The Blockchainers + Max Kordek: Interview Transcript
While touring through Asia in September, Max had the chance to do an in-person interview in Seoul with See Eun and Younghoon of The Blockchainers, who host their own blockchain technology-focused channel on YouTube. They covered many subjects in this hour-and-a-half interview, including the Lisk strategy, the ins and outs of the Lisk technology and blockchain in general. Please note that this interview was conducted in September. Numbers and figures cited below are very likely to be outdated.
Thank you to See Eun and Younghoon once again for giving Max the opportunity to discuss what we’ve been up to at Lisk, as well as what our bright future holds. They were integral in helping Max and Thomas better navigate and gain a foothold in the South Korean blockchain industry. We look forward to being part of many more interviews next year.
Read the full interview transcript below.
Y: We heard you’ve been on an Asia tour. Can you briefly explain to viewers what you’ve been doing in Asia?
M: We’re a European project and we’ve been to America a few times, but we feel that Asia is huge. It has the largest population in the world, as well as the biggest blockchain adoption in the world, and we decided we want to make a market entry to three new countries: Japan, India and Korea. We also want to strengthen our bonds in China. That’s why we planned this Asia tour: four cities in two weeks with Marketing Lead Thomas Schouten. We basically want to spread the word, get listed on exchanges and want to get people involved in the Lisk community. Maybe, we can even find Lisk ambassadors, so if you really are excited about Lisk and want to help out, reach out to us and we might get you involved in the whole process. Until now, it’s been a blast — very exciting but also tiresome because it’s one meetup after another and so many interviews. It’s great.
S: I know you guys recently got new members. Your team is much bigger than how I remember. Tell us a bit more.
M: We started hiring early this year. We started back in the day with just Oliver and I. Since then, we expanded the team one by one, and over the last few months, it’s really gone crazy. We’re now 20 people. We just hired our first Student Assistant and we now have a marketing team of five people. Our Twitter engagement exploded, and I think it’s because of the amazing work of our marketing team. There are so many developers joining, I don’t even know the number — it’s about 14 technical people on the team with five new developers. DevOps, QA and Backend developers already signed contracts and will start in a few months. We’re running out of office space. We’re currently renting four offices at WeWork Sony Center. We’re just going to hire more and more people to support this amazing project we’re building.
S: This is kind of a random question, but nowadays when people are weighing out projects, they go through teams and look at how many members there are. When did you guys know it was the right time to expand? Your focus is very different — you have a pretty heavy marketing team too.
M: Sometimes, it just doesn’t make sense to expand the team because it is much quicker to go for one or two months with the team you have — with a very small, core team that pushes forward really strongly. But there’s also a time when the work is just so much that you need to expand, especially in a blockchain project where you have a large number of different products in the pipeline. For example, at Lisk, we have Lisk Core, Lisk Explorer, Lisky, Lisk JS, Lisk Nano and we have the SDK coming up. There is a really big number of products and it’s not like one or two developers can work on all of these different products. You have to have a team dedicated to different technologies and projects, so that they can really get deep in it and deliver a good project and product. The way to know when it is the right time — it’s really difficult to know in my opinion, but you will feel it when the work gets to be so much that you’re sitting in the office for 16 hours a day just to get stuff done, and you’re like, “Oh man, now I need more people.” For marketing specifically, we felt like our software has now reached a point of maturity and stability and now it’s time to market it — to start the marketing machine so that when the SDK comes out, we have a strong foundation to build upon. I don’t think it makes sense to have zero social media activity. We have a great product; that’s why we started hiring marketeers.
S: The Neo Effect is a popular topic of discussion on Facebook. The idea that every time there is a rebrand, the price of the coin just skyrockets. Not only will you have people who love the company, but honestly, I think companies may get more haters. People may say the product is overhyped and what not.
M: I don’t think so. I think haters are often louder than ones who like your product. I don’t know where the incentive or motivation comes from to hate our product. But this happens every day on the internet, it happens on BitcoinTalk and Twitter, you see it all the time. We are doing a rebrand like NEO did and in my opinion, it isn’t a hype element, it’s just a move toward a more professional brand and outward appearance. We merged the rebrand together with a better user experience; I don’t think a rebrand should be used as a hype element.
Y: I have a question for you about your team. You have a very strict requirement that in order to be a full-time employee, you have to live in Berlin and work with you in person, like geographically centralized. Is there a reason you chose that policy and what are the benefits of this policy?
M: So, it’s definitely essential for a team to be in one place, in my opinion, because a centralized team is the most efficient team you can have. People all over the world spread over different time zones, only communicating on Slack or Skype, it’s never as efficient as a team coming into an office at 9:00 AM, working together all day. If there is something which needs to be defined or discussed, someone can say, “Hey, we need to talk” and they can do it in two or three minutes. They can define everything, discuss it and that’s it. We make it mandatory to be in Berlin because Berlin is the blockchain capital of Europe. Even though the legal foundation of Lisk is based in Zug, Switzerland, it is very hard to get developers to move to Zug. It is a small village. It is an amazing and beautiful place but it still very hard to get the best of the best developers, designers and marketeers to Zug. They all want to come to Berlin because this city is just so amazing. If I’m talking about the pros, you can really build an amazing team culture. It’s not just about working and contributing on GitHub, it’s also about fulfilling a new life with what you’re doing daily. It’s also about having an amazing culture of team members meeting after work, drinking beer, playing ping pong — that’s just amazing. Everyone comes into the office in the mornings just wanting to work. They’re not sitting at home alone and just think “Today I need to do two more pull requests.” Regarding cons, we sometimes can’t get certain talent because not everyone is willing to move, but I think there are so many great people out there that if someone is not willing to move, we’ll find another.
