AMA: 3.6.2018 -- Video AMA #2 with Matan Field


#1

Video AMA #2 with @matan on March 6th discussing reputation mechanics, use cases, GEN token features and more.


#2

I. ON IMPLEMENTATION AND USE CASES


Josh: Hello, Matan. Thanks for doing this once again. We really enjoyed last week and looking forward to today. Let’s start with a question that we didn’t get to right at the end of our call last week, about implementation and use cases. This is from CryptoDanny. On the forum she says, or he or she says, "We’re interested in building DAaps on top of DAOstack. Will there be documentation in the library to help us, will there be tutorials? Written or video? And how do you envision collaboration with your busy core team happening?"


Matan: Firstly, thanks for the question and for wanting to collaborate. Firstly, in terms of use cases, maybe just to mention. There are many, many different use cases and probably, most of them, we haven’t thought about yet. Nevertheless, I think from the tens of use cases that I’ve ever thought about or spoke with people about, I could put all of them into four categories. It’s good to have in mind those four categories.

So decentralized applications I am aware of are either of the type of collaboration - so decentralized organization that collaborate and build something together. Open source developers like, Wikipedia-like, etc. The second is basically management of assets. In the first, decision is about the distribution of ownership and power and budget to different agents, different actors that collaborate. The second is about asset management so decentralized VC, decentralized pension, decentralized insurance network. Their decision is about allocating funds for certain purposes.

The third category is curation. Google is a curation engine for websites. Yelp is a curation engine for restaurants. TripAdvisor, booking. All of those, even to some extend Reddit. I mean, some other applications also have curation components to them. Decentralized curation networks is the third category.

And the fourth category is marketplaces. Although the fourth category doesn’t have to have government system in order to operate. So everything that is marketplace is basically peer to peer. Ride sharing is a marketplace. E-commerce is a marketplace. So everything that’s a marketplace doesn’t have to have governance. These are kind of just a map for use cases. Just to have in mind.

Now, in terms of integration, the way that we’ve designed the architecture was exactly to put in front of us this notion of making integration as easy as possible. The whole philosophy was, it’s not that we’re not also building product, but more than we want to build products we want to enable building products. It’s more, I would say, prior. For that we have put a lot of emphasis on the easiness of integration. That’s why we built Arc.js, which basically makes everything that can be done with Arc, which is the core of the governance engine, collective decision making engine, organization engine, this whole piece of DAO that we mention all the time.

So everything that can be accessed on Arc can be accessed through Arc.js in a much simpler way. Both simpler in sense of JavaScript rather than Solidity but also simpler in terms of not needing to understand the architecture of contracts that you actually need to understand when you interact with Arc.

In terms of documentation, for a long, long time we’ve worked on code with none. Just in the recent, I think, two, three months, we put a lot of emphasis on documentation. I can guarantee one thing. The derivative looks very well, which means that I think there is like ten X progress in documentation than we had three months ago. Which is to say that we hope and we’ve tried to make it another ten X in the next three months or so.

It’s getting better and better but also, right now, to me, it’s not as it should be. There is very good documentation but perhaps no perfect “Hello World” guiding tool kind of thing that helps integrate on your own. In terms of collaboration with our dev team, which isn’t that busy, but that’s a good question which I think we should put a more attention to and I suggest, actually, to ourselves. Take it as an action item. What would be the right channel for that? Probably telegram’s not the right channel. Or the forum. Maybe we can make a slack-like kind of channel for that.

Like, people who actually want to directly interact with the team to get more information about the integration. I think that could be a good action item for us right now. Was there another question in there? Remember Josh?

Josh: I think you covered. Tutorials and collaboration with the busy core team?

Matan: Yeah. I think the only way, actually to scale up information is by really. I mean, it’s kind of like an exponential curve, right? Maybe two people know a lot of information then they transfer this information to another four but then these four transfer it to another eight and so on and so forth. In the beginning, it goes slowly but at some point it will accelerate because it’s exponential. I think we are close to the knee of the exponent. Where more and more people understand architecture and everything.

