 Hello, I am Dr. Chris Demchak, I am the U.S. Naval War Colleges Hopper Chair of Cyber Security. What you're about to see is the first of four very basic lectures, given in the spring and summer of 2021. These lectures are intended to provide you a very basic fundamental understanding of four key technical issues in national security. The first topic today is hacking and cyber conflict. The second one is artificial intelligence and machine learning. The third is blockchain technology and Bitcoin. And the fourth is quantum qubits and implications for the future. These are at a very basic level, but the intent is that you will have a fundamental understanding so that you can read articles, understand the language, and have some grasp on the national security debates attached to these technologies. Thank you very much. One of the underlying points of the talk today is that this problem is in a large measure our fault. It is literally the United States is very technical or technological bad parenting that has led us where we are today. And this is what we told the world everyone was going to get if they used our internet in the 1990s. It was going to be universal prosperity, universal love, universal knowledge, anyone could know anything. And by the way we wrote about governments fading away laws disappearing because they won't be necessary. And if any of this sounds like what you've heard recently about blockchain technology and smart contracts and we don't need judges anymore. It's quite similar. Unfortunately, this is what we got. We get no global village we planned we got no universal prosperity. We didn't get an open free safe and we did not get this universally democratizing internet. None of that happened from what we had planned. So I'm going to make a couple of arguments to you today. The first and fundamental one is hacking starts with just because they can as Keith Alexander once noted and that's because we did that we made, we chose to and made this incredibly insecure substrate, and it basically penetrated everywhere in across our societies. And this insecure substrate provided five offense advantages which I'll explain to you. It didn't cause the rise of China but it accelerated the rise of China. It encouraged a new kind of conflict between peace and war we often call call that now the gray zone. It induced a different kind of cyber West failure it induce borders and cyberspace as people start to want to protect themselves. And now, all of this together, this combination of this large scale adversary the shoddy cyber street, and this very fragmented response across democracies, many of whom's leader can't get over the 1990s promises as producing what I'm calling a great systems conflict, where we're in each other's systems in each other the socio technical systems of each other's society. Now for us in democracies, it is nothing less than our own future well being at stake. And at the end I'm going to talk a little bit about clouds, because clouds are one of the great promised solutions to this problem. And I think it's worth knowing something about them as we do this. All right, let's get started. This is the basics code is everything. It starts down here. And you've heard about ones and zeros and you've ignore them now we don't need to know much about them but what you need to know is that everything we do ends up somewhere, turning on a circuit, either closing or opening a circuit at the bottom. And that's all the machines understand is do I open the circuit do I close the circuit. It's translated into machine languages, and that's where their ones and zeros become. Now you will note that this set of ones and zeros is just a stream of ones and zeros. The important part about this is that those ones and zeros have to be divided up in some way for us, the humans to understand what the machine is understanding. And so from here on forth it's important to understand that all of these rules of computing are rules we agreed to. No one in a physical sense is making us agree to them. And so if you come along and you know the rules, we've agreed to, but you break them. And you use that to your advantage. That is part built into the system, because you just said the consensus doesn't work for me anymore. So what we did is we agreed that we would break up the ones and zeros into various chunks and bits and then bites, and we would start make those we would make those chunks something humans could read. And that led us to assembly language where we started telling the computer to do things by using these chunks of bits and bytes monthly bytes, and we could read them because they would reflect back to us on a computer, and we could tell a computer to do something. So first that meant you had to memorize all of these basically these commands, and after a while, we developed what they call high level languages, so we can write complicated long streams of code. And as we calculate and we send it down to the computer it makes a lot it's able to do a lot of complicated commands, and that created languages there's lots of ways to do this. But what's the other important thing if you all you see is the code I have up here. You may or may not understand how that code translates down to what that machine is doing. And so you find that people have to start learning more layers if they want to hack into they can hack into it up here, but then if they can get lower into your stack. They can make you download a program that works down here and hide. You might not necessarily ever see it. And remember, it was always consensus. The second important point is that sloppy code is now everywhere. Right. In the 1980s, in particular, we started having choices. The first sort of languages at that mid the higher level, particularly the mid level were fault intolerant that meant that you had to actually code it very cleanly very precisely. If you had errors in your lines of code, the program would fail. It would be pretty much hard to hack but it would be really slow. And that meant