#237 Matan Field: DAOstack – An Operating System for Collective Intelligence

Transcript

[Music] this episode of epicenter is brought to you by shape-shift oh the easiest fastest and most secure way to swap your digital assets don't run the risk of leaving your funds on a centralized exchange visit shapeshifter IO to get started hi welcome to epicenter the show which talks about the technologies project and startup javac decentralization and the global blockchain revolution my name is Sebastian Quechua and my name is Brian Pavan crane we're here today with Martin

field he's been on the show before we were just checking before and it was in January 2016 and he was working on a project called back feed back there and so you're gonna speak about that a little bit but mainly we're gonna speak today about his new project which is a project called dollar stack which is trying to build kind of tools and frameworks for decent trust organizations so thanks so much for joining us today Martin thanks Brian really pleasure to be here again so maybe we can start there I mean so you tell us a little bit about how did you first become involved in the blockchain space and how did that lead to initially starting back feet sure so yeah I've been working as a theoretical

physicist as my you know it's my profession then my bachelor's in physical mathematics and then PhD in string theory so that was like kind of like when one side of my of my career and then in parallel to that I was very interested in social entrepreneurship and mostly around alternative organizations and cooperatives so I founded a food cooperative organic food cooperative and a community garden and such stuff and when I was doing my postdoctoral research at the Technion that was kind of like yes the end of 2013 and during that I had an

idea about social ride-sharing and I started to work on that idea with some friends to build a social ride-sharing app and via that project we quickly discovered a book chain we we had a need for generating cooperation at scale and and and we I kind of like he knew that we need to contact somehow track values our value of people and to have some sort of token I didn't know about blockchain back then and then someone told me book it look at VidCon look at this thing Bitcoin and even look at this thing here that was the same months when Vitalik published is his white paper no but I think was November

2013 and when we looked at that it was kind of like instantly clicked as a technology that we were looking for and quickly the the project become became decentralized ride-sharing application on the blockchain back then it was a big Bitcoin blockchain I think it was one of the first applications on the Bitcoin blockchain and and that kind of like accelerated very fast I quickly quit the Academy and focus full-time on that and but as time went by I was more and more interested in the core problem with our how coordinating large number of people and January 2015 I quit project at quickly Sue's the

ride-sharing project and founded my own company backfield to solve that problem so to build protocols and platforms for coordination of large number of people or for doubt so that was the goal for of battery founded around January 2015 it was early days we still thought that ICO is too far in or too far and risky and all that we raised the regular legacy capital from angels and have been working for 18 months basically just ended our our our last a work around the explosion of doubt that we'll probably speak about later and at that point we kind of like we

felt that we need to start to hold hold a moment for a moment and reassess where we are at reiterate on them on the focus of the product I one of the most like the urgent burning thing was that I didn't have a technological partner and I didn't want to continue further without finding that partner and then I took the time to do some research both finding that technological partner and focusing the product that we want to build after like 18 months of building several iterations of several products yeah it took it took six months to make this focus and mostly to join forces with my main partner Adam levy who was

also a PhD in physics that I knew from the academy but is also high very highly talented technologists and around January 2017 we found in Dow stock we started we thought a lot whether to continue it as backfield or found a new company but after some iterations and checking ups with you know with external players we realized that it would not be feasible to continue under the same it was to kind of like to long has been passed it would not be feasible for example to raise money to the old company so we started as a new entity we also started with food with a new strategy and new codebase and everything but the idea was similar to build the

platform for a central store most organizations who are fantastic now so you worked on on back feed right which is a similar problem and then you switch you started tall stack are there some kind of big insights or learnings that you took away from back feed so you said like okay these things I'm gonna have to pay special attention to these things I'm gonna have to do differently this time absolutely so I mean when I started back feed I was kind of like more thinking along the interface like how the interface did for those looks like only after a few months I realized that the big big problem is not even the inter I mean the interface is a problem

but there is a like there is a lower problem of the protocol like how those are operating game theoretically and already we didn't back feed I I kind of like shifted for focus on the interface to focus on the protocol one of the lessons from back bid was that these this protocol space is enormous and and there's it contains a lot of big challenges and more so that different use cases and different elements in the Dow landscape if you wish will require different protocols so one of the main understanding from backing was that it's not maybe right to focus and build a protocol but rather it's build a better build a framework

from which you can build any Dow protocol and let those protocol evolve under economic evolution so if you want a get ready for building an infrastructure for neck for the ecosystem and then let that infrastructure evolve over time so that that was the one of the biggest shifts made from backfill 2000 us that we instead of tackling a protocol or a use case we firstly tackle we firstly build a whole framework for governance which scuf like we'll probably get back to that later but we call that wordpress Firdaus so a framework with which you can be many gavest protocol that was one big lesson for bad fit maybe another big

lesson was that in that field we kind of like started very theoretical and you know we firstly wanted to solve everything theoretically and then and then to start building things physically and even if we said we first a bit of shame because they told you was not mature and we said yeah we'll get to do to be launched and only when the technology is maturing up further and in back in in down stock i realized that we have to develop with the technology we have to mature up with the technology so we started coding our code based on the blockchain from day one it was really important we take it with an aim with a very specific aim to produce a very specific product and get it to market adoption and later we also