Josh: Yup. Okay, alright, thanks. Just as far as how we’re thinking about this comment-

Matan: Can I add one more thing? Actually, I want to use this as invitation. Anyone that wants to take it as a journey. For example, going with a team, working with a team, deeply understanding the architecture and everything that needs to be understood in order to integrate and then to do the best documentation. We would see that of great value and we would offer in return a great value back. So I would like to use it as an invitation.


Josh: Okay, great. Along the same lines about use cases and implementation, this one is less about app integration but more about using the DAOstack to set up a collective. What are the best resources to learn how to do that?


Matan: That’s exactly the purpose of the first interface that we built. Alchemy. Firstly, to put out an app as soon as possible. And secondly, to direct to a very specific use and specific audience. Although we want to develop a full suite for foundation and management of collective, we started with a little bit narrower use case of decentralized budgeting or asset allocation for open source projects.

That is to say that Alchemy would be that tool to found the collective. It will be partially suitable for that in the coming months and more so, I would say, in the coming three months. Beyond that, one way to start preparing or moving forward with this would be to understand the governance protocols. There is a large dimension of space of governance protocols and I think it’s good to understand the domain and then to start prepare or choose different governance protocols that suit the different collectives.
We are right now mostly designing protocols for large organizations. So the ability to interact with thousands of open source developers. Effectively. But of course, it’s very different if you want to make collective of 20 people that are running a food cooperative, for example.

Josh: Right.

Matan: Eventually, not very far in the future, all of this will be part of the application. You will be able to configure a governance protocol with a few clicks of a button. But right now it’s good to kind of like learn the domain and make the right choices. Interact with any of us to make these choices, learn about the existing resources, and prepare for that. We’ll probably be more mature for these kind of ventures in about three months from now.


#3

II. ON ARC INNER WORKINGS AND UPGRADEABILITY


Josh: Great. Let’s move to a couple questions in the area of the DAOstack and product. Speaking about Arc, “Dawn” asks, "How is Arc upgradeable? Is there a vote? Who has the say in that?"


Matan: Oh, that’s very good. Is it possible to share a screen or not? Let me just pick couple of slides. Can you see that now?

Josh: Yes.

Matan: So this is Arc. Well, this is actually not really Arc but this is the thing that can be built on Arc. This is what we call smart company. Or if you want, the atomic unit of governance. Let me tell you a little bit about how it works. Each component here is actually a smart contract. This is like a bundle of smart contracts and all these together is the governance protocols. Basically, every governance can be divided into Dos and Don’ts. The Dos is basically if such and such would happen, if, let’s say, 50% majority of reputation holders will vote yes on a proposal then the Proposal will be executed. This is a Do. The Don’ts look like, no matter what happens, we’ll never print more than million tokens. Or no matter what happens, we’ll never burn more than 100,000 dollars a month. And then the Don’ts could even define in them, which Dos can change them. So no matter what happens we’ll never burn more than 100,000 dollars a month but governance scheme logic number six can change myself.

For example, only if 80% of token holders vote yes I can be changed. So we call them governance schemes and global constraints. These are the rules of the organization. Then there are the actors of the organization. Again, this is like a governance unit. A DAO can be a large multiplicity of those. The top actors are the internal actors. The above one is the token, in a way distributing ownership of this governance unit.

The second one is the reputation, distributing voting power in this governance unit. And then the third is the external agent. The external actor. It can do anything that can be done with blockchain. That contract, on behalf of the governance unit, for example vote in another contract, another unit. Or call any function on any contract in blockchain. So basically it’s most general.

Then the controller is just the intermediary between the logics and the actors. It also includes the logics that these governance unit is subscribed to. Now there are three dots on the bottom of the controller. These are three functions that can also be triggered beyond the token reputation and what we call the avatar, the face.

