Grand Tactics and the Thin Skin of Civilization
A TECHNOLOGY FOR GLOBAL SECURITY PODCAST
Featuring Paul Bracken, Peter Hayes, and Philip Reiner
Listen to the full podcast here: https://www.tech4gs.org/dr-paul-bracken.html
OVERVIEW
In this podcast, Paul Bracken analyzes the big structures and large processes of nuclear multipolarity. A framework for analyzing this global system is developed, one made up of national command and control plus the “system dynamics” of their interlinked behavior. The paper underscores how advanced technologies — cyberwar, drones, and anti-satellite weapons (ASAT) — affects NC3. The structures include the national command and control of at least eighteen countries, to include nine nuclear weapon states, “shared” weapons in NATO, missile defense, and key intelligence nodes in select countries. Processes include the delegated flow of launch authority, innovation, and digitization in many forms.
The podcast specifically assesses the two big structures that are forming in Europe and Asia. NATO is attempting to modernize its nuclear deterrent for the new realities of European security. In Asia, a pentapolar structure of major powers (United States, Russia, China, India, and Japan) has growing nuclear interactions: in missile defense, cyber, space, and in upsetting the U.S.–Russia strategic balance. The critical importance of information transfer for bolstering a coalition member’s ability to target its nuclear forces is analyzed as an example of the “new” dynamics of multipolarity.
Paul Bracken is professor of management and political science at Yale University, and the author of The Second Nuclear Age (2012), and The Command and Control of Nuclear Forces.
TRANSCRIPT
Philip Reiner [00:00:10] I think what I was interested in doing was maybe kicking it off with a bit of a discussion that you begin to talk about in the paper that we really seized upon in the workshop which is this notion that nuclear commanders may not actually even understand their own system as it is and how the introduction of these emerging technologies whether it is in the cyber domain or then exacerbated by the introduction of AI techniques how it compounds that set of issues and really wanted to dig into that a little bit further and build out your insights on that and we can kind of start the conversation with that.
Paul Bracken [00:01:00] OK. Well a couple of things on this. Not understanding your own forces. First of all if it was true back in the Cold War where at least on the American side I would bet on the Soviet side. We did not understand our own forces in terms of what they would do. This is for a couple of reasons, one American nuclear forces never went on high alert. The number of alerts was actually quite low. And every time we did one of these we found some incredible new discoveries. So sort of that’s one dimension.
Paul Bracken [00:01:37] The other dimension bringing it into the 21st century is the radical nature of these new technologies which are being acquired from the bottom up. And let me just use the term advanced technologies to cover things like drones, AI, cyber warfare intercepts and such. First of all most of these programs are highly compartmentalized. They are new. You can see the organizational restructuring already going on in the United States with NSA transforming into NSA and Cyber Command as one example. But we could go into Global Strike Command and what’s happening is from an academic point of view is that very different kinds of information are being collected and absorbed and compared to the Cold War. There’s vast increases in the amount of information. So I frankly don’t think it’s at all surprising that the United States doesn’t understand its own forces.
Paul Bracken [00:02:47] As you probably know at Yale, I teach in the business school and as well as in political science and I’ve been working with a lot of companies and you find the same phenomenon for large global corporations. So I’ll just give one example very recent example, data analytics, big data is one of the hot areas as you know along with AI and the companies that I’ve worked with are all supportive of it. There’s no bureaucratic resistance to it but they simply don’t know what data to collect. And so they how they don’t know how to link this with strategy. So this is a general problem. I’m not I’m sort of not agreeing with those who blame all of this on Pentagon bureaucracy. It’s just a hard problem and not surprising.
Philip Reiner [00:03:38] Yeah it’s interesting to note that one of the ongoing conversations that I am familiar with along those lines is when it comes to the Pentagon trying to introduce some of these AI techniques into their systems…yeah it’s a data management question at the outset. They’re sitting on so much information they don’t even know how to deal with it.
Philip Reiner [00:04:04] They don’t even know how it could be useful. And they’re really there turning to public industry or private industry to try and help them figure that out. I think they’re starting to make some inroads in trying to answer those questions. I’m curious. One of the one of the elements along the lines of what you’re just describing that we’ve engaged with some others on when it comes to the introduction of some of these new advanced technologies is their assertion that it really doesn’t fundamentally change much it may speed things up a bit. There’s more but we’ve also got systems that can process it faster.