built on top of that infrastructure we also build applications or right now one application for now so it was kind of like in a way I would say the different the order of thing was kind of like upside down so I'm interested in in this approach that you've described to building this product so you mentioned that when you were building back feed that you you were building specific use cases and that you sure shifted in endow stack to building a more generic platform is that correct correct that is that correct I mean eventually we go back to use cases but we had to take one step backwards to build a more robust framework both in order to you know to be able to produce

a vast number of products and also to enable others to build product but also from the understanding that those products will have to evolve rapidly and you need to be ready with the infrastructure that can allow you to evolve then rapidly so how much of your of the work that you're done on back feeds or is now then poured it over into Java stack in terms of products and not necessarily I mean applications like any product of code base you know protocols or anything you can name its products a almost zero we completely kind of like started from scratch code base protocol

everything we start from scratch but of course I would say that we started from scratch in the right place because of the lessons learned before so how you measure that in percentage that's it for you to decide so let's speak a little bit about dollars in general can you first of all define what is a Dao sure I mean sure it is an overstatement that's that's a quite quite hard thing so firstly firstly the central organization is something that I mean an organization generally organization is something that gets input from agents right members organization and then spells out output product decisions transactions and so on

so forth now this is the one first difference of Dow from regular organizations that the rules of processing inputs into outputs those rules are coded on the blockchain there are coding on and trustless synthesized technology that's the first defense the second difference is that those rules allow for a organisations to grow and scale wide while become decentralized which means that this decision-making process is not held by a small number of people or agent members in the organization's that's another way - or another criteria if you want

so this of the synchronization now you can also ask like what is the how they look like not like the Phoenicians but rather like what is the flavor of those so I would say that the characteristic or flavor of those is that since since decision-making is pretty distributed one outcome is that these organizations are much much more scalable much more scalable and still much more edge and a same time remain remaining much more agile so if you there is there is actually very very well-defined way to correct corrector is that so every organization the planet has a common

factor which is that they become less effective per person as they grow up this scale up so I imagine that dowse is something on the contrary more similar to the way that network effects and free markets and internet works is that these are structures that make decisions and make products and coordinate but at the same time become more effective when they scale up so it's kind of like the opposite of the way that regular organization and behaves when when you grow them up and we that it is also the prediction that once you created that creature it will basically grow up exponentially explode in that domain in its domain so can you perhaps give some

examples of what are some decentralized organizations that exists today or that have existed in the past that you know don't not necessarily in the blockchain space or that use blockchain technology which is generally speaking what are some good examples of decentralized autonomous organizations that don't function with the central governing structure so firstly the reason though you know full blown down thou does not exist I mean I'm claiming but there is you know there is a need for that I think it will explode once it is but it doesn't cease yet however we have many examples of things that are almost douse or you

know seem to be close to or reminding us or giving or hinting us how those will look like so for example in the virtual space so blockchain itself right blocks itself is in a way an organization it's a coordination of large number of people producing something the difference though is that it's producing very very specific thing it cannot do anything else a deteriorated blocks yourself but actually the network around the project you know including the developer community and all that is you know getting closer to something that looks like that but not yet you can also look at examples which are the central

structures but they don't get again they don't have a decision-making power a capacity such as the Internet itself or BitTorrent or you know open-source organizations etc but they don't have a decision made capacity that we are speaking about maybe the closest thing that I'm aware of right now is something like actually I've met a CEO of our loop I think what's most exciting projects that I've met it's the is the team that was winning the the Ilan mask competition about generating Hyperloop technology and while they you know other competing team that did not win the contest raised two hundred and three hundred million dollars the winning team was building a working technology with you

know half a half the speed and fifty scale with a budget of $200,000 1300 distributed engineers over 60 countries all of all of which hold a jobs so that's like this the closest thing that I've heard of to the centrist organizations and but it's the decision capacity is there in terms of budgeting and all that was still fairly centralized like I guess and also was very much based on good faith like volunteering but if you would like to reproduce this kind of behavior a on you know different causes not based on volunteering and on it on a more general domains and at a larger scale you

definitely need to have a systematic decision-making engine that can scale up indefinitely when we think about das or when you think about dollars is this something that you see as just the next form of organization you know maybe had like different forms of organizations in the past and now there's gonna be dolls which is basically you know structured organizations where you have an explicit processes for making this decisions you know that kind of scale you have these network effect so do you think that down the line you know most companies will be sort of dolls maybe countries or like large social organizations or local organizations and

maybe nonprofit communities well they all start to resemble dolls or is dolls more a solution to a specific problem which applies to you know maybe a small number of abuse cases that's that's a really good question so I think the answer is yes or let me be more specific I think the advantage of the DAO is at large scale you know if you if you operate an organization of 30 not necessarily this centralization is superior to centralization but if you operating an organization of 100,000 people then yes I think the decentralization is vastly more effective than centralized organizations and thus just like the the way that

