 I call this talk streaming money, and if you're watching Bitcoin from the outside, if you're engaged in Bitcoin, but don't have the time to spend to watch every new technical innovation that is being created with Bitcoin, it's hard to see what's happening behind the scenes. And what's happening behind the scenes is a lot. A lot of very interesting work. Bitcoin today is not Bitcoin as it was in 2009. It's continuously changing with new technologies being introduced. And the pace that new technologies are being introduced keeps accelerating. One of the most fascinating aspects of Bitcoin that was introduced towards the end of 2015, which I'm very, very interested in, is the introduction into Bitcoin of a time function. So what this new invention did is it introduced time function to transactions, the ability to control the timing of when a transaction can be redeemed, when it can be spent. That particular invention is called check lock time verify, which is a very engineering word for something very powerful. And at first, when you look at it, you think, okay, great, so I can put my money, I can lock it, and I can say this money can be withdrawn for 90 days. Okay, if some of you have problems with addiction to shopping, or a materialist consumption kind of attitude, and you can't save money unless you lock it away, that could be useful. Okay, you could use it just like that, just lock my money away for 90 days. And the nice thing about the Bitcoin network is that when you put a condition like lock it away for 90 days, it's locked away for 90 days. There's absolutely nothing anyone can do to undo that particular constraint. But if you look at this time dimension purely from the perspective of locking an individual amount of money, you're missing the point. Because what's really interesting about this is that it creates a whole new set of applications that allow you to manage the time dimension of money. And this is a game changer. And most people haven't really noticed that things are going to get very interesting, very fast in some of the new developments in Bitcoin. One of the first applications that uses check lock time verify and check sequence verify, which are the two time-based constraints, is a technology called state channels, or payment channels, or more broadly, lightning network. How many of you have heard of state or payment channels here? Okay, so that was about 15 or 20 hands. How many of you have heard of lightning network? Okay, great. That was about 75% of the audience. And how many of you understand lightning network? We got a lot of this, myself included. Okay, very complex technology. So what I want to do is briefly describe this technology and then give you a perspective on something much, much deeper that can happen with this technology. So bidirectional payment channels allow you to do transactions between two parties that are not recorded directly on the Bitcoin blockchain, but essentially mean exchanging promises that have a time dimension. And then setting up this bidirectional payment channel where I say, okay, I'm going to put 10 euros away to communicate with this, let's say it's a bartender, it's a bar, right? So I'm going to put 10 euros in a multi-sig account. And then we're going to set up a payment channel between us. And we're going to say, okay, I buy one drink. That's let's say one euro. It's a very cheap drink. And so now I'm going to sign a transaction that says of those 10 euros that we have in a multi-sig together, I'm going to give one euro to you and nine euros back to me. Hold on to that transaction because I'm not done drinking. Great. Five minutes later, I say, that was a lovely drink. I want another one. I'm going to make a new transaction. This one is going to cancel the previous transaction, which we still haven't broadcast. And it's going to say, well, now I'm going to give you two euros out of the 10 and I'm going to get a refund of eight. Now hold on to that transaction. It invalidates the previous one. The bartender, if they want to submit it, they submit it. They get their two euros. But when they do that, I also get my eight euros change. So I'm happy. They're happy. We can both walk away from this transaction anytime we like. And we now have transferred this money, but none of it's yet on the blockchain. Now this is a really, really good night. So I'm going to have another drink. And now I sign a new transaction that says I'm going to give you three euros. You give me seven back. And we keep going back and forth like this until eventually I say, okay, I want to close my tap. The last transaction we have, maybe four or five euros in drinks, five euros refund, is the transaction that actually gets recorded on the blockchain. So we did four transactions, but only one to start and one to end were recorded on the Bitcoin blockchain. And that's really interesting, because in the process of doing that, I could create as many transactions as I wanted, and I could make these as small as I wanted, because we're not paying a fee for them. We're only paying a fee for the final balance. So I could be transmitting very, very small amounts in this payment channel. Bidirectional payment channels are really interesting technology. But when it gets even more interesting is when you combine multiple bidirectional payment channels to create a routed network. So I'm sitting there with my friend and I'm having drinks and they're having drinks and we have two payment channels to the bar, right? And right now I owe