Y: I have some follow-up questions. The point of being an open source project is that you can try to get contributions from developers all over the world, so if your project is too centralized, can it undermine the aspect of being open source? You have to decentralize at some point. The foundation cannot take care of everything after five to 10 years, so when is a good time for you to decentralize development to voluntary participation?
M: First of all, open source means that the code is open. In my opinion, we are trying to be as open and transparent as possible, but we are not a project in which everyone can do whatever they want to do. We are a startup, we have a roadmap, a vision, the guts to fulfill this vision, and we can’t just accept every pull request. It is very hard for people who are not sitting in our office to understand 100% of the vision and with the plan we have, we sometimes just have to say, “Sorry but no, we can’t accept your pull request, it doesn’t align with our vision.” This will increasingly become the case because, for example, with Lisk Core, we want this ultra lean and simple core and it means secure. If people start to add random API calls for example, which may not even follow the RESTful standard, then it often means bad code. We also have so many developers sitting in our office, it sometimes takes longer to review a pull request than it does to add the necessary functional and unit tests to know that it really performs as it should. It sometimes even takes longer to do this than the developer just writing it himself. We always welcome community contributions, but as a startup, it’s very hard for us to just accept any pull request. If we are now speaking about the user interface, it will be an absolute no go. We can’t do it because every element we will add needs to go through a rigorous scheme of conceptualized thinking: about the user experience, the user interface and then the design. We’ll then maybe even do an A/B test to see if it is really working. We’re becoming quite rigid in that sense, but I feel that to really get the most out of the project, it is necessary.
Y: It might make sense when you’re actually bootstrapping the project, there are a lot of things to do and when you have limited resources, you can only focus on things that matter the most. There will be a point when you’re going to have to decentralize it; you can’t control and rely on it staying this way forever.
M: The great thing about Lisk is that it is modular. Later on, anyone will be able to add any feature you might need or want with the sidechain. If someone wants privacy features, he can implement them as a sidechain. The time when we’re ready to decentralize it, give it out to the people will definitely be when the SDK reaches a certain point of maturity. It will be when the core is so damn stable and scalable that it can support thousands of transactions per block — then we’ll give it out. At the end of the day, as I said earlier, it is open source, anyone can fork the project and make his own chain. But definitely for the first five years, we need to focus really hard on the product to make it really nice.
Y: Is it more that Core is a tightly controlled, clean, simple and secure, but you can do all the experiments with the sidechain?
M: Yes, exactly, sidechains are extremely useful for experimentation because in case you break a sidechain, you won’t break anything else. But of course, sidechains are also extremely good for real, useful features later on. What we will definitely build later is an identity app, so that if someone says, “Hey Max, why don’t you integrate an identity system into the mainchain” I can say, “No, listen, this would just clog up the mainchain, it would make Core more complicated, we’ll do it later on a sidechain.” We are here for a long time, we now have 78 million euro in capital, we will build this project for years.
S: The thing is sidechains came out in 2013 and although I think one of the biggest weaknesses people point out when they talk about sidechains is how in the world are you going to get miners, at least in the original Bitcoin iteration, to secure the sidechain. It’s a big enough problem to secure the main Bitcoin chain, but you guys have a Delegated Proof of Stake (DPoS) problem. What was the realization when you said, “Hey, if I choose a DPoS, I’ll have less of a problem if I choose sidechain + Proof of Work (PoW) vs. DPoS + sidechain”?
M: First of all, when u say sidechain came out in 2013, it was maybe the idea. Sometimes, it takes years to implement an idea. With PoW, I think there was a solution with mining where you are mining the sidechain and the mainchain at the same time. But I am not a big fan of mining, we implemented DPoS for several reasons and one of them is that it will be much much easier to find a few delegates renting 20 $30 nodes and this way securing a sidechain. In most cases, a sidechain will not require heavy computation. That means you might even be able to run two or three sidechains on one node and if the sidechain becomes very popular, the amount of transactions and computation grows. You’ll need to ramp up your server and probably just need one chain per server. But with PoS or DPoS, you definitely remove the high-cost factor of securing a chain as well as the entry barrier. Buying a miner or switching your entire infrastructure from one blockchain to another, which might not even be that profitable, requires a bigger effort than just renting a server and running the sidechain or mainchain on it within a few seconds. You saw it recently with Bitcoin and Bitcoin Cash — that because of the difficult adjustment in Bitcoin Cash, the profits were sometimes enormous in Bitcoin Cash and the miners were switching like a binary shot, 80% on one, 20% on another, then it switched back and forth. Luckily, we don’t see that with DPoS.
S: People are still gaining a lot of Bitcoin Cash. It takes the whole meaning of having difficulty adjusting an algorithm that basically broke the system. Maybe I’m just not curious or familiar with this. Let’s say I’m a regular person running my node, and then simultaneously I can also be running another node to secure a sidechain — what will I need to pay for it? Will I just need to stake some more LSK?