internet so internet replaced know kind of like global wide distribution networks completely but then you still have distribution networks and you know locally and inside the internet so in the same way I think that large organization would be completely decentralized in the future but you'll still have companies operating in that you know global decentralized network but but but but still it's scale I think that those will be orders of magnitude more effective than regular organization and that's just evolutionarily I think every organization will either need to decentralize itself or basically be out competed vastly so does that mean I

guess the implications of this would also be you see in the future kind of the demise of you know the organizations that we know today whether that would be something like you know the US government or Apple or or maybe even United Nations like all of those very large complex centralized organization you think those will essentially be out competed and just not be able to adapt and respond as fastest as dour run organizations I mean I think it's it's a process it's not to be a long process and there are domains where those are kind of like it's harder to decentralize

them so for example anything that that its core activity is essentially physical and geographically local is a bit less fit the model that describing so just because of that naturally I think States will must be much harder than centralized also you know anything that the main assets that the organization is managing our physical assets for example and not only I mean also the ownership structure is kind of like so for example factories factories are maybe last to be decentralized you need to decentralize the technology of manufacturing in order to decentralize them I mean you can decentralize the

ownership of the factories already today but the center is the actual production will require you know for example Google to be based on 3d printing and etc so or micro factories so yeah I think it's not like black and white and it's not that Google will disappear tomorrow but definitely what will happen is that those organizations will will also gradually move towards greater degree of customizations so what do you think it's the positive outcome if this really happens like that and we're gonna have these super well working large number of decentralized organizations at scale like well what will the world look

like that's a really good question because also it communicates with the question of good and bad right so you know one way to look at things is simply about efficiency I mean things a lot of things will become much more efficient the market will become more efficient opportunities etc but that's that's just one aspect and I think not the most interesting one the more interesting aspect you know I think and I think so ever more and more more so every day I mean if you look almost every place that you look at the world you know the systems when you look at a system how things work you know what motivates agents in the system and how they behave you feel that everything is broken like literally

everything is broken and when you look inside you understand why they it's broken is boring for a reason it is broken for the good reason or not good with for the real reason that the there is no good alignment incentives like you know I just talked about it today with a friend just take just take it any topic of your own on your mind like take the topic of the garage like my garage owner doesn't have the incentive to actually repair my car he actually has incentive for my car to be as broken as it can or the health system doesn't have any same thing for us to be healthy it actually has incentive to keep us sick and that's built-in in the economy economics and

then it arrives the rest and part of that crowd of incentives is the is the ability to capture the power so the power is is constantly concentrates in you know in the hands of few and then and then the interests are aligned around the interest of those of those individuals so or centers not necessarily individuals so one I think the big ship will be that power will be much more distributed on one hand a number two is that encrypt economics models will design a much much greater I'm of interest and on general I think the global and you know network economy will be much more not not only much more efficient but also much more you know

I'm trying to be careful here not to say fair because I have in my mind the word fair but I think it's not just about fairness it's about more resiliency and sustainability and and you know harvesting more of the power of more people because there are more involved in greater alignment and eventually I should also what I don't want to advocate for that but I also think that's the only way to actually resolve the you know the big week we could problem we could problem such as climate and and and another pilot that we're facing that's the actually only way to solve that I think without that trying to resolve climate while there are major gigantic corporations that have incentive to actually make our situation

worse is just like fighting you know Don Quixote in the windmills so in the white paper it says that the Dow stack is sort of WordPress for douse and that one would be able to create it out just as easily is one can create a blog on WordPress can you explain what this means what this will look like for the end users so whether these are developers or entrepreneurs or actual users of the douse that will be built on this platform absolutely so when you say it Dow you can look at it from several different levers so let's see and like the first level the first step is the level of the

game right when you say it Dow you say there is a game there are certain rules people can play people can input whatever they want into those rules and things happen right that's that's the game now of course the rules this debt is one there by the way there is a layer below the rules which is the blockchain of course this is the layer that facilitates those rules so there's the blockchain and then there is the layer of rules and then you also have higher layers such as interfaces right the way that people actually interact so if you if you zoom in for a moment at the layer of rules when I say that you can you know establish it down and call you know and and operate it out easily like yet like you would build a

blog from a WordPress in WordPress the meaning is that you you have you already have we already have that you have a framework of rules it's a framework of rules for coordination of people that you don't need if now I want to establish it move down with its own rules with own Governors protocol I don't need I don't need to code that from scratch and rather I have many modular many modules of rules for coordinating people and I can build you know a few models one to the other like plug and play combine them and you know boom I have a centrist organization and I want to have slightly different rules great I can pick up a different volume

plug it in maybe configure some parameters and I have another organization and so forth so I don't need to code the rules each time and beginning but that's also why it's very easy to upgrade the rules you know and evolve over time so that's that's at the rules level now at the same thing you can imagine at the interface level but I mean of our focus right now isn't that level to enable others and all for ourselves but mostly to enable others build their own interfaces but then you can think about the same ideology at the interface level you have certain you know components you can put inside and build an interface for it now so as a user if I'm building a down to have stack I would have access to interfaces