five euros to the bar and my friend owes maybe six euros to the bar and we decide to play a game of pool. And so I start a game of pool with my friend and we place a bet and we say whoever wins, it's five euros. We're going to place the side bet on our game of pool. So we start playing pool, great. And I lose because I suck at pool, right? So I lose badly. My friend might also be a hustler in pool and hiding this. So I lose very badly. Okay, so now I owe my friend five euros. Well, I could pay directly and start a new payment channel with my friend. But here's what I could do. I could go to the bar and say he owes you five or let's say he owes you six and I owe you five. How about you change it so that he only owes you one and I owe you 10. Great. So now we've got two payment channels. I close mine paying extra to the bartender and he closes and signs another transaction to pay less for my friend. Essentially I've paid my friend but without having any direct connection with them, right? Now take that and imagine connecting tens of thousands of payment channels together on a network that is routed where I can basically go out and discover the network and say, that guy, I want to give him a tenth of a Bitcoin. Now I'm not connected to that guy but he's connected to that guy. That guy is connected to this guy. This guy is connected to that guy and I am connected to that guy. So I will give you one euro but only if you give it to him who only if it gives it to him, only if he gives it to my final destination. And when the euro arrives at the final destination, you get paid, which then means they get paid, which means they get paid, which means they get paid, which means I paid them, right? And that's Lightning Network in a nutshell, right? Lightning Network is just a series of Bitcoin transactions that we exchange among each other. And here's where it gets funky. Because the really interesting thing about this is that the speed at which I can process these transactions, these transactions that are fully formed Bitcoin transactions guaranteed by the Bitcoin network, any one of the parties can walk away at any time. We don't have to trust each other. We can take the last transaction, which is the most valid one. We can submit it to the network and close all the channels anytime we want. So we always have Bitcoin acting as a judge to ensure that all of our transactions we've exchanged are secure. But now we can do them as fast as we can process elliptic curve signatures as fast as we can process transmitting these payment transactions and how fast is that? Milliseconds. And we can do them for amounts that are as small as one Satoshi. So now I can transmit Satoshi's in milliseconds across a network of tens of thousands of participants that are all connected at a layer above Bitcoin. Stop there for a second. Take some questions. That's really good. So in the legal terms, it's an assignment of claims. It's a series of IOUs. It's a series of forward-looking promises. But the thing is that if a party doesn't deliver on their forward promise, they can't collect on the promise that's coming to them in a routed network. So you extend these promises out to your final destination. They provide the unlocking code that rolls it back so that everybody gets paid. So you don't have to trust anybody in between. If someone doesn't fulfill their promise in between, you just start a different route to get to your destination. No one can take money without fulfilling the terms of the contract. It's a system of smart contracts. So you don't need to trust any of the other participants. In fact, if this is properly implemented, you have no idea who the other participants are. You just say, I'm paying Alex a tenth of a Bitcoin. Find me a route. Great. It takes 233 hops to get there. I don't care. Just like you have no idea how your TCP packet actually got to Google, you don't care. It's the same system. In fact, it's better because the first implementation of Lightning Network that we're building is based on onion routing like Tor. So every connection is encrypted, which means that when you receive a Lightning Network promise, you have no idea if the person sending it to you is the person who started the transaction, or it's someone who's just relaying it from someone else. And you have no idea if the next person you're sending it to is the end of the transaction, or they're going to relay it somewhere else. You have only one hop information. And so it also massively increases privacy and anonymity. Let's take some more questions. Yes, this is a smart contract. This is smart contracts using three technologies in Bitcoin. One is multi-signature technology. The other one is check lock time and check sequence verify, mostly check sequence verify, which is relative time from the previous transaction. And a new invention called hash lock time contracts, which is a way to forward a promise that can only be unlocked by a secret. Yes. So these are smart contracts using Bitcoin. Correct. Why would an intermediary want to do this? For a number of reasons. One of the reasons they would want to do this is because using this also involves participating in the network. So you just do it because you want to use the Lightning Network. Another reason you want to do it is because you can collect a fee. So part of the option there is to make very, very small fees payable to intermediaries if you want. And you could then select the route that gives you the lowest fees, or the route that gives you