M: For the mainchain, you don’t need to stake any coins, what you need to do is collect votes from other people on the network. Everyone who holds a Lisk tokens can vote, if someone owns 100 tokens, he owns 100 votes, and now he can vote for so-called delegates which are registered accounts on the network. By voting you are not sending the money to him, you’re just pegging your number of votes to his account. You can also rent out your own stake to others, but there are no registered delegates. In the Lisk consensus algorithm, those delegates with the most votes, 101 of them, become active delegates and they’re generating new blocks. If the sidechain would use DPoS, it would be very similar, but of course you can also change the system on the sidechain, you could run a regular PoS like Nxt does.
Y: You have sidechains, and they said previously that sidechains can implement their own consensus algorithm. What if sidechain implements PoW, will it work?
M: Yes, it is better to have its own hashing function, otherwise someone points out a few miners to it and creates a 51% attack, but of course you could also implement PoW into sidechain. You could even merge-mine it with a Bitcoin blockchain, but I think no one is going to do that because PoW in my opinion is obsolete, it is just not efficient or easy enough for sidechain. If you implement nice incentive structures into a huge sidechain, which does an IPO and collects millions, the miners will all jump onto it, then it will work, but is it really the best system? I don’t believe that.
Y: What is the main functionality of the mainchain? What is its role?
M: We already said earlier that the mainchain is the super lean and simple version of a blockchain with not many features. The mainchain is a sort of mother chain which all other sidechains are connected to. That means the mainchain acts like a basis of the whole platform which you can send transactions on back and forth to other users. You can register your own sidechain on it or initiate a transaction to a sidechain to transfer tokens to it. Later on, we are also going to implement ICO mechanisms, so that you can issue your own token on the mainchain, do an ICO, distribute it to your investors and then develop your sidechain. You can later transfer the token to the sidechain. When I’m talking about transferring or sending to a sidechain, it is more like a replication gets locked on the mainchain created on the sidechain, then u can freely send it around. If you want to send it back or withdraw it from the sidechain, then you basically destroy and unlock them again on the mainchain. That means the mainchain acts as an in-between settlement layer. You could also register delegates and so on, but that is for the consensus on the mainchain.
Y: The mainchain contains a blockchain?
M: The mainchain is a blockchain. It is the blockchain of Lisk itself and then let’s say it’s in the middle, then you can create new blockchains, a sidechain which is running independently, which has a link to the mainchain, so you could potentially list all sidechains in existence because every sidechain has a registration on the mainchain. This means an entry in the mainchain in the main blockchain.
Y: The state of the sidechain is committed to the main blockchain?
M: Potentially, that would be possible, it has some scalability problems. If you now have 10,000 sidechains and they are all reporting back, yes, this would have to be involved with a fee. That would mean you could secure the sidechain on to the mainchain. Take a block safe as hash on the mainchain — this would fix this block in time. I personally believe that it is not necessary, as long as the sidechain is big enough to attract enough delegates, forgers and miners.
S: I remember there was a time, when I got into Bitcoin, sidechain used to be the buzz word. Everyone was talking about having parallel chains, but now you have other scaling solutions, like lightning that appeared in Plasma, Cosmos, and Polkadot with their parachain, why do you think sidechains died out for a second there? I started picking up the sidechain paper again because of Lisk.
M: I don’t know so much about PolkaDot’s parachain but if you take a look at Rayden, Lightning and Plasma for example, or even the solution Cosmos has to offer, it is just that they are basically outsourcing it from the mainchain. They’re putting it into state channels, payment channels, sidechains or these other chains you just named. I think the blockchain space itself has a problem of too much theory. There’s just too many mathematical people in the space and they just love to create white papers, but they don’t love to actually work on the solution for the problems they describe. They just grind the stuff down and think about all the mathematical implications and so on. You studied math, and I think there was just no one who picked it up and realized how to do it. We knew the concept for a long time and we think it is a great system. We also think this is not the scaling solution. It is more that at one point, you probably have to implement many different scaling solutions to reach this point of scalability, which is enough but it’s one and it’s a relatively easy one in terms of scalability. You split the data into multiple blockchains, so it’s just a logical one. Maybe the people in the space are too intelligent that for them it’s just too easy of a solution, why split it into multiple blockchains if you could split the data amongst millions of nodes. Not every node sees every kind of data they probably think it’s better to focus on more complicated solutions than grabbing the low-hanging fruits first. We actually think it is an amazing way in the beginning — then you can implement the other scaling solutions individually on the sidechain.
S: I think another argument that people make is Bitcoin already set sail; it’s in the middle of the ocean, weathering the storm. All the others haven’t set sail yet, they’re chilling at the port. Maybe some will say Lisk hasn’t gotten to that size where you have to really start thinking about scalability. And then you’ll have all these scaling options and you have to pick which one is the best one to implement at this current moment.
M: I think if you’re at the point when you have to think about scalability, it’s already too late. You have to think about scalability before. For example, Lisk had to artificially limit to the number of transactions to 25 per block. We’re not at the point where we have to scale up. At this point, we could easily scale up, but we don’t see the need. In terms of scalability, it is always something you want to have leeway with, you don’t want to reach the tipping point. If you do, it’s already too late. I agree that Bitcoin is sailing in the sea already, but so is Ethereum and Dash. There are many blockchain projects out there with significant transaction volume. I don’t know from which speech that argument came from, but I admire him (Andreas Antonopoulos), so it might be a bit old already, but by now, it is not true anymore.