through which I could very easily as you mentioned sort of plug-and-play you know build this this Dao sort of like with Lego blocks right like bringing all the different rules and components and governments governance structures as and schemas that I need for my organization I think maybe where the challenges and where I'd like you to address how - like what addresses is as as a user building endow perhaps I need some guidance as to what types of blocks like what which plugins do I need basically - that - that are better suited or better adapted to the

type of organization I'm building let me maybe give you an analogy if if I'm building a company that's meant to scale to many employees you know hundreds of employees sort of spread out across the world you know with some venture capital from different places around the world and certain types of corporate structures like you know there are sort of established ways that you can start a company for that or if I'm starting a small build a small business that's only going to operate in one country well you know perhaps there's a certain type of legal structure that I need or if I'm building a doubt well maybe I want to have a Swiss Foundation and a you know a a company in Delaware similarly for any type of decentralized organization and

one can presume that there will be certain structures that would be adapted to certain organizations how will Dow stack or how would a user be sort of educated and gain the knowledge necessary to build the right type of doubt for his organization great question so totally so the whole idea of the stack is to provide different solutions at different levels the different you know level of users so for example if you start from top down if you start you know with the average user then instead of working with modules and plugging them in you'll have completed templates so I have one two three four templates you know with names for different use cases in other different

conditions and then you can just click a button that's it you have a template you just need to feed in the parameters for example the founders of the DAO you know the voting power date each of them hold or something like that the talking fusion etc so you'll have kind of like completed templates then for a slightly more advanced users that you won't come back play play themselves and you know and kind of like a experiment with different governance protocol you would have modules models they could combine and get it out for slightly more advanced users or developers you have the library of modules but then it's an open source library and then any development can come up and and yet another module a new rule that you can play with and so on so forth so so yeah

so you have we have a solutions at different levels for different audiences for an advanced user or for the area users this episode is brought to you by shape-shift the world's leading trustless digital assets change quickly swap between dozens of leading cryptocurrencies including Bitcoin ethers e cash gnosis manera golem or cur and so many more when you go to shape-shift dot IO you see if I select your currency pair give me receiving adders sent the coins and boom shape-shift is not your traditional crypto currency change you don't need to create an account you don't need to give them your personal information and they

don't hold your coins so you are never at risk from a hacker or other malicious actor shape-shift has competitive rates and is even integrated in some of your favorite wallet apps like Jack's so you can swap your digital assets directly within your wallet just as easily as putting on your slippers whenever you see that good-looking fox you know that's where shape-shift is so to get started visit shape-shift yo and start trading and we'd like to thank shape-shift for their support of epicenter can you walk us through the different layers of the platform the difference of layers of the stack and the role of each layer as I mentioned there is that of course the blockchain and then above the blockchain there is the rules for governance

organization maybe this is a question on that so block blockchain does this mean like hiriam or like a new block chain or will does that kind of be able to operate on many different block chains or how is that gonna work yeah good point so right right now all coded on the etherium blockchain thank you for that over soluti in the future there is no reason why not to be interoperable over several block chains but right now the only book and it makes sense for that is aetherium so yeah that's that that's the current situation we probably for you know foreseeing future so we have the blockchain there you can also have a you know the central database clear such as ipfs we actually

have our earliest version was coded on ipfs as well but but right now where we are still database wise we're using centralized servers and then you have the rules layer so we call it Ark Ark is the framework for governance which which you could basically architect be from diverse protocol basically any governance protocol so kind of like like somewhat analogous the way we'd say okay we determine can I can I can program any any a smart contract then with is with with Ark it's a higher level framework on facility with which you can program any governance protocol so this is the next layer basically yeah it's a

question on that so if you say any governance protocol do you mean like let's say different types of voting systems or maybe something else where you have some kind of right right right so so what what I mean by any governance protocol so I would say that any govern SportAccord as I mentioned governs protocol is simply that logic that collects input of agents and translate that into decisions outputs whether the decisions are transaction funds budget or whether decision are registering results on a databases or maybe operating yet another function so generally this engine of like collecting inputs and making decisions you can and that's what we call governance you can break it into two categories of rules to

kind of rules the do's and the don'ts so those means that you know if if certain things happens then we do such such for example one one example that is voting systems you know if someone makes opposes and then certain number of people say yes then under some condition let's execute that proposal so this is a voting system voting logic and that's a yes that's a dues kind of rules another another kind of dues can be a token say you know if someone is sending such as such ether to that contract under some conditions the contract will automatically print new talk and the saying that that's a there's also kind of dues then the other family is the don'ts which means that that's for

example crystallized organization with statement no matter what will ever happen the organization will not print more than one in tokens so that's it don't no matter what this organization that managed 1 million dollars would not spend one with more than fifty thousand dollars a month there's a burn a burn rate that's another don't and you can the architecture is as flexible as it can get so you can actually combine the doors and the tools for example no matter what you will not burn out burn more than fifty thousand dollars a month but scheme number six that it can approve in the voting's with majority of 70 percent of reputation can change myself can change the domes so you can

combine those things and basically create any sort of logic for those rules or if you wish any governance system so that's there that's there is a cruise it's called arc then we realize that if you want to be in an ecosystem if you want now to allow for you know thousands of for node for front-end developers to work with this engine and to build collaborative essentials applications we need to make it much more accessible so then we've built a arc Jas which is a JavaScript library over web three that allows you to basically operate our build build organizations configure them operate them vote anything directive with with commands over JavaScript there