The three dots, one of them is the function to change the governance rules, the governance schemes. If the right ones are the governance schemes, the first dot can change them. Basically, which means that there are rules to change the rules. For example, only governance scheme number four can actually register new governance scheme or unregister old ones.

The second dot is the changer of global constraints. Again, there might be particular rules to change the global constraint. Finally, the last dot is the thing that transfers ownership of the actors. These actors are completely owned, it’s called owned in the solidity. Completely owned by the controller, which means that only the controller, that’s the only contract that can command in these contracts and call them and basically tell them what to do.

This controller can decide to transfer that ownership to a new contract. Perhaps a new controller or completely different architecture. By definition, any governance unit has the rules to change the rules but also has the rules to transfer ownership into a new completely new architecture or perhaps a new controller and by that make a full upgrade from the system.

In that sense, the rules themselves called also the rules to upgrade the governance protocol as well as upgrade the technology itself. The architecture itself.


Josh: Okay, alright. Thanks Matan. A related question maybe is, “if developers can contribute governance modules, can you explain how they will be rewarded for their contributions?”


Matan: Right, that’s a good question. There are generally two simple ways to think about it. One of them is already built in the system. The other one will be, I think, developed in the future. DAOstack, is a project that wants to be a DAO itself. The first thing that we are going to do is to found a DAO on top of the DAOstack. We are going to call it a Genesis DAO. It’s going to use the GEN tokens.

Now, this DAO, its purpose is to build the DAO ecosystem on top of the DAOstack. One of the ways that is does so, not the only way, but one of the ways that it does so is by rewarding any sort of I call it anti-rival contribution to the ecosystem. So anyone that makes a contribution of components to the stack or even founding a project that enhance the ecosystem, the genesis DAO would basically, I mean, anyone can make a proposal for any reward, but basically the Genesis DAO will reward that contributor for their contribution of value.

That’s already right now or soon as we found the genesis DAO in the coming months. This is the immediate, I think, way that developers will be contributed. The genesis DAO will basically manage both Ether from the fundraise as well as GEN tokens. 40% of the economy is reserved for that Genesis DAO to manage.

The second way is of course by more market forces but that may come up in the future, where there will be like an inherent business model. Think of the possibility of having this kind of app store of governance modules in a way that if you produce a governance model there is some business model around it. Right now we’re not starting with it.


#4

III. ON GOVERNANCE


Josh: Okay. Alright. Great. This is a question form Clive that is about the DAOstack but also transitions us into more of territory around governance and theory. “What safeguards will be built into DAOstack to protect against the tyranny of the majority?”


Matan Field: Firstly, in any DAO, the best protection from the tyranny of majority is called Fork. Basically, if you don’t agree with the majority you just need to fork out. Now, you can fork a network but you cannot fork a fund, right? However, if a large number of agents that also hold funds or tokens, if they fork they can basically sale their tokens in first DAO and then fork into a new DAO.

I would say that the way that comes in the DAOstack is that, you know, in a lot of previous communities or projects, fork is considered the bad thing. Something to be avoided. We are thinking about it just the opposite. We think that fork is a good thing. We should make fork as easy as possible. I’m not speaking about fork at the blockchain level, I mean, that can also happen but it’s not our deal. I speak about the fork of a DAO.

So if there is a DAO, we would like to promote the forkability to be an easy function. It doesn’t mean that the thing will dissolve fastly because if you fork and nobody follows you, clearly it’s not effective. But if you fork and half of the DAO or even one tenth of the DAO is following you, then you should fork. I think that’s the best protection against tyranny of majority.

Beyond that, it’s basically really acknowledging that there are different values systems and they are both to good live in parallel to each other. There’s no reason why to enforce one or the other. They can even cooperate. I think we can stay with that.


#5

IV. ON TOKENS AND REPUTATION


Josh: We do have a number of questions coming up about reputation and I thought maybe first you’d like to say something about token and reputation.