It’s just as complex as it was for senior level decision makers back in the day as it is now the time frames were so condensed. Even then it’s really not going to change all that much. I was very curious to hear a little bit more granularity from your perspective as to how you see it having actual how these technologies may have actually had an impact so far and how they’re going to change things going forward.
Paul Bracken [00:05:15] Sure. I have this view that there’s a lot to be learned from the Cold War. Of how the organizations responded to the radically different environment. Clearly the technologies especially in the information technology areas are radically different and the international politics are obviously quite different but in what’s going on now is they in the absence of understanding how to use these things they still invest in them. And so what you’re seeing is what I would call grand tactics instead of grand strategy by which I mean these technologies are coming into the force and the systems very much from the bottom up.
Paul Bracken [00:06:07] They are not top down directed. There’s a very interesting book that Freeman Dysonrid did 10 years ago.
Paul Bracken [00:06:17] I think it’s called the “Something in the Genome” where he makes this point about scientific discoveries. That some of them come from new theories like Newton sits down or Einstein sits down but others come from the invention of the telescope or the microscope and the effect of these things is to do is to introduce information and knowledge. You never had before. Therefore you couldn’t theorize today what I’m saying is you couldn’t strategize about it. So I would view this is to get to your point, is it different or is it largely the same. I think there’s two columns.
Paul Bracken [00:07:02] One column is things that don’t change very much. Second column is that things do change a lot. Now I can easily construct that chart for the early part of the Cold War. How much did nuclear weapons impact things. And something simply did not change. But other things dramatically changed. Most of us would be familiar with what they were. The point I would make here is also that it took 10 to 15 years for the people at the time to figure out what was on which column of the chart and more sort of in year two or three for the twenty first century version of this story.
Philip Reiner [00:07:49] I appreciated that reference in your paper in terms of thinking of it almost as an adolescent where it’s untested and unstressed. And in this more novel environment we really haven’t had the crises that were necessary for leaders to do to begin to understand those systems decades ago. That really hasn’t happened so far.
Philip Reiner [00:08:18] It’s a curious thing too for us to try and because the timescales are so compressed and because if something were to actually arise in today’s environment to begin to stress these systems it’s really unclear how the escalatory ladder can be constrained.
Paul Bracken [00:08:43] I mean correct. And no one is suggesting let’s have some crises here so that we can learn about these systems. I mean it you do get insights.
Paul Bracken [00:08:54] I mean the fact that the crises of the Cold War sort it started out small and then only got big with the Cuban crisis and that was a matter of political luck. But who knows what the story is going to be now. In any case we don’t control those things. But I think the absence of a learning curve or the fact that we’re at the very beginning of the learning curve that the technologies are being introduced from the bottom up so that senior commanders themselves who grow up in a different era and the other thing I would point out is that we had the strategists come out of the early part of the Cold War who had a very interesting role. I mean this is Tom Schelling people at the RAND Corporation in the 1950s and the intense blood on the floor arguments about is Albert Volstead are correct when he says something. I guess I would make a different point that all of them had the effect of ending the government’s monopoly on the debate of these issues and looking at them in new ways. And that by itself was an extraordinary contribution even if they were wrong in their particular arguments about different issues whatever they may be. And I see definite signs here of younger scholars younger people being attracted to this field once again after for a long time not very much was happening. So I’m hopeful that some research institute or collection of you know government officials scholars think tankers will make progress on here.
Paul Bracken [00:10:43] One more. It’s very relevant to this too is that an interesting twist on today’s situation is the diversity of the countries with nuclear weapons. So back then Soviet Union was not particularly conducive to innovation on the strategic level for these things. The United States was and there was a lot of people who wanted to put like the Rand Corporation out of business. But now you have. So what I’m saying is countries like India maybe Pakistan or people from another country could actually make major contributions today. I think there’s a much greater appreciation of the dangers which would invite them into applying their creativity in these areas.
Philip Reiner [00:11:39] And one of the interesting elements of that. I was just spending some time with some folks talking about this yesterday was the strategic thinking it’s almost required for some of those other nations is constrained by the fact that they don’t necessarily have some of the same advanced technologies or access to them. So for instance South Asia, India, is I think the one that I would think of when it comes to artificial intelligence for example they are just beginning to think about how to really push the fold when it comes to developing their own internal ecosystem if you will for developing AI related technologies and then integrating them into the defense establishment. So beginning to wrap their thoughts around how it’s going to impact things like nuclear command and control is such a step further down the line. I would agree with you there. We can hope that that sort of strategic thinking comes out of these scenarios that are driven more by necessity right. They have to come up with ways to think about it to avoid catastrophe. But they’re also constrained by still being a little bit behind on the learning curve if you will when it comes to the advanced technologies themselves.