is some other layers that maybe are secondary in importance and then of course at the end you have been there of the applications so many different applications that's due to the fact that are all writing on all over the same open protocol that all interpret so different dolls can use different interfaces infinity do you mind speaking a little bit about the genesis now and what is the role of the Genesis style in this ecosystem of different tools the idea was that we have to we will have to experiment I mean we have to do some dog food yeah we have to eat our own dog food so while we have a company that works on you know

further building effectively right now the Dow stock platform we want to have also a parallel experiment a parallel decentralized experiment that kind of like aims at the same goal say mission so the Genesis that will actually so maybe you know maybe it's here a moment to mention we done like a token sale initially privately and eventually so some of it was publicly and we've raised a back then at the date was thirty million dollars and and the idea is that majority of that fund in addition to majority of token distribution will be

managed by that decentralized file so Genesis is that centralized fund that operates on top of dowel stock with majority of that capital in and in that empowers and engines if you want the growth establishment of the growth of the Doucet ecosystem so you know rewarding and investing and supporting projects that support the Dow static assist them in one of three ways either projects the that build infrastructure components of the of the stack more tools more more plugins more modules more governess modules or projects that are building more interfaces to stack

more applications so we've been the first application the native application alchemy and now there is a bunch of other applications interview stock so companies are building interfaces in integrated the stack and finally projects that are themselves douse that are using the stack in order to you know yet do something else maybe build something else or organize or manage collective collectively manage assets or collectively curate databases or whatever so that that's the idea of the genesis down so this first application that you've built this this alchemy application what does it do precisely right so welcome alchemy was I mean initially we wanted to build the first interfaces for that for those so how do

you make a large number of people organize and build something together but then we decide to focus to make it slightly more focused and answer it real me that people have right now and so it's a decentralized budgeting and rewarding system so which answers the the following pain so today you have in the blocks of space you have a lot many examples of projects that have very large capital so for example the you know the the flagship example would be the Tyrian foundation detail project itself so the team project let's say I manage you know just trainee number managed a 1 billion dollars in capital so 1 billion dollars is roughly infinity right it's much more than you can spend

in a short timeframe so it has infinite capital so ok so there is no there is no limiting factor in terms of capital now we also have thousands of developers if not tens or hundreds of thousands that could work on getting a macro system so there's also north shortage of human capital so what is the limiting factor what is the actual limiting factor right now for solutions for producing solutions like why haven't we solved the stability problem is do we lack duty theorem foundation lack capital the answer is no the rotation lacks developers I mean you know in the broader system the answer is no the actual answer is the decision-making capacity to wisely deploy capital into

human capital and produce solutions and that's exactly the role for alchemy so decentralizing that function the decision-making functions so it's a it's a you know people can get into the system they can make any puzzle to use of that fund and then people can vote on that proposal and produce decisions in large numbers and effective in capacity and effectivity yeah I'm super excited about that because this is one of the things that you know concerns me a little bit is that all of these projects have started right and and they raised money and maybe into foundations and now they meant to launch these decentralized networks but they were actually operated

or controlled in a fairly centrist way and now I think that's there's not the fault of these people or you know it's not the fault of these projects it's just that the tools aren't there yet to kind of build and decentralized way from the ground up but I'm you know I'm unsure how well it will actually work to over time transition to a more decentralized organization or whether this will not be sort of fatal for many projects that they will never be able to dissent us anymore so I think when people when purchase confirm the very start like you know deploy funds and manage and grow in this collaborative decentralized way I think that would be extremely powerful yeah I agree with you

but I have to admit that I'm quite surprising also by the big project how seriously they're they're they're desiring to decentralize themselves and more so willing to do real live experiments with real funds to check that up like I think a lot of the big biggest project in a space right now sincerely look for down solutions to manage their hundreds of millions of dollars they just are waiting to see that technology comes up and once technologies is ready and it's almost ready we are launching the pilot these days and they're actually experiment with real projects like such as North East for example I think then we can explore you know maybe not with two

hundred million dollars but maybe with you know hundred thousand dollars first and a million dollar then but once we can show we can show that we have a technology for for producing large number of decisions effectively by the professional crowd and then we produce much much faster growth and innovation and I think once once project will see that you will see massive adoption so let's talk about consensus the white paper states that the Dow stack uses a consensus model called holographic consensus can you describe this consensus model and how it's different from the ones that we know like rivers sake proof work right so firstly just to just award on

proofs like for it it's a different kind of consensus so they are basically there are two kind to families and consensus actually we can discuss more than two but let's let's focus on two you have you have consensus about I call them objective realities so blockchain is a consensus engine for about objective reality so objective realities are realities that computer can decide you could see if you see yes or no so so for example they whether to add a block whether the block is illegitimate is an objective yeah right program can read that and say yes or no so objective consensus are falling under the title block sheet that's what I don't you is it's a it's a the centrist consensus