the lowest latency, and you can use a whole marketplace of services to implement that. Yes, this would introduce transaction fees, but you can talk about transaction fees on a whole different scale. Because you're not required to fill a capacity that is very limited, which is a one megabyte block, the fees that are likely to be used for something like a Lightning Hub are going to go close to the marginal cost of delivering that service, which is tiny, tiny. So you're going very close to zero. Anybody who charges high fees will get pushed out of the market by people who charge much lower fees, and if nobody else wants to do it, I'll do it. Yes. Yes. Well, you don't settle any of these hops on the blockchain because actually what you're doing when you're using a big route, so the question was, sorry, let me repeat the questions, if you use a route that's less costly but has more hops, won't that introduce more settlement costs on the blockchain? No, because as you, in fact, when you create lots and lots of hops, what you're doing is you're cancelling out bilateral obligations between the parties involved in the hops. So if now instead of, you know, Party 2 owing Party 3, now they owe a bit less, so you can actually balance out all of the settlements, it will actually reduce the settlements on the blockchain. We don't know yet how this is going to play out in terms of scalability, but I'm very excited. Let's take two more questions for lightning. Without on-chain scaling, won't this push a transaction fees away from the miners, and therefore compromise Bitcoin's long-term decentralization and scaling? No, it won't, because for every payment channel, you need an anchor transaction, which is the one you set, you use to set up the payment channel, and you need a settlement transaction if and when you decide to close the payment channel. And so you still need transactions, what it expands is the ability of doing a lot more transactions in between those two. Yes, so there is the possibility that it's too expensive to go between channels, in which case we're also going to need to scale up the core Bitcoin blockchain layer, which I think is inevitable. We're going to do both. This is not about a choice between scaling on one layer or another. It's about scaling at every layer. What these technologies do is they offer you leverage. So when you scale one megabyte on the lower layer, that effect is multiplied by a thousand on the layer above. So you can have this multiplier effect. Last question on lightning. Will lightning network be compatible with Bitcoin Unlimited? That is entirely up to the Bitcoin Unlimited developers to choose whether they want to introduce that. If they introduce segregated witness, and they've already introduced Check Lock Time Verify, then yes, you can run lightning network on top of that. You can run lightning network on top of Ethereum. You can run lightning network on top of any cryptocurrency that enables the three basic primitives of checking hashes, multi-signature contracts, and lock time, time-based controls. So it's a network that can be overlaid over anything. All right, great. So now we have a basic understanding. I hope everybody's got a basic understanding. The lightning network is this thing that can be layered on top of Bitcoin, creating these bilateral obligations that allow you to stream money at a different scale in time. Here's the bit that no one's noticed yet, which is that when you change the medium, the message changes. When you change the granularity of time, very strange things start to happen. Right now, we think of money primarily because of the containers of money. The containers of money control the way we think about money. So each type of container for money imposes certain restrictions on how that money is used, and we tend to think of money in terms of the containers it comes in rather than as its pure form of transmittable value. All of that to mean this. If it takes three to five business days to make a bank transaction, as it does in the slowest banking country in the world, which is the United States, if it takes 24 hours to make a banking transaction, if it costs certain money, those things affect the behavior of people who are using money through the system. How many of you here are paid by salary? How many of you receive your salary automatically through a bank account? That's almost three-quarters of the audience. Great. How often do you get paid once a month? Why? Now, this is a question we haven't really considered. What is the nature of paying? Why does it occur at monthly intervals? Why do we chunk money at monthly intervals? This is a very good reason. This is money acquiring the characteristics of its container, the medium of banking payments, the accounting systems, the ability to pay employees, are constrained. It gets expensive to make transfers more often using the banking system. Let's look at some parallels in history before. Right now, we are living on the internet in the era of streaming. Streaming has become one of these enormously powerful concepts that is changing the way we consume various things, the way we experience various things on the internet. For example, MP3 music is disappearing. Why? Because I don't want to hold 30,000 MP3s on my mobile device when I can stream them in real time from a provider. How many people here have stopped storing MP3s on their mobile devices and used the streaming service? That is 75% of the audience. This concept didn't exist 10 years ago. A