WBD095 Audio Transcription 

WBD095+-+Interview+with+Sergej+Kotliar+(Banner).png

Sergej Kotliar From Bitfrefill On Using Lightning In Retail

Interview date: Tuesday 12th April 2019

Note: the following is a transcription of my interview with Sergej Kotliar from Bitrefill. I have reviewed the transcription but if you find any mistakes, please feel free to email me. You can listen to the original recording here.

Bitrefill is leading the way in building support for Lightning in retail, both by accepting Lightning payments but also with their Thor product, selling channel capacity on the network. In this interview I chat with their founder and CEO, Sergej Kotliar, we discuss the challenges with Lightning, custodial v non-custodial wallets and UX.


“You can’t moralise for people who want to take short cuts when that means you can get something in their hands today.”

— Sergej Kotliar

Interview Transcription

Sergej Kotliar: You are doing a great job at getting people to speak. You're doing 100% the right thing. I would like the other opinions to come out, but maybe this is my chance!

Peter McCormack: Well, yeah, here we are! You've got a chance to speak. Obviously you know I'm doing a Lightning month so...

Sergej Kotliar: But there is an interesting aspect around the Lightning is that the same people that say they think that Bitcoin is a store of value, are the same people that are excited about Lightning, which is very contradictory!

Peter McCormack: Yeah, it is and it isn't. This is one of the problems I run into doing this, is that there are so many strong opinionated people with contradictory opinions and trying to get to a single truth is very hard. But one of the things I'm kind of getting to, myself personally, is actually I only really care about what I want to use Bitcoin for and Lightning for.

So right now I use Bitcoin and I do use it as a store of value, I hold value. But I do use it also to pay people or people pay me. I issue invoices to some people in Bitcoin, which is useful. It's much better than bank payments. I also am starting to think how I would use Lightning, but I don't have a huge use for Lightning right now. I guess less so outbound? More inbound, like if somebody wants to send me a tip or subscribe, I have a use for Lightning. But I don't have an outbound use for it just yet.

Sergej Kotliar: Well maybe we can provide one for you over time! I remember you bought the gift card for your daughter, I think?

Peter McCormack: Yeah, my son. I did! But I guess that was a novelty to try out what you were...

Sergej Kotliar: Yeah I know

Peter McCormack: ... Before I interviewed John. But for Lightning, I think my main problem with using Lightning at the moment is a UX thing, rather than a desire for the technology. But great to have you on, man! Thank you for coming on. I've wanted to talk to you for a while. You've obviously got a long rich history in Bitcoin. You've got a very well respected business.

You now employ one of my favoUrite people in Bitcoin, which is John, who I think is amazing. But it's great to talk to you. I think just to set it up for everyone, it would be good for you just to give... You don't need to do a full backstory, but just to let people know who you are. I think more importantly tell people about who Bitrefill are because not everybody who listens to this will know Bitrefill.

Sergej Kotliar: So my name is Sergej Kotliar. I'm the CEO of Bitrefill. What we do is we try to create, we try to make Internet money useful by buying things in the real world. We want to bring about that vision of cryptocurrency, of Bitcoin, of billions of people paying each and getting paid with these coins.

So concretely, what we do, and I think what makes us different from many other companies, is that we start from the customers, rather than like in the past, a lot of it has started from merchants. We start from the customers and say, "okay, but what is it that we can help the customers achieve with Bitcoins today?" So on that path, we started four years ago now doing phone refills all over the world.

So in most countries of the world, it's still the case that Bitrefill is the only thing that you can buy with Bitcoins. Then gradually since a year or so, we've been doing more and more gift cards, which appeals more to a Western audience and we do a lot of Lightning stuff. We don't actually build our own Lightning implementation or anything like that.

The analogy I tend to make is that while we don't build the motorcycle that is Lightning, but we're really good at riding it and we know everything about it, tuning it, tweaking it and so on. We're building for a vision where there's a billion people, or even 7 billion people living and using Bitcoins every day.

Peter McCormack: Yeah. So that's one of the things I'd noticed, also with obviously your tagline, "living on Bitcoin", is it? Remind me what the tagline is.

Sergej Kotliar: We've had some different ones. Yeah, we've had "living on Bitcoin". We've had one, "live on crypto". We've had playing around with the Internet money concept, which is sort of very value neutral and it doesn't bring up Bitcoin and so on. But yeah, that's I guess the red thread of it all and we're all mostly passionate about Bitcoin.

Peter McCormack: Yeah. What I can see though with the vision for you is, or what excites you is, I think a lot of us when we are using crypto or Bitcoin, we've always got the price of our local currency in the back of my mind. So if I buy something with Bitcoin, I'm still thinking, "well, what does that cost me in pounds? That costs me properly £20, £25". I'm like, okay, I get that. I do the conversion.

Then sometimes when we travel, if I go to somewhere like Thailand, I have to have my calculator ready to do the conversion. Not so much in the States, I'm kind of getting used to that. I guess you're looking forward to a vision where people just start using Bitcoin and they're not really thinking about the conversion? It's almost that Bitcoin becomes its own ecosystem and people aren't worried about going in and out of fiat right?

Sergej Kotliar: Right. Well the conversion question, that's the unit of account setup which is the one thing that we can all agree that we haven't reached yet. Eventually I think it will happen. There are slight tendencies of that. So for example, I would argue that the number one thing that people buy with their Bitcoins today is buying other coins. Other coins are priced in Bitcoin a lot of the time. I was just checking the other day Veriblock and reading up about that and I know that one Veriblock coin is worth 1,800 Satoshi.

So then that's my reference point. I don't know how many dollars it is. I know that it's 1,800 Satoshis. It is gradually evolving, all of that stuff. It's just starting from a very low level and I think that we will eventually get there, but as always with this stuff, it's not going to be the Westerner customers first. It's going to be in a different order of things. I'm pretty sure that, everybody brings this up in as well, but I'm pretty sure that there's a big pretty good grasp of, among the Bitcoiners at least, what the value of a Bitcoin is and they don't need to do any calculations.

Peter McCormack: Yeah. So also whilst you're a Bitcoiner yourself and predominantly a Bitcoin company, you are also accepting alt coins. I guess if people want to buy things with alt coins, sell them to you and you can dump them, you're kind of coin agnostic there?