engine for objective realities and then you have to there you have prefer for above stake and so on so forth now the interesting stuff or the next interesting stuff is about reaching consensus about subjective realities so for example whether this claim insurance should be you know handled or not whether you know that that article is good or bad whether we want to approve budgeting with ten thousand dollars this or that task so this is subjective because a program cannot read that and say yes or no because by definition the decision is subjective it's not objective the first step was to build consensus engines for subjective realities and that's what falls on entitle the centralized governance

now what is the problem of the sentence governance and it's actually the exact analogy neurologic problem to decentralize coming to consensus off objective it's the same problem which is the scalability problem so just as we have scaled apartment block change you have scalability problem of the server's governance systems what is the scalability problem it's very easy to understand you have one thousand let's say let's make it easy you have 1 million voters 1 million agents in an organization and let's say that there are equal voters so each of them has equal weight no voting one now naively if you want if you want to be resilient like if you if you want to

make sure that the system is not manipulated by a small group of people that the system is well represented by the majority you know the decisions that are made are in line with majority you you would need to require that each and every decision is looked by you know the entire by a majority by approved by a majority organization but of course that completely doesn't make sense I mean you you would you wouldn't be able to produce a single decision and definitely not a million decisions a day or minute decision a month so you see there is a strong tension between scalability the ability to produce more and more decision effectively and resilience the incorrupt ability or non money polity of those decisions by small minorities and even more so having those

decisions be representing really the you know the mind of that the high this hive mind of the doubt so to solve that problem that I think that's was the biggest actual problem of of those apathy today so to solve that problem with design holographic consensus now there are some alternative solutions one of the hottest I think of them is for example VCRs and talking registries but then I can also we can also dig into why I think those pure economic governance models will not work well and the the route of holographic consensus is basically combining two different systems inside a government system one system is the reputation system votes

are weighted votes of our weighted agents votes are weighted by the reputation these systems are highly resilient are not very well manipulated but they have products capability and all the other on the other hand markets like prediction markets for example they are highly scalable but their problem is they're a highly manipulable so we combine those two systems together you have both a prediction engine and padishah market and reputation system and together you can show that you can generate both scalable and resilient interaction at the same time you're basically translating the tension between skeleton resilience into an economic problem that you can always grow if you are willing

to pay more costs and basically the prediction engine so the people who are placing the prediction they are not answering like a future key for example they are not answering what is the right answer like what do you want to be the result but are there actually placing predictions about but what do you think that the reputation system will decide so that's that's that's the core essence of program concepts and in essence maybe just say what it allows you it allows you to make decisions fast decisions by small minorities but at the same time ensure that those decisions will be in line with a greater majority with a doubt majority and the way that it ensures that is that you're basically

creating and a critic anomic incentive to identify possible mismatches so anyone that can identify mismatch between a process that seems to go to that way and what people think that the dow would vote would be that way any way that can identify that mismatch can make profits by predicting that means that this is actually quite interesting could you elaborate a bit more maybe walk us through how decisions are made with holographic consensus like if it's a concrete example of how so if you marry this reputation system and this prediction market and who are the actors basically taking part in this consensus right yeah let me give you a concrete example by the way I also need to kind

of say that this whole graphical sense is really a concept just as much as a blockchain score is a concept and off chain is off chain compeition is a concept actually there is a close analogy between the object reputation and no graphical sensors and I'm saying that it kind of like explained that once you understand the concept you can have many different protocols that's for under that title so it's not another protocol it's a concept now the ballot but but nevertheless let's go through a complete example so let's say that let's focus on the simplistic version which you have a dhow that makes a yes/no decision so anyone can throw on this doll proposals I propose to do such as such you propose

to do such a search and the Dow the collective mind just saying yes yes no no yes no yes now so anyone can open up a puzzle now then there is a reputation system that a way the voting power of agents now when when a proposal is being opened is it open into a queue so just a list of proposals now by default if your proposals on the list on the queue needs if you want to execute that proposal you need to have absolute majority of reputation holders supporting you so actually 50% of the entire reputation of Dow need to needs to say yes in order to execute the proposal this is completely

resilient it's very coherent and unbelievable the progress of course is that it's not scalable so now the next step is that you would like to allow you would like to allow to boost to accelerate the process of decision making at some under some conditions so what I mean by that is that you you would allow you would like to allow decisions to be open for finite time voting so let's say that I take a I wanna I want to boost a decision which means they open that decision for one week and then whoever vote on that decision in that one within that one week makes a decision so if the majority of those voters says yes it's been you executed majority says no it's not it's rejected now that's of course very

scalable you can you can process any number of decisions you want but of course then again it's not resilient because now I can I can attack the system by opening millions of boosted proposals and and then most of them will be remaining i unnoticed because the collective tension would be completely diluted so how do how do you fix that so now there is the second system comes in play while those decisions are in queue there is a second thing that you can do with that and is to make predictions about this about the fate of those decisions so you can come into Institute I'm proposing a decision to budget my task of producing the next feature I need $20,000 and I'm proposing to allocate to me $20,000 now you're coming

