FTP016: Vitalik Buterin – What is Ethereum and How to Build a Decentralized Future

Mike: [00:01:00] Hey Vitalik, thanks for coming on our show. Why don’t you start by telling us a bit about yourself?

Vitalik: I basically don’t really have a history of my own before I first joined the scene when I discovered Bitcoin, back in 2011. I was still in high school at the time and, at some point, I was browsing the internet. Actually, the first time, my dad told me about Bitcoin [00:01:30] at some point early in March 2011. I thought, “Okay, it’s a currency. It has no intrinsic value, there’s no way it’s ever going to take off. A few weeks later, I heard about it again and I realized, “Hey, if I hear about something from two completely different sources, it’s at least worth investigating.”

I did. I started browsing the Bitcoin forums. Came upon a guy who was paying people $5 an hour to go out to write articles for his Bitcoin blog. [00:02:00] I started doing that. I sent him a message saying, “Hey, I’d like to write an article for you on microtransactions.” I did and a week later he sent me 5 BDC, which, back then, was $3.75. A good wage for three hours work. Continued for a while. Eventually, he ran out of money, so I invented this cool business model by which that particular Bitcoin blog would stay running. The way it works is that I would write two articles a week. We would [00:02:30] release only the first paragraph of each one and we would say, “Okay, these articles are for ransom. We are only releasing them if people crowdfund 2.5 BDC into this Bitcoin address.”

That actually worked. That was my first interaction both with that particular decentralized currency and with the general idea that these kinds of platforms are cool for doing this kind of really rapid [00:03:00] social and economic innovation. I was in university studying computer science full-time. At some point, I realized that I was doing cryptocurrency related things for basically over 30 hours a week, so dropped out, went into doing that full-time. Spent about six months traveling around the world, talking and visiting various different… Back then were Bitcoin communities because, back then, it was still very predominantly Bitcoin and there wasn’t this consciousness that crypto and blockchains are [00:03:30] an interesting thing in and of themselves.

I spent two months in this weird abandoned factory in Spain. Travelled around Europe, visited Israel. Eventually came upon a group of people that were actually trying to use blockchain technology for a much more broad set of applications. They were looking at escrows, very specific kinds of financial contracts. There’s this protocol called Mastercoin, which is trying to do a lot of those thing. I actually briefly [00:04:00] joined the Mastercoin team as an informal consultant, came up with two new features for the protocol. Eventually, what I realized was that it really doesn’t make sense for the protocol to be coming up with these features, because people are just going to keep on coming up with new use cases every month.

It would be simpler to just have a programming language, so people can basically write whatever features they want. That was the core idea behind Ethereum. At one point, when I was in San Francisco, [00:04:30] the thought came to me. I took a long three hour walk, formalized a lot of what were the basic concepts back then. Wrote a white paper, emailed it to a bunch of people. Some people came back saying, “Hey, this is a really cool idea, I’d like to join and maybe help out in the end, start actually developing the software for this.” The team coalesced over a couple of months. I’ve been basically doing that full-time ever since.

Euvie: Can you tell our listeners in plain language what Ethereum does?

Vitalik: [00:05:00] I’ve actually been changing the explanation every few months, I guess as we’ve been changing the emphasis on what Ethereum is about. Originally it was just meant to be a decentralized platform for financial contracts. Lately, I think we’ve been expanding in the thing that we’re trying to be is something closer to… Actually, one of our communications people, Vinay Gupta, describes it as being the world computer. The idea is that there exists this magic [00:05:30] computer in the cloud and anyone can send programs to it, anyone can run programs on it. Those programs can talk to each other and, particularly importantly, you can trust that this computer will run the programs in the exact way that you specified them to run.

Very simple, not even all that significant use case of this. Suppose that you want to be reminded of something in five years. You would create [00:06:00] a program which, after five years, would be programmed to automatically release some kind of notification. Take this program, put it on the world computer, pay some kind of transaction fee. It’s guaranteed that five years later, it’s actually going to send you the notification. That’s guaranteed to happen and that’s even going to happen if any particular entity shuts down. It’s not the cast that, if Google shuts down, then the whole thing is going to break.

There is nothing that any particular entity [00:06:30] can maliciously try to corrupt the execution in some fashion. It’s also completely transparent. Everyone can see everything that’s going on in the system. Just specific kinds of use cases for something like this. Currency is just a simple one, because it was the original blockchain use case. What is a currency? Basically, it’s just a program that maintains a balance sheet. The balance sheet might say something like, “I have $100 and you have $50.” If I want to send $20 [00:07:00] to you, the I send a transaction and the program interprets the transaction, which is my balance to $80 and yours to $70. If, instead, I try sending $300, then, of course, the program’s going to reply, “Nope, you don’t have enough money, fail.”