Sergej Kotliar: Well, yeah, I mean if people come to us and want to buy stuff, then that's great. I mean for a long time we were a Bitcoin only. Then during the whole fee crisis of late 2017, we kind of had to do a bunch of different things. One of those things was that we went from just Bitcoin and we added Litecoin, Dash and then eventually Doge coin and we tried to reach broadly I think.

I think if we zoom out from the rest of the world, we here, we can discuss like, "oh, what's Bitcoin? What's Ethereum? Those guys, they don't understand the proof of stake, blah blah". But half the world, they just see this as Bitcoiners fighting about different things.

At the end of the day, I think what unites all of the different coins is a vision of sort of a free internet money, free as in speech, that is available for everybody to use. It's convenient and censorship resistant and so on. Then there's like this little schisms and differences and so on and it gets very political around that.

Peter McCormack: So what are the main challenges that you face as a business trying to build a retail operation using crypto?

Sergej Kotliar: I think the biggest challenge is that the crypto world has evolved in a super unaided way. It's sort of like a teenager hits puberty and the ears grow faster than the nose or whatever. But this is like extreme. So we have the speculative use case, whatever it is, has so trumped everything else! If you look at the speculative market of crypto, it's in the tens of millions.

But if you look at the people that are using Bitcoins in the sense of earning them somehow, of paying for things or paying people, that number's a lot smaller. So it's sometimes tricky to navigate because we're trying to build this thing and I think everybody agrees that this is important and this is where it's going to land eventually. But it's also a little bit tricky when there's like these periodically huge things that pop up and sort of distort the conversation and someone shows up and does a huge ICO that does something similar to us and then they spread a bunch of money around.

Then a year later, they run out of money or they disappear and then we have to sort of react to all those things. It's a very uneven market. I mean, take the technical issues as you've discussed in the past episodes. There is a lot of UX things that needs to be a lot better. We're still super early days, so there's so much stuff to do!

Peter McCormack: But what about actually handling crypto? So one of the things that I imagine might be quite difficult, is you price with the local currency as the unit of account, but you can pay in Bitcoin. But what happens if I buy something from you for say $20... You do zero confirmations?

Sergej Kotliar: Yeah. I can discuss that as well if you want.

Peter McCormack: Yeah. So I'll come to that. But so you do zero confirmations, what happens say in that hour, if there's a 20% price fall, what happens? Are you out of pocket there?

Sergej Kotliar: But here's the thing, over time Bitcoin goes up more than it goes down. This is what makes the Bitcoin companies different from non-Bitcoin companies. We understand that. For us and especially like with the phone refills where we actually have a bunch of currencies from all kinds of different countries. Bitcoin is not one our big problem currencies. We've held stock in things like Nigeria Lira and so on.

That's more like... They can talk about some volatility! But so to answer your question, like yeah, we take exposure, we keep a good chunk of our capital in Bitcoin and we have it on the balance sheet, as we have 9 or 10 currencies, both fiat and cryptos. So yeah, sometimes it goes up, sometimes it goes down. There will definitely be cases where we'll have lost on some orders. But since the orders come in, they're relatively even trickle over time. We benefit more than we get hit by it and we've learned to navigate and surf these volatile environments.

Peter McCormack: Okay. How often do you cash out? Are you cashing out instantly as a new receiver or on a daily basis? I mean, how do you tend to play this?

Sergej Kotliar: We have so many different products and suppliers. So the answer to all these questions is it depends. But usually we don't... There's not an immediate sell when you buy something from us. The sell would happen at a random time when our CFO, head of treasury decides that now it's time to fill up some balance somewhere.

Peter McCormack: You have a treasury process for managing this?

Sergej Kotliar: Yeah and we keep stock for a lot of things, like an inventory. So when the stocks are low, we refill them.

Peter McCormack: So the zero confirmations was very cool because obviously I wanted to try it out. I bought a voucher for my son and it came straight away and I was like, "what? Hold on, what? How did this happen?" I hadn't read the terms or anything and then I realised that it was zero confirmations and I guess you've done a risk reward for this?

Sergej Kotliar: Sure.

Peter McCormack: You've looked at it and thought, "well how often do payments not confirm?"

Sergej Kotliar: Yeah. Often actually never, you have to be aware... It's so funny that people go around and say that it's impossible to do unconfirmed and here we've done it for four years. We've had a lot of problems and that is one in our company, because it also has to do with the kind of transactions that we're doing. So it's relatively low value. We have a decent margin. So if someone says that, "hey, I can defeat Bitrefill's algorithm 2% of the time", but we have maybe 4% margin on that product, that's 8%, so knock yourself out, have fun!

But the truth is that... And we don't just accept any transaction unconfirmed. There is a secret sauce here to how we check them and if it's a regular transaction from a regular wallet and so on. Then yeah, it's relatively safe to process them. It's a trust model. This is a different trust model compared to Bitcoin, you have to be very aware of that.

So we trust in a lot of the things. We trust in to some extent the miners and the nodes to be running in a certain way, we trust in the wallet that the person is sending from. Concrete example, if we pick up that the transactions coming from Coinbase and it's a purchase for a $20 gift card, what are the odds that they are going to be double spending us?!

Now people are now going to be screaming and complaining about that! Here's something people don't understand, that the people that we discussed all this stuff with on Twitter and I'm active on Twitter as well and so are a lot of our team members. But the users on Bitrefill are to very large extent different from the ones that you have on Twitter. They don't know about unconfirmed, they don't participate in this conversation. So when a transaction comes in, when they make a payment, they expect it.

We have had the periods, where for different reasons we had to tweak the algorithms or we have to turn it off or something and people will complain because their expectation is to get the thing right away. So that's I guess how we do it. I've considered to put out a proper challenge for people to try and double spend us. It still hasn't happened. It absolutely that it could happen some point. It is possible, but yeah the risk/reward is not even close to compare.

Peter McCormack: Right. So what do you think of most people with Bitcoin using the standard six confirmations then? Do you think there's a little more flexibility for people to maybe move lower, or do you think that's not something that we should be discussing and it's too risky?

Sergej Kotliar: Bitcoin again, everybody has to do their own thing. If you're an exchange and you and you accept unconfirmed and they let people withdraw for example, if you get beaten 2% of the time, then you're broke, your company's dead! So it's different. They have a different security model than we do. So I'm not saying that just because we can do that, everybody can do it. You have to be aware that Bitcoin is a lot of things for a lot of people and to these companies. People will use it differently.