you had to have people checking code after code when a huge program would fail you'd have to have alpha beta and Charlie sets of testers. Right. So in the 1980s particularly turning on to the 1990s there were other languages. These languages actually developed closer to the 70s but nonetheless let's see other languages began to dominate. Because they can be very quickly coded they don't have to be checked every single step, and they will run anyway more or less. So errors can be left into it, and you can finish making the code you can put out that application. It will more or less do what you want it to do and sloppy coding or errors. Well, until somebody finds them. They may or may not ever cause a problem. Of course that meant it is now much, much easier to hack. If I can spend the time to find those errors and get around you. Well then 1990s forward. That's what I'll do. And so from the 1990s forward we start seeing that these easy quick fault tolerant languages, see for example, go forward and they build the world you know today. They build our e-commerce. They're the basic construction of the Internet. And of course they translate very nicely into all what they call the Internet of Thing devices that you have today. So all the things that were already hackable are now spreading into all the things that are all part of your society and your life everywhere. The only thing to know is that access is everything. And the three main ways in to get into whatever it is you have and you care about and you need. Right. First off there's brute force. This is the old fashioned way where we spend our time guessing your password and there's a number of ways to do that. And then we just pretend to be you and we get in. It's not as common anymore because there's other ways easier. The second one is the type error exploit. And that of course is the one that we just discussed where you've left errors in the code. You have a whole line of code somewhere where you did not specify exactly what needed to be in that particular section, and basically a hacker finds it and they specify something else. And they change it. Now they can either do it to just disrupt your code or they can do it to have your code allow an application to install someplace else. And the final and now pretty much the biggie is what we call fooling or fishing. This is the one we all know right this is the one where you get the email with the please click here, and please go ahead and you know, or the, the false ID, right email, your account is closed unless you tell me this, that sort of thing. And if you can fish your way in through a document that you know people can almost can't stand not to open. Then the document can carry the malicious code that the computer reads as an executable and goes, Oh, okay. Off it goes, and it drops something in there. All right. That's what the world is like, hacking. It's insecure. It's everything you use. How did we get to such bad parenting. Well, for one, we were greedy. So when you talk to Bill Gates from the 1980s and 90s, you know, this, this was great that we were going to spread the computer all around the world and make it great, you know, but they want to make a buck on it. To make a buck they needed to use these cheap languages, and they needed to connect it everywhere but even more, it was naive. There was an expectation that we could code it everywhere, and the world would be completely happy not to take all our mistakes and use it against anybody else. It wasn't that they couldn't expect someone might try something. They just didn't expect them to, to succeed or be at the scale it ultimately ended up. But at the time was, remember, governments would fade away. The worst thing would be for governments to have regulations or constraints on anyone because that would kill the innovation egg. And you don't want Rick, you don't want regulations because it just kills the innovation egg. We believed it on our policymaker side, we believed it on a commercial side, and to this day, you still say, well, we don't want regulation because that will kill innovation. This is a dominant line everywhere. Even, even when we do have examples of regulation not doing so and we see no relationship between them. That is a dominant line. So now even to agree, let us spread this everywhere. One of the differences for us was that we opened up ourselves to the world and handed them five offense advantages that the average every man with hardly any money never ever had before. And from here forth. Anyone could create a botnet, effectively an army, the size of which only emperors could have paid for if it were humans, or your close neighbor was the only one that could put that level of hurt on you. Secondly, we didn't have to be anywhere near them anymore to get high level intelligence. Again, only emperors superpowers could afford to do that or your, you know, your neighbor next door. Thirdly, we didn't have to commit to a particular weapon with all our money in it, right, because now we can the huge underground cyber market that has grown up along with this internet. We basically we can buy a tool that will attack our good friend Laura, and half of Detroit. We can buy another tool that will attack me, you know, I don't know, Switzerland, and a bank there, and we can run them all simultaneously. We can run them for a little while we can stop and go home, you know, have lunch to start up. We can make any choices we want in multiplicative. In the old days when you wanted to actually harm someone, you generally, they generally knew what you were using if I'm going to use a missile on you you got a sense okay that was a missile I got that part. Right. But now we don't want you to know what the tool is in fact we want you to lose time hunting for what it is and how I used it and how I got in. Of course, this is part and parcel of what you've heard is zero day exploits. So if I've got a program out there it's been out there for years and it's spread all over. And no one has yet found that