Matan: Sure, so token is a fungible token. You can transfer it. If you own it, you own it. Nobody can take it from you. It can reflect anything so, if it has the market, if it has a liquid market then, in a way, it reflects cash. On the other hand, if it’s tied to a business model of an organization, then it’s kind of like reflection of ownership in a sense but not really. If it reflects usage of an app it reflects utility etc.

There are a lot of different ways to reflect value with token so there’s no one singular way. Reputation is similar to a token but it’s not fungible, you cannot transfer it. So you can accumulate it on an address but you can never transfer it to another address. Then also it can be taken from you. If the organization decide that some agent should be taken reputation, they can do that. It’s like token in a sense that it keeps it’s balance, there is a balance to each agent. One agent has six and other has eight, etc. so there is a balance that everyone can see but it’s not transferable and it’s not protected.

The organization that manage that balance can also take it away from you. These are two differences. Generally, we think about reputation so maybe it’s a little bit confusing. For a long time we tried to come up with different names, also inviting anyone to think about it. But basically when we say reputation we really have in mind something like voting power. Power inside governance unit.


Josh: Maybe you’re already answering a question that Ezra had. Ezra, I see you have a live question but first going to ask Matan something that you left ahead of time because it’s relevant. Ezra says, “Reputation seems like an endless positive feedback loop, reputation rich people will be incentivized to vote for decisions that will grant themselves even more reputation and they’ll have the power to do it. Too much lopsidedness here could even endanger the decentralization of a DAO as it would, in effect, be centralized around the few super-rich reputation holders. Possibly a solution to the could be limiting how much power reputation can provide.” Any thoughts on that Matan?


Matan: Yeah, firstly I want to say that, I mean, probably it’s not the first … It’s definitely not the first time, probably not the last time. The reason why we built Arc the way we built it is just because we believe there is no right, single protocol. There is no such thing. Different organization will need different protocols and more so, protocols will need to evolve over time and fine tune themselves.

Yeah, some organization might limit reputation and that’s why we built Arc the way we built it. Because it’s very easy if you just want to add that it’s super easy. It’s almost trivial to implement it with the architecture that I showed before. Probably some organization will limit reputation, some others will not. It’s true that it might create a feedback loop. I would say in different words that basically, it will trigger, or it will move towards increasing alignment. Those who are aligned with the majority will become more aligned, or I mean, will have more reputation, and that’s more important in the consensus and this increasing alignment.

Those who are not in line with the majority will have less reputation and thus will basically pushed out from the alignment. That’s not necessarily bad because, again, if you are one of those who are not aligned with the majority you can just fork out and just have a new value system where you are the majority, initially, and other people are following you or aligned with you. Then there is a multiplicity of vectors, multiplicity of value systems that really think differently and then compete in a way, with each other, about achieving goals in the world.

Maybe they want to achieve different goals or they want to achieve the same goals in different ways. In that sense, yes, if you are creating increasing alignment it’s not necessarily bad because then, forkability will create multiplicity of increasing alignment. We can discuss even more that.

The other force that is balancing increasing alignment is growing the organization. The more people are coming. On one hand you are creating increasing alignment but on the other you get more and more agents that think differently so it’s kind of like balancing out these forces as well. Then again, some organization will decide, maybe because they don’t want to encourage forkability and they don’t want to encourage also increasing too much homogenization of the organization and thus they will, for example, limit reputation. Yes. You can do it in many different ways.


Josh: Couple of questions that are similar. “Can reputation be staked since it can’t be transferred?” Somebody else asks, "Can someone vouch for one another, thus transferring, yet at the same time staking his reputation?"


Matan: I’m not 100% sure what’s the intention of the question. What each of these words meant. Again. That depends on protocol. Different protocol behave differently. Under some protocols that will be available on the stack. When you vote, you actually stake your reputation. Whenever you vote on something, you put your reputation on stake. If you are voting according to the majority then your reputation will increase. If you are voting not according to majority, your reputation will decrease. In that sense, you are staking your reputation.

In terms of vouching, I’m not completely sure what is the meaning of that. I mean, if you’re asking if you can delegate. If the question is whether you can delegate then, definitely, yes, you can delegate to someone else your reputation. It’s not transference of the reputation, just delegating the power under certain condition or certain time frame or whatever is decided.

Oh, maybe the meaning behind vouch is that someone doesn’t have reputation, then I vouch for him and he gets reputation. Again, that will be defined in the protocol of that organization. However, know that just me vouching someone else is not a resilient protocol. I mean, I can vouch to another agent which is me and then I vouch another agent which is me etc, etc. and just increase my reputation.

If that’s the meaning then it’s a really good question. Some organization would like to build vouching mechanism. It’s a challenge to build a decentralized vouching mechanism that’s also [inaudible 00:31:16] resistant. So resistant to this kind of attack that I just mentioned. But it’s possible. Yeah. Some protocol, for example, will have this kind of decentralized vouching and [inaudible 00:31:27] resistant protocols.


#6

V. ON PROPOSALS


Josh: Okay, so Julien asks, “If I submit a proposal and I have to pay in GEN for this proposal to be reviewed by the community, what’s the advantage of submitting a proposal if I have to pay?” And maybe even saying something about GEN as relates to submission of proposals.


Matan: Yeah, there is no need, at least right now. And again, that may vary, but with the decision of people, but basically, right now there is no need to pay GEN for submitting a proposal. So submitting of proposal is free. Again, that can be changed or decided differently by any organization, but right now the default is the submission of a proposal is free.

So now the question of what is the role of GEN? In order to answer the question maybe we need to understand the role of GEN more deeply. Basically I described to some details, not fully, last time, the problem of decentralized governance and the solution we are suggesting and that we call holographic consensus. Let me try to summarize it in two sentences. Organizations are this kind of engine to make decisions, right? And by that coordinate a large number of agents into co-action. When you speak about decentralized organization, you think about distribution of power for those votes. The probably is that if you distribute the power to influence the vote then if you demand too much of majority, too much of attention, collective attention, to approve each and every decision, then you have a resilient system but it’s not scalable. It cannot grow effectively the organization and process more and more decisions.

But then, if you only require very little reputation to approve each and every decision it’s very scalable but not very resilient. Small group can basically kidnap the DAO if you want or just make wrong decisions like decisions not aligned with the DAO. On the other hand, markets are very, very effective. They scale very well and they are becoming more and more effective as they grow. Just the opposite from organizations. Because of that people realize that maybe you can relate markets and decisions and make like that decision, effective decision making. And people have explored that using prediction markets for decision making. It’s called Futarchy.

The probably with Futarchy, as I see it, is two fold. Firstly, because markets are pure economics, it cannot capture different value systems. Some value systems would like to put in values, not just value. Like, different values and pure economic markets cannot capture that well. That’s the first downside. The second downside is that markets, well, they are more manipulable. Because you can purchase tokens and then you can manipulate the market. That happens all the time.

In a decentralized organization, there is no regulation. So if it’s manipulable, it can be re-attacked. So the point is to realize that in governance, there are tow different functionalities. There is the decision making and then there is the sense making, the signaling. Market are really good in sense making and signaling but they are not very well for decision making. Organization are very good in decision making but are not well in effective signaling so what we did, we separate the two. There is both governance that is decisions that are made by voters, by reputation holders, and there is prediction markets.

There is a prediction market, basically, for each DAO of people that can make prediction about the fate of proposals but those predictions are made by agent does not need to be part of the DAO. So anyone could come into any DAO and make a prediction about any proposal. By that, he is willing to put a stake on it. He’s willing to put 200 dollars that this decision will pass or fail.

By that, firstly, if he’s a good predictor, he’s making profit. If he’s a bad predictor he’s losing his money, his stake. But by doing so, basically he’s signaling the network for the interesting proposal. For the interesting decision and by that making the collective decision much more efficient. It can navigate more securely and effectively in order to derive the decisions.

It’s both serving the DAO and serving the people that are doing that. That’s because of the holographic consensus. This whole protocol I’ve just hinting for, I’m not describing the whole protocol. It’s important to say that for all of the DAOs there is like a single universal background of prediction market going on top of them. This whole prediction market is basically done in GEN. You don’t need to pay for making a proposal but if you want to accelerate your proposal you can promote it by putting a stake and it’s a stake because if eventually you are right, if eventually the proposition will pass, you will even gain more back.

You can think about it also a different direction, you can think about purchasing the collective attention. But if you are purchasing collective attention, then using it improperly so that eventually the proposal doesn’t pass, then you are losing your stake. This is the use of the GEN token.


#7

VI. ON GETTING INVOLVED


Josh: Yeah. Just want to make sure, did we fully answer that question? I think so. Okay. Great. Let’s then talk a little bit, just taking a live question here from Ezra. "I’m not a developer, and investor, or a journalist, and there’s no way my company will be converting into a DAO in the near future. What is the best way for me to get involved in DAOstack and the movement it represents in general?"


Matan: Well, you named some of the core contributions, investment fund, developing, writing articles, but you can participate in any other functionality, right? There are few more functionalities in the system and basically, any of them. For example, if you really believe in the project and want to spread it, just telling your friends about it is just maybe perhaps the best thing that can be done.

Being on telegram. Asking questions, actually, just asking that question is also contributional. So asking questions on the forum. Any sort of education. You’re educating yourself or educating others is super useful. Sometimes the best contribution of someone is just bringing another person that is just the perfect person to the project. Suggesting ideas, learning.

I think, eventually, everything starts from learning. If you are in the process of learning, then by definition you are also in the process of learning how you can make many contributions.


Josh: We’ve gotten a similar question in the past, Matan, and one thing that you’ve said is that small organizations or individuals that are not in position to form themselves as a DAO are still very eligible to participate in the DAO ecosystem and the DAO mesh.


Matan: Sure. As a user, you can definitely promote DAO ecosystem as a user. And DAO, maybe it’s early days, well, we haven’t yet launched. We launch these coming months. Even when it’s launched, I assume that in the early days it will be more for early adopters. We’re actually aiming to make it pretty useful also for people who are not developers. Yet, I am ready to … I am taking assumption that early days are more for early adopters that have maybe a little bit more tech background but not necessarily. Definitely in the very near future it will be accessible for a much wider audience. One great contribution, just using the system in different ways.


#8

VII. ON INTEROPERABILITY


Josh: Okay, so this question is about the DAO mesh and the interoperability of DAOs. “Assuming that DAOs can use schemes, offers, or proposals, or whatever we decide to call them, to make deals and communicate with one another, what is the actual advantage of this DAO interoperability versus two employees at different legacy organizations communicating with one another?”


Matan: Yeah, that’s a good question. Firstly, let me just mention that when we speak about interoperability, there are two levels just to look at. There is the interoperability of DAaps and the interoperability of DAOs. There is the stack, right? There is the Arch and the ArcHives, the registries and the Arc.js. And then on top of that there could be a very large number of Daaps. That should be interoperable. This is that analogy to what happens in the internet, right? Nobody tried to build the internet. People just built HTTP. And then the internet emerged.

Just as all web applications on the internet are by definition interoperable because they are all communicating with one universal open HTTP protocol, in the same way all different collaborative decentralized applications on the stack will be interoperable thanks to Arc and also the ArcHives. So it’s two different layers of interoperability.