You can take this kind of program, you can upload it to the world computer, and everyone who sees the program can be guaranteed that the program will continue working in some particular way. [00:07:30] It’s like a very general and abstract metaphor for a very large number of categories for decentralization.

Euvie: Do you want to explain why decentralization is really important? Obviously, a lot of people talk about it, but what are the actual benefits of decentralization versus a centralized system?

Vitalik: Sure. I think it’s a very important question, because now that we, all of a sudden, have the opportunity to decentralize a whole bunch of things, [00:08:00] with blockchains you can decentralize the running of any kind of application. There is a bunch of other technologies, like torrent networks, that do the same thing for file sharing. People have been getting excited about things like the sharing economy, which is, “Anyone can be their own taxi company or be their own hotel in their spare time,” almost. First of all, there’s two kinds of decentralization that people need to think about. The first kind is what I call architectural decentralization, which basically means [00:08:30] that the whole system isn’t technologically reliant on any single component. The way that the system runs is actually shared among a whole bunch of different technological parts. The second kind is political decentralization, which actually means that the system isn’t just controlled by a whole bunch of different pieces, but those pieces are actually under the administration under the local rule of a whole bunch of different people.

Architectural decentralization without political decentralization. [00:09:00] One example is torrent networks. For example, the original torrent networks, things like BitTorrent and Kazaa, they’ve been around since about 2003. They’re really good for sharing files. Things like BitTorrent and Kazaa, they’re both architecturally decentralized, because the files are spread among a whole bunch of computers and this makes the network extremely robust. It actually makes it more efficient. It’s also politically decentralized because there’s no single thing that you can attack in order to shut it down. The interesting thing [00:09:30] is that torrent networks as a technology, it turns out that they’re just so efficient for certain use cases that they’re sometimes even used in a completely centralized context.

As one example, there’s a whole bunch of online games like World of Warcraft, for example, where World of Warcraft periodically send patches – which are basically just big software updates to all of its users. Each patch might contain several gigabytes, because there’s often a whole bunch of new content. The question is, how do these patches get distributed? [00:10:00] One option is everyone just downloads it from Blizzard servers. The problem is that would be really expensive for Blizzard. What Blizzard does is it uses the torrent network. Of course, it’s completely centralized, Blizzard decides what content gets distributed from a political standpoint.

From an architectural standpoint, every single peer, every single computer in the system is helping in both the downloading and the uploading. That actually means that you’re taking advantage of all of the network’s resources. That actually allows the whole thing to happen much more [00:10:30] efficiently. Blockchains are basically politically decentralized and architecturally decentralized. I guess the important difference is that, unlike torrent networks, blockchains don’t’ become more efficient by decentralizing them. You can’t actually get a higher transaction throughput, at least right now.

The big difference here is all of the other advantages of decentralizing – things like reliability and things like transparency and, what I call, political benefits that there’s no one particular group [00:11:00] that you need to rely on for the whole thing to continue to work, those are the ones that take centre stage.

Mike: I think this is really cool that it takes the middle man out of a lot of different types of transactions and interactions. I’m curious as to how it could benefit certain types of middle men that actually need to be there that make the system better. I can’t exactly think of many examples, maybe you can. Can you think of examples where the middle man is actually important and how it could benefit them?

Vitalik: Practical example: [00:11:30] Uber. It’s one of those social network services. The difference between Uber and taxi companies is that with Uber anyone can be a driver, anyone can be a consumer. It’s highly open on both sides. The middle man still exists. To some extent, the middle man is a bit of a greedy monopolist that’s taking far too large a fee for its own good. To some extent, also, [00:12:00] it’s providing a very legitimate service. In this case, the specific service is basically that it’s being a reputation system. It’s providing a filtering service where it makes sure that customers are only interacting with drivers that are reliable, that have good ratings, that are not likely to be criminals and so forth.

That’s a service that people need. You can design a decentralized version of Uber. There is at least one project [00:12:30] called La’Zooz that tries to do that. The thing is that in that kind of a system, you still need to have a way of actually performing that middle man functionality. You still need to have some kind of reputation mechanism, or some kind of quality assurance, or filtering, or whatever you want to call it, in order to buy a system to reward customers to being able to interact with drivers. Another more purely technical example. If you consider Skype. Originally, it was a fully peer to peer system. [00:13:00] More recently, however, soon after Microsoft acquired Skype, Microsoft moved to this architecture where they had a bunch of super nodes. These super nodes are politically controlled by Microsoft.

Architecturally, of course, they already are decentralized and spread around the world. Depending on some degree of what your political ideology is, there’s two ways of looking at these super nodes. One way is that this is just Microsoft being corrupt and inserting themselves at the centre in order to be able to spy on everyone’s data more easily. [00:13:30] The other is that, as it turns out, extremely decentralized networks actually aren’t all that efficient, because you need to send messages through a whole bunch of hubs. Having these really powerful hubs be involved in the process speeds things up. If the second viewpoint turns out to be closer to correct, then in a good decentralized replacement for Skype, you’d still need to have some concept of super nodes.

