Decentralized Autonomous Society

What follows is an edited transcript of Decentralized Autonomous Society, a conversation between Vitalik Buterin, inventor of Ethereum—a platform and programming language for cryptocurrencies and decentralized applications—and Triple Canopy senior editor Sam Frank that occurred in December 2014 at the Swiss Institute in New York. Buterin and Frank speak about Ethereum and cryptocurrencies such as Bitcoin, the theory behind cryptoanarchism, and the lessons of the online drug market Silk Road. They go on to speculate on the potential of Ethereum to provide replacements for corporate and governmental structures and services ranging from Dropbox to Social Security, and, ultimately, on the possibility of a totally decentralized future built on the blockchain.

Sam FrankI’ll begin at the end then move back to the beginning. On Reddit, someone asked Vitalik about the future he sees coming out of the technology he’s working on. Vitalik described a scenario:

“You wake up, and see that $17.27 was automatically deducted from your primary wallet, as you had authorized to happen every day, to pay the rent for your apartment; if you canceled the authorization, then, after a warning period, ownership in the land-registry contract would automatically transfer back to the landlord, and the door lock would no longer recognize signatures signed by your smartphone’s private key as valid for letting you in. Of course, your landlord is bound by the same restrictions. If he shuts off his account that pays the local government $6.60 land-value tax per day, then he loses ownership and the contract automatically switches over so you are renting from the government instead. The government itself is simply a large decentralized organization, and you can see in real time the $6.60 moving on the blockchain and eventually getting into an account to pay for a medical-research program trying to extend the human lifespan from 170 years to 230.

“The Internet that you are using to access this information is based on a decentralized and incentivized mesh-networking platform; you paid $0.0009 to access the information, but your laptop also earned $0.0014 transmitting other people’s packets at the same time.

“You get into your Mastercar self-driving car to go to work (originally, all self-driving cars were made by Google, but Master Corporation, a decentralized autonomous entity that automatically uses a combination of futarchy and liquid democracy to determine how the company should spend its funds each day, proved that its governance mechanism was so efficient that it overtook Google on some core services within three years, and alt-Mastercorps took over most of its other operations). You get in, and Mastercar runs a optimized version of the A* search algorithm (for which James Wilbur automatically got a bounty of $782,228 worth of MSC from the Master Contract) to determine the optimal path to your primary workplace. Given that your self-tracking app has detected that you value your own time (or, rather, the delta between time spent in a car versus time spent at home or work) at an average of $14.18 per hour, the Mastercar’s algorithm chooses a route that takes an extra eleven minutes in order to avoid road tolls and also on the way moves a shipment from one side of the city to the other. You drive out, and thirty minutes later you have spent $1.04 on electricity for your car, $1.39 on road tolls, but receive a reward of $2.60 for moving the shipment over.

“You arrive at work—a location which is a hybrid living/working space where “employees” of five different alt-versions of Master Corporation are spending most of their time, except that you chose to live at home because you have a family. You then get to work, running simulations of a proposed new scalability algorithm for the now community/DAO-driven Ethereum 6.0.”

So this is the future—maybe Google will get us there—but now I’ll give a little bit of the start. Vitalik, you’re twenty; you moved from Russia to Canada when you were young, and you’ve programmed since childhood.

Vitalik ButerinSince I was about ten or so. For half a decade I mainly programmed video games that I ended up playing myself immediately after I wrote them.

FrankWhen did you realize that Bitcoin was important and interesting, and not just a fake currency for nerds?

ButerinMy dad told me about it. And my immediate reaction was, “Well, it has no intrinsic value, it’s backed by nothing, so it’s never going to work.” I just ignored it for a few weeks. Then I heard about it again from a completely different source, and if I hear about something twice it’s probably at least somewhat legit, so I looked into it. Once I started, I got in pretty quickly.

If you use email, you know it just works well. You type in your message, you hit send, and it gets sent. The kind of online financial system we have today only kind of works. From a cryptographic perspective, it’s backward. What you’re doing when you’re paying with your credit card is giving the merchant the equivalent of your private key. You’re giving them all the information that they need to make payments on your behalf themselves. The way payment works is that instead of your typing in “OK, here’s where I sent the money to and how much I sent,” you send this form authorizing them to do whatever they want. Hopefully they won’t do anything too nefarious. With Bitcoin, here are the fields where you put in whom you’re sending to and the item you’re sending, and here’s the send line. You click it, and that’s it. It makes sense. It’s exactly how a financial system should look. I also liked how it’s more developer friendly. It doesn’t stop you from actually hacking into the lowest-level components. Some people like to handle things like cash and gold because they feel tangible. To me, Bitcoin feels tangible in the sense that I can touch it on the command line.