So I'm not going around saying that others should be using our stuff or doing things the way that we're doing them. I don't know. There's so many different variables in this stuff. I can't give a general recommendation but it's absolutely feasible for accepting payments or you know the usual case of if your buddy is paying you back for a cup of coffee. Do you need to hold your body by the hand until it's confirmed six times or are you going to accept that your buddy is not going to run the double spend script on the side? It's a simplistic example.

Peter McCormack: Another question before we start talking a little bit more about Lightning. Why do you think there has been so little merchant adoption or merchants have adopted and then kind of rolled back?

Sergej Kotliar: There was a ton of merchant adoption and the problem was that there wasn't customer adoption and it's simple as that. All of the technical issues, and it's not like there weren't any, but end of the day had there been millions and millions of people using it, then the technical issues would have been ironed out or worked around and so on. I know because we did a lot of that working around for us.

So you can only imagine for a big company like Steam, and "how many it should have been" and "what links they should have been going to," but they didn't because the level of customer adoption is not at that level yet. That's why something like our service, which starts from the consumer and can serve a service, the sort of active community of cryptocurrency, which is a lot smaller and it doesn't require the merchant to go through all of these different hoops and processes.

Also like the conversation that you and I just had about unconfirmed! Try to explain that to a merchant and you have to! That's the thing, something like BitPay can never sort of... There's still always going to be stuff that the merchant needs to do. So that's why we start from the customer and because we have the extra step of the unit to redeem the gift card and that which is slightly annoying, but on the other hand you get a lot fewer errors because, we control the experience. That makes sense?

Peter McCormack: Yeah, it does. But I guess this therefore comes down to... There were times I could have bought for things in Bitcoin, but I'm like, "I want to hodl my Bitcoin. I don't want to spend it". So we have that kind of challenge of, I think you alluded to it, of a store of value versus medium of exchange. If it's a store of value right now, the primary use case, sometimes people are scared of spending it because they don't want to lose value. Also then add into that, there is a potential that Lightning could be a bit of an anticlimax because people still don't really want to spend their Bitcoin.

Sergej Kotliar: So here's the thing, it's not one thing. Again it's one thing to you. In your mental calculation, you have your regular money, you have your credit card and you have some Bitcoins. You're like, "you know, I'm going to hold my Bitcoins". That makes sense for you right now. There are other people right now for whom there's a different calculation.

Those people whose income isn't Bitcoin and they want to buy food, very simply. To them it becomes, this whole thing is going to be irrelevant. That's what I mean by the... The transactional use is a lot smaller, than the store value use today, there's no arguing about that. But it's not that it's just one thing. All of these things are developing and they're developing at their own different paces. There's even the situation of like, imagine Bitcoin went to $50,000 tomorrow. I bet you are going to buy some stuff?

Peter McCormack: Yeah, but I'm probably just going to sell as quickly as possible, sell some Bitcoin on an exchange and then...

Sergej Kotliar: Maybe! Or maybe buy something really nice for your kids, I don't know, someone else might. Then the only thing that needs to change, that in your mental calculation right now, your perspective is, "relative to my wealth, I have less or as many Bitcoins as I would like to have." But then suddenly boom Bitcoin moves, "oh now I have a little bit too Bitcoin. Maybe I should spend my Bitcoin and keep some of my regular money". There is this dynamic that is going on I think and what we look for is people with coins to spend.

I'm not going to go around... I know some companies are doing it, trying to convert hodlers to spenders or whatever. I think all that stuff is super cheesy. Let people do whatever you want. You can't moralise people into doing things. There is enough people today that are looking to buy stuff with the coins to make a business out of it. That number is growing super fast and there will come a day when all of these millions are going to show up in that category and that's what we're building.

Peter McCormack: So let's talk a bit more about Lightning because obviously this is a Lightning month. So you were very early in adopting them, back even before there was really wallets, back in March of last year. So essentially a year ago.

Sergej Kotliar: Longer than that. I think we did it on testnet in Summer 2017.

Peter McCormack: So what does Lightning mean for you? Does it offer an improvement of a service or is it essentially just another option for payment?

Sergej Kotliar: Well, yes and yes. I mean Lightning is a technical feature. In the end game when we hit the point where we have billions of users, they will not know what Lightning is. This is my guess. They will just buying stuff with Bitcoin. So in some sense, yeah, it is a technical feature that suddenly like has gotten all of this traction behind it for... A lot of it is historical reasons. From my perspective, it's always been known that something like this is needed.

So a little backstory is before I started Bitrefill. This is mid 2014, I was in conversation with a Swedish company called Strawpay. They identified back then that there's going to be a scarcity of transactional space on Bitcoin and there's going to be a huge mess around it and the solution is going to be building networks of payment channels. They got into that and they started building it. Now their solution didn't have some of the fancy stuff that Lightning has and the stuff that requires SegWit.

Then there's some other OP codes and some other things that didn't exist back then. But conceptually, that was already figured out. It's a long story why it ended up that I didn't join them, but I kept contact with them, since already then knew that something like this is going to show up. So when Lightning did show up and it's all over this mindshare with developers and so it was sort of a relative no brainer to hop on.

At the time it just so happened that we had a summer intern that was Justin Camarena, who was fresh out of college but spent all of his time reading about Lightning stuff and we were fiddling around with the nodes and so on. We were like, "let's fire up a testnet Lightning", because there were these star blocks and these pretend stores. We were like, "no, let's make a real pretend store, so you can go through the real experience how it actually will be!" So we kind of already started testing in that sense.

Peter McCormack: Yeah because I guess with your zero confirmations, I've got no reason to use Lightning with you because essentially it's just as quick, unless I've already got a Lightning wallet to hand, which has capacity. There's no reason.

Sergej Kotliar: But that's the steps that it needs to go through to reach adoption. So there's a new S curve of adoption inside of Bitcoin. There's another S curve that is Lightning and in the beginning, obviously it's going to be people that try it out for novelty or because they're interested in the tech and some people are actually going to be using it. I think there's already reasons for you to be doing it that way. For one thing, you will be certain that your transaction costs are instant and without any hiccups.