Paul Bracken[00:12:56] Yeah I’ve been following India and I would come to exactly the same conclusion you just articulated. I would just agree with everything you said. They have the Indian establishment has recently discovered the potential threat that a company like Wahway building the 5G networks in India. MOD in India is is is just woken up to this threat. It doesn’t appear that they’ve ever thought about it before they’re considering constraints on Wahway way this gets into the whole U.S. campaign against Wahway but it illustrates the point. The after inventing the I.T. outsourcing industry for a lot of reasons I don’t understand it sort of stagnated over there and they’re woefully behind in AI.
Philip Reiner [00:14:03] The folks who have been involved in their senior most processes to try and wrap their heads around what needs to be done will admit it. They’ll tell you point blank.
Philip Reiner [00:14:13] They will also tell you though that they’re going to at least this is their assertion that they will approach it with the Indian way that they see ways in which the Defense Base will benefit from the technology institutes and all of the rich human capital that they have to bring to bear on it. But it’s more of a wait and see approach at this point for them and that part I don’t necessarily think is going to be advantageous for them in the end from a national security perspective they’re gonna be kind of caught flat footed. Perhaps if I could go back to where we started with this in terms of how the commanders may not understand their own systems. I think what’s really instructive at least from my experience from spending a lot of time with people who work on AI is that even the people working on cutting edge techniques don’t understand what they’re what they’re building or how it works. And so there is a lot of I think discussion within Strat Comm and obviously just within the Pentagon more broadly about how to begin to understand what the range of systems are that we have that that need to be updated today. But then you introduce as you’ve written about fairly I think incisively the cyber element. But then you compound that with the AI piece of this. I’m curious as to how you see the AI element of this as making it that much more complex. Even up and above and beyond the cyber complexity that’s being introduced already.
Paul Bracken [00:16:03] You know these are these are big questions in the people who do this as you point out I don’t understand how they get the results that they get. And that’s characteristic of as I’m sure you know neural nets and deep learning in particular.
Paul Bracken [00:16:23] And that way of doing AI. Where this all goes is something I’ve been trying to think through it’s very difficult and I don’t have any definitive answers by any means but what I see happening is that some application area is going to come up which is if you will a killer app in the sense there’s widespread agreement that this is a good thing to do militarily.
Paul Bracken [00:16:52] So we, and it will probably be the United States or perhaps China and India will will see a great opportunity here and there’s a couple. Well cyber war is clearly to shut down command and control system is getting a lot of attention that would be on the list. The other one I would commend is the hunt for mobile missiles if that because the the way to find a mobile missile is almost surely going to be with a collection of very diverse information streams that when you look at collectively you see certain fingerprints. The amount of data there is simply too vast for the human mind and AI could do that.
Paul Bracken [00:17:43] But either a hidden Markoff model or something like that is it is really. Well that’s gonna get a lot of resources.
Paul Bracken [00:17:59] There an altogether different dimension of this and let me just raise it for your consideration and we can go there with questions if you like. And that is what happened to the Boeing airplane. Which it now looks like was put on a sophisticated AI driven command system that the people operating the plane were unaware of.
Paul Bracken [00:18:28] The command and control implications of this are immense and in pretty negative so that a lot of the ships now are like in the Western Pacific or simply on put on a very sophisticated kind of autopilot it’s pulling information from overhead in every which way. And if recent accidents are any indication of in 2017 the technology is racing ahead of the seamanship of the of the people running the ships. I don’t know if you follow this but the checking autonomous systems software is just about impossible. Because there’s so many branches to it and there is a whole industry that focuses on this and they’re getting better. But as far as I can tell and I’ve asked very senior people in the Pentagon they’ve never even heard of this field called model checking. They think it means checking the model it means something very technical precise as you can look up. I may mentioned one implications of that it’s being able to go on alert safely even though no one is interfering with you is now a the major issue in my mind even compared to the Cold War days.
Philip Reiner [00:20:09] Honestly I had not made the connection with the Boeing software piece and that element of it.