FrankPeople have been writing about the decentralized database called the blockchain, and about the idea that Bitcoin is a way of establishing consensus without a third party like a bank or a government.

ButerinA blockchain is a database or ledger that updates itself according to certain rules. In the case of Bitcoin, the database stores how many bitcoins everyone has. The rule is that if you have some number of bitcoins, then you have the ability to reduce that number by some amount—but not by more than what you have—and increase the balance of someone else’s account by the same amount. It’s a mathematical formula that’s defined by three lines of computer code.

On the financial side, going beyond just Bitcoin the currency, you could build various sorts of economic platforms using a blockchain. If you look at Kickstarter, where someone wants to raise money, and the way it works is if they get to the goal by the time limit then they get the money, but if they don’t get there then people get all the money back—once again, this is a very small amount of programming logic. You should be able to write a platform that lets you do that without too much difficulty.

FrankYou were trying to build on top of Bitcoin and you got frustrated, right?

ButerinI was working on a project called Colored Coins—a way of creating your own asset and using the Bitcoin blockchain to track it—and another project called Master Coin, which was trying to build financial tools on top of Bitcoin. And I noticed that these different approaches to extending Bitcoin’s functionality were trying to increase the number of features. Imagine if you had a calculator and then people realized that, oh, hey, people might also want to use the device in their pocket to listen to music, so let’s take a calculator and tape a Walkman to it, and oh, by the way, people want a clock, so let’s tape a watch to the side as well. Meanwhile, Apple comes out with the iPhone, which can do just about a million different things. Ethereum is the idea that, instead of taking a calculator and taping a Walkman to it and so forth, you just have a computer, and that computer runs arbitrary program code, which you can use for whatever sort of app or feature you want. Bitcoin was never designed to be the foundational base layer for everything.

Frank“Everything” meaning “decentralized computing”?

ButerinDecentralized computing, decentralized name registry, asset management, and so forth. I kept on noticing various technical issues. We were trying to strap on functionality to something that was never meant to have functionality strapped on. People sometimes call Bitcoin the HTTP or TCP/IP of money. HTTP and TCP/IP are foundational base Internet protocols that are behind just about everything. But Bitcoin is more like SMTP, which is a protocol for email. Imagine that email is the first thing that came out, and people started trying to implement PayPal inside of email, or Netflix video streaming on top of email. People would be sending thousands of emails back and forth, and then you’d have something that would be trying to translate those emails into a movie. I realized it would make sense to build something—which I named Ethereum—that’s properly designed to be a foundational platform that makes it possible to do just about anything decentralized.

First of all, it makes sense to have a new, non-Bitcoin blockchain for it. Second, instead of continually tacking on new features, it’s better to have a blockchain with a built-in programming language so you can write programs that go on the blockchain directly.

FrankThat was about a year ago. In the past year, you’ve had a multimillion-dollar presale of ether, which is the currency of Ethereum, but it functions more like gas or oil because it pays to run cycles of computing. Ethereum hasn’t been launched yet, but the blockchain, the genesis block, is going to launch soon. When Ethereum launches, what is it going to be?

ButerinEthereum will have a browser. You can think of it as being similar to a browser that you use to access the Internet. The difference is that when you’re looking at a webpage on a web browser on the centralized Internet, your browser sends a request to the server—“Please give me the website to Google.com”—and Google servers reply with the data that gives the website. There’s HTML code that depicts what the website looks like, and there’s javascript code that dictates what the website does. You click search, and then there’s javascript that takes what you search and sends the search query to Google, and Google replies with whatever the results are. With the Ethereum browser, basically it’s all the same except instead of javascript calling the server, javascript can call the blockchain.

FrankLet’s say that people start developing for Ethereum and increasingly use it. A few years down the road, what does that look like? Are we going to re-create the Internet we have now, except decentralized?

ButerinI can see services emerging that replicate what Dropbox does, what Uber and Airbnb do, and so forth, but in a decentralized format. The benefit would be that you don’t have a company that charges 20 percent. You could argue, of course, that if these monopolizing middlemen are charging 20 percent and then a decentralized solution comes out that does equal the functionality at the cost of 3 percent, the centralized middlemen will move their fees down to 3 percent, and they’ll still be around. But people like decentralized systems not only because they charge 3 percent and have these particular rules now, but because they commit to having those same rules forever. There are a lot of cases of companies being nice while they’re the underdog and then once they get large enough they just don’t need to care all that much anymore. Here, you know what the rules are, and you know that the rules are going to continue being the rules for the foreseeable future.