Our zero confirmations is not perfect. There's a lot of cases where we won't accept your transactions over zero confirmations and you'll have to wait for one confirmation, which can take, depending on when you make a transaction, can take a long time. Transactions on Bitcoin can have fees. It's not unthinkable that one day... And we have fees, because we need to consolidate all of those little coins that we get, 20 bucks here, 50 bucks there. We need to put them together into big ones and that has a cost as well. It's in discussion, but it's not unthinkable that we're going to expose the cost to the customers to premiere the ones that use payment methods, like Lightning that don't have that cost us.

Peter McCormack: What are the challenges that you face by adopting Lightning? So that are different from say accepting Bitcoin payments.

Sergej Kotliar: Well, it's the same but different. There's a new node that you need to run and it's new, unstable and it will break. It's not as tricky as people will think. There's also certain security aspects that you need to be aware of running a Lightning node and how you configure it and so on. There's been some mishaps were that has gone wrong for some companies.

So there's things like that, but on the whole... And there's the question of channel management, which we've only kind of started hitting at a little bit and that's probably going to be a bigger thing over time. So there is stuff like that but on the whole, I would say that we already at the ballpark equal amounts of trouble from Lightning that we have from something like Ethereum and the Lightning one is decreasing.

Peter McCormack: Does it add another level of treasury management because essentially you've got your Bitcoin wallet and then you've got your Lightning wallet and if you want to convert your Lightning back to fiat, you have to go to Bitcoin. You get an extra step.

Sergej Kotliar: There's been some explaining with my non technical co-founder that does the accounting about what the different things are and what the channel is and so on. That stuff can be a little bit tricky, I guess. Understanding that channels have a cost to close and this and that and how to re-balance them and so on. We tend to address those things as they come.

It's been going relatively well so far. We gradually ship things and we sort of ship first, then we figure out what the problems were and then we fix the problems, because very often the problems are not the ones that you thought were going to be the problems. So that's one of the philosophies that we have here, is "ship it now", even if it's not ready. If it sort of works, we ship it and then we'll find out.

Peter McCormack: So are there any other challenges with Lightning that you as a retailer face that people wouldn't be aware of?

Sergej Kotliar: Well, there's the whole question of the topography of the network. If you want to talk about that kind of stuff, that people are sort of gradually realising, if I should be the critic here, but not all nodes in Lightning are equal, for example. It's also the case on Bitcoin in a sense. But what I mean by that... So imagine we have a channel. Someone wants to open a channel with us. That someone could be Pierre Rochard that's connecting to a bunch of things.

All right, great, there's probably going to be routing on that channel. But what if that channel is some dude with his android somewhere and he just opened a channel and bought some stuff. Now he has an empty channel and he turns it off. That's capital that we have locked in the channel that we can't easily get out. So there's aspects of that, that there needs to be... Since Lightning is still in a very early adopters phase, it means that this stuff hasn't crystallised yet.

You can't just open a channel with just about anyone because eventually you're going to run out of money and you need to be somewhat selective about that. This is something that I think is going to start crystallising itself. So we built this Thor product, which basically is that. It's a service where we open a channel to someone, the user in the basic Thor service, the user has no funds on their side.

So we have all of the funds, but we have the channel and we guarantee that we're going to keep it open for at least a month. That part is important because without that, then the rational thing for us would have been to close it. So we guarantee that it's going to be open and we charge a fee for that. Those fees still needs to be figured out a little bit. First they were too high and people gave us shit for that and now we've lowered them.

But there's also certain bottom levels, because we need to make Bitcoin transactions for those so we can't really go lower than that and so on. But one of the purposes of creating it was, for example, to create a debate or a discussion about the fact that someone needs to handle the contact between the well connected parts of the Lightning network and the edge nodes. So there is a role to be filled. I sometimes describe Thor as an ISP, an Internet service provider.

On one hand it connects to this well connected network of the Internet on the other side goes out to people's apartments. For that they have to be charged a fee. There's something like that in Lightning as well, which they need to figure out a little bit. I think that we've already gotten quite a couple of steps in that process. The next iteration is probably is going to be getting this service integrated into wallets to make sure that this stuff gets set up right away and so on.

Peter McCormack: Yeah. I guess if someone has a channel open for them by you, it's got to be a channel that's been used quite regularly to justify the fee. Otherwise it doesn't suit being used for one or two payments. Because otherwise you might as well just do an on-chain.

Sergej Kotliar: Right. So who's you in this example?

Peter McCormack: So say for example me, say I needed to, I don't know, receive a Lightning payment. Say somebody who wants to subscribe to my podcast and they want to do it with a Lightning payment and I open a channel with you. If that's the only payment I receive. Or even if it's only like two in a month, I might as well have just done an on-chain payment, because your fees are likely in total going to be higher than the fee I would have paid for an on-chain payment.

Sergej Kotliar: Sure.

Peter McCormack: So channels are only useful if they're being used regularly.

Sergej Kotliar: Of course. I mean this is the same as when Bitcoin was new, people would say things like, "but I need to buy the Bitcoins at the exchange, then pay a fee for that. Then when I sell them, I need to pay a fee for that." Yeah, but that's in the setup phase before things start spinning inside of it. There was one of your recent podcasts, I think it was Pierre Rochard or if it was Jack Mallers, I don't remember who said that all Bitcoins are going to be on Lightning. I'm maybe not as optimistic, but I would say like this, "there's going to be cold Bitcoins that are in cold storage.

But the hot ones, they're going to be on Lightning," and then suddenly it's going to be just that. At some point you're going to get to a point where certain services, will probably stop accepting on-chain payments. Because remember also that if you send a tip, if you send a dollar to some other podcasts that you like and you send them a dollar tip. Then you pay the 25 cents in transaction fee, that's already high for that.

But they need to pay 25 cents to collect the fee as well! Then suddenly there's a 50% transactional cost, split evenly, but it sucks for both of you. So eventually it will just start spinning. I think it already has. I think there is already a layer of Bitcoins that is spinning around Lightning, that is used in tipping, that is used by enthusiasts, that is used with us. We've reached a level of, if you count transactions, 6% of our transactions last week came with Lightning. That's not nothing.

Peter McCormack: No, no, of course. But I'm just thinking, for example, how much does it cost to open a channel with you?