Then there is a second notion of interoperability which is about DAO interoperability. As you mentioned, like, how different groups could communicate and make business relationship. Now, true, this is not, philosophically if you wish, it’s not different from different businesses in the world making communication. What makes it different is a couple of things. So, firstly, because of the fact that the rules of the DAO are pretty much programmatic. Not pretty much, are totally programmatic. So you can have a DAO and within DAO there are, like a fractal structure. Like companies and sub-companies and sub-sub-companies and it doesn’t even have to be a tree structure, it can be a totally generic graph network. But each node can be autonomous. So a unit over here, even though it’s part of a larger DAO can just decide to interact with a unit over here.

That’s not exactly like, if you are a department in a big organization, not necessarily you would be able to, regulation wise, you would not be able to maybe interact with another department of another large organization. Again, that’s because of the centralized structure, right? Everything critically needs to go through the center to be approved and go back. That’s why it doesn’t happen. Also there is no strong way to align interest, where here the whole intention is for different DAOs to actually have strong alignment of incentives because they are part of a bigger DAO ecosystem that is constantly promoting anti-rival economy. Anti-rival behavior.

That’s one way to think about it. The second way to think about it is simply because you have, like, a whole internet of work it also makes the discoverability, like 100x more efficient. So today, I mean, there are like tons of units all over the world but you are not aware of that. There is no Google for the world of work, right? But here that is exactly what’s happening. We are building an internet for work that will also be that Google of work. So discoverability will be highly enhanced. In that sense it will also be different. I think it will make a big difference in terms of interoperability of workforces.


#9

VIII. ON CURATION NETWORK MARKET


Josh: Part and parcel to that discoverability will be curation. So let’s go to this question from Capetown9: "From what I understand, with DAOstack it would be relatively easy to set up a curation market?"


Matan: Yes. Basically the whole purpose of DAOstack is to make it easy to build any sort of collective decision engine. However, I will say two things. Firstly, curation market is a hot topic these days. But on the other hand, there is no one definition that goes under curation market. There are many different things that people call curation markets. So that’s one comment.

Secondly, there is curation market and there is curation networks, which behave differently. Again, as I said, markets, if it’s literally a market, a pure market, then by definition it has no collective decision making. It makes decisions purely economically by the market. It is purely peer to peer and in that sense, you don’t need, really DAOstack as much as you would have otherwise.

If it’s a curation network that can make collective decisions about curation, then absolutely, yes, you can easily make it in DAOstack and you probably need something of this sort anyway. So this is like difference between curation networks and curation markets. And more so I would suggest, just as I said about the holographic consensus that for a long time people have looked at just consensus or just economics, Futarchy, and I argue that the solution is actually a hybrid of them. The same thing I would say about curation market.

I think just curation market or just curation network will not be… one would not be effective and the other would not be resilient. And the right solution would be a Curation Network Market. Basically, the analogy of holographic consensus for curation.


Josh: And so, just to reiterate the question specifically, would you say that with DAOstack it will be relatively easy to set up a Curation Network Market?


Matan Field: Yeah. More so, I would say that the whole purpose in that it it’s been built in a modelar way, in general way, such that I would assume that, although we haven’t yet done that, that’s definitely on our roadmap, but I would assume that the structures that are already put in place for collaboration will almost be identical or easily tweakable to get to the curation category. Yeah, it’s almost given, I would say.


#10

IX. ON EARLY ADOPTION


Josh: Yeah, related then might be something like Travis’s question, "What would be the advantages of beginning with the stack when the use cases and the ways that teams and nodes operate and interact isn’t yet clear but is rather being defined?"


Matan Field: Yeah, I would think that this is the question for any new technology, right? The downside, of course, is that there is much more friction. The upside is that you have the huge advantage of first mover. You could say the same thing on Ethereum, right? Like, those who tried to code anything with Ethereum a couple of years ago, three years ago, it was super hard. It wasn’t clear what can be done or how can be done. There was no integrated environments. It was super hard.

Then, also, the progress was very slow. Then, of course, they got so much educated that when things got easier they had a huge advantage and nowadays they are leading the space. I would think that, that were true for any technology.


Josh: I love this answer but I also think that Travis is actually asking something slightly different. I think he’s asking about when, in his own organization, the use case, the ways that teams and nodes operate and interact isn’t yet clear. Is it still, does it still make sense to begin on the stack?


Matan: I see. Thanks for the clarification. That’s a good question. I would say that the whole power of programmatic operation is that there is much less to discuss about. There is a rules for the game and other people want to play with it or not. If you start an organization with those rules and if people don’t play, they don’t play. Then, you know, you are left with a smaller organization. If people play there is nothing to discuss about.

Basically, ideally, it’s kind of like stigmergy: ants are not discussing how to build their way from food to nest. They are just algorithmically cooperating. That’s kind of like the same. I would argue that small scale corporation and large scale corporation are dramatically different, physically different. I say physically different, I mean having different rules of physics, laws of physics. So I think that basically small scale corporation should be social and you know, lot of discussions, let’s say. And large scale corporations should be asocial. It’s more like economics. It should be no discussion. Should be a lot of people expressing their opinion and something happens. In that sense, programmatic operation fits, we don’t need to define, you don’t need to understand anything. You can just start it and either people play the game or not.

And then anyone can say, okay, but I don’t like those rules of game. I am make new rules of game and great, now we have two core organization. Maybe later they can merge. Any sort of evolution can happen.


#11

X. ON DIVERGENCE AND CONVEGENCE


Josh: Okay, live follow-up question from Cleave and we’ve got two more, this one included. Thank you for your earlier answer. This was the answer in which you mentioned safe guards and tyranny and forking. The question is, "In situations where a DAO has forked, is there a mechanism for them to rejoin one another? For example if a minority solution is proven correct, can the majority dissent be reversed?"


Matan: Yeah. Definitely. Yeah. Any sort of logic that you can imagine can be implemented, right? If you just said that in terms of, in words of rules, that those rules can be coded. And are easily coded on the stack. That’s why it’s built modularly. Yes, the answer is yes. You can, just as you fork organization, you can merge organizations. And you can combine them in different ways. You can combine them in a way that just add, like a total sum, you could combine them, you know, in different combinations that maybe keeps some of their autonomy … Think of these two organization, you can combine them in a way that there is a larger organization that both of them become a subsystem of the larger, or you can literally just combine them into one organization.

There are more ways you can think about it. Answer is yes and everything you can think about, just know that you can translate to words and you can translate the words to code.


#12

XI. ROADMAP AND CUSTOMIZATION


Josh: Okay Matan, last one for today. From LivingBeings, "When will DAOstack launch in a form that I can use to build a custom DAO?"


Matan: We are launching this month, I mean, we are launching a pre-pre-MVP this week, on the main nets. This coming week. By the end of the month I believe we will have, like, an MVP version. For some time we try to think whether to firstly go after the customization of DAO or firstly about the usability of fully customized DAO. Initially, we went with the first direction then we turned to the second direction. Now we are in the second which means that the first application, you can still open a DAO but it’s fully customized. Like, you cannot find it in anything. There is one kind of DAO that you can build. One protocol. Then that DAO can do some things that we think are already valuable for real organizations. Real, existing organizations. Basically decentralized budgeting of open source projects. Then the next step is exactly, well, one of the next steps is exactly to add this configurability. I think it’s the very near future. I speak about one month ahead, two months ahead, three months max.

The beauty of it is that the decentralized budgeting engine, Alchemy, we can use it to build more stuff. You think that configurability, actually, I will agree with you, that configurability is one of the most important feature of next. You can just make a proposal to the DAO that will be founded on Alchemy with real budget, with real money and you can make a proposal: I suggest to do the next feature, to do configurability and to allocate such and such funds to that.

Maybe I am also willing to build that, etc. Then, if that approved there is also a budget for that and people will start working on it. That’s exactly what I’m planning to do. To just make a lot of proposals. For next features. Once Alchemy is launched on the main net with the real money.