and you are looking at that decision and you you have familiarity with that Dow and you're basically making the prediction you're saying okay I know that this guy's has a reputation I know that this this proposal makes sense so I know you know I know that the way that the people here things I know some of the reputation holders I predict that this decision is going to pass and I'm also willing to stake some tokens over that at the station so I'm willing to stake you know $1,000 that this decision will pass in the Dow if enough people look at it okay so now basically when you do that you are putting your your capital at stake so if if eventually you will be right

you will gain more capital more tokens and then if if you will be wrong you will lose your stake now the way that this prediction engine connects with the voting system is that the only the only condition the only way that a proposal can be boosted is if enough predictors are predicting that it's going to pass so that's that's the first instance and then once an inferred for without without too much channel so without too much of others predicting that it's not going to pass so the relative people that think that don't pass with respect

those between is not the past is crossing some pressure and then once it did cross some threshold then the proposal is being boosted and then people are voting on it and still people always can always predict against the concept against the status so if the status of the decisions days ago it seems to be passing you can bet against it by predicting it's not going to pass and if it seems that it's not going to pass you can bet against it by predicting the east to us and then all those who predicted place the predictions they also have incentives to kind of like hold the process and make sure that those who needs to vote in order to reflect the truth that they think they exist to make sure to call them out and make and and

and remind them that they need to vote because it seems that the reality right now is incorrect and then of course the voters place the ball in the subjective we requiring the subjective majority and if it's being approved the the thing is executed in an otherwise and not excluded and rejected otherwise and now I can get into details and the game theory and show that actually this system can and also I skipped you know a few parameters and so and so on but you can see that the system is is fully resilient in to manipulation and you can also see that we can process more and more decision in fact in definitely providing some cost and the cost comes in in the fact that the Dow itself needs

to be put a bounty that is basically distributed to successful predictors for successful proposals so you're basically essentially you're trusting the tension between stability and resisting the cost of the management cost if you wish yeah this is great I you know we did podcast before on future key multiple before with Ralph Merkle and and and Robin Hanson as well and and it's it's a it's a really cool concept but it seems sort of you know hard to make work and maybe not so realistic in the short to medium term but this to use prediction markets the kind of managed you know priorities and you manage to the scale at which a

doll can process proposals and decisions I think is really great and very elegant and seems quite quite simple and and so yeah so that's that's great it would be very interesting to see this in life with with an actual doll cool so let's let's move to to another topic which is a little bit about the sort of economies and economics of this one thing that is written about in white paper is that the Dow stack is based on this concept of circular token economies can you describe what that is sure it's actually not the offense concept it's it's quite

trivial and and you all know it and just we just you know and add it there in from readers that are not familiar with your space very well but basically it's just the regular that economy where you have a token you know a token economy where on one hand you're printing and distributing tokens to convey the economies of value for example in the blockchain it would be miners right but now since you have a general-purpose decision-making engine you can you can distribute those tokens to any computer file I mean someone write is writing so right now in the Bitcoin or in deep you know if someone is writing a code and place a pull request successful pull request he will not be compass you know

compensated by coinbase big right additional Bitcoin but you can actually make it out that print new tokens for any sort of confusion of value just need to agree about what is value and what is not so on one hand tokens are being produced two contributors of value and then on the other side of that loop tokens need to be consumed to consume the value that is being created so for example if it's a blockchain and the value is created is the blockchain itself then if you want to use the blockchain you need to spend the same tokens that were you know distributed to miners then you need to spend them in order to use the kinetic effect that is created by the miners which is the book chain so and that that same circular economy

I mean it's critical to have that circular economy is critical in order to build an echo effect and correctly and I think that every dot every dark economy should have that super economy in place and if it doesn't it has a broken bottle so yeah now we may ask what is the circular economy I guess of doubt slack is it correct yeah yeah let's let's pray please speak about that so there Jen token and kind of how that works in this circular economy so I think I think that's really interesting and and to be honest it took us a lot of time to crystallize that model and I think it's quite hard to build a really robust model so the idea is that

identifying you know what is valuable right now so I you know in in our ecosystem in the telco system the claim is that you know the hardest things is coordination at scale that is the thing that is the thing right I mean of course you need contracts and all that but it's not enough and then I told you that there is a core problem to generate cognition scale the the core problem is that there is a tension between scale and resilience right that's that's that's I'm arguing that is its universal practice of my problems it's any any you know any project will try to tackle that that front will enter the same problem and then I also told you that I think there is kind of universal solution to universal problem this universal

solution is not a protocol it's concept as much as I would say there is a scale problem for for from blockchain and there is a universal solution we called day-off chain computation as universal solution well it's not totally new vessel is also unchanged solution there is non chase solution but it's kind of like almost universal in the same way I'm arguing that it is kind of like universal concept of holographic consensus which allows you to make decisions in small numbers but then I mean 10 rosiness or make sure that those decision in line with majority by crypt economically incentivizing identifying the mismatch so that's programming consensus and I also told you that this or graphic consensus requires to have

certain prediction game right the prediction game of the entire ecosystem is done with the gem token so while you can have many different applications on top of the stack on top of arc and you can have many many different dolls using those many different applications all of the prediction games that that is critical to facilitate to enable large-scale decision making processes is being done with the gentoku kind of like somewhat analogous but in a very different way to the Athenian being the gas for consensus over the chain here Jenny's the gas for large consider making race it's basically the token the staking token for the prediction game over those that existed so if you're gonna have these