What really excites me [00:14:00] about this field of blockchain technology, and what I’ve been starting to call crypto economics, is the idea that what we’re actually doing is we’re combining a decentralized architecture with an economic system. The advantage of an economic system is now… Originally, 10 years ago, decentralization basically meant the only people participating in a system are amateurs. To some degree, even things like Wikipedia, which you can think of as being partially decentralized, have this ethos [00:14:30] behind them that a really large collection of amateurs, the wisdom of the crowds and so forth, can do almost everything that a slightly controlled small group of experts can.

Sometimes that’s more true, sometimes that’s less true. The nice thing about crypto economics is that, in some circumstances, you actually have the ability to use economic incentive. People actually do have the ability to basically run some equivalent of a super node as a business. Inside of something like a decentralized Skype [00:15:00] replacement, you could potentially architecturally set it up in such a way that anyone can run one of these super nodes and they’d collect some very tiny micropayments in exchange for running it. The system would, architecturally, look exactly the same way that Skype looks now, but from a political standpoint it would still have the benefits of being completely open source, being a completely open standard, and an open network, and not having the risk that one particular company [00:15:30] is going to end up changing the protocol from under everyone’s nose after the fact at some point.

Something like Uber, it’s in some ways similar. You recognize that this desire for some kind of filtering system is a legitimate thing and you come up with ways of actually getting that information more efficiently from I guess decentralized sources. The most common idea behind a reputation systems is that if I trust person A and person A really trusts person B, [00:16:00] then even though I might never have heard of person B I still actually have some proxy evidence that person B is reliable. There’s been this entire field of reputation systems for about 10 years, that tries to figure out how to most efficiently and unexploitably take advantage of that kind of data.

Euvie: Yeah. Actually, in one of your other talks that I saw you talked about this, having some sort of a blanket reputation system that could apply to different tools and different platforms within Ethereum. Right now, for example, you have [00:16:30] one reputation system that’s your credit rating, another one that’s your Uber rating, another one that’s your Air BnB rating, another one that’s your number of Twitter followers.

Vitalik: Exactly. The problem, of course, is if at some point you want to switch over from uber to Lyft, because Lyft might have lower fees or whatever else then, guess what, you lose your Uber reputation. These kinds of really high switching costs are potentially one of the problems and actually why I’m interested in this whole [00:17:00] specific idea of trying to decouple reputation from everything else.

Euvie: Yeah. I think also this platform specific reputation like we have with current systems is actually helping those systems monopolize people’s attention and their money, because the cost of switching can sometimes be very high, so people stay with a system that’s increasingly inefficient or charges high fees every year.

Vitalik: Yeah, completely.

Mike: Are you trying to create systems to transition [00:17:30] people from these current monopolized systems into Ethereum, or is that something you’re just hoping that people will figure out on their own?

Vitalik: It’s interesting. We ourselves, as ethereum.org, can’t do everything, but we are involved as gardeners of parts of the system in a very weak sense. Things like the core for the Ethereum protocol for Mist, that’s something that we’ve been developing ourselves and that we’re going to continue to develop ourselves. Things that are built [00:18:00] on top, things like reputation systems, some of these different decentralized applications, we personally know a lot of the people that are building them. A lot of them are building on Ethereum, a lot of them are sitting back and trying to be as generic and multi-platform as possible until they see what they should throw more weight behind.

In some of the cases, we’ve been supporting some specific efforts. In some cases, we just as individuals working inside the Ethereum project [00:18:30] are interested in these areas and we help out on a lot of them. It’s a multi-stage thing I would say. Right now, the large part of our focus is on just building a platform and making sure that it’s good and that it’s easy to develop on. Once people see on top of this thing you can actually build your own Kickstarter clone in 20 minutes, then that’s when we see a lot more developer interest. As far as transitioning goes, to some degree we’re facing the same network effect problem that just about [00:19:00] everyone else is facing.

There was a Facebook alternative called Diaspora that started appearing around 2010. For a brief period of time, there were people who were posting stuff on it and, for a brief period of time, it worked decently well. Eventually, it fizzled. Part of the reason was that they just didn’t have the ability to overcome the network effect. There’s multiple ways of dealing

with that problem. One of the ways is that, because we have Ethereum and we have this decentralized platform that’s trying [00:19:30] to, to some degree, facilitate everything at once, it stands a much stronger chance than any particular isolated effort of having all these network effects reinforce each other and take over at the same time.

That’s one of the hopes that we’re putting our weight behind. In other cases, there’s all of the same ideas that just about every other case of network effect [inaudible [0:19:51], which basically involves making it easier for people to interoperate, making it easier for people to generally use multiple services at the same [00:20:00] time, even providing some kind of transition paths. I would actually say that one of the benefits of decentralized technology generally is that when we actually start talking about how do you even move between one decentralized platform and another, because eventually people are going to start building continually better and better and better ones. Then, just because these platforms are so open, it’s very easy to make the connections and the links between them, and transition people over fairly easily.