lot of us who were involved in the internet recognized that at some point, the value of storing the data versus streaming it live would switch, and no one would store their own music collection. In fact, now, a lot of the value comes from the editorial, the curation, the DJ, the playlist. If I take my 30,000 MP3s on my laptop and I hit shuffle, bad things happen. Why? Because I have a very broad multi-genre music collection. I may go from Tchaikovsky to Iron Maiden to Justin Bieber in three minutes. That may damage my psyche, especially if I end with Justin Bieber. Okay, I don't actually own any Justin Bieber, but just as an example. The bottom line is that we now value not the permanence of music, but the experience of music. It has changed how we experience music. The same thing happened with videocassettes. Remember when you had to go to the store and get a videocassette? For anybody who is on the 30, a videocassette is this plastic thing that you had to rewind, kind of like a DVD, only suckier. Now, if you experience movies that way, it changes your experience, right? You have a limited catalog, you have to consider very carefully what you buy, you experience it in a different way. You actually sit down through a whole movie and enjoy it. Now we have streaming video, we've changed our expectations. It hasn't just changed our expectations in terms of using Netflix. The really important impact is with the emergence of things like YouTube. We have gone from experiencing video content in an hour and a half to experiencing video content in 15 minutes to now experiencing video content in 30-second vines and Insta videos. That experience is completely different. Something happened there when you changed the container that changed the actual experience of video. Now you could start experiencing video in much smaller amounts. It could be created by a whole lot of people you've never met who didn't have any production values and still sometimes be quite surprisingly good. Streaming video changed the nature of video. Streaming music changed the nature of music. What happens when we start streaming money? Now think about this for a second. If I can do payments that are on a millisecond frequency and that are as low as a Satoshi, why not get your salary paid every minute? Now, this has some really important implications. If you think about it purely from the perspective of salary, now you're working real-time. Money becomes a real-time thing and its nature fundamentally changes. I watched this video the other day, which was a team at a university that created a camera that could take one trillion frames per second. Then they shined a beam of light through a plastic bottle. You see that video and the beam of light suddenly changes into this chunk of light that is moving through the bottle. You can actually see individual photons clump together a pulse of light moving through a bottle. Our experience when sped up is quanta. It's discrete units, but in our everyday experience it's not. It's a wave, it's a flow. Nature is like that. What happens when you change the time dimension of money and you convert it from something that is packetized, that is discrete, that we've been used to dealing with for generations now in chunks of monthly payments and quarterly accounting, and maybe for lucky daily payments. Now take it down to milliseconds. When you do micropayments over milliseconds, the term cash flow takes on a whole different meaning. Cash is a flow. It's a continuous stream that has no meaning as an amount. Imagine doing accounting in businesses on a real-time basis based on flows of money coming in and flows of money coming out. We have not even scratched the surface. Until now we appreciated the fact that Bitcoin and systems like it converged all of the networks of payment that we have in our world. Just like the internet converged all media onto a single network, Bitcoin and cryptocurrencies converged payment networks. Today we have payment networks that are big for governments to pay each other, payment networks that are small for us to pay each other, payment networks that are for consumer to business for retail, payment networks are for small amounts, payment networks are for large amounts. If you were here before the internet, and I certainly was, we had communication networks for sending photos, we called them facts. We had communication networks for sending letters, we called it telex. We had communication networks for sending voice, and we called that the telephone. And the internet brought all of these together. Bitcoin combines the ability to do transactions from the level of a microtransaction all the way to the level of a giga transaction. So it's compressed the space of payments so that a single network can support making payments worth billions to making payments worth pennies. That's compressing space. Now with this new introduction of a time dimension into Bitcoin, we are going to compress time. We are going to open up a completely new dimension to money that allows us to address money as minute flows that flow continuously and can be aggregated and split as streams. And when I say streaming money, it's going to take 15 years for us to really understand what that does. What that does to human payments, what that does to corporate payments, what that does to border payments, what that does to the nation state. I don't know what that will be yet, but I do know one thing. It's going to be big. And that's what streaming money is. Thank you.