Sergej Kotliar: So a regular Lightning channel, if you want to do it the good old fashion way, doesn't cost anything, but we've set our node to have certain limits that it can't be smaller than that size. Otherwise you can't open it, you can open it for someone else. If you want to open a Thor channel, where we guarantee that we're going to keep it up and all of that. Then that has a cost and there's a fixed cost because it needs to pay for Bitcoin transactions and there is a variable costing from a couple of percent of the value of it.

Peter McCormack: So about what approximate cost would that be?

Sergej Kotliar: So I think right now, if you get the biggest one then it costs, something like 3%, I think of the amount. If you open a smaller one, then it's going to be more because the transaction is going to be part of that and then... So there is already incentives for people to try to open bigger channels.

Peter McCormack: No, of course. But I think the point I was trying to make is that, if I was considering opening one with you, what I have to do is actually see if there is the cost benefit of having that channel open with the costs, because Lightning network is there for faster payments of course, but also to save costs on transactions. But I've got to have enough transactions going through that I save more than I spend on opening the channel with you.

Sergej Kotliar: Okay. Let me curse in church some more! You know which kind of wallet doesn't cost anything for making Lightning nodes?

Peter McCormack: Blue wallet?

Sergej Kotliar: For example. Yeah. But if it's a custodial node, then yeah, you don't control your keys but then you can, you can ride on their channels and so on and have someone professional that manages it. Now, that's not the Bitcoin we came for and it's not ideal, but it is there.

I think that too, brings up the fact that probably, I don't know if this is a law of nature or not, but probably it is the case that the more decentralisation you want, the more you will need to pay in terms of both money and user experience and it comes at a cost. I think that in some ways it's reasonable because it is more tricky. By definition it's easier to do something centralised than decentralised and so the decentralisation is going to be more expensive.

Peter McCormack: We can discuss custodial versus noncustodial but what I'm saying, it's not a criticism, but I'm just trying to get my head around this. It's almost like today is very early. People are playing with it, getting their heads around it. But longer term, I guess my vision that I would use, probably something you would share, is that I will have a cold storage wallet and possibly like a semi hot wallet for Bitcoin and then a super hot wallet, which is my Lightning. I optionally choose to use whichever at the time and ideally, long term, a lot of these channels won't close. I'd want to keep my channel open.

Sergej Kotliar: I think there won't even be a middle, there will only be the two. When we get back to the way things were, sort of when people would pay each other back for lunch with Bitcoin. I think it's starting to happen again now with Lightning. Today, if you're a regular complete non-technical user, probably the best choice is to use a custodial wallet. Now over time, me and many others are working actively to make wallets that are not custodial, that don't have these problems and that solve these difficulties.

But out of what we have today, yeah it might be... Unless you enjoy tinkering with the stuff, which I think a lot of the current Lightning adopters do and that's fine. To me I think that one of the core principles of Bitcoin is that you can do whatever you want and such I think it's good that people care about the core principles of it. But I also say that, can't moralise for people that take shortcuts, especially when the shortcut means getting something in the hands of people today. Because that allows you to understand what problems are really there. I think that has a big value in itself as well.

Peter McCormack: What are the reasons that you're seeing that people would open a channel with you, say, as opposed to create their own channel and creating their own capacity? Is it really just for speed?

Sergej Kotliar: So you're talking about the Thor service?

Peter McCormack: Yeah.

Sergej Kotliar: So the way the Lightning network is today, is that there is a core that is very well connected. First of all there's the basic convenience, that either you ask someone, email them, they get you back etc and it takes a while. Whereas here it's like "click, click, yeah, it costs a little bit, but whatever, you got it right away". So I think there's quite a bit of that going on right now in this initial stage. But then you have to remember also that you want to be connected. If you're on a channel to someone else's Android, it's going to be offline most of the time.

Then what good is it? So you want your channel to be connected with someone who is well connected with the rest of the network and that has good liquidity, so they put some money on the line on those nodes. Our node is one of the more well connected. It's been that way since the beginning. So obviously if you can connect to one of the very central nodes, people want to do that. From our perspective, we still kind of do that as a service to the community.

Other than they say that it's Bitrefill connecting to it, we don't get so much out of out that. If somebody wants to open to one of the other well connected nodes then that's great too!

Peter McCormack: So these well connected nodes are also "mega nodes", "mega hubs", whatever you want to call them. Do you think these are important then for Lightning?

Sergej Kotliar: In what sense?

Peter McCormack: As you said, if you're just connected to somebody who's got an android and it's offline, it all almost makes sense that most of us have a channel open to one of these larger, more centralised hubs. Because with that we've got a high chance of our payments going through.

Sergej Kotliar: Yeah. What's the question?

Peter McCormack: So do you believe therefore that these "mega hubs" are important and they're going to be an important component of Lightning? Some people I've heard argue that they lead to a little bit more centralisation. Is that really centralisation or is that just a reality of the success of Lightning?

Sergej Kotliar: It's some centralisation, but again it's a matter of definition. It's not black and white here. There is obviously some trust with the nodes you connect to, because what if they go offline, it's going to be inconvenient for you. So you trust them that they're not going to cause you inconvenience by their incompetence, for example. You don't trust that they're going to steal the money, but there is still some trust.

I don't necessarily think that we're going to see "mega hubs" in that sense. I think we're going to see probably some natural, bigger hubs are going to emerge and I think it's going to be companies like ours. It's going to be some of the custodial services. It's going to be some wallets around hubs like ACINQ runs one for their users. I think that exchanges when they start getting on board of the Lightning train in a big way, it's going to be relatively natural that they have those hubs.

Then there's always going to be enthusiasts. There's always going to be people that just say, "hey, I want to run a node and I want to spend time building it and getting it connected". That's kind of what we have. I mean if you run Electrum wallet, it will connect to an Electrum server. What's that? Someone is running it. Is that centralisation? Very small way, but yes somewhat. You can run your own and no one stopping you from running your own. I think this is the beauty of Bitcoin, is that you can run your own, you can run your own everything.

With Lightning especially, you can run your own mega hub and there's several companies now that that sell Lightning servers in a box. They can run at home. They will always be online and they will do all the right things and who knows. Maybe that will be the case and then it will be that my Lightning node at home. Is going to be the one that powers my friends and my family and my neighbours maybe. I don't know, we don't know that stuff yet!