FrankYou have an idea that a lot of the functionality of the financial system—which is quite opaque in certain ways—could be replaced with little bits of code that are much more transparent and affordable.

ButerinThe financial system is very complex, and not many people understand exactly what’s going on. It’s not really clear under what circumstances some financial derivative is going to collapse and you’ll end up with absolutely nothing. With centralized stock exchanges like NYSE and NASDAQ, high-frequency-trading companies are setting up servers right beside the exchanges and have a 200-millisecond advantage over anyone else. If the same thing were implemented on a blockchain, then there is no server, and while a better Internet connection would give you a better time, there’s nothing to huddle up to. Another advantage is that the whole thing is completely transparent. On the blockchain, you can see it. You can do statistical analysis on the whole thing in real time. If you are going to get some particular financial derivative that is inside Ethereum, you can see what it’s backed by, you can see under what conditions you’re going to end up getting what, and so forth.

FrankThe economist Steve Randy Waldman recently gave a talk about the possibility for decentralized, opt-in social insurance using Ethereum. Will it be possible to replace governmental functions with decentralized structures, or fill in services that might be lacking as governments cut back?

ButerinThe idea behind the presentation Steve gave is that even back before the government started getting seriously involved in welfare, at least to the extent that it is today, there were things like fraternal organizations and unions. You would have a bunch of people who would agree to participate in this organization, they’d all chip in some amount, and if anyone needed money because they got into hard times, they would get it. It’s like insurance except it’s a decentralized, not-for-profit model where everything that goes into the system ends up directly going to help someone.

FrankSo you pay in, and you can look at the code and see that your money’s there in case you need it.

ButerinYou can take alternative economic models that have existed for hundreds of years in a lot of cases, and by porting them onto the blockchain, it’s more scalable and transparent. Potentially, you get the benefits of it happening across multiple countries. If you do it locally then the problem is—well, if you have one of these things inside of New York City, and there’s a big flood going through New York, and suddenly everybody needs all of it at the same time, then we have a problem. Whereas if you set one of them up with some people you know on the Internet, and they’re in ten different countries, that creates something a lot more robust and reliable.

FrankThis puts me in mind of the privatization of Social Security, where you have something that the state can potentially do rather efficiently, versus putting it all on the individual—is that the way you see this going, or is that not the right analogy?

ButerinThere’s a lot of ways to do privatization. The ones that end up going horribly wrong are generally those where you take the existing government thing and sell it off for 10 percent market value to your buddy. Private things are sometimes less efficient in practice because you have middlemen who end up taking profits. Whereas here the point is to get the best of both worlds by having an opt-in system where you can try different models but in a decentralized, community-based fashion.

FrankAlthough still, some people are going to make better decisions than others, and opt in to things that work versus things that bankrupt them.

ButerinThe idea that people might end up doing something stupid, I suppose, is an argument against any system that empowers people.

FrankBut you see more and more functions that we’ve traditionally thought of as central-government functions being able to be decentralized in a similar way. There’s a lot of futurism around the idea that the state is going to be hollowed out by these structures that allow us to do things on a small-group level that we formerly needed a big, centralized institution to take care of.

ButerinThe state is a model of economic organizing. Big companies are a model of economic organizing. So if this new model of economic organizing turns out to actually be practical, it probably will hollow out both the state and big companies to a partial extent. You have to look at the economics and see if it makes sense in a particular case.

FrankCan you describe the idea behind a decentralized autonomous organization, or DAO?

ButerinThe idea behind DAOs is that instead of having an organization that’s managed by a few people, we have an organization that’s managed by an algorithm. It’s like a computer program that has the ability to hire and fire people. It has customers and all the same kinds of relationships that you would expect a company to have, except it works on its own and is incorruptible and so forth. There are many different things that you could think of as being either full DAOs or partial DAOs. You could even think of Bitcoin itself as being a partial DAO. One service that Bitcoin needs is people to secure it. The Bitcoin protocol has a built-in mechanism where if you are a miner and you help secure the network, then you automatically get paid 25 BTC every block. There’s already an entire industry in which people are organizing on their own to figure out how to best fulfill this functionality that the Bitcoin network wants them to fulfill. DAOs can create these sort of large mechanisms that create incentives to build some kind of shared infrastructure.

FrankI know you’re interested in basic income as one plank of your ideal future society. Is there a way of potentially building that on top of Ethereum? Can you create a new currency, like Bitcoin, ex nihilo, that has an egalitarian or antipoverty function?