Y: You said there is a lot of academic research on scalability solutions, but there is not really a solution that stands out which has been implemented right. I think it seems there is a really high barrier for actual implementation of those solutions, what do you think is the biggest obstacle for implementing the scalability particularly the sidechain, what are the major obstacles you will face when implementing your sidechain?
M: I think it’s going to be a mix — at one point we probably will see that it might just be a combination of all of them plus increasing capacity to really just reach the level of scalability, which is sufficient but again you could also argue that an application could get so many users that scalability never stops.
Y: What, in your opinion, what will be the major obstacle to actual implementation of your sidechain?
M: I think the biggest problem will be forks. Imagine, on the mainchain, you’re on block 1 million, and on the sidechain, you’re on block 200. Now it’s going okay, the blockchains keep going and suddenly someone sends a transaction to the sidechain, it registers it, creates a token on the sidechain and on the mainchain, the fork just happens a little bit after, and the whole blockchain moves back behind the point where it transacted to the sidechain. The sidechain doesn’t know it, so the token exists on the sidechain but on the mainchain, the user also has it. So there is a mismatch. The biggest obstacle for sidechains, there are different solutions for that, like you could always do an atomic swap, which means someone is actively selling it and buying it on one of the chains, but it is also a matter of how long you wait. If you’re doing an amazing mainchain, which is super stable, has very reliable delegates, in our case, then such a long rollback will not happen. You just mentioned you got your sidechain paper out, so if I remember correctly, it says you have to wait one or two days if you send it in, and another two days — this is, of course, a huge amount of time. I think with Lisk we don’t have to wait so long, in the beginning we just make it extremely long for security reasons, but once you’re on the sidechain, you’re on the sidechain. An exchange could potentially just say, “Okay, we accept the token from the mainchain and the sidechain,” so if someone wants to sell it on a decentralized exchange, he would not have to send it back, for example. It’s not that huge of a problem that it takes one day or two days to send a token to the sidechain, to make sure there is no rollback happening because two day rollbacks will never happen.
Y: You made a very interesting point about the inverted incentive for Core development, how can you incentivize core developers to contribute to Lisk Core and what percentage of your development team is spending time on Core?
M: We have five developers on Core — DevOps and QA, very, very soon, and they’re all exclusively spending their time on Core, but we want and need more, because testing takes quite long to write and peer review all the code. How to incentivize them? With a good salary, an amazing team culture, an amazing office and a nice Lisk bonus, which is being paid out over years. That is a pretty good incentive, but this is just monetary. I think the best developers don’t do it solely for money they do it for passion as well, and to be honest, I can’t imagine a better job than working on blockchain technology. If you’re a JavaScript developer, Lisk is your only serious option in the space. It’s a pretty good job and they love it. I can fully understand that, it is not so hard.
S: Didn’t Lisk have a network halting problem, I don’t know if it’s a forking problem but called a halting problem?
M: It’s not such a serious problem, it’s just bad PR.
S: Serious enough to write a blog post about it?
M: Of course. We are fully transparent. We have a non-profit and we always do as much as possible to stay transparent, so of course we issued a blog post about it. First of all, it was a halting of the blockchain and this is super bad PR. Super bad for marketing, Thomas would agree to that, but it is, in itself, not a huge problem in a project that is really early stage because it just means the blockchain stops. It stops at one specific block height, then we would have to fix the bug and release a new version. The delegates will have to run and install it on the nodes and then the blockchain keeps going. It is very normal that a very early product has some bugs, you know that’s why we are focusing so much time on tests like unit tests, functional tests, behavioral tests and on quality assurance to make sure there are as few bugs as possible. While it is software written by humans, it is good that it stopped and not forked because that means all funds are safe. In our case it always stopped, there are some rollbacks with a few blocks, but it is normal, that’s why we always say okay please wait for like six block exchanges, Poloniex has 300 confirmations required to accept a transaction, but like, small rollbacks always happen on the blockchain, like one or two blocks and that’s not a problem if there would be huge rollback of course that would be a real big problem. What’s happened basically is just that there was a backward crash on a node and this bug was being created by a transaction and now we had a node that was forwarding this transaction to another node. It received it and crashed while verifying the processing here and it also sent it further, so it’s created a chain reaction. The whole network shutdown but within a few hours, we released a new version. That is the advantage of having a centralized team, or having a team at all, we issued a blog post to be transparent, we openly communicated this months ago.
S: Bitcoin had its fair share of accidental forks in the past. Forks have basically become a swear word, like the worst thing to say, anytime anyone were to suggest hard or soft fork.
M: It is mostly about hard forks, which do not fix a bug, but introduce something new, making it more complicated. I think it is good to establish in the early days of a cryptocurrency, and early days for me is three years, because it is complicated technology. To establish a culture of hard forks, we call them upgrades by the way, the community accepts change and it should be this way. Technology can’t move forward if it doesn’t change. We see it in Bitcoin’s case, it is Bitcoin Gold. It doesn’t really move things forward. I am so glad that segregated witnesses updates have been accepted, I hope that in Lisk’s case, everyone will just accept the great things we’re working on so passionately every day.
Y: I think you slightly touched upon the very interesting topic of hard forks. Which one is your preferable option of evolving your blockchain platform? Hard vs. soft fork — what are the benefits and the right balance of those two?
M: It often is being introduced with a milestone on a specific height, if you don’t update, you’re landing on another road, another chain, basically, a fork — it depends which side you’re on. For me, a soft fork is better because the implementation itself makes it unnecessary to introduce a hard fork. I think both introduce new features to the system, but if it is possible to do a soft fork, you should do a soft fork. If it is not possible then you just have to introduce a hard fork.
Y: Do you think a hard fork is a necessary evil?
M: It is not an evil at all. If you upgrade software in a way that you put so much work into it that it gets so much better — tests, many new and cool features that it’s just not backwards compatible anymore, then it is good, not bad. If you have to delete all transactions or put back into action what happened in the past, then this is a little bit shitty, but sometimes also necessary.
S: How do you get your delegates to cooperate during upgrades?
M: I think we have one of the fastest upgrade cycles in the industry. After four or five hours, every delegate has updated hundreds of nodes, mainly because delegates are earning a lot of money and they don’t want to miss it. It is also because people believe in what we’re building, they’re expecting and looking forward to any small update being introduced to the system. For example, in Bitcoin’s case, there is no leadership involved. It is a completely decentralized project and there is a group of people which just evolved into publishing Bitcoin Core releases. In the case of Lisk, there is a whole legal foundation, a team and founders behind it, pushing it forward how they see it fit and right. If people don’t agree with the vision, they can get off the project and create a copycat, but in Lisk’s case, for example, while there are many good reasons to follow because we have the funding and the team to make it a reality and a great project.
S: It’s a double-edged sword, I think, those hardcore “everything should be on the protocol, done algorithmically” are going to hate DPoS. There is just such a random variable, it brings in all the goods and evils from past institutions that we’ve seen onto the blockchain. I read somewhere that you guys are planning to change your consensus algorithm later on in the future, that you might be considering others, in terms of the social aspect, where is your end state?
M: First of all, these hardcore people who want everything algorithmically solved, there needs to be currency which delivers that and I hope that Bitcoin is that currency, like a digital gold that people can always rely on, which introduces updates nearly never, which just runs. But for many other blockchain projects, it’s still too early. Maybe in a few years they can demand it, but not in the alpha or beta stage, it’s just too soon. About the other part, yes, we’re thinking about changing the consensus algorithm. It’s very likely that it will evolve. It is not a radical change from A to B or black to white, it will evolve and after several steps, it will likely be completely different. But we will likely start working on actual implementations somewhere in the middle of 2018, because it’s working and running and we don’t see a big need to change it at this point. This is especially with delegates watching. In Bitcoin’s case, the price just shoots to the moon, I’m looking forward to see what happens in Lisk’s case. I think it’s great to have a social element in a social blockchain, so as I said, Bitcoin has to be this algorithmically solution, which people can just trust. They just trust mathematics, they just trust the underlying system, it comes down to a social element, but we are not trying to be that kind of a system, we’re trying to be a project built for people to get introduced to blockchain development which should come up with great use cases and there’s always some kind of a social system involved. I think we don’t want to take it too far, but the way it is right now is quite good. It is about federated delegates in a system, which can be anonymous or known, but they’re real people. It’s not just some mine running somewhere, one of 10,000 mines in a huge hall. It’s always a real person taking care of the delegate. We were just in China where we met ten delegates. It’s been amazing, it was real people who were taking care of the nodes and believe in the system, they’re really regular people, not some sharks, just normal people who are supporting the living network.
S: Is it hard to run a node?
M: No, it is not hard to run, but hard to master it. At the current time, it is just setting up a node, typing in a few commands. We made it super easy. Isabella did a great job on the diploma script, it’s so stable, you just need one node and it runs for weeks. Back in the day, in the super early stage, people came up with failover scripts. They ran three nodes and when one crashed it switched, this was the case once a week but now it’s just running for months with no problem at all. We are Linux and UNIX only, so not for windows, because servers are running in most cases on UNIX-based operations. What I meant earlier with hard to master is there will be a time when there will be so many transactions on the mainchain, like hundreds and thousands, maybe millions every day, where you need quite powerful and secure hardware. That means hardware which is being protected from DDoS attacks, hardware which is so powerful that it can sustain this mess of transactions and then it will become a business. Right now, the delegates are running them for $70,000 to $80,000 a month, and I think later on at some point, they will maybe earn hundreds of thousands a month. That means it will evolve into a real business.
S: We gotta become a delegate. But how do you stop delegates from colluding, working for the benefit of one another, or would you view that as them working for the benefit of the Lisk chain?
M: They already work together, trading votes with each other, campaigning together, with banners. Two or three delegates are putting money together, starting a campaign together, so that’s already happening. They’re doing it in order to stay where they are, they have the biggest incentive of everyone to just keep going with the blocks because they’re earning a lot of money with it. This will still be the case until we let the consensus algorithm evolve. Even then, the incentive will, of course, be super strong, and that’s why it’s so secure. Would you give up on $80,000 a month?
S: Hell no, it sounds pretty good.
M: That’s like 101 people saying “Hell no!”
Y: But here’s the thing, because delegates have the incentive to try really hard to get votes and they’re campaigning with each other, it sounds to me like an election campaign.
M: It is similar.
Y: And then people are economically incentivized by the delegate, so isnt it like a delegate explicitly buying the vote from the people? You might see corruption and some bad things being done, like in politics, in this system.
M: What you mean is that delegates are giving back forging rewards to voters to incentivize them to keep on voting for them?
Y: It sounds like “If I get elected, I’m going to give you $1 dollar to each of my supporters.”
M: Yes, that is happening in most cases. Delegates make it dependent on how many votes you have given them. Basically, you see corruption everywhere and in the whole blockchain space, you see all this corruption. But you also see it in economic crises, pensions trading, everything that happens in the real world also happens in the blockchain world on a smaller scale. I think a basic law of the universe is that these things happen. While the world still keeps spinning, Bitcoin keeps running, even though there is huge corruption involved there every day. So, yes, it happens, but it doesn’t destroy the system. It is a self-balancing system; if corruption gets too high, or some problem grows too much, the system will automatically balance it out by maybe kicking the delegate out. The number one delegate has around 30% of votes — that means 30% of the whole network voted for him. The whole community, all Lisk holders, can get together and do something.
Y: This all depends on the assumption that people will find the right balance for a system, the whole point of blockchain is to minimize human intervention.
M: The whole point of PoW is to minimize social intervention. The point is to implement some sort of social aspect. I tend to agree that maybe the social element is a little bit too strong, and it will probably evolve into a consensus algorithm where the social element is not that strong. But for now, it’s totally working fine. The price is a little bit too high for my taste, I didn’t expect for that to happen so fast, but that’s how it is and it seems to work great. Again, it’s because the incentive is so high.
Y: It’s an exciting experiment, it’s interesting to see how people react.
M: We are now approximately one and a half years old. Ethereum is much much older in blockchain terms, and some people from Ethereum say it is one huge experiment — a social experiment, a tech experiment. I say the same thing about Lisk. I say we are in a better stage of it, with our main product not even having been released yet. People should not expect something finished, they should expect an experiment at this stage, which will evolve to become an amazing product in the future. But, at this time it is still high risk and high reward.
S: In your opinion, what’s a healthy node ecosystem? Some people might say that everyone should get one chance to be a delegate, to taste that $80,000 a month. Everybody should get a fair distribution of the chances.
M: All these crazy examples of re-voting every three months is quite dangerous because if people don’t re-vote the whole security of the system goes down, basically now with 30% for the first delegate, it becomes more and more secure over time because more and more people vote. They figure out, “Wow, what’s this voting system? Oh, I can earn Lisk?” so they do it. There are many use cases where people just find out about this voting aspect and that means more LSK in general are voting. And that means the whole ecosystem becomes more secure.
S: That’s truly interesting.
M: It also becomes more rigid, but I mean, if it’s running great, it’s running great. At one point, as I said, we will change it, but up to this point, I don’t see a problem with it.
S: Even before you get there, you can imagine — you don’t know whether the 101 nodes are run by 101 individuals. What if there is a delegate who owns five nodes?
M: There might be, but I know nearly all the delegates.
Y: Is there a delegate locking period?
M: So there are 101 delegates. Block each take ten seconds, that makes a round approximately 17 minutes long. If you are now voting within this round for a new delegate, the changes will only be put into effect once the current round is finished, so every 17 minutes it checks what the current state of votes are. It would change it so in the worst case, that’s like 101 blocks. Within one round, you can vote for one and then immediately, in the next block, remove the vote and it would not change anything because the change will just be registered at the end of a round.
S: How do you tackle that ease effect where the rich get richer and the poor just continue to get poorer?
M: This seems to be a normal element in this world, even with no proofs, in the real world, the rich get richer but in Lisk’s case, it’s DPoS, so it is not necessarily the rich get richer. It’s the most popular ones get richer, it’s not that you need a lot of money, you just need a lot of votes.
S: So it’s like high school.
M: Being popular always helps. It also helps in Lisk, that means the Matthews effect doesn’t catch here. It doesn’t grip here, but I still think it’s normal. We see it in Bitcoin, the rich people in the beginning could buy more mines, they’re getting more rich now, they can buy more mines with the money. They’re getting more rich. Like a few years ago, people who invested in real estate in some cities who could invest made more money. And now, they have even more money, which they can again put in other projects, it’s a normal cycle.
S: Maybe it’s unrealistic to think that somehow algorithms can solve all these things.
M: Let’s say you have $100 and you invest it, you make 10%, $10. If you have a million dollars and you invest it, and make 10%, it’s $100,000. It’s normal, otherwise you’ll have to create an unfair system where the rich get less and that’s not how the world works.
Y: I think it’s more of a problem with capitalism, blockchain actually boosts capitalism. You mentioned weak and strong leadership once, so can you elaborate a bit more, when weak leadership is required and when strong leadership is required?
M: I mentioned it in our first interview. I think Lisk is a startup and a startup always has strong leadership; someone or two people who say where the ball is going, where the project is moving. I came from the Nxt community where there was no leader at all. It was a completely decentralized project. The initial founder was anonymous and he ran away, same as Satoshi Yakamoto, somehow no one really evolved to become the leader. There was some effort, I put a lot of work into it, Icreated a Nxt organization, I made many efforts and campaigns. Of course, I was not the leader of Nxt, there were a lot of people who did a lot of work, but they were not leaders. The project succumbed because there were no clear path. Everyone wanted something and they implemented everything, it was just a mess. With a very clear and strong leadership, you can build this lean and useful platform, which attracts millions of people, but this of course makes it mandatory that the leadership is a good leadership, a strong leadership that people trust and believe in the vision of. I am a big fan of that. I think decentralized projects need this kind of leadership in the beginning to really find their own way, or else they will just give in to the many voices on the internet and this will kill any project.