I personally think it would be really cool if there was a cypherpunk device and a bunch of people have them at home and go, "I have a friend with a cypherpunk device and it has all of this stuff. It has Lightning, it has Tor," whatever he wants!

Peter McCormack: Yeah. It's funny because I'm almost trying to imagine what it will feel like in five years. Because obviously we're very early and I said yesterday when I released my episodes, I think I'm guilty of high expectations and treating Lightning like a finished product. But actually it is very early and lots of the things that I'm coming into or the problems I'm having are things just because it is early. As an example, when I had the trust chain, I wasn't online when I received my payment and I was like, "what? I thought I was meant to be online!"

It turns out my blue wallet is custodial, so I didn't need to be online. Which to me was a benefit. But then I spoke to Jack Mallers and he obviously said he thinks these custodial wallets are actually bad, because it's like leaving your Bitcoin on an exchange. It's not your keys, not your Bitcoin. So then I became really conflicted, but really in my ideal world, just as simple and dumb as this sounds, say I've got my Ledger or Trezor, I would love to just be able to open it up and send a payment in exactly the same way I can send a Bitcoin payment and not have to worry about channels and capacity.

It just all just works without me even thinking about it, because for me that would be great UX. But I've starting treating it like, "oh, Lightning doesn't work properly because this doesn't happen". But really it almost probably will in a few years?

Sergej Kotliar: So it's very clear. Everything that you said is true and Jack's critique is fair and balanced in every way. I actually think that... I don't know if we're already at a point where it could just take off, we speculate here a little bit and try to imagine when the Snapchat moments comes. When there is some social use case and it's going to be something social in a way, that's going to just take off like mad and it's probably going to be teenagers, they're going to do something that is super unreasonable to all the adults and this is going to take off like mad and we're going to have a hundred million users.

What is that going to look like? Personally I think that almost certainly when that happens, there will be people that will be screaming, "we're not ready yet!" That's been the history of Bitcoin since Satoshi was complaining about Wikileaks, '"we're not ready yet, don't bring it". Then someone is going to bring it anyway and then it is going to be chaotic in that sense. There is going to be quite a lot... I'm pretty certain that whatever that is, the thing that triggers the Snapchat moment, a lot of people on Twitter are going to be super upset about that because it's going to violate some of the cypherpunk principles that Bitcoin has been built around.

Maybe it's going to be that there is some custodial functionality or I don't know. It's too early to say. But from my perspective, we kind of found out the good old fashion way of doing Bitcoin, it doesn't scale. We kind of always knew that. But we also experienced it firsthand, what that looks like. So there is that sentiment in the back of our head. Now that stuff is gone and it's gone for several different reasons, Lightning is one of them.

There's also other things that have solved the scaling issue. We don't have to get into that. I don't know if we're already in ready or if there is a couple of pieces. There's so much exciting technology coming out of the Lightning implementations now with the AMP stuff... There's a lot of cool stuff that's coming. But I'm pretty certain that when it takes off, people are going to say, "we're not ready yet", because usually good enough, is a lot earlier than what people think that it is.

Peter McCormack: I pretty much predict that the majority of early use case, if it does blow up, will be through custodial services. I just think the ease of use there for new people to come in and start adopting and using Bitcoin... Because it's quite a challenge to set up a node, link your node to your device, open channels, understand what that means, understand what opening channel means. I think it's way more complex than people appreciate for the normal user.

Sergej Kotliar: Sure. I tend to agree. I still think it's worthwhile for us to work to reduce the level of custodial or centralisation in every way.

Peter McCormack: I don't disagree.

Sergej Kotliar: I mean if you look at the status quo, disregards Lightning for a second. If you look at the spiky period around the late 2017, early 2018 during the hype, the bulk of all transactions, what was that? Well, it was someone buying coins from Coinbase, sending them to Poloniex, sending them to Bittrex, sending them to Bitmex, then maybe maybe withdrawing them to a ledger. But most of the other steps, it was centralised services, custodial, sending to other centralised services. I'm not saying that it's good, I'm not advocating it in any way. I'm just observing that that's what this activity actually was. Again, on Twitter, people are arguing that, "oh, my Electrum wallet is giving me the wrong fees." Yeah, maybe, I don't know.

But that's not... Most people they just use the custodial stuff and they don't understand the other stuff yet. I think part of the nature of this stuff is that there is certain tendencies for the draw towards centralisation and there's people and companies and things that we do that try to counteract that. To say that one of these sides will win, I don't know, maybe it's going to be the way it is right now.

Maybe this is an ongoing sort of battle of forces. For me the big value of Bitcoin is that you can do whatever you want. Not that you always have to do the most cypherpunk version of every saying. Always. Because if that's the requirement, then pick any Bitcoin service, wallet, company and so on, I can point out the way in which they're not perfectly cypherpunk. So it becomes this, you are without sin, you can cast the first stone. Now there's obviously shades of this.

There are fully custodial services and so on. There are the ones that do KYC. I mean we all agree that less centralisation is better and that's what we're all here for. It's not under discussion, we're not debating that a non-custodial wallet is, all things being equal, going to be better than a custodial wallet, because it is.

Peter McCormack: So what I'm thinking though is say the prices of Bitcoin start going up. I'm going to be down the pub on a Friday or Saturday and I'm going to have some friends say, "oh Bitcoin, I need to get into it. How do I do it?" My spiel is pretty simple, it's "right, okay. Go and buy a ledger. Once your Ledger gets delivered, go on Kraken, buy some Bitcoin, send it to your wallet and that's going to be a great first experience.

Just buy $100 worth," and they're going to do it. Then they're going to come to me and they're going to say, "oh Pete, what's this Lightning thing? How do I do that?" It's going to be like, "well, you need to buy yourself a $300 node or whatever, you need to get that set up, you need to connect that to... It's already at a point where it's like, "no, it's too much."

Sergej Kotliar: No, that's probably not what it's going to be. Probably what it's going to be, is a feature in the Ledger app. I'm super simplifying here, but Lightning UI, you click on a coin, then it does things and probably connects to the node, run by the Ledger company or something like that. Again, there are some points of trust here, but it's still virtually trust-less.

Peter McCormack: But if that's a node run by the ledger company, is that therefore a custodial service?