Philip Reiner [00:20:18] And it was the space where I had been thinking quite a bit about it was in terms of the decision support systems and the potential for the manipulation of those systems because of the black box, right. So the ability to actually perhaps poison a system or perturb it in a way that that wouldn’t be perceptible and thus perhaps sending a ship off course or what have you.
Philip Reiner [00:20:49] There was actually an allusion made earlier this week in a conversation we were having with someone about a company that is that is specifically building their business model around the ability to perturb neural nets. In other words, to break the Nets ability to detect just by introducing small elements that really it still is it’s a technical problem that even the world’s leading research scientists have not been able to figure out why it actually happens or how to stop it.
Philip Reiner [00:21:37] They’re talking about the ability to do this in an offensive way and not just with images of perhaps things as simple as stop signs but also with with text and being able to introduce just small perturbations into these systems in order to get them to essentially to fail catastrophically.
Paul Bracken [00:22:04] Yes, so are you referring to GANS.
Philip Reiner [00:22:07] It has. Yeah. It has to do with that and then some. Right. So what’s been curious for us is thinking about and talking to folks at fairly senior levels within stratcomm and elsewhere about what their consideration of the integration of these things and I think the default that most have have come back with is, there’s really no consideration of putting weapons systems into the hands of automated controls or decision making. But I don’t know that that’s necessarily the problem.
Paul Bracken [00:22:49] Obviously that debate has its own merits but it’s really everything that leads up to that that really concerns me even more.
Paul Bracken [00:22:57] Right now, Boeing is the undergraduate version which is that if no one is interfering with you it still may not be safe to raise the alert level. What you’re introducing as the graduate school course in which somebody is intentionally interfering with your processes. And we know that that goes on. We know that’s a major focus of the Chinese program. And so it I mean it really raises large problems it’s not only the problem of you can’t go on alert safely so then you what do you do. You fall back and you really don’t have a lot of options.
Peter Hayes [00:23:45] Paul — I noticed in a recently issued Strategic Command instruction to do with how the early warning sensors on satellites for missile liftoff which are always of course staring down at earth looking for the various thermal and other signatures of such an event which could of course be information signaling to a country including the United States that someone was about to fire ICBM at them or is firing ICBM or submarine launch ballistic missiles at them. That they actually have in the procedures that you can’t test the operational sensors directly because they’re actually on duty. And if you tested them without extra ordinary care and even basically taking offline the downlink of the streams information that could lead to tripping the early warning system itself. So it essentially says that it’s almost impossible to do tests of the fielded systems.
Paul Bracken [00:25:07] I teach a course on technology and global corporations. Totally nondefense. One of the reasons GE is in so much trouble as you know they tried to move into this complicated predictive analysis field for MRI machines that they sell or aircraft jet engines. In one of the things they completely missed was that the customer would not let you test your systems on their operational systems because it disrupted their performance when they had to be an operation. So GE had to actually go build replicated offline GE owned power systems down in South Carolina to test whether their predictive analytics would work. This is something I’m fascinated by. A lot of these problems show up both in the business sector and in the defense sector but there is very little dialogue going back and forth about lessons learned.
Peter Hayes [00:26:14] So Paul and Phil can I just take us back to what I call the genesis of these issues will sort of mid Cold War. When you were first writing about this at Yale, Paul in your book on nuclear command and control. And it basically suggested that there was this there was a global system of systems that together even though they were set up to annihilate each other they actually constituted even then a system that was bigger than either of the separate national systems.
Peter Hayes [00:27:08] Now today we have nine such systems. Included of course the allies as part of that system and we’ve noted already in the discussion that nuclear commanders may not understand their own system from a sort of introspective self-referential worldview but I wonder if we’re not actually now in a system of systems where there each serve roughly one ninth of a global system and they don’t even understand that that’s what’s happened to them that the interdependence and linkage because of the information flows and the links between the different systems via the early warning communications and intelligence connections are already so intimate and so instantaneous that they actually form one system and the emergent properties of that new system at a global level and not even perceived yet alone understood and yet they actually are already in effect in a relationship with their adversaries that they’re going to have to manage in the real world.
Peter Hayes [00:28:40] It’s sort of like Karachi is really already in some respects a suburb of New York and New York is a suburb of Shanghai and Shanghai is a suburb of Tokyo and at the bottom level you know that municipality and local government mayoral level they were already working on that but at the national level they don’t understand that and there’s this mismatch of governance at the different levels in that system.