place that sloppy code line in there, and somebody finds it. They're not going to go out and go look I found it. They're not going to wait, or they're going to use it for themselves, or they're going sell it on the cyber crime market, or they'll store it away and go waiting for the day and then I get in I'm going to make sure that they won't know what hit them, because, of course, I don't want them to know. And finally, opaqueness and conflict before, if I had a conflict with you eventually figure it out. You know, I'd show up with my tanks or my guns or whatever, or my bully boys but I'd show up somehow you'd know. Now, we don't want you to. We don't want you to know exactly who it is and you've heard all about attribution how can we know it's the Russians or not know it's the Russians and the Russians say no why. It's very clear why. If I'm going in at you that I'm connected to the internet as well. And there's a good chance you can come right back at me in the same way. In fact, you can go to the underground cyber market and find ways to get me so I would prefer to, you know, maximize my return without having you know what I have. And who I am. Right. Now we've made it worse. Completely unanticipated by this, this community building this shoddy cyberspace already we had complexity problems already we had large scale enterprises that were very complex, and they would have nasty surprises and they would deal with them and they'd have security issues and reliability issues. And these enterprises would, would link up into critical infrastructure we talk about that, and they would have interdependent surprises from the complexity and society had to deal with it anyway. But now we added a fourth, if you will layer of this huge malicious malicious community of bad actors across the world, able to use these five offense advantages and reach in and cause us harm. And to add to that, on the top of this crew, if you will, in terms of skill sets, maybe 10 to 15% of the super scary smart ones, and they tend to be hired by our adversaries, or by criminal organizations, and these guys are super scary smart. And they are not deterred like the regular bad actors are, and they are also using those offensive advantages and access into us. And they're causing major problems for society. So here we are 30 years later, and we are paying the price of our poor technology parenting, right. Our consolidated democracies are losing on average one to 2% a year. That estimate was before COVID, of course, we have fragmented government responses and, interestingly, because of that naivete that it technologies only for the good and if someone misuses it what's a few people here is not too bad. Our private sector is largely missing from the defense of the entire society, they defend themselves, and they'll try and defend their clients, and they'll try and bolt on defenses on top of that very insecure substrate that we put together, but they are not taking responsibility for defending the country. So whole. And this new form of cyber conflict is mutating. It's looping across all of these sectors, as the internet itself, and now cyberspace spreads, and has different manifestations, right. But some at the end of the day, the consolidated democracies, particularly us who are responsible for this, turn out to be pretty weak cyber powers overall. We have poor resilience and even if we have a cyber command, they have limited missions, limited ability to respond. And as I said, our private sector isn't even involved. Let me sort of roll you through this very quickly, because I just an academic talk you have to have a quick model. And this is the model this is the virtuous cycle we are stuck in. Over here, you have the offense advantages, and they just keep rolling along. Here you have the four layers of systemic surprise, and the offense advantages just keep feeding them. You have weak systemic resilience and if they're lucky, maybe some forward defense. And that ultimately means that for cross our socio technical economic system were weak cyber power. But of course governments want to respond so here's our governments responding and because we're democracies. When they respond they try and keep themselves to the lowest level of cyberspace, the connectivity, they look for bad guys there. They're not trying to change the content, and they're certainly not trying to change the cognition side note, the adversaries of course play along all three of these. And then here they'll focus on the wicked actors and adversaries. But the loop just continues. It just continues going forward, because we don't understand it, and we don't break into it. And it did not make the rise of China. It accelerated the rise of China that's an important distinction to China was going to be an adversary. It was going to be a peer problem. But even for the Chinese, the transfer of wealth that was possible transfer of wealth by the way includes intellectual property that been possible with this shoddy cyberspace would never have been possible otherwise. And it certainly has advanced their ability to not only match that technological advance, but to using government subsidies and their sheer scale everywhere around the world to be present where they can remove IP or compete with those subsidies with IP that was borrowed from someone else with extra knowledge, they can take over markets and was never supposed to happen this fast. Our own estimates were that the Chinese GDP would match the US in 2030. The Chinese themselves were working on the expectation it would be 2049. And yet we have pretty substantial data that if you test if you basically assess GDP as purchasing power parity PPP. And Chinese already matched us in 2014 and in 2019, where I got this data, they were already 20% ahead. This is before COVID. So it didn't make the Chinese problem but it accelerated their ability and indeed we're seeing a commensurate response on their part. A certain amount of expectations Americans. So, their rise is contributing. They have lived