Euvie: Yeah. I think that’s really [00:20:30] cool that you guys are inviting a lot of developers to come and build projects, so it’s not just one program that’s running, it’s a whole bunch of people doing all these different things. Which kind of makes it more resilient to change. I think people are more likely to stick around if they can improve on things as they get dated.

Vitalik: Yeah, I agree. One of the comments that I sometimes make about Ethereum is that, if you look at Bitcoin, the Bitcoin wallet, and the Bitcoin client, they’re synonyms. They refer to the exact [00:21:00] same thing. In Ethereum, the Ethereum wallet is just one of the many applications that runs inside of the Ethereum client. The vision here is something much more abstract. We also have a lot of components other than just this world computer. A long-term vision is actually to build out something much closer to a full-scale decentralized application platform, and that would include things like the blockchain, it would include different kinds of decentralized data storage, [00:21:30] decentralized messaging. The idea is to present all of these in a cohesive, combined package that makes it really easy to work with all of them at the same time.

Mike: I was looking up Ethereum the other day and found that you guys were, I think, the number three most funded crowdsourced campaign that has ever existed before. Can you tell me a bit about how you guys got your funding and how you posed it to the community?

Vitalik: Actually, when we first finished that campaign, we were number two. I think the Pebble Watch [00:22:00] just beat us only last month. What we basically did – that actually we didn’t pioneer ourselves, it was originally the Mastercoin, the project I’d been working on before. What they did was they said, “Okay, we’re launching a new protocol and inside this new protocol there is a currency.” They basically said, “We are going to be issuing these initial master tokens by just selling them. You can buy 100 Mastercoin for a Bitcoin.” A few people had the ability to do that. [00:22:30] There was a month-long sale during which they did. That’s how the entire currency got issued. Unlike Bitcoin, which issued everything to miners, Mastercoin actually lived on top of Bitcoin, so it didn’t need any miners.

You just had to distribute it somehow. For them, it raised $500,000 and then, because Bitcoin went up in value, it quickly grew to five million dollars. It works really well for them. We basically decided to apply the model for ourselves. In our case, back then, even still now, we’re using [00:23:00] mining as a consensus algorithm. The idea here is that you have a whole bunch of computers everywhere in the world that are solving very hard – useless but very hard – mathematical problems. The sole point of these problems is to deliberately make it computationally very hard to add a block to the history of the system. The idea is that there exists this chain of blocks and each block represents the current state of the world computer [00:23:30] and if you want to participate in updating it then you have the ability to just produce a block by spending a large about of time computing.

It’s the same model that Bitcoin used. Lately, we’ve actually been thinking of moving away from that model to something which is much more energy efficient, but that’s an upgrade that’s probably going to happen in maybe at the end of this year. We had mining but, even still, we decided that, “Hey, it’s a good idea to get money for this kind of project by selling the coins.” [00:24:00] We basically said, “Anyone can buy Ether at the rate of 2,000 Ether per Bitcoin. There’s no limit.” We originally intended to start the sale in February, 2014. We ran into a bunch of legal issues, making sure it doesn’t fall to foul of things like banking and securities laws. It took us a while to do the legal work to make sure that we’re okay there.

We were thinking the sale would be in two weeks for a really long time. [00:24:30] Eventually, it was actually in two weeks and the sale started at some point in the middle of July. People got the opportunity, they had the ability to go to our site, download our wallet, which is basically just the file containing the cryptographic key that you would use to later access your Ether. You could purchase Ether into your wallet and we would provide a Bitcoin address, you would send your Bitcoin to it. You wouldn’t get your Ether immediately, because Ether wouldn’t exist, but you would get [00:25:00] a record in the blockchain and later we would run a script that would compile all the records in the blockchain to figure out how much Ether everyone has.

People piled in. In the first day, I think we got a few thousand Bitcoins. By the end of the first two weeks, we got up to a total of somewhere around 12.5 million dollars. This, back then, actually made up the second largest crowdfunding campaign ever by a small margin. Eventually, it lasted for another four weeks. [00:25:30] We received around 18 million dollars. Because the Bitcoin price was falling at the time, we never really had anything more than about 15 million dollars in our Bitcoin accounts. We had the money, we started using it immediately to pay salaries, pay developers, massively expand the team and working on getting Ethereum 1.0 out the door as soon as we can.

Euvie: Very cool. Actually, one thing that you said got me thinking of a question. With Bitcoin, the way that I understand, the [00:26:00] transaction is near instant. I wonder if, with this system, is it possible to have a delayed transaction? Why I’m asking that is because is debt possible in this kind of system, or is it intrinsically debt free?

