn it, would do I want to invest the time to a learn it and then study um every protocol change that is being discussed or do I prefer to to just trust the opinion of someone else? I mean blockchain client is a completely completely different thing but so I'm mainly thinking about smart contracts which are more isolated but but still there I mean so who who of you who of you has ever used a smart contract and who read the full code from my side I was preparing to focus on the next step because if I invest time for the to get rid of the food code I need to invest time and I lose the focus on the um result and I prefer to focus on the which one is the next step. So I trust in someone and then I think about which one is the implication to the next step. Doesn't the same go for a legal contract you read the full legal contract that was signed? Yes, I fully agree. Well, I think um the problem isn't really in reading the the whole contracts and not being able to trust anybody because life is always partly about trust. The pro the like the difference between blockchain and the old world is not that now in the blockchain world you don't have to trust anybody because in the end you always have to trust people because you can't read every single smart contract and and every single um program based on that smart product and so on. But the difference now is that you can choose whom to trust and whom not to trust. In the old world you have to trust banks because there's no alternative. Now nowadays you can be your own bank and probably even in 50 years when everyone is using blockchains not everyone will will be its own bank because people will choose to trust centralized services but if you really want to you can be your own bank and with smart contracts is the same. If you really want to you can read the whole contract and you can understand it and you can um maybe program your own smart contract. So the the problem is not in that you like you always have to trust someone. The the difference is that nowadays you can choose whom to trust and who do not contract. But how is that different to what he said before like in many cases for example a legal contract you're talking about the bank now but a legal contract I I read it I think I understand it but it's it's quote it's just fuzzy semantics where other people studied seven years and became lawyers and then they had to practice another uh a lot of years. So right now in theory I also have the possibility in theory I should understand what a contract says but in practice or I could write my own let's say contract if I want to buy or sell a house yet I don't do that. So how is this system that we're creating now so much more different uh apart from where it really disintermediates um and makes um um clearing um um um it there are certain instances where blockchain really intermediates. It gets rid of certain I would say agency problems, principal agent problems, but it creates new principal agent problems. And I guess this is what I was trying to say. We shouldn't be so naive as to think that we're this is the uh the tool to solve all uh problems that we have around human conflict. We will be able to radically reduce many many many transaction costs but we're creating other principal agent problems that uh we're very often probably not even aware of. Yeah. Yeah. Yeah. There are a few. Let's start with you. Um is there any test of testing approach to verifying um whether the system as a whole um is okay or not? Um for example if I use a a secret software or trusted software some software existing software where goes through some kind of certification which on which level you mean the smart contract level then you should level uh on the blockchain level uh I don't so on a smart you want to really it's uh I do formal verification test foundation um it's really hard to test attacks before or your contract is attacked because you don't know how it would be attacked so you cannot test it. [Music] That's no no right. Thank you. Yeah. Yeah. So suppose we all agree that we need some governance. Uh now the question is do we agree on what governance what kind of governance do we need? Is there any obvious solution that wouldn't maybe be complete but is better but it's obviously better than what we have now. I mean fluid democracy but you probably have other ideas uh liquid democracy I guess we're I mean blockchain really is in its very very very step. So what you're referring to we need more sophisticated governance rules on chain and offchain as well. I think everything that you do onchain especially if you do it on the consensus layer is is is too rigid. So I think we need to learn from the current political world or system that we have where we have a constitution and the constitution uh just in most countries is a very like has a is a very limited rule set defines a very limited rule set the basic rule sets of how to interact with each other and everything else built on top of it. So I guess this is what partially the the blockchain ecos the Ethereum ecosystem uh resolved trying to make the simple laws on the smart contract layer and the and the uh the code um um the the consensus layer of the Ethereum protocol is kind of the constitution. The question is uh and I'm not expert enough whether the protocol could be even regulating even less because every time we need to make a system upgrade, we need um supermajority consensus which is really hard and um some people say that uh we should uh yeah um forks are good. Um so uh the minority protection happens uh by by fork. The people who don't want to to um to upgrade they stay on the old chain. But in reality that is not an option when you have a vested economic interest on one side of the chain. So a lot of people will go on the other side of the chain in spite of the fact that they're against um maybe the decision because they have vested economic interest. they would not stay on the other side of the chain. And I wonder how like in a minor where do we protect minorities on the blockchain? I don't see that yet. Liquid democracy is the thing that you were referring to. I think we need to to have more liquid democracy tools in and uh on top on top of uh blockchains. I yeah we will see that happening. I mean for the DAO there was a liquid democracy proposal uh that never went through because the DAO then never happened but we're already seeing the first um proposals to have such installations. There was another question around here somewhere. No, I had a kind of a rhetorical question to the audience. Who uh did who remembers performing a full uh technical inspection of the aircraft before boarding it? H. So, I'm not sure if anyone does that. Well, yes, that's true. Yet, um, no, no, the aircraft is a different system. It's a it's a different security. It's a different proposition and so on. So, well, the thing is it's not only a different system. And the thing is if you really don't trust the airplane, you don't have to take an airplane, but a constitution of a network. So I think here's where we need to start to get a bit more realistic of what's happening. So in the very early stages of blockchain, it was an optin system and it still is. So first there was Satoshi. He wrote the white paper and for many really uh for many reasons uh many people thought that was a cool thing and they opted into the system and it was better because it was money without banks, without bank managers uh without central banks etc etc etc. But as the community is growing in the Bitcoin network, in the Ethereum network, in many other blockchains, you have more and more people who are now part of the system, their stakeholders, and they opted into uh into a certain code. But when we're talking about protocol upgrade because time changes and there conditions changes or all of a sudden um there is a bug. Now you have to come to a consensus. It's not an opt-in process anymore. You're already there. And how do you reach consensus? And how do you make a decision in a system where you're already you're already in the system on the airplane? You just uh if you don't trust the airplane or the technology behind it, you don't take the airplane. But when you're already in the system, when you're already on top, like blockchains are the virtual nations uh of our future, I would say. and we're probably going to be like citizens slashconsumers of uh various blockchains and uh I don't think that the decision is quite fair. But and then I' I've got a question because you said we clearly need um better systems and so my question would be um looking back at the DA and the post DAW Fox for do you really think that the um that how the events unfolded was such such a bad system because um in in my view we we had quite a constructive discussion 35 days or whatever and then after that we had a convergence on one solution M there still were some people who were not um um fine with that and they stayed on their old blockchain and now they have Ethereum Classic and while I would never use that they still are happy with that and that's fine. So so we even had you know this so so in the end everyone was was happy. So is it such a bad system right now? No, I'm not saying it's a bad system. I'm not saying it's a bad system at all. Um I think it's it's we're experimenting with a great tool that has a lot of merits but it also has limitations. So what I would like to discuss more are the actual limitations the system has in order to be yes uh in order to also build good workarounds or improve the system uh recognizing the limitations that it has and having better governance uh tools and information aggregation tools around it. So when we need to make decisions about protocol change, we can actually make a decentralized decision because the promise of the blockchain was to create a decentralized world. Yet we're seeing that it is much less decentralized that than we initially would have liked it to be. It's true for Bitcoin uh for other reasons than it's true for Ethereum. Ethereum um many people would um criticize or had criticized the Ethereum Foundation to make unilateral decisions and um that decision- making was intransparent and it was done behind closed doors and I think uh I don't think that it's true but I think because of intransparency an average um token holder or an average user or an outsider could have this impression. So we need to make a better job on making these decision- making um um processes more transparent, more inclusive uh and then it's really also more decentralized and we really need to uh rethink the um token governance rule sets on if we want to create them identity based or um uh token based because if we create them uh token based we're creating a world where more money buys more influence. So this is definitely a discussion I would like to have because this is not decentralized in my opinion. Who says decentralized is good? Just buy a gun and distill your own book. Well, isn't that the claim of blockchain? It's useful, but not necessarily good. Well, I'm not saying it's good or bad, but it's the claim of a blockchain to have a an unstoppable machine that no single entity or no small group of people can take down or manipulate. I mean, this is what I understood was the claim. I'm not saying it's good or it's bad. I'm just questioning whether the reality of state of the art blockchains um uh is true to its promise, which was actually to be very decentralized, let's say. Well, the reality is I'd say it's equally possible to use it for surveillance and control than for more fair and inclusive than my system. Yes, I I fully agree. But I'm not sure how that that's a different discussion. I feel I mean I I really fully agree with you. That's that is a bit I think a different discussion though. Yeah. Uh just go back to that uh decentralization was positive pressure on acting more democratically. Can you repeat that? I didn't uh decentralization put a positive pressure on acting more democratically in in in the blockchain world. You mean just generally? Yeah. Uh but also what I wanted to go into is uh so you said that like two or three people understood like what was going on with it. I mean fully I'm not as well as most people are in terms of watching but two or three people really understood that what was going on hard for software with a no for this is my estimate five around how many how many users and how many services currently operate on on the Ethereum network I think you're better it's hard can answer that question let's see what do you okay like individual services that built on blockchain using smart contracts kind of thing things that rely on that kind of governance. Not not too many yet. There's mostly proof of concepts, but I guess you want to have a ratio of how many users or how many stakeholders do you have versus how many people decisions and the thing is the problem with the Dow app is that they managed to exploit the smart contract in a way that it was unstoppable because it was already, right? Yes. So surely any debate about any constitutional structure for Ethereum, the Ethereum network should be one of the most debated things in history. Well, it was very debated during um these 35 years. It should be far more debated and far more well understood before anything is an act. it. This is what I'm trying I mean this is one of the reasons why I'm writing this because I feel that we are creating this very powerful tool which can be used for good and as um SA has been just stating it could also be the ultimate control machine and um I I feel that we're not and and there's already so much economic interest building on top of it currently we're everywhere in proof of concept that stage there is no really uh running applications that are critically running applications that I know of uh running on top of Ethereum or very little. I hope there will never be a critical system. But I depends, right? I mean, people people use ether as a currency, although it's extremely volatile. I mean, isn't that stupid? Yeah. And it's not even a current or it was not intended to be a currency as opposed to Bitcoin. So, what is happening right now uh and is that a lot of money is pushing into the let's say Ethereum network. In that case, for the sake of this argument, I feel that a lot of companies are pushing a lot of money and they're building more and more. Currently, they're just testing it. But as they're building critical applications on top of that um and it's um Yes. And we're we're not even done uh building the constitutional layer or the foundations. Yes. But lastly, is is is the is one potential answer to move towards a multitude of base blockchains. And the more base blockchains there are, the more options people have. This is a possibility. So I guess nobody knows um what will happen. But the the way I'm seeing it uh um unfold is that we will probably um live in an internet of different blockchains and there are already some startups building some um uh blockchain interoper uh operability cosmos polka dot so we will see but probably it's going to go into that direction but there consensus is even more critical. And yeah, and uh I don't know who was at the Polka Dot or at the um uh blockchain meet up a few weeks ago when Cosmos and and Polka Dot was pres were presenting. I felt that basically only Gavin understood what Gavin was presenting and two people understood 50% and the rest maybe 10%. So I don't know. Yeah, it doesn't need to be an understanding, right? I mean the whole of blockchains is that I can I can I mean you mentioned earlier about how when I when I buy into when I sign up to a blockchain or buy some currency I am agreeing to that consensus but this the great thing about blockchain is it's very easy for me to convert to another one to convert my currency to another one. So the whole of internet blockchain is that I can easily if I don't like what the consens how the consensus are changing I can easily go and change if you if you only have uh cryptocurrencies um yes you can easily change if you're a company but if you're I don't know if you're a company um I'm thinking of consensus one of the biggest players in the Ethereum ecosystem they have a lot of startups building on top of Ethereum so I would I would think that when a hard fork happens and um the more um critical those startups or the more advanced those startups become and the more critical the services are that they're building on top of it, I think they they won't be as flexible as a simple token holder who can opt in and opt out by simply selling the currencies. So there are different network participants with different vested interests. They can build their applications on blockchain agnostic way. I mean they they consistently have chosen to focus on Ethereum blockchain but they could easily focus on I'm not quite sure how easy that is. Yep. Yeah. Very easy. Obviously rooto allows you to compile code running on Bitcoin blockchain. So okay but in general I think what you're saying is you have two big questions. One, how are decisions made in groups of people? This is an age old 2,000 year old question, right? And I think we're still figuring that out. The other thing is what you can do about that decision, the execution of that decision, right? So as gentleman up here brought here brought up uh you have choice in the system. So for example, if uh you know the network fees go up, which can be looked at as a tax on your transaction, right? Uh when that cash goes up, you can switch to another network, which is a lot of what we're seeing happen right now with Bitcoin and Ethereum. Now, if your cast goes up on your property and you live there, which you brought up, you have to move to another country, much harder to do. Yeah. So, if you drop these systems in your level of choice and competition against the world that we're coming from, they're vastly superior. And that's why we see continual growth. And of course, the decision-m aspects still need to be uh some level of improvement, but with