off of these offensive advantages. They are using any of those connections any of that coercive power that comes from it. And they are determined to rise. And they are determined to be the central node in future global business now their demographic scale would lead them to that no matter what. And now it's leading to them much faster than they expected to be. And there's a certain amount of arrogance that's going with it, right. And they're of course using that scale to undermine that what we thought were the permanent rules of the international system. So now we are looking at the way cyberspace was built, how it enabled an adversary, and now we're looking at the fact that we can't control the international system anymore as we thought we would. So the Chinese are a key economy you can expect as ours did, as ours did when we were the key economy in the world, that their business practices will dominate the digitized world. And that means for a lot of reasons their scale the strategic coherence but particularly their wealth on their, you know, what do we call economic state craft. An interesting example of economic state craft we can talk about it's what China's doing to Australia today. Now one of the things I wanted to make sure that happened today is we talked about levels or areas in which we don't understand cyberspace beyond hacking, but very related to hacking. And one of them is what is the actual structure of the internet, and why this matters is because people, very often people who are making policy, people who have the mission to defend us in cyberspace cannot tell you how the internet is constructed. Right. My, my joke is it's it's like handing someone a fleet and say please go to war. And it doesn't matter that you don't know how ships work or how the ocean works. Right. So you, this is a sort of fundamental understanding part that is necessary, and very often missing. And part because it grew up like topsy. It didn't grow with a plan. It grew up like topsy. So let's look at this for a moment. This is a nice one. It comes from the from Wikipedia, I like it very much because it simply captures most of what you need to know. So here you are down here the user, and you're typing on your little application, and your application is being translated into the ones and zeros and sent across wires. And at the other end, someone is accepting those ones and zeros well they're actually translating it into packets, which themselves are in ones and zeros but they take your message and they cut it little packets, and they send it out. And it comes over to the next place where they accumulate and move the packets on and then finally at the recipient puts the packets back together, and you read the message at the other end. That's how it all works. It comes down into little packets with headers on them, and every node they pass a little header is added, we're taking off and it looks that way. But importantly, this is not magic that when you send yours to your ISP you signed up you're paying for that service so they are you are paying them to take your packets. They then pass it on to someone else, whom they also pay to take their packets. And then those people up to the tier two networks if they're not appear will send it on to what's called the tier one of backbone networks, and they will pay them to take their packets. It's very top with so many packets passing forward so much traffic, the peers, what we call it backbone peer networks. They don't charge each other. What they do is their peers, I'll take yours you take mine, we're making our money on what comes up to us, and we'll send it across. So, okay, this works like this. And these little blue things are all what they call internet exchange points, and an internet exchange point is not quite like a box in your house. An exchange point is a huge building generally speaking with tons and tons of servers. And all of these ISPs are buying server time or they're positioning their servers there and they own it or they have rented it, and someone is making sure the building is cool and and then everything goes in and out like that. These guys, of course, post anyone who wants to come in. So here you are you're sending your traffic this looks very good. And anyone who wants to offer a, I'll take your traffic, and I'll send it on buys a space on that. And when they buy it in a large in large chunk that's called a point of presence. I'm here and China telecom, and I now am on these servers so if you are a China telecom client, you in particular I want like you to come and send your traffic for me, and we have a contractual relationship. Okay, no big deal. So how does the internet itself work I'm sure you've heard that it was built so it would survive in a nuclear war, not entirely true. But it would capitalize a little bit later. reality was it was built because they were first sending these packets, this information across crappy telephone lines in California. And so they needed to have a way to make sure that the message actually made it if you send a whole message, and there was a break in the line or some glitch pieces would not make it. You send the packets and you sent it down. And then what happened was, it figured out that the best way to get the packets there was to give the packets a choice. In a sense, not a choice, but a way to make a route that would go around broken bits and the wires. So you end up with what's called TCP IP, where the packets are sent along paths, but not all packets take the same path. And they are often sent because people declare I am a neighbor of this particular autonomous system or that particular Thomas system, and therefore the shortest path is the one the packet should go along. It makes it fast. So if you normally go through me over here, but I'm not working well and I'm not the shortest path, then you can go over here. Now this doesn't work very well doesn't. Let's say, it's not showing it well I presume there's one over here, and I can take another path. The Internet