Peter Hayes [00:29:07] The lower levels understand that they’re already part of a global system and have to respond to it and work across borders with their counterparts. But the national leaders don’t understand that.
Peter Hayes [00:29:20] Can you speak to you know how you see the global system of systems of nuclear command and control at this point.
Paul Bracken [00:29:30] Well I think that is precisely what is happening right now.
Paul Bracken [00:29:34] That we are seeing a fantastic technological version of the Congress of Vienna early 19th century would legitimize the post Napoleonic multipolar system confined to the geography of Europe for the most part with some interesting international issues involving the British Navy and that’s what is being built today. So each nation sees this system of systems from its own perspective in terms of its own enemy.
Paul Bracken [00:30:20] Back then they were aware that they were building a larger macro system of systems which required a different kind of diplomacy namely power balancing than some other rapid coalition formation of anybody got to advanced. This governed Europe until the catastrophe of 1914. I’m not saying this is a desirable system but that’s what we’re building in the 21st century. The absence of anybody’s ability to stop North Korea, Pakistan, and Israel. Smaller countries from getting nuclear weapons is I mean it’s just catastrophic it seems to me. I don’t have any good ideas for stopping North Korea. You know I would tell the White House if I did it. But but the incentives for these countries once they realize the technological and AI driven environment that they’re in is to build up the size of their arsenals to blast through any conception. When I talk to American scholars, it’s well know we wish North Korea didn’t have nuclear weapons but all they need is 15 to 20. Well I’ll tell you where I go what this is they’re going to blast through the minimal deterrence strategies because they’re gonna feel threatened perhaps correctly so I think correctly so because we have not yet really…we’re at the very beginning of in this instance the hunt for mobile missiles. We don’t know where it’s going to go but the U.S. tends to be good at things like this. Lot of things were bad at. We tend to be very good at this. And who knows how you measure good or bad if they have any sense they’re going to presume that we’re good at this. So I think we’re building this very poorly understood system of systems which has to add the political layer that the people at the top don’t understand the technological dimensions of it. But the politics still turns out to be quite important both the domestic politics of nationalism and who happens to be in office and so it’s a very different world than was envisioned in the late 20th century on the future of nuclear weapons one I think is just could get very dangerous.
Philip Reiner [00:33:05] Yeah. If I could kind of pull that thread a little further.
Philip Reiner [00:33:09] One of the the perhaps instructive and well may not only instructive in potentially sobering elements of some of these advanced technologies there may be a counterintuitive here which is thinking about from the cyber perspective how it could potentially play out in the command and control space. There’s some research that’s being done today that questions whether or not having that capability would actually lead states to execute on it. In other words the ability to exploit and be in a position where one could intentionally offensively damage the command and control capabilities of an adversary. That states would actually stay away from that because of the likely potential catastrophic results. In other words we kind of almost see a back in you know previously it was a mad scenario because of the nuclear weapons and the number of them. But now we find ourselves almost in a mad scenario where states may have cyber capabilities but they won’t act upon them because of the horrific damage that it may result in. Curious as to your thoughts on that potentially counterintuitive element.
Paul Bracken [00:34:42] Yeah I’m not sure it’s counterintuitive since it was this has been the argument about why there’s been no nuclear war since the first papers came out on the subject in 1946. But I think we need to think through where it takes us. But let me sort of legitimize the hypothesis first though. You don’t see China or Russia trying to or even North Korea trying to attack the US electric power grid for the purposes of bringing it down. For fairly obvious reasons. And I think there’d be as much reluctance to all out cyber strikes as there would be two large scale nuclear strikes. This drives you into a world where in the 21st century a lot of your conventional forces are going to be nullified. Particularly if they’re in theater or close to other countries who can because those forces can be disrupted with these technologies and so we’re driven into a position of limited war with a hope that the threat of further escalation either to all out cyber or certainly to nuclear use imposes enough restraints so that you can win locally. And this is called the Cold War because everybody said strategy we’d had a dead end and wars would be over with because of nuclear weapons. We know after two Asian ground wars for the United States this was not the case. I don’t mean to be negative here.
Paul Bracken [00:36:28] Overall I can see some real positive developments in the world. Which lead us away from conflict limited or otherwise. And we should get to those. But this re application of the mad argument to the world we are in. Also to me there’s something very dangerous and weird and bizarre about a world that rests on this security structure. I would be the first to admit it’s better than the alternative of nuclear conflict and I have no better ones for it. But I’m also impