EthCC 2: Martin Lundfall

hello everyone thank you for coming back down all the way down from outside and someone said that that things were literally on fire even before my talk started so I'm happy about that I'm gonna be speaking you speaking with you today about subjectivity and objectivity in decentralized systems and really perhaps a more accurate description of what I'm actually going to be talking about is the values or value of the value instruments that we are building so it's gonna be a little bit of a more philosophical governance ideology oriented talk then then the more technical talks that I've been given before so if this is disappointing to you and you really want a technical talk then I can give a little announcement that there will be an impromptu formal verification workshop happening two o'clock tomorrow so if you're interested in that and get in touch with me and I'll tell you where that's gonna happen it's gonna be a little more intimate so justjust get in touch with me animal I'll tell you the location so what inspired me to give this talk was a presentation that I saw at Chaos Communication Congress on ethics in mathematics by Maurice Giotto he was speaking about the enormous influence that mathematics is having on the world and how little mathematicians are willing to acknowledge this fact you know a lot of mathematicians are really happy with solving problems of geometry and statistics and analysis and they don't particularly care for whether what they're working on are going be used for calculating the jail sentence of some number in a database to them whether their work is going to be used to calculate the trajectory of a ballistic missile or whether it's gonna be an algorithm that decides who's gonna hook up and potentially create the future generations so this is obviously an enormous impact that people have an in any profession you shouldn't simply disregard the ethical implications of your work and I think in the blockchain space we're more willing to acknowledge that there is an ethical dimension to what we do in fact I think that's what attracted most of us in the first place to the space but when you speak about ethics it's really hard if you don't really have an underlying theory behind what what you're reasoning with so I want to present a theory on value that I found very helpful for analyzing value instruments the type of value instruments that you can build in decentralized systems and go through a couple of case studies so to hopefully make this more concrete instead of just talking about this value of theory from an abstract perspective so let me begin as I said what we're doing is constructing value instruments so really constructing something that we need to understand in order to constructing in order to construct value instruments we need to understand what value is as these instruments that we're building have the potential of having tremendous impact over people's lives in some cases we're building weapons of death construction in other cases we're building some we are derivative of some possible future value and we're constructing really dreams about what the future can be and then we need to be aware of the methods by which we create the future that we want so we're doing this in a setting where we're renouncing traditional institutions and usually we don't speak very clearly about what particular aspects of these institutions that we don't like we just say well all these centralized institutions they're so bad because they're controlled by a few actors but I think we really have to be more sophisticated in our analysis of precisely what features of these institutions it is that we don't like and what are acceptable and which ones of these features we want to bring in to this sandbox little utopia that we're building on the blockchain and I also think that it's important what language we use to describe these systems I've really a hard time with a lot of the the double meanings of some of the words that we use like consensus like when we speak about decentralized governance and oftentimes when when people are referred to system that's democratic they're not really democratic at all we speak about incentive reasoning and we speak about things being valuable I think we should be more granular about these things we should really talk beyond incentive reasoning I mean crypto economics is really just a very narrow view of what economics is is this micro economic perspective that really is just game theory rather than the more holistic view of the resources in a system and and how people function to distribute those resources together which is more macroeconomic perspective in a more sociological behavioral economics never really talks about that so here's a little bit of how I would like to approach the matter I'm gonna be spending most of the time talking about the method which is based on an anthropological theory of value mostly developed by David Faber or at least articulated by him in the way that I understand it and then I'm gonna apply this theory value on a couple of systems and there's gonna be some opinions in this talk sometimes I'll make a big red warning when they are but since a lot of this talk will be a little bit abstract I thought that I would just begin by stating a couple of concrete's opinions so that we can really start engaging in a discourse about this so here are my opinions the first one is an opinionated question which is how decentralized are plutocratic systems really because I really think that most of the systems that we have built upon the blockchain are plutocratic right now and I don't think that that's as good as we can do why we want to build decentralized systems and as a result of that there's a significant lack of democratic input to on chain applications and that's obviously a hard problem but I would like to give some opinions on how I think it can be approached and finally tokenized securities represent legacy financial values and we should not embrace them without a critique of what these values are we should take this opportunity to reconsider which ones have these contractual obligations that the state has invented we really accept this legitimate so the method here as I said is based on an anthropological theory value that is really embracing the double meaning of value as it is being used in the economical sense the value of a particular asset and values in a sociological sense when we speak about values like friendship or honor you know depending on your culture you have different opinions on what these values are but what unifies these two notions of value is that they're ultimately speaking about conceptions of desire that influence the decisions we make in our lives and one way to think about how this bridge between value and values work is that it really has to do in what direction the analysis is happening so if you're a micro economists then you might start from utility function and model people as trying to maximize their utility and so you have a predefined notion of what they value and then you try to get a model for how they're gonna behave whereas in anthropology you would start by looking at what how people behave and then try to figure out what their values are based on how they act and it becomes really apparent if you try to reverse this analysis if you ask an economist to explain the act of gift-giving for example then then they start saying well they're really trying to give a gift because they might get something in return and then you're like well what about anonymous gifts and I like well they're really trying to maximize their good feelings or self-worth or something like that and once you start talking about those values then you're really not that far from when this sort of soft values that people speak about in the sociological sense but there's another way to understand the double meaning of volume and values that goes through the concept of tokenization and i mean this in a broader sense than just the tokenization we do on the blockchain is really taking some abstract value and manifesting it as a token of value as a representation of this value so some examples are the value of education being tokenized in degrees or certificates from a particular prestigious school the value of science being tokenized in the form of grants citations or publications the value of art being tokenized in art reviews or awards or artists retrospectives or ideas being tokenized as intellectual property and whenever this acts of tokenization happens there is a curious phenomenon that happens with humans which some people refer to as a Phet association of of these manifestations we start pursuing the manifested token in and of itself instead of pursuing the token because it represents some value that we might value so people start to go to university not because they want to get educated because they want this degree or people start only caring about their citations if they're doing science and so on it happens in every regard and it's a little bit related to the to the concept of gamification whenever there's like in high score you're gonna pursue the high score instead of actually pursuing whatever the high score might represent and this is particularly obvious in the case of money because money is also representing something I mean it's a little bit complicated to decipher what it represents because it's so fun jabal and so detached from the origin of the value where it was created but I think there's no clear case of this objectification and fetishization that happens than in the case of money where people really start pursuing money for money itself rather than the things that they can get with it or the values that it represents so it really has to do a lot with fungibility as well and we have to realize that what makes money fungible is the fact that other people are also acting according to this common hallucination that it actually represents some form of desirability some form of value and it's it's very easy to break down this fungibility if you go to some place where people don't accept your form of currency then you know it doesn't matter that that that your actual cash bills or your little bitcoins are actually fungible because they don't actually transfer the value that they are meant to represent if people don't care about this particular currency that you have and this also gives an interesting view or I think a helpful way to think about the different values that fiat currency and cryptocurrency represents because the value arises pretty much not only from the various ways in which people use it but also the systems and the mechanisms by which it is created in the way that it can be recognized by institutions and fiat money is getting a particular kind of encoding with the contractual rights that it can represent and other things that you can buy with money the fact that you have to pay tax with fiat currency and so on makes it encode a particular kind of different value and then cryptocurrencies do so if we want to think about what value is cryptocurrencies represent then we should look at the underlying values of how people created them and why people created them in the first place and it should come as no surprise that the the most tangible arguments for why they are being created is in the argument of decentralization and I think people understand this concept fairly well so I'm not gonna spend too much time on talking about it I just want to highlight another double meaning that is in decentralization from the perspective of the system designer the system architecture and the centralization in a more ideological sense where power is being diffused but this double meaning is kind of useful because the diffusion of power happens through the technological means so so I'm not saying that we should separate the two but I'm just saying that when for example people say that users don't care about whether your app is decentralized or not then it is true that they don't care about where the servers are being operated or whether you're using the blockchain or you know the technical infrastructure they don't care about but they do care about or they should care about the power structures of the system that you're creating so they do care about decentralization or they should care about the centralization in the political sense and so this diffusion of power comes through open participation you have permission lists networks and open non participation the right to to fork and to do your own thing the right to opt out of the system completely and if you don't like the way in which the platform is being governed you can just create your own platform and when we start talking about platforms I want to emphasize a difference that I think is is useful in order to understand this non participation because it's really a distinction between non participation in a protocol and non participation in a platform and it has also has to do with with governance so I like to think about protocols as rules of interaction and communication protocol and platforms as places of interaction and the key distinction here is is what happens if you want to fork I mean if you fork a protocol and and start acting to the protocol in some other place then arguably you're in the same protocol still you're acting according to the same protocol but if you fork a platform and start doing your own thing a different place then you're on a new platform we can think of a theorem as a protocol but the particular aetherium blockchain that presumably everybody here cares most about is a platform and if you start doing your own chain then that would be a different platform so platforms have an objective state sometimes maybe not global but they at least have a state that defines the platform whereas protocols they have more views and and particular subjective understandings about what's going on and so when you encode value instruments in a protocol or on a platform the value instruments are going to inherit the qualities of these underlying infrastructures so let's look at some examples let's start with a protocol a very nice protocol in my opinion is secure scuttlebutt a gossip protocol so each node keeps a log of all of the messages they send out they link them together with hashes and they Sinese each message and they send them out and start gossiping these messages around and you select your friends and your friends start storing your messages and you start storing your friends messages but there's no consensus there's no objectivity in the state of the platform there's all of these messages that you and your friends have sent out but it's but it's not a consensus protocol in fact many of them many of the people that are participating on secure scale about are very much against consensus but it's still possible to build a lot of interesting value instruments even if you don't have an objective view of the system you can for example build reputation systems quite nicely because reputation I would argue is very much a subjective phenomenon you know in some cases there might be it might be useful to to have a little bit of a higher degree of objectivity and the reputation but it's really something that I decides how much I value all of you and what reputation I am willing to assign to you it makes the most sense I think that's a subjective measure another example of a value that can it.you can encode in the system is a mutual credit system so if you just start issuing IOUs to people then you don't necessarily need a global objective view of where all of these are you sly because it might not even be the case that you recognize they are you use of everybody but you're presumably gonna be recognizing the i/o use of your friends and it's really easy to keep track of where the IOUs are I mean of course everybody can double spend in a protocol like this but if you double spend some I use then that just means that you have now given debt to more people and there's no there's no limit to the amount of debt that you can give to yourself I mean at some point maybe your friends start knocking on your door with a baseball bat and go for your kneecaps but that's up to them I mean everything in a protocol like this is very much ad hoc and designed by the nodes in the network in the participants so it's arguably way more decentralized than hammering an objective system so another case study is douse and in particular I'm gonna be focusing on plutocratic tiles so douse where you have some token holders and they're voting on what the Dow should do or how it should operate and the vote is being weighted on the number of tokens that the participants have because that's basically what we can do when we are operating on blocks right now so they form an objective decision based on the subjective views of the participants but the subjective views by of the participants are being weighted by the amount of tokens that they hold and the level of decentralization that you can really achieve where the Dow is completely dependent on the token distribution this is often not known to the majority of your participant in the network because it's really hard to say who owns what or associate identities with accounts in the blockchain and it's gonna get a lot harder once we get better privacy techniques so pretty soon we're not even going to be able to say whether a Dow is properly decentralized or not because there's no way of reasonably knowing whether it is so if our critique of the centralized systems and the centralized institutions that we're trying to escape from is that their governance mechanisms are opaque and they're being controlled by a select few then we're likely not going to get away from this by using plutocratic dots because they have precisely the same properties so I think this is one of the most pressing issues that the blockchain is facing right now because we have all of these ideas on how we can bring more democratic inputs and and notions of identity onto the blockchain but it's really not happening as quickly as we are building up these plutocratic systems and and we really have to act quicker I think if we hope to have democratic control over these on chain applications so the question is how do we build mechanisms by which we can have democratic input to the on chain systems well people start talking about decentralized identifiers the IDS and and they're making progress and it's really interesting but it's very slow and I haven't seen anybody using them on chain yet at least not for any substantial Dow so I think we should start experimenting with our own solutions it's very easy for example to imagine that you and a couple of friends go together and you start identifying people and you know you can use Twitter you can use key base and you can use all these ways in which people have been building up webs of trust off chain and you can just feed that onto the blockchain and start experimenting with giving people identities on chain I think this is something we should do even though it is riddled with problems everywhere because I really think that by experimenting we'll we'll get a lot of value out of it so if there's a hackathon coming up I would love to see stuff like this you can plug it in to plutocratic systems you can bundle your tokens together and you can say that these tokens instead of being simply voted on by individually by the token holders you're going to decide electively according to this identification protocol that you're good but as I said there are a lot of problems with on chain voting systems and to build one that is successful really have to think really carefully about how to avoid coercion into voting in a particular way or vote buying and stuff like that but buying by the way in the worst case degrades to what we have right now so maybe it's not that bad but paper voting is still a pretty safe methods so I think we should be looking more into how to have Oracle's inputs the result of paper voting onto the blockchain and of course that relies on having some institutions that we need to trust but they can still be competing for our trust we don't have to rely on any one of them because we can still fork if something goes wrong but I really don't think that we can avoid the problem of having institutions in in the in the generalized sense of what an institution might be you know it can be a process that is carefully thought through so I'm gonna do a final case study which is gonna be the case study of maker down and what I find to be particularly interesting if we apply this analysis or this perspective that the value of a token is encoded in the process by which the token is created and the language we use to describe how this token is created then we can start by saying some pretty concrete things about the value encoded in die first of all we can say that the means of value production are distributed according to the token distribution of the underlying collateral so the collateral that is being used to create die that the distribution of this token is going to influence the distribution of the mechanisms by which you can produce die so if you have a bunch of eath whales then now you just produce a bunch of die words but also we need to think more I think about not only the price of the assets that one might put into a system like that but also the value of the assets that one puts into the system so tokenized securities coming up again here as a particular kind of resource or an asset that might have a high price and so from the pure micro economic perspective it's it it is a temptation to add it as a form of collateral whereas I really think that it would be detrimental from an ethical perspective to add some of these things so we can think about land natural resources or intellectual property being tokenized and used as collateral and maybe depending on your particular view on on these contractual rights then you might be fine with that but we can start thinking about oil drilling rights in the Arctic being tokenized and used as collateral I really think that this is a horrible type of tokenization that we should actively be rejecting I think it was in 2014 Martin shkreli paid 55 million dollars to purchase the marketing rights to daraprim and AIDS medicine and 55 million that's that's perfect you know think about how much die one could be able to issue a house that I really think that this type of value is not something that we should accept at all in these decentralized systems so that's about all I have to say let's see if I have some conclusions I'm missing an emoji here I really wanted to do this this handshake between social science and technology but apparently there wasn't support for this emoji and the PDF when I downloaded it but let me just let me just read out these conclusions I think we should not cannot avoid thinking about the ethical implications of decentralized technologies and we should recognize that right now as these systems are still being built and experimented with we have the potential of the most impact on the structure of these systems there are as most malleable right now as they are still nascent and we should emphasize which aspects of the current system we're trying to improve upon and share ways of thinking about how to achieve those goals and finally we should embrace the opportunity to reconsider all of the contractual rights of the legacy financial worlds as they are being introduced into this sandbox that we have as tokenized securities and we should think about which one's of these contractual rights we can recognize as legitimate in our brave new digital world thank you very much do I have time for questions or hey with the launch of multi collateral die and the proposal to add a set of new er C 20 tokens as form of collateral do you think this will be a very pressing a pressing issue in the near term for instance I would use my MKR to vote against Iran or EOS being added to the system and all those debates going to happen and how do you think they'll play out so I think debates are definitely going to happen and like tokenized securities are not a theoretical concern plenty of people are designing the the legal infrastructure for how to bring them into the blockchain I can't really speculate on how they're gonna be relevant to to make it out at this point a great talk Martin really great so I wanted to point out like it's really thought-provoking what you were saying so for example one of the thing which I found made me think again and again is the dichotomy that is used in the language in our ecosystem so on/off chain centralized decentralized or subjective objective and you know the truth always lies somewhere in the middle so for example to be more concrete say if I like you I like this talk a lot that's subjective right now but when you create a system of liking like a reputation system it is inter subjective and then when you get your point of kind of a web of trust you could argue this proto objective it's not objective and objectivity still we are looking for is debatable there exists or not so that might be an interesting way and I don't know if you have thought about it and similarly like I just try to change the conversation from subjective objective to this for any spectrum similarly on ovchinin I think because we use this wrong I mean limiting terms we have debates around should be of Oracles like the example you gave which I'm really agreeing with so I don't know if you have thought about and one of the thing is when you talked about the eat whales becoming die whales do you see a world where it could be like a plug-and-play collateral it won't be like just each people would be able to choose their collateral and there will be some sort of a I don't know social consensus or I don't know what which will decide whether it is ethically okay to have this collateral you know that decision will be based on case basis yeah yeah I definitely think of objectivity and subjectivity as happening along a spectrum of intersubjectivity and really yeah and and you know you can speak of a cultural hegemony on the objectivity axis where you really just recognize one form a value and that value happens to be dollars because that's a cultural hegemony that we live in but still that's that's not completely objective I mean it's not a fact of the world it's just a common hallucination among people that we are willing that we can't escape and I really think what you say about social consensus is is way more effective method of achieving a nice plethora of different kinds of values being represented and and qualitative measures of arriving at those values so I happen to think that the governance of etherium development is a perfect example of how good governance looks like you know there's this long meetings everybody brings their point across and it's it takes a really long time and everybody gets really upset but this is like how functioning consensus based decision-making works and it has for a very long time been working this way because it really allows you to bring these qualitative arguments into consideration and I think we need more of that and it just is difficult to bring that qualitative decision-making on to the blockchain right now and that's where I think Oracle's can play a big role thank you I don't have more time for those guys so thank you [Applause]