Y: I guess it’s really strongly based on your experience at Nxt. You said that many people want different things, what is the criteria? Perhaps in terms of what code to merge and what code not the merge?
M: When we receive a pull request in the middle of a sprint, everyone is excited, because everyone loves open source. But then we remind them, “Hey guys, we’re in the middle of a sprint, put it on the backlog, then we take a look at the backlog and what is left.” Then we analyze what it is trying to solve and figure out if it is good code. Does it make sense to implement it? In most cases, maybe one of the four criterias are met, it’s just complicated to write good code, it’s nearly impossible to know where we are headed to in terms of small details, if we need this API code now. No developer likes to write tests, so that’s often also missing. It’s very hard for us to accept a pull request and we really don’t like to see that, but it’s the harsh truth that we have these super high demands and that it really has to meet all criterias. We can’t just merge something because we like the guy or the idea. We are a startup; some people may not get it, we can’t just implement everything. Blockchain is open source, it has this aspect mainly because Bitcoin didn’t have strong leadership or a clear roadmap back in the day, so people just assume it’s like that with any new project. We see it more and more, the latest blockchain projects are real startups with funding, a team, a brand, a vision and a roadmap.
S: Ethereum has consensus and there are several Ethereum-based development production studios around the world. People are excited to build on this platform together, do you foresee something like that for Lisk in the future?
M: People wanted to build blockchain technology but they just couldn’t do it. They can now build smart contracts, but they can’t build their own blockchain because it’s super complicated to do it from scratch. When we release our SDK, it will become possible for the first time in history that developers can build their own blockchain — that a university can experiment with consensus algorithms on their blockchain in a way where everything is nicely encapsulated in an SDK. It will be modular and you can replace various functionalities, go into the code and change things, it’s crazy. We’re already getting contacted by so many people who want to build blockchain apps on the Lisk platform, that want to come up with their own sidechain. There is the Lightcurve Foundation, which was founded by Oliver and myself, which is the contractor of the Lisk Foundation, which takes care of the development and marketing, where there will definitely be some kind of a consensus. It is an incubator where we will develop our own apps, we will make it possible for team members later on to build their own startups and get help. We will say “Hey, I think your idea is great, do you accept $50,000 as a seed investment, we will invest in your startup for this amount of tokens” and they can become entrepreneurs and repeat the Lisk story. I foresee this within Lightcurve and for many other people.
S: I think it’s awesome, just the nature of JavaScript, it is very easy to write. User interface-wise it is easy to create something comfortable.
M: Not much will change for developers, it’s just some principles in which blockchain technology is unique. You have an immutable database. At the end, it’s just JavaScript code and people are used to that. In our case, a developer working on Lisk Core, at Week One or Week Two, he started to get productive. After one or two months, he was deep in it. He didn’t need years of blockchain experience, they just need to be talented and have experience in JavaScript and Node.js, focus on the code and he gets it very quickly. It’s much more complicated to introduce updates, but it’s not rocket science.
S: It seems like modular programming is the way to go.
M: Definitely, this way later on you can set up teams working on specific modules or may take the decentralized approach and say, “Okay guys, we’re done with the project, you can now work on your own consensus algorithm.” We plan to allow every developer to create his own modules.
Y: What do you think is the best thing to have — single or multiple language implementation?
M: There are basically two camps — the Buterin camp saying different implementation is good and the Todd camp that says it isn’t good. Buterin says consensus failure is less likely this way whereas Todd says more likely. I agree with Todd, but we can’t stop anyone from transitioning from Node.js to C++ but we at Lisk definitely concentrate on one language. If we are talking about the SDK, I can see developers creating different implementation in different languages even if just for fun, but that is less of a problem. I am just focused on what I’m doing and I think developers need to be focused. We are now more than 20 people, even though we are up to here in work. It’s just not feasible to split the whole team to do it in different languages, it would create a lot more problems than solutions. I definitely tend to learn more toward Todd, but I say both ideas are sound. Both reasonings are sound, both are great minds, but the Todd way is the Lisk way.
Y: If someone comes to you and says, “I’m going implement Lisk in a different language”?
M; We wouldn’t stop them. We just need to stay focused. We do all our hiring on Node.js.
S: This is for people watching from home — in Korea, we get a lot of questions from non-technical people saying, “Do we need to learn how to code?” or “How much does making a mining rig cost?” We get those questions the most. The image is “I really like blockchain tech and crypto and I want to contribute in some way, but it seems like the only people I see contributing and getting economic benefits are the developers or the miners. If regular people want to contribute to Lisk, what are some ways do you recommend getting involved?
M: I think someone who asks how much a mining rig costs is clearly not deeply involved yet. I think, first of all, yes, my background is in electrical engineering but I dropped out for Lisk. I never felt like an engineer, I felt more like a business, startup guy, I contributed so much to different projects in the space, doing PR and community management. Anyone can contribute — I know an artist and he contributed by painting pictures of crypto symbols. Anyone can do their fair share of work, no matter what you’re doing. Let’s say you’re someone who is making tables, you could support your local blockchain startups by giving them the best tables in the world for a discount. Everyone can contribute and do something, it’s just the naysayers who say they’re talentless and can’t contribute in any way.
S: Do what you love and get creative.
Y: The impression sometimes is “I can never get to the core and make the greatest impact.”
M: But that’s greed. Is it so bad to do marketing or community management? Everyone has different talents or interests. If everyone says they really want to go to the core, maybe he should learn coding. Everyone can contribute — they could follow us on social media, help us with transactions, answer questions, it’s just a matter of doing it. Every single retweet, or like on a Twitter post is amazing, we consistently get 300 retweets, our marketing team is doing an amazing job, everyone can do his fair share of work. It’s about becoming creative and thinking of ways to contribute.
S: Right here, at the bottom we should have a link for the “I love Lisk” girl, so that everyone knows there is a girl in Japan who absolutely adores Lisk.
M: Yeah, everyone knows her, that shows how everyone can contribute. She did it in a way that everyone could see. I invited her to our meetup in Tokyo, she is from Japan, she needed a few LSK to pay for the train ticket. I think she did a great job, so why not support her? I think people with creative ideas should be supported and this was definitely creative.
S: Final question is — how can developers from different countries contribute — what are the first steps to start contributing? In Korea, we do have a language barrier, it’s hard even if you’re a JavaScript developer. It’s hard to understand Lisk and to figure out everything.
M: First of all, every JavaScript developer speaks a common language — JavaScript — I said it in the beginning, we’re building a team, we’re building the software in the first and foremost way. Developers can contribute. How can they get into it? The documentation needs to be top-notch, I think most developers know English. Maybe in Asia, not so much, you probably have a lot of local resources. I think if it’s a good developer, he can definitely read the code and know the standard. We saw it in the past already that people who barely speak English contributed with pull requests. The descriptions were bad but it doesn’t matter, code is code. We also accept many pull requests, so it’s always a balance. I think the best way that people can contribute later on is by writing modules for the SDK by making things better and better. By forking any product we have and just working on it to make it better. Maybe creating alternative systems for example, an alternative user interface. We’re also really actively trying to integrate developers into our community. That’s why we take on this whole modular approach, even Lisky, which is a command-like interface where you can use Lisk functionalities within the terminal. Even there you can write your own plugins and transact the functionalities. We’re trying all kinds of things where people can contribute in different ways as a developer. So there are some things, if you talking about documentation being translated, and it already is really hard to come up with English documentation.
S: Really good documentation is hard to come across.
M: That’s why we focus on English and try to get it right. And then maybe we can try it out in other languages.
S: We covered a lot, I guess.
Y: Please tell viewers what they can expect from Lisk in the future.
M: Currently, Thomas and myself are in Korea. The next target is Tokyo, Japan, after that, Mumbai, India. Korea, Japan and India are first-time market entries, so you can expect more activity within these countries in the future. Our upcoming rebranding, which is introducing a whole new user interface, design, positioning, and user experience for all graphic user interfaces we have: the website, Explorer, Lisk Nano. And of course, the logo. The name will not change. That’s one of the major upcoming things. Maybe just to quickly explain what Lisk is doing at the end of this tour — Lisk is a platform which enables JavaScript developers to create their own blockchain applications. We achieve that by developing an SDK which makes it simple, which makes it easy for developers to handle blockchain technology. Around that of course we are building a platform, we will basically take a developer by the hand or an entrepreneur and inspire and give you the necessary tools to come up with whatever you are dreaming of doing in the blockchain space. We will help you out. We will actively support the project, we will help with marketing later as well. We’re written in JavaScript. We started in 2016, we currently have funding of approximately $78 million. Just recently, we did a major push in hiring, I mentioned it a few times during the interview, we are now 20 people. Sitting in Berlin, Sony Center, in four offices in WeWork, but quite close of course. Basically, we want to make a difference in this world by just making blockchain technology accessible for regular developers. We think the technology itself is great, it’s a paradigm shift, we really want to see more experimentation happen. We really want developers to play around with it and create amazing ideas and startups which evolve into powerful corporations later on. We are big supporters of this idea of a trustless network, a decentralized network. Lisk is the way we help out to achieve a future which is decentralized.
S: How can people contact you?
M: First of all, our website is www.lisk.io. The Lisk social media handles are the following — Twitter is LiskHQ, Facebook is LiskHQ, Reddit is Lisk. You can contact us on our website in the Support Center, if you have a question or a problem with our products. If you have a business inquiry, you can contact us with the Contact Us form at the footer of the website. My Twitter handle is MaxKordek. Please don’t bother me on Facebook, Twitter is enough. I really have to say we are really open and public with the community. There are always some guys from our team hanging around on Gitter or on Lisk chat, so just go to our website. You will find all the social media channels, you will find ways to contact us, so if you need something, just contact us, but please only serious matters.
S: Thank you Max for propagating straight from the center of Seoul. We’re the Blockchainers and we’re here to tell the world and the internet about everything that’s going on in Korea, about the blockchain ecosystem, bringing you news, interviews and bringing you all that general goodness and Korean soul to the world. If you’re up for any collaboration or projects, we’re always open to that.