different dolls on you know let's say maybe hiriam let's say you have different plasma chains or different maybe on cosmos differ the ether main change that run some of these different dolls right because of scalability you know they probably won't all run on on the fioor main chain how could you then still use the Ghent open across all of these different would you have how would that work that's not critical firstly we can you can still have the prediction game on the main chain but I think you need to because you can have many many different sub Dawkins on the sub chains there I mean when you work when you work on a sub chain doesn't mean that you cannot use either right you can still you know you

can use tokens or the sub chain that reflect the eat there or reflect some other tokens so it's not that the off chain layers will not be interpretive token economics I don't think I don't see or could you have like one chain where it's like a prediction market chain and all of the predictions and economics for all of the different doors that may live on different chains take place there yeah sure we you know if we just are we this current version is already or domain it I mean this current version of the stack is is fully on chain so it's not scalable if there's blocks of scalability while I think you can already use it for some level of organization maybe not millions of

people but maybe thousand of people and maybe not you know daily decision will be weekly decisions and then already they like next to next version we are planning to completely optimize that platform so let's talk perhaps a bit about roadmap and where the project currently sand so you mentioned earlier that you've done a crowd sale and that you'd raised at the time around thirty million dollars so where are you currently at in terms of project development and and when should we expect to see the do is the Dow stack platform launch right so we are launching the stack and

alchemy is the first application I mean literally these days it's it's ready for launch we actually we're just postponing a bit just to be organized after the sale we have a lot of things to wrap up and then write right away after that we are launching the pilot the first pilot so no June it will be already running the pilot of the Genesis tau so the first you know use case will will actually hold the real fun Demeter and let a large crowd to comment and that's fun around the dancer project that's like the pilot of pilot then we want to have like a you know a few iterations on that pilot and also including pilots with some other teams such as I will

tell you about few use cases in particular good gnosis which are close partners for us actually closest we are we are running two experiments one experiment is that gnosis themselves are launching their this centralized exchange a Dutch X and want to have that touch X completely managed by a douse that I will run outside platform and the second experiment that you know crosses themselves they want to reward incentivize developers to build on top of the grossest platform application on top of the post platform with the using their prediction market and then reward them with witness tokens so that that inward process will also use alchemy or at least partially use out to me in

experiment with that two experiments with with gnosis and one experiment in-house with the Genesis fund firstly in alpha and later in the you know kind of like public beta and then we have few other applications already in process of integration so the idea is that the first if you want q3 will be dedicated to piloting and firstly our own Genesis da but then also a few other maybe two or three other projects and mostly piloting and stabilizing the system in terms of you know operation gas cost a scale a protocol parameters vector attacks

sorry attack vectors and so on so forth so stabilizing system thank you for actually launching a you know real real full products again including the chance is fun but with you know ik much more significant amount of funds as well as most experiments and then a few at least one at least one other application this is already in process so that's that's in terms of like I would say six months a timeline in twenty I would say that we would like to have a few live experiments by the end of this year and 2019 our focus would be on one hand to a widespread the

usage of this of the system and by that we're not going to you know develop many many applications ourselves we are putting a lot of efforts and focus on building the technology in a way which is very easily usable integrable by others so whether you need a whole gal for a server or maybe you have your system but you just needs you know component a central's component then you can easily integrate integrate with the stack so yeah the idea if the ideas have to build horizontal a platform an X system for many different integrations so while once one one front to proceed with tonight 2019 next year will be to

widespread that adoption the second will be to iterate heavily on the technology right now the gas costs are still high it's fins on chain I mean we want to optimize that everything is all chain want to optimize everything or most things integrations with you know with with with with other interface other other other tools such as identity you port kind of thing integration with the central databases such as like the affairs integration with external services such as slack and Trello and so on so forth so so maybe you can intern actually separating three into three categories one would be like market

adoption like massive of those right we we actually believe that this would be the killer I mean does this VR potentially would be the killer apps on the theorem so one would be massive a community adoption another would be a heavy iteration of the technology and third would be I would say UX so kind of like integration with existing tools and so on suppose oh well thanks so much Martin that was super interesting I'm really excited about Dallas tack where it is and where it's gonna go I totally agree with you def I think finding new ways of collaborating of organizing of building structures and organizational

systems is one of the most exciting applications of blockchain and I can't wait to see how this is gonna turn out in real life thanks Brian I'm also like really looking forward to see this you know taking shape in reality thanks guys for having me here yeah absolutely and of course we're gonna have links to to a bunch of resources about our stacks if people want to check that out and they can check out the show notes and otherwise thanks so much for a listener for once again tuning in so we're putting out new episodes of episode every week you can scribe the show on iTunes on SoundCloud your favorite podcast applications or you can also watch the videos on youtube.com slash

episode and Bitcoin and if you want to support the show you can leave us an iTunes review and that helps new people find find a show and it helps us keep this going so thanks so much and we look forward to being back next [Music] [Music]