also works on something called autonomous systems. Now why this is important I will am about to show you when it gets hacked when it started, it started in universities. When you're inside the university, everybody's account in that university goes up through the servers of the university, and then out with your little university account. You don't go out unless you go up through their servers and then out. It's like a tree. So at that point, all that network is called an autonomous system because there's that clinch point and everything underneath it has to go through that clinch point, or gateway choke point. And then autonomous systems then became the fundamental way you send emails. It's like having cities. Imagine that you can't, you can't go to another city and post a piece of mail. You have to be in your city to post your mail it has to go to the main thing, the main post office in your city, and then it can go anyplace else. But that makes it a little easier, because then I can send my packets to autonomous system one autonomous system to autonomous system three that's how I can label them. And then they can drop back down to where they need to go. Right. And of course, if one of the autonomous systems isn't working well. I can skip it. I can go to the first one and find out the one I wanted to isn't working well or not accepting traffic or something. And I can go to another one and go around it. Right. And that's the fundamental, we can survive nuclear war is the assumption that one of those nodes is not working well and will send it around. All right. Now just look at here for a quick minute. And say, All right, I'm on six. And I need to get to 20. And I'll go 123 hops over to 20. Keep that in mind. That's my three hops. I'm not going to go over to the number as three, because that would be 123 four hops. Keep that in mind. No one is restricting that route. I can't assume it's going to go this way. But someone can go in here and force my traffic to number three. It was never supposed to happen that way. And the routes are advertised by something called border gateway for protocol. It's a terrible name. It just means these are all the routes you could take. But the routing table that the BGP actually announces is what tells you what are the shorter routes. So if I want to play with this game, I might want to change your traffic from two to three. So I get a copy. Or I might want to change the routing table at BGP. So let me show you some story, quick story. So another thing that is often forgotten is that 95% of the internet travels and undersea cap cables. This is the internet is fully a maritime problem. And one of the interesting things about all those cables is every single one of them shows up on somebody's sovereign territory. Not only do we have an internet that is peer or pay, it's consensually agreed upon. There aren't rules saying yes, you but not you and so on. But it comes up on someone's sovereign territory to transit to someplace else. And anyone along the way can make rules about what goes and what doesn't go. So let's give you a little example of someone who understood that we didn't understand our own system. And China telecom is one of the major. It's still one of the top eight or nine telecommunications corporations. And it is Chinese. It is, is definitely a Chinese origin is not a commercial I mean it's a commercial endeavor but it's a Chinese subsidized. And it took this concept of points of presence just just to remind you. And it dropped 10 of them in North America, and all of them are very strategically placed right by major traffic hubs, where traffic needs to have a place to go. And so China telecom can easily insert itself into the BTP table saying oh if you want to send traffic, I'm here, I'll pass it along with anybody else. It's those major exchanges, and it offers to be, you know, a neutral party passing traffic, except it doesn't necessarily. And here's a good example. So in 2018, a colleague of mine you've all shall be actually late 17 came to me and he goes Chris I've got this data and he's a BGP expert and what he does is he looks at these flows of traffic. He's talking about one little packet. We're talking about flows of traffic. And he looks at these flows of traffic, and he tries to determine where the routes are that they're physically falling not what was planned, but that they're physically following and he started to see this funny pattern. And here's one he tracked for six months. And this is where you have the government of Canada is just sending something to the government of Korea for six months, not one thing but a stream of traffic. And here's where it's supposed to go, you know, Toronto to LA to Korea, and this is what he saw that in Toronto. It hits a China telecom point of presence. What if the CT point of presence just sending it up to that peer network where it's cheap and free, just go work between peers and do what you got to do. The China telecom kept it in China telecoms own set of connections, and it went from Toronto to a China telecom pops over here in China, went to several of them, and then finally ended up in Korea. And you say why didn't they notice this, because we are talking about streams of traffic with tiny little legs. And if you're not paying attention, or you don't really want anyone to pay attention. Then you make the lag small, but you're nonetheless you're listed there. Right. This one on for six months here's another one I just like. If you look at the, and I can give you the link to it with peace we wrote on this topic. There's a few more, and I can also tell you about some more if you want. But anyway, here's another one I just like this one because it went the other direction. So here's Houston to New York over to Italy, and it's just a bank. 2018 lovely bank. What did it do hit a China telecom pop in Houston went to one in in the in California, and then it goes to China, and then it just goes lost. It never gets to Italy. And this, this continues for a while and you can