Sergej Kotliar: Well... I mean that's depending on where you draw the line. You could also say that, "not your keys, not your Bitcoin" and then "your keys, your Bitcoin". If you have the keys then maybe that's enough for example. A Ledger wallet I'm not super familiar with it, but I'm pretty sure that the Ledger app talks to a server that's run by the Ledger company, I just assume. Now the keys are still on the hardware, so it is your keys, but they do provide some infrastructure to facilitate.

I mean this is already sort of status quo. This is how it is. If there is a lot of market demand for people to be more cypherpunk about it, then great! Then maybe they will add a field that, you can now run a wallet server on your cypherpunk device and you can type in the IP address of your buddy in Bedford that the runs one and you connect to that. Maybe that's okay. It's a buddy and it's not some faceless company.

Peter McCormack: Can you see what I'm getting at with this line of questioning though? It's the level up from Bitcoin to Lightning. It's a steep learning curve and also it's not always as practical as Bitcoin. Now, look, I'm not saying it shouldn't be, I'm not saying this maybe is the only way it has to be. But at the same time it's just that these are the things I am now wrestling with and I'm thinking about in terms... Because I always think of the point of the user. I always think of the point of the newest user possible. How are they going to understand this? Right now it's tough.

Sergej Kotliar: What I can say is I think you're setting the bar too high. You're doing the fallacy of assuming that Lightning will solve all of the problems and that every transaction ever and all of the Bitcoins are going to Lightning.

Peter McCormack: No, not that. Definitely not doing that. All I'm talking about is, if I've got a friend and they want to get into Bitcoin, I can get them into Bitcoin very easily. If they then start asking me about Lightning, the level up is really hard because of some of the nuances of Lightning.

Sergej Kotliar: Today! The way the wallet's are today. But today the wallets all say that we are beta and it's reckless. So it's also not fair. Is it unthinkable that your wallet is going to do a splice out and splice out a new channels for your buddy? No, not unthinkable. Just that nobody built it yet because there wasn't that big of a demand yet. When it comes, then there will be someone who will build it and they will be maybe even three and they will try different approaches and we'll say that one of them was better. Then the others will copy it and then it becomes standard.

There's a hundred ways you can solve it. What if a services... So we have the Thor channel service. Imagine a world where there's 50 of those and one of those integrate into the Ledger app. Because probably you're not going to want to have your pocket Lightning coins on a Ledger for long. You probably want to have them on your phone, so you can easily scan. I don't know, I'm just speculating. In that case there's a button that says "purchase Lightning" and "copy paste your address" and then it's copy paste address, scan a QR code and then you're up and running.

Some of the stuff will have some trusted parts for certain things and so on maybe. Then someone would make a more decentralised version of it. I think that this stuff will sort itself out. It's not even thinking about never just discussing it. What if the wallet just actually buys a Thor channel for your buddy and then your buddy scans the QR code and gets it. Then suddenly he has nice channel right away. Then you have exactly that already, but with Lightning. So there's a lot of ways of solving it.

This stuff hasn't been built yet. There's a lot of unsolved questions in Bitcoin, even in regular Bitcoin. I mean, we're still arguing about zero confirmations, replace by fee and all that stuff. It's still unsolved, but different people have different views and arguments and so on. So there is a lot of stuff that needs to be worked out. I think that from the perspective of Lightning, I think that we will get to a point where a lot of it can be abstracted away and hidden.

I think we're not there yet. I think we're right now at the point where it should be very easy to "pop the hood". So that you can quickly get feedback on what the hell went wrong. So right now we're in the "pop the hood" phase and I think that it is important that we expose all the inner workings a little bit and confuse people with the mess and force this learning.

But eventually it will be the case that your buddy is not going to need it. Maybe you will be the expert because you were around when this stuff was built. So you will you probably know how to troubleshoot a Lightning channel, he won't! Maybe there will be some button somewhere deep in his wallet, that is like pop the hood. Then you can look at it and your like, "oh, okay. I need to do this and that" and so on.

Peter McCormack: I mean this is why I'm talking about it because there are a lot of complexities around Lightning. What it is now, what it means to people, what it will be in maybe two, three years, how some of these things may be abstracted away. It's a really tough thing to try and balance to judge and use it now. But also, most of the issues I've run up against, it's like, "oh, this will be solved in the future". So therefore, trying to picture that future and it's not even a criticism. If anything, it's a criticism of myself for having the wrong expectations and treating it like a finished product.

Which is totally the wrong thing to do and I think I've come around to that point, is that I think I understand now, these are building blocks, so we've almost got the foundations in place now. A bit like when Bitcoin was command line. There was a time that Bitcoin was just command line. This is the building blocks. Like Nic (Carter) said, the experience of buying Bitcoin with the Cash app now is way different from 2010 and I think I've come around to that point. I go, "okay, I get it now. I should stop judging Lightning as a finished product. This is the first couple of building blocks in place.

As people use it, as people start to play with it, there will be things fed back to the developers and the developers will work on new things". What I can't do though and I'm not able to do is envisage what a future Lightning network looks like and what a future relationship between the base chain and the Lightning will look like in three, four years, because I just don't know and I guess nobody knows really.

Sergej Kotliar: You're absolutely right that nobody knows really. There is a pretty clear idea. I think that some people have. There is the question of the custodial and the noncustodial and so on, that probably will be even bigger I would guess. We can already see if you look at the download numbers of different wallet apps, you can see that the custodial ones are far ahead in downloads, because it's easier. This is the nature of all of this stuff.

People go to the easiest option out there and this is the argument that the people who say that the easy option shouldn't be there. Their argument is that if there is an easy option, people are going to use it. So let's not have the easy option, let's force people to use the good "version". I think that, depending on how much centralisation and custodianship and whatnot you tolerate, you could argue that we're almost there. If you use one of the custodial wallets and you receive a $5 tip from someone and you go and Bitrefill and buy something for $5 with it, it's a really good experience.

It's not the case that this could be built with a database. But no, it's still a Lightning transaction. It goes over a decentralised network. The sending entity and the receiving don't know anything about each other and have no relationship and so on, it's still pretty fucking cool! But it's not the end goal. But already there we can picture it, we can feel it out, we can start solving the UX issues. If you tried out these custodial wallets, you'll see that their UX isn't perfect.