Vitalik: You can definitely have a delayed transaction. I guess one really nice use case of that is recurring payments, if you want to pay someone $10 a month. In Bitcoin, the only way to do that is basically by using centralized services. Those things are like Coinbase that [00:26:30] basically build recurring payments on top of Bitcoin. The way to do that is you send your money to Coinbase and Coinbase just shuffles it on for you. In Ethereum you could just write a script and your script would automatically send $10 every month. As far as debt goes, the problem with debt is that the system is fundamentally anonymous, anyone can create and account. If you were to create and account and your account had 1,000 Ether in it and you were to make a delayed transaction that would send 1,000 Ether later on, [00:27:00] then there’s basically only two options.

One option is the 1,000 Ether is actually locked in that contract until you send it and the other option is that you have the ability to take your Ether out during the meantime but then, if you do, then once it comes time to send the 1,000 Ether then there wouldn’t be enough Ether and it would fail. If you have a debt to someone then, unless you have some identity outside the system that you can use in order to promise to repay, then fundamentally there’s nothing to stop [00:27:30] you from just running away and starting a new identity.

Euvie: How would you deal with this kind of problem?

Vitalik: There’s a bunch of groups that are trying to loans on the blockchain. Most of the time, they actually do involve things like reputation systems. Trying to use measurements that are… You can think of them as being like decentralized credit scores… Use these measurements to figure out which loan recipients are [00:28:00] trustworthy and which ones are likely to pay you back.

Euvie: On that note, I wanted to ask what other kinds of problems you envision happening in this sort of system, like fraud, for example, or companies trying to control large portions of the nodes. I don’t fully understand it, but like the 51 percent problem.

Vitalik: Yes. Yeah, as far as the bottom level consensus goes, the centralization issue is a huge issue. People are very concerned. [00:28:30] In Bitcoin there’s about five or six of what are called mining pools, which are basically online entities that automatically produce blocks for you and basically tell you what block to mine. The reason why people subscribe to mining pools is, in part, because they provide more constant payouts. Instead of getting a very, very tiny chance of getting a really huge 25 BTC reward if you manage to mine a block, if you mine into a mining pool then you just get a very constant stream of payouts. There’s a few of them that [00:29:00] control over half of the network.

It’s a risk but it’s a smaller risk because if those mining pools end up doing anything bad, then people can just switch to other ones. The larger risk is the fact that, now that Bitcoin mining is basically controlled by specialized hardware, there are these devices called ASICs, application specific integrated circuits, that can mine in Bitcoin 10,000 times faster than any normal computer can. Just about every miner that wants to be profitable is basically running one of these [00:29:30] specialized devices. Problem is there’s an incentive for these ASICs to [inaudible 0:29:35] in data centres because, if they’re in data centres, then the owner only needs to hire a couple of full-time employees to manage all of them.

Those particular employees could become much more efficient at their job, because they’re doing it all day, so it’s much more cost effective than just having a whole bunch of people run each one separately. The pressure for the system to be very efficient only increases, because the price of Bitcoin right now is dropping, [00:30:00] so a lot of miners are unprofitable. Right now, I think my estimate is that there’s somewhere between 10 to 30 data centres that control over half of the network. Theoretically, the number could continue to decrease. There’s the data centres then there’s also the manufacturing companies. I think the manufacturing companies are even more centralized because just the sheer amount of existing factory and infrastructure that you need in order to actually build on these ASICs just costs huge amounts [00:30:30] of money.

There’s a bunch of factories in China where a lot of the activity happens. Last year in May I visited a factory which, at the time, was producing a quarter of all the new Bitcoin miners that were coming into circulation at the time. There’s a couple of issues. One of the issues is that these companies might eventually decide to collude and start doing things like censoring transactions and even double spending. Double spending is basically this attack where, if you have more computing power than the rest of the [00:31:00] world combined, then you can go back to some previous point in history and you can start mining. Eventually, your history will have more computing weight behind it than the original history.

People will basically switch to yours and, that way, you have the ability to go back in time back to whatever point in time you want. One risk is that companies will just collude to do it. One reason they might collude is to try to reverse their own transactions. One option is that you take 10 million dollars, send the 10 million dollars to a Bitcoin exchange, [00:31:30] buy 10 million dollars’ worth of Litecoin. Then you go back in time and you cancel the transaction where you sent your 10 million dollars, and now we have your 10 million dollars’ worth of Bitcoin back and your 10 million dollars’ worth of Litecoin. Second option is that you could short Bitcoin on an exchange. There’s exchanges that let you basically bet against the Bitcoin price.

If an attack succeeds, then you can basically earn a huge amount of profit from this. There’s those possibilities. The other possibilities are that just some guy [00:32:00] hacks into a lot of these farms at the same time. Third possibility, theoretically, the Chinese government might be able to just do a 51 percent attack by giving all the data centres within China a warrant to basically tell them, “Hey, cooperate with us or else.” It’s a lot less decentralized than it seems. The bigger problem is that we can’t really quantify even how decentralized it is, because mining is completely not transparent. It’s very hard to [00:32:30] try to even trace where all the miners are located, where they’re coming from and so forth. Which, to some extent, makes it harder to collude but, to another extent, means that we just can’t know how bad the problem is without extremely serious analysis.

Euvie: Do you see Ethereum having these problems too, or are you taking measures to prevent that?

Vitalik: We are taking a lot of measures. One of our initial ideas is basically coming up with an algorithm which is much harder to make more efficient with specialized [00:33:00] hardware. The theory is that people would be much more able to run the software on normal computers. It could end up being high powered gaming rigs but, at least, there’s hundreds of thousands of people around the world that have gaming rigs. It’s going to be much harder to collude. The strategy that we’ve been moving to lately is this idea of proof of stake, which basically means that instead of the consensus being who has the most computing power, it’s a process where the people [00:33:30] that hold the coins inside of the system collaborate on making the next block. Instead of using computer power as a proxy for what your vote is, you would just have the amount of Ether that you have the thing that decides how much you get to participate.

Euvie: Do you think that might encourage monopoly though?

Vitalik: Yeah, it’s a common criticism. I would say that if we compare it to Bitcoin back in 2010, it’s an extremely legitimate point. If we compare it to Bitcoin in 2014 [00:34:00] then the problem is that the current system encourages monopoly even more. There’s this economic concept of economy as a scale where, in the current system, if you’re 10 times richer than someone else then, first of all, you can produce 10 times more but then, on top of that, you have the ability to make more efficient capital investments. You might even be even more than 10 times more effective. Whereas, in this system, it’s fairly simple. If you’re 10 times bigger, you do 10 times more.

Against the backdrop of this more hyper specialized mining industry that we see right now, it’s probably [00:34:30] an improvement. Against Bitcoin as it was ideally meant to be, it’s not but, unfortunately, we can’t guarantee that it’s always going to be this idyllic thing where everyone just contributes a bit on their own laptop.

Euvie: Let’s backtrack a little bit. Talk about some of the everyday applications of this kind of system if it becomes ubiquitous. You guys did a Reddit AMA and there was one response that I particularly liked where you described [00:35:00] a guy from the time that he wakes up to the time that he gets to work, all of the different things that happened within Ethereum. Do you remember that one?

Vitalik: Yes, I do.

Euvie: Could you explain it to our listeners?

Vitalik: The idea was that a guy wakes up and whatever amount that he’s supposed to pay for rent everyday automatically gets deducted from his balance. Whatever amount from that that’s supposed to go for property taxes is automatically deducted from the government. You could use the blockchain in order to [00:35:30] trace where the government tax money goes, you could even see exactly what programs it ends up supporting. Then get into your car. While you’re in your car, first of all, we have a decentralized self-driving car Uber scenario where… What just about every car company is doing is they’re building cars and then they’re just releasing the cars into the wild.

Each individual car would just be pretty much autonomously running around looking for people that need to get from point A to point B. [00:36:00] It’s using some standard computer science search algorithms to try to figure out exactly how that particular car can get the most revenue. The person would go onto some kind of decentralized market place and he would say, “Okay, I want to go from where I am right now, here’s my work location. Please post bids on how much it would cost to send me over.” A bunch of these autonomous cars would post bids and the guy would pick the lowest one and the car would start driving him over.

[00:36:30] At that point, you’d have the option of either taking a road that has heavy traffic on it but where there would be some kind of dynamically adjusted toll, so the road would be more expensive. Or, taking a road that nobody else is using right now and that would be cheaper. Eventually, the guy gets to work. Chances are, he might end up working for even some kind of decentralized organization that would rate his contributions and pay him for his contributions automatically and so forth. The idea behind the vision is that, [00:37:00] first of all, with these different kinds of decentralized market places you have the ability to really reduce the barriers to entry to different kinds of industries. Right now, if I wanted to start a taxi company, I don’t just have to start it up, I also have to go around and market it and try to get it adopted.

I’ll run into network affect issues because I’m only in one place and so forth. Whereas here, there’s one platform and it’s a completely decentralized platform and anyone can participate on it [00:37:30] on either side. Theoretically, it would interact with things like roads. Roads would be enabled in order to incentivize people to, for example, drive less during peak hours and drive more when no one else is driving in order to minimize traffic. You’d have a whole bunch of these different systems for just about every aspect of work, city planning, architecture and so forth. Once the guy gets to his company and starts working… The other really interesting question is to what degree can you actually decentralize [00:38:00] the way that entire companies are governed?

Ultimately, what some of these decentralized platforms do is they basically replace companies, to some degree, and they just completely swap out their functionality for the system where anyone can come in, anyone can participate. Trying to figure out to what extent you can do that. The line is that the existing kind of automation that we’ve seen tends to be the kind that automates the way the people at the edges – people like factory workers, [00:38:30] different kinds of manual labourers and so forth. This is kind of a different kind of automation that maybe paves the way for automating away certain kinds of managers.

Mike: In recent times, it seems like the focus for a lot of entrepreneurs is to create platforms and launch these platforms with the idea that they’ll be the middle man or they’ll create the system, then people will participate, not unlike Uber. If this kind of system is going to change it, where do you recommend entrepreneurs focus in the future if everyone’s using this [00:39:00] type of platform?

Vitalik: First of all, I think that one of the answers will just unfortunately have to be there is less need for that particular kind of labour. The effort of trying to push a platform out, trying really hard to market and monopolize the industry and so forth, there’s just going to be less need for entrepreneurs in this particular space. Just because all these decentralized blockchain technologies just make it so much easier to build platforms. On the one hand, the profit opportunities are going to be lower [00:39:30] but, on the other hand, it’s also going to be much easier. It’s this interesting dichotomy in that it reduces both at the same time. The best analogy might even be to something like the internet, where it, to some degree, just let a lot of people enjoy themselves without spending that much money at all.

It probably ended up killing a whole bunch of business opportunities. On the other hand, it helps people save a lot of money. Particularly, it massively reduces barriers to entry for people to create [00:40:00] new kinds of businesses. In terms of the kinds of businesses that will continue to exist, in the 20^th century the paradigm was, “How do I provide a whole bunch of things for a really large mass of people?” What I would describe as a transition phase. It seems as though it’s switching to, “How do I provide a platform for a whole bunch of people to help each other?” To some degree, it is that but, to some degree, it is still centralized. Platforms like Uber, they’re still providing [00:40:30] the interface, they’re still providing particularly reputation systems and so forth.

I think the future opportunities are going to have to do pretty much entirely with various different kinds of data aggregation. The thing is that all of these intermediaries that we have right now are basically informational to some degree. There might even be opportunities to earn a profit by somehow just being an entity that provides a whole bunch of data for specific people that they need. I also think [00:41:00] that the move to this digital realm, it doesn’t even have to do specifically with decentralization itself. It’s been happening for 20 years. It’s also creating all of these different concepts of virtual property. People sometimes think that Bitcoin is the first really peer digital asset. That’s not really true. The first peer digital asset is Domain Names. It’s a completely emergent form of digital property that just somehow emerged out of nowhere.

As it turns out, [00:41:30] it’s a multibillion-dollar industry. Advertising is a thing that the beginning of this particular century invented. The challenge would be trying to figure out what exactly the next decade’s equivalent of things like advertising and maybe some form of virtual property that just somehow emerges out of this particular industry is going to be. I’m not completely sure about exactly what that’s going to be. If I was, I’d already have [00:42:00] a billion dollars of venture capital investment. Right now, I don’t. To some degree, it will probably just see a continued shift in industries that will it end up looking more like finance industry. If you look at the finance industry, generally, financial firms they don’t really serve customers, so to speak, directly in a lot of cases.

All they do is they buy things and they sell things. What they’re doing in the process is they kind of are being informational intermediaries to some degree. They’re [00:42:30] trying to quickly analyse a whole bunch of data about the world and they’re trying to quickly figure out, “Okay, is this data going to lead to this particular thing being more expensive or this particular thing being more cheaper?” If they find an opportunity, they seize on it. It’s this service that they end up providing on this extremely abstract level, without really interacting directly. I could easily see other industries in the reputation space moving [00:43:00] toward that kind of model. Altogether, it’s hard to say.

Euvie: On the opposite side of that spectrum, there’s a DAPP that I read about called Crypto [inaudible [0:43:11]. It’s built on Ethereum and it would allow content creators to get paid for really great content and for talent scouts or trend hunters to be able to get rewarded for discovering this great content early on.

Vitalik: Right. It’s an interesting project. [00:43:30] I’m personally a bit sceptical of the whole, “Hey, let’s tip people for the content they provide,” thing, because, in practice, there are platforms that let people do that. The tips generally end up being on the order of five cents. First of all, my one point that I was making is that the opportunity to make massive profits by being an intermediary or being a company, will go down. But if you provide something valuable, there probably will be more opportunities to make money from it [00:44:00] just as some kind of an individual. In a while, I think that the idea of, “I’m going to create some great thing and people are just going to donate enough thousand dollars to me every month, in order to let me survive,” is a bit overoptimistic.

We are going to see, I think, more instances of, “Hey, can you do this particular thing for me and get some small amount of money off from that?” That’s a trend that even exists outside of the decentralization industry. [00:44:30] There’s a whole bunch of companies that are jumping on this whole idea of quickly pay someone some number of dollars to do some particular kind of thing.