understand the bankers are calling each other going to see what happening and not yeah I don't know whether. I don't know a lot of you send me another one. Right. And this happens across very distinct, very distinct email exchanges that he was able to know. The future is coming very fast for us now. We have to understand what we've built. We have to understand what is being built. So, knowing what we just talked about now, move it forward and understand that the technology coming out of China is not even it's not not more perfect. It has connectivity and embedded in it it has presumptions about how governments and other people can use that network, or that application, or that tool. And this is the Chinese spreading technology they fully intend and the new by the way the new date for them is to be the technology dominant giant by by 2030. And I want you to pay attention to, as time goes on, the smart cities. It sounds like the right thing doesn't it crime happens in our street. We want to have those, you know the the equivalent of the modern ccts. You know cctvs that are all over the UK, we know that those are really good things. But think about that for a moment, where can that facial recognition. Where can all that data be used particularly if the tools you use connect to the internet, and you don't have full access to everything that's programmed on the inside and they're spread everywhere. I can talk about that as well. And also what we know is happening to the rest of this open free internet is borders are rising and internet that countries are trying to protect themselves and we know the adversaries we're going to do it from the beginning. And remember at early on it was the argument governments are too slow they're too stupid they can't control anything what are you talking about don't even think about it. And now we have some lovely examples of governments controlling the internet, not being too stupid, and not being too slow to do so. And we ourselves are now developing strategies where we talk about France's internet in Germany's internet. So we ourselves are in this game of making. And of course that's going to make it even harder to understand who's translated what to do what at the other end. So we've throw a few words in about clouds. clouds are neither complete solution because remember they have to rest. They rest on top of this shoddy substrate we built. It's also wrong, because it does allow us to sort of organize and look at, if we're allowed, what's going on underneath it so it's going to depend on our parenting going forward whether clouds or a solution or snake well. So here's our lovely picture. And now let's do this. Here are our clouds. So what we see at the moment is that the clouds are part are provided by owned by maintained by designed by commercial entities that have no obligation to share any of that information with us. If you put clouds all over this lovely picture I showed you below. And then you tell me, you can't know how we designed it, you can't know who designed it, you can't know the structure within it because that's proprietary. Then how am I to know whether or not an adversary has gotten into your cloud is one of your engineers is moving material out without your view of it because these are huge phenomena. So we are moving down to the last mile. And that's a different thing that having AT&T set up your telephone wires. So we have to ask ourselves what kind of parenting we want to engage in going forward. I take a socio technical systems approach. I'm looking at all of this. I argue we need to be good parents, as well as good defenders and to be a good defender we need to have systemic resilience across all of our society. But we have no robust cyber powers today. We have no democratic nation that alone that can defend themselves. And none of us will survive without the others. Start with recognizing the scale of the challenge. So last couple slides here. What do we have to do first. We have to clean up our own house we have to understand the fundamentals the fundamentals of the Shawty cyberspace how it built out what's engaged. Why AI is going to rest on it and is a side note on giving three other lose on artificial intelligence blockchain and quantum largely because that's what I get asked about. But that's also what we need to talk about the fact that the offspring of this thing. And we need to have our government and our government leaders and our own senior military leaders fully understand this. And then we need to involve the private sector particularly in for this problem, our telecommunications corporations. They lean towards us but they want liability issues as all bunch of things, but they need to be involved in our systemic defense. And then we need to have our own it capital goods involve corporations involved, and this is going to take war time temple R&D. We are here we have to break into these five offensive managers, we've got to take them apart. And to do that we're going to have to rebuild, transform the underlying substrate. Right. And to do that we're going to have to basically put money out, and we're going to have to give money to universities to labs and to the private sector to come up with something that we can replace it with. And it's doable. We've done this kind of thing before. And to do that, we didn't get a handle on the these surprises, and we are ready to not defend alone. All of us, then we work with allies and we create what I call the cyber operational resilience alliance that we build each of us build with our other Democratic allies, so that we become in a sense to the Chinese one great big cyber pier. And we demand peer difference. We have strategic coherence we have scale. And of course, we're making it much harder to play the game that has been played for the last 25 years. So this is our only feature in my worldview, we have to do it collectively. And if we do not do it collectively, both internally and with our allies, then as Franklin said we will certainly hang separately.