Even though they took a whole bunch of shortcuts, their UX isn't perfect, but it's a lot better than the non-custodial ones! But it's still not perfect. They work from there and they work from a lower level and we're all building towards the same goal and the same user experience. Eventually it will get figured out. But I still think... So we mentioned, for example, that Thor is a potential ISP sort of type model, if that's the case, well first of all, we're not an ISP company currently!

It's a little thing that we put out that was cool to build and some people found it useful and now we built another one of those. But if it is the case, then these companies will be forming entities or services and they will be the central option. I don't know, it is tricky to know how exactly it will all look in the end. I mean that's part of the whole Bitcoin thing. We don't know what Bitcoin looks like in the end. What does the "hyperbitcoinisation" look like?

What does it look like when everybody in the world has access to Bitcoin. We don't know, it's going to be strange and confusing, I can promise you that! I think it's okay. I think that it works today. I think that's what I want the take away to be. It works today well enough, so that a meaningful amount of people already today use it to pay for stuff and to get paid for things and pay each other. I think that that's pretty damn cool.

We're still super early and we're walking into the unknown. We can only see so far ahead. From the stuff that we see, there's a bunch of issues that we've already had and there's a bunch more that we know of, that are being worked on by brilliant people. There's also a bunch of issues that we don't see yet because we haven't gotten there yet. When we get there, we'll figure them out as well. But it is working, it is being used!

Peter McCormack: It does! I have used it and the times I have used it, don't get me wrong, it's been an absolute pleasure. It's amazing to be able to send something and it's instantly there. Or send somebody an invoice, they tell you they've paid you and boom, it's there. That to me is absolutely amazing and I'm a big fan of it. I'm probably just a really annoying UX person because that's what I did way back at the start of my career and whereby was a UX designer. So I always look at things with a critical eye.

Sergej Kotliar: If you want to help us, we'd appreciate it!

Peter McCormack: Yeah, well yours is pretty good as it is actually, the experience of buying on yours honestly was fantastic. Okay. So look for you, with Lightning, what are the most important things that you think need to be worked on and what are the things that you are most excited to see come to Lightning?

Sergej Kotliar: Well from the technical level, there's a lot of cool stuff, like the Multi-Path throughout thing and that kind of stuff that will solve a lot of a lot of problems. The technical stuff is kind of like... I'm mostly excited, a little bit like you, I'm excited about trying new UI paradigms and the new ways of doing things. I'm more excited about integration with browsers and things like that. We can talk about Lightning improvement proposals, but you have better people to talk about those things.

So I'm more excited about a world where there is a wallet in the browser and you can just the click it to receive and click it to send. Come to a website and you click the wallet, and it opens the website. This is sort of a little bit what the Web3 vision is on the Ethereum side and there is a version of that on Lightning called WebLN, which is really cool. That stuff, I'm excited about that getting even easier. I'm excited for having noncustodial wallets that don't require a running a full node, for example. This is controversial I guess these days, but I'm excited for that.

Zap released a version with the mainnet over Neutrino to try out. So we're already at the point where you can run a Lightning wallet, that is "your keys, your Bitcoin" on your desktop. It plays well with the Joule extension and you can imagine that being iterated a couple of more steps and living inside of the browser. I think that that kind of stuff is cool. With all this, I'm excited about non-technical stuff. I'm excited about services that will let people earn money over Lightning.

The large part that that was one of the motivations for us when we built Thor. It's hard to receive payments for your earnings if you don't have a channel and how are we going to get a channel if you don't have any Bitcoin. That was the reasoning. Now we're already at a point... Just picture this. The way that every company has a jobs link on their website. You can click there, you can apply for work. What if every company has a microtask link, you click it, the task shows up, you do the task, you get the money and that's it. There's no log in. No send your personal information. There no nothing. You are being identified by your node ID and that's enough and you do your stuff.

Right now we have billions of people that would so eagerly, hop on and do work from microtasks to larger tasks. What will the world look like when any company can instantly pay $1 to a thousand people to do a couple of minutes of work? What kind of things will that enable? When that starts circulating, and I think I really think that we're a lot closer to that than people think. I've had a couple of meetings recently, where it could be the thing. Probably not yet. Probably a couple of more iterations, but it could spark it. People will be taking shortcuts along the way.

But it's still pretty cool! I'm excited about that stuff and I guess I'm also excited about the things that are real. Right now we're in this paper building games and toys and this and that. You can send five Satoshis to each other and that's cool. But I'm also excited about things, where money is exchanged like money for goods or services or whatever. When the intent is to actually transfer money these types of usage, I'm super excited about seeing that evolve. Whenever I get attention, I always talk about earning and how that could happen. There's so many projects that we're working out right now, I don't know if anyone's cracked it yet. I hope they will soon.

Peter McCormack: And for you guys, what's coming up for Bitrefill? What should we keep an eye on?

Sergej Kotliar: First and foremost, we're constantly adding new products and countries and so on. So there's a constant flow of that for a year ahead of things that you can buy it with your coins. When it comes to... Which I think that's super important because you need... In order to trust money, you need to know that I can buy something with it and so that's one of the motivators.

When it comes to the Lightning stuff, we just released the second iteration of our Thor product "Turbo Thor", which is the one that lets you, without any confirmations suddenly have a Lightning channel, with money on it that you can use to buy stuff. We have a couple of more products that are in the same category. they're utilities that bridge the gap between the old world and the new. So like backwards compatibility, how can we make it so that you can go from Lightning to non-Lightning and vice versa in different ways.

There's quite a bit of stuff like that that's going to be coming out from our side. Working with wallets, figuring out new ways of making things work better. Figuring out ways how we can integrate services like Thor into those things.

Peter McCormack: Cool! How do people keep an eye on what you're doing? How do people keep an eye on you? Who do you want to hear from?

Sergej Kotliar: You can keep an eye on Bitrefill, at bitrefill.com and on Twitter @bitrefill. My Twitter handle is @ziggamon. We are always looking for good people, especially when it comes to the UX side and front end and so on. But good people that are passionate about this stuff and about doing something that is really useful for people, today and now! That's what we're looking for.

Peter McCormack: Brilliant, well listen look, thank you for coming on. It's been great to talk to you about this. I can't wait to get this one out. I might even put this one out a day early. I think this has been a really useful interview and yeah, thank you for coming on Sergej!

Sergej Kotliar: Cool man. Thank you so much. It's been a pleasure.

Peter McCormack: Take care!

Sergej Kotliar: All right, bye bye!