ButerinThat distributes units of currency equally to each individual person? Yeah, you can. The number one problem is that you need a secure mechanism for figuring out that two different people are actually two people and not one person. The second problem is, is the extra revenue from producing currency units actually enough to pay everyone and help everyone survive? It might not be, in which case you might have to supplement it with both decentralized insurance and traditional government programs. Or we just might have to wait half a century for technology to advance even more.

FrankTimothy May wrote his “Crypto Anarchist Manifesto” around 1988. Is that a tradition that you feel like you’re part of?

ButerinThe idea behind cryptoanarchy is basically to create a society where the economic mechanisms are all based on the Internet, and anonymous, or rather pseudonymous. Each identity is protected by public-key cryptography. Not just the more expansive concept of regulation, but even antifraud laws, would be substituted by this identity and reputation mechanism. If you screw people over, they will report that, and your account will become disreputable very quickly. So you have a system where government intrusion is not possible because all these interactions happen purely online in this convenient, encrypted form. On the other hand, the government is hypothetically not necessary because even though things are anonymous, you have a much larger amount of information available. Mechanisms like reputation worked perfectly fine in traditional societies where people know one another personally, but started working less well in larger, more anonymous modern societies. It might start working again on the Internet, because even though we have lots of people we have lots of information again. That’s the party line with respect to how cryptoanarchism is supposed to work. In practice, I think, various experiments will provide a lot of educational value on what the actual benefits are.

FrankSo far there’s been a Bitcoin drug market on the dark web: the Silk Road.

ButerinIt did pretty well, I think. Hundreds of millions of dollars of trading over a couple years. We saw a practical implementation of close to pure cryptoanarchist ideology. It used public-key cryptography to encrypt messages, it used Tor to encrypt the source and destination involved, and it used Bitcoin to make the currency anonymous. It was an online market that looked very nice and respectable, like eBay, but instead of the headings being “Home,” “Furniture,” “Kitchen,” and so forth, they’d be “Ecstasy,” “LSD,” “Fake Passports.”

It’s nice to have a warm and fuzzy in-person economy where people are shaking hands, but the problem with drug markets is 1 or 2 percent of the time that hand has a knife in it. While Silk Road was around it probably improved the safety of the drug market substantially because people could get their stuff without having to expose themselves to that risk. Another point is that it showcased a free-market alternative to product-safety regulation. First of all you have this really well-developed reputation system to distinguish good products from bad ones. And maybe a few months before the collapse, there was someone running a volunteer FDA. He would order one thing from a bunch of different vendors and do a full chemical review of it and then post the results.

Silk Road created a nice playground where people were, for the first time, experimenting with ideas of how the economy would be run that before were just people writing on paper. In practice, it does have a lot of costs. It turns out that a problem with cryptoanarchy is that—well, in some ways, you could argue that it’s a benefit—in the normal world, the larger you are, the more powerful you are. In situations where violence dominates, if you’re large than you’re powerful and you can protect yourself and buy politicians. But in cryptoanarchy, the larger you get, the more information you release that might be used to track you, and the more people trying to get you. The larger you are, the weaker you get. That’s finally how Ross Ulbricht ended up going down.

FrankAt the beginning of this talk, I quoted your vision of the future. That was the idea that micro payments happen all the time, and you’re tracking the money in a government that’s essentially a database, and you can see it going into life-extension research, and so on. Do you see things actually moving in this direction?

ButerinIt’s moving in that direction in some respects already. For instance, it’s becoming more efficient to pay for resources on an ultra-metered basis that captures exactly how much you’re consuming.

Ultimately, our lives are going to be mediated by software to a much greater extent than they are today. Humans are filled with preferences, such as time versus money. The system might try to guess what your preferences are, because what we say we want isn’t always what we actually want. You’ll have the computer optimizing a large number of the tiny details of your life for you.

FrankAnd what’s the governance here? Is this democracy? Is it something else?

ButerinIt’s hard to say. Efforts at creating some new model of running things rarely end up completely winning and replacing things 100 percent. Democracy is going to continue existing in some form for a very long time. Sure, its decisions aren’t always good, but there are many more protections against decisions that are really bad. You could see other forms of decentralized governance emerging that don’t necessarily look like people voting on everything. There’s been interesting research done by people like the economist Robin Hanson on using prediction markets to make decisions. DAOs could have other models. The answer’s probably different for every system.

“Decentralized Autonomous Society” was published as part of Triple Canopy’s Research Work project area, which receives support from the Andy Warhol Foundation for the Visual Arts, the Brown Foundation, Inc., of Houston, the Lambent Foundation Fund of Tides Foundation, and the New York City Department of Cultural Affairs in partnership with the City Council.