Euvie: Like microservices.

Vitalik: Exactly, microservices. The potential that decentralization has to expand opportunities for people who are performing microservices is specifically this idea that you can jump between platforms and you can carry your reputation across them.

Euvie: What do you think about the near zero cost economy? Do you think [00:45:00] that a platform like Ethereum would be conducive to that? Because if a lot of thing are decentralized, if they’re not really managed by people or managed by people to a very, very small degree – maybe just for troubleshooting – would that make a lot of different services cost close to zero?

Vitalik: Yeah, I think so. I think there’s a bunch of things that will just continue going down in price. There’s a lot of intermediaries that just end up charging 20 or 30 percent. [00:45:30] If the concept of decentralization takes off and does well, then those are probably going to also decline to near zero. Anything that has to do with various kinds of informational arbitrage will go down. It’s almost a matter of asking what kinds of things aren’t going to be zero cost anymore. That’s, to some degree, a hard question, because you have to figure out what is something that is going to continue to be valuable [00:46:00] and that can’t be provided for free in some sense.

There’s a couple of ways to answer that question. The first one is that I was mentioning, that I think that a lot of these new digital platforms, they reduce the extent to which people care about physical property but they also increase the role of various new kinds of digital property that they create. The idea of domain names selling for millions of dollars is just one of them. I think that, in general, there just are things in human society that are [00:46:30] zero sum, they are a status competition to some degree. Those are just things that aren’t going to disappear. If old ones disappear, then new ones are just going to keep popping up. People seem to like competing for status, all just for its own sake, in a whole bunch of different areas.

There’s always going to be things that have a price. If you can figure out that kind of thing, then I guess there’s an opportunity to be selling it. In terms of everything else, at least things that can be digitized are [00:47:00] just going to continue to become cheaper. The things that aren’t going to go zero quickly are things that basically things that can’t be digitized. Food. Okay, food is going to be cheaper. Eventually, we’ll figure out how to do this whole automated food thing and produce extremely high-quality stuff at lower rates. But it’s still not going to be zero. Health gear, not going to be zero any time soon. It seems likely that I guess things that you would categorize as being essentials [00:47:30] would also remain non-zero for some period of time.

To some degree, the digital economy can’t magically make them much, much better all by itself. There are gains, you can make it more efficient, but there’s always limits. Those kinds of things all continue having a price. Information is the one thing that’s much harder to have a price and have that price be defensible. Things that aren’t informational in any way, they’ll probably remain non-zero for some time, at least until the whole concept of a post-scarcity economy hits and we have robots and so forth, [00:48:00] but then who knows what’s going to happen.

Mike: Do you think this platform has the potential to create a vibrant artist economy?

Vitalik: Possibly. One point that I’ll have to make is that, once again, I’m a sceptic of the whole tipping concept, but that’s a side point that I think people don’t even realize the extent to which that’s only one particular way that you can tip artists. If I’m an artist, then I have a bunch of things I can do. First of all, you can make personalized art for some specific person [00:48:30] and for some specific group. The demand for that could very easily increase. These kinds of platforms actually make it easier to do things like that. That’s one path.

The other path that I’m interested in is that, going back to the whole status symbol concept, at least back in the physical world people seem to have an interest in buying things that express their support for a particular artist. That would be things like t-shirts, things like CDs, [00:49:00] even just going to see them at a concert and so forth. The specific property that those kinds of things have is that they’re very public. It’s not just something that you’re just buying as private donation, it’s something that you’re buying specifically to express the degree to which you’re interested in a particular artist to other people.

There’s opportunities to do something in the digital equivalent of that. Come up with some kind of badge that would [00:49:30] exist in the context of a whole bunch of digital platforms that would somehow get attached to your profile and that could be used to represent your support in a digital realm. Maybe that’ll end up as being a source of income for artists. Doing something like that would probably be a very significant behavioural economics and psychological challenge, but it could easily be done.

Euvie: In closing, I just wanted to offer one hilarious bit to our listeners. It’s this thing called [00:50:00] Inthereum. It’s a parody of Ethereum. You’ve seen this Vitalik, right?

Vitalik: Yes, I have.

Mike: Are you worried about the Bobchain becoming a major competitor for Ethereum?

Vitalik: Bobchain is never going to scale but Bobchain 2.0 seems a bit dangerous.

Mike: How do you guys need help? How do you hope the community will help you guys to achieve what you’re trying to do?

Vitalik: Right now, I think there is a lot of people that have ideas and a lot of them [00:50:30] are really cool ideas. There’s a lot of people that are interested in developing. I think there also might be a lot of people that might be attracted by the whole concept, by the whole idea of developing in this decentralized application style. To some degree, it’s just a matter of reaching out to them and trying to make sure that people actually find out about what we’re doing and, ideally, start building something.