 First question comes from Luca, your thoughts on DAO software. I've been recently testing out Aragon, and yesterday I met the wet-onomy guys. They all seem to bear great promise. When I think of what this can become in five to ten years, it blows my mind. They appear to be early-stage, so it's probably too early to use their software. Do you know of other similar projects? What's your general view on DAO software? I'm honestly not that up-to-date on the various software projects out there. I do know Aragon is a governance project for running DAOs. I know that a bunch of other projects are using it in order to build governance into their own applications. This is one of the interesting things that's happening in Ethereum, which is the maturation of these composable services. People are building special-purpose, single-focused smart contracts to deliver a single service. Those can be composed by other smart contracts into more complex services. For example, there are stablecoins that are decentralized and backed by a reserve of Ether. There are governance projects like Aragon, reputation projects, and identity projects. A lot of Ethereum projects are now creating these composable applications, where they put together the best of breed of each one of these categories in order to build a new application. I'm very excited about DAO software. I think I've said this quite a few times before. To me, governance and the DAO, if you like, or decentralized autonomous organizations, this concept of building organizations that are governed by the members, or that have a degree of autonomy based on the rules in the smart contract, or some mixture of autonomy and governance, I think this is the killer app for Ethereum. I've always thought that Ethereum primarily enables these rich governance applications. It may produce other things too, but a lot of the talk in Ethereum today is about decentralized finance. If you look closely, I think a lot of the decentralized finance projects are more about governance than anything else. There are financial instruments built with decentralized governance. Very interesting stuff. The concept of a decentralized autonomous organization, or any kind of decentralized organization, adhocracies, or special purpose project-based collections of people who work together, or systems of software that set governance rules, all of these are fascinating topics. They're not mature yet, and that's one of the big caveats. You've got to recognize that in a fast-moving space, a lot of these projects are still experimenting. But that's the interesting thing about this space. There's a lot of experimentation going on, and we learn a lot from each one of these experiments. I'm trying to follow these as closely as possible, and learn as much as possible about these projects. I'm looking forward to seeing more. Kevin asks, Hi, Andreas. What do you see as the biggest obstacles for decentralized autonomous organizations, DAOs, right now? I'm particularly interested in solutions to the civil attack problem. To me, it seems like a practical solution to that problem will open the door to a lot of really interesting concepts, like on-chain elections and universal basic income. Do you agree with the importance of this issue, and what other issues should new DAOs be focusing on? Great question. The civil attack problem, which you can actually read up on by reading Sibyl on Wikipedia, is spelled S-Y-B-I-L. The civil attack problem is a problem understood in distributed systems that predates Bitcoin. It is the fact that you can create an unlimited number of fake virtual computers online. Any kind of mechanism of truth, or voting, or any kind of mechanism where you're using some form of identity... to count something, or distribute something, can be easily gained, where one person can create a thousand, a hundred thousand, a million fake nodes, or fake IDs, or fake avatars, or fake Twitter accounts, or fake whatever. And then basically break the distributed system. So let's say, for example, you're doing on-chain elections. Well, if I can create thousands of voting identities, and you can't have any way of verifying those voting identities, then I can vote a thousand times. Universal basic income. If I create lots of fake people, I can get paid universal basic income lots of times. Within decentralized systems like Consensus, the reason we have proof-of-work is because proof-of-work... is a direct solution to Sible problems, which is, how do you know the miners are real? How do you know someone's actually put any effort into mining? And the way you know is because there is proof of the work they did in every block. So that means you get the one CPU, one vote, or in this case one ASIC, one vote. And that solves the Sible attack problem, because you can't spin up a miner without proof-of-work, and therefore you can't create fake mining. Your mining has to be real, because it has to contain this proof-of-work. So Sible attacks are a big problem. They're a big problem in a number of distributed systems, and so far one of the best... there's basically two ways to solve that. One is centralized. So I mentioned previously elections or universal basic income. How do we solve Sible attacks in things like social security systems, or welfare systems, or elections? We solve those problems very poorly through centralized identity verification mechanisms, like providing a national ID, or dipping your finger in indelible ink, or some kind of biometric, or something like that. All of these are flawed systems, because with biometrics, the biometrics have to be checked against the database, and that database can have multiple fabricated entries in it. You could have, with identity, if you compromise the identity mechanisms, then you can create fake people. You can have dead people who are apparently alive and voting freely, as the joke goes in some parts of the world, where every year the number of people who vote far exceeds the number of people who are actually alive in the country. You can do all kinds of things. So even these centralized solutions, which are very costly to maintain, are subject to Sible attacks. And of course with decentralized solutions, it's even worse. If you can't see the other person, if you can't have any form of verification, and you don't want to use a centralized solution, then this problem becomes very, very difficult. I can spin up thousands of nodes on cloud servers to pretend to be doing things. If you want to do a vote or poll or a survey on Twitter, or you want to troll people, it's very easy to spin up thousands of accounts and then vote thousands of times, and skew the outcome of any survey poll, voting mechanism, or something like that. So that's one problem with a DAO. Now, so how do you solve the Sible attack? Interestingly enough, one way you solve it is proof of work. So theoretically, at least, you could have DAOs where in order to participate in the DAO, you would have to submit proof of work. I imagine basically an autonomous organization that's based on a decentralized system, some kind of network, that is used to do some kind of function, let's say payments. And in order to participate in this decentralized autonomous organization, you have to demonstrate proof of work through mining. Wait, I just described Bitcoin. Well, Bitcoin is a DAO. So once you realize that Bitcoin is the DAO that solves the Sible attacks for proof of work, this all suddenly makes sense. We don't have other solutions for Sible attacks. One other proposed solution is to use proof of stake. So if you put real money up, then it's difficult to fake having multiple participants. But at the same time, then it means $1 one vote. And $1 one vote is a system we used to have. It's called aristocracy. And it's not a very good system. In fact, it would go even worse. It would become a full-on plutocracy. And so, again, this is a difficult problem to solve. I think an equally difficult problem to solve with DAOs is the oracle problem, which you might find interesting. And the oracle problem is the fact that a lot of the most useful things you want to do require you to get information from outside the chain. And then how do you verify that that information is correct? And how do you make sure that incorrect information can't be injected into the DAO? That's a very difficult problem to solve. And of course, the kinds of platforms that are experimenting with this are extremely interesting. It's one of the reasons why I'm interested in Ethereum, is because it provides a platform for experimenting on exactly these questions and trying to find solutions on these questions that is happening at scale with real money at stake and adversarial conditions. And that's not really happening on any other chain.