that choice, you can do what he's saying there. You can you can move to another network. No, I fully see also I mean blockchain definitely is an improvement over the current state systems we have here as well. So to your point about the the reputation, right? Uh since there's two basically two layers, the smart contract layer and the consensus layer. Maybe a lot of people can read a smart contract and it's easier to understand, but far less people could understand the what's happening in the consensus layer. So to your point it it's probably going to keep centralizing in that fashion. So what what are your thoughts on the how reputation should work in that sense? In the sense of in the sense how how do how can we is it a subjective thing that we uh that we trust someone because maybe we read their blog post or something or is there any thoughts on making that reputation like well I mean um I mean the the cool thing about blockchain is that we can now start building identity based reputation systems but this is all in also in the making so I'm hoping ing that as we have better reputation systems, we will have better solutions for web 3 decision making. Um, but I don't distinguish between the consensus layer or the application layer. We will need I mean a lot of the problems that we have in the web 2 nowadays, all this debate around fake news is because we don't have a reliable reputation system. So, we need to really start building them. I mean the internet already promised decentralization and it did to a certain extent decentralized information. Um but now we need to start uh building this other layer and reputation is key to that. Are there any ideas of how well consensus like like there are a lot of identity providers that are already also building on um on um uh uh reputation systems and Yin are you here? Yulcom has he left yet? Okay, you should talk maybe to to Yulcom later. They're also building it's it's not pure blockchain based. Uh it's linked data and blockchain but uh reputation is probably also something. Yeah, you should you guys should talk. Yeah. Okay. Uh last question. Is there any other question? One more comment question and then maybe we should move on in terms of uh reputation and voting. Um I would suggest people look at uh a sort of hedge fund called Bridgewater Associates is the largest hedge fund in the world. They have a great system where they have a matrix of believability based on track record. Obviously that requires you to have identity. So over time as people make decisions or make statements their believability is increased or decreased and that becomes a waiting factor on the votes essentially. So if you think about we have to have BM on one side uh to make decisions about governments and then we have schools on the other which are essentially people who know nothing on the internet don't know anything about what's going on. They have a very good system which has helped them become the largest hedge fund in the world where they people's votes are rated based on their ability but obviously you need to do that. Uh what are they called again? Bridgewater Associates. Bridgewater Associates and Dalio has a paper mobility matrices for systems. Okay, so just a short I guess that works well with hash where you can directly or actually can you directly assess the impact of a a decision on some goal. That's my first comment. And the second comment is what would be the goal be in a blockchain at all? I mean, what do you want to optimize? Where do you want to go? That's not clear, right? Well, it's simply making a decision, right? So, if you're a hedge fund, you say, you know, do I bet that the euro is going to go up? And there's a group of analysts that come together before this decision is made and they say, I believe the the euro is going to go up. That's my vote. They record that vote and then as time pass the euro goes up and those analysts over time develop a track record of being accurate and thus their believability that increases in their vote waiting that increases too. So it's a system where you can determine whether a person makes good predictions. In that case with financial markets, yes, predictions. But even if you're an inventor or a software engineer, you have a design decision, right? You have a, you know, conversation that goes on with other engineers. Do we implement this feature? Do we do this? Well, I believe we should do this because of this. I believe we should do this because of this. And over time, people whose decision-m process or their their talent or skills could be either proved to be um precient and their believability they move up in the believability matrix but logically makes sense but um how do you stop associally becoming something like black mirror black mirror watch the uh black mirror now but basically I mean we're Right. So from a track record standpoint, this is Bridgewwater Associates, the largest hedge fund in the world, most successful hedge fund in the world. So I'm not this is like I'm not talking about, you know, this is like Michael Jordan investing, right? So you know his processes might have produced good results over say 40 years. So I don't know about that. I think he's talking about the notion of murder that you basically rape each other constantly and it becomes a more social thing of raping a person and therefore you can destroy a person's reputation social engineering results and it could be possible for a financial stuff if you have decisions inside [Music] Yeah, I mean in this situation obviously all the analysts who go to a meeting or people who bet on it, you know, they all they know that something said X Y and Z. In fact, all of their meetings recorded and the tapes are made available to everybody. That's for consensus, but obviously it's within that range of this specific hedge fund that was, you know, if you had the entire internet, you know, ranking or somebody down, you don't know, yes. Okay. Well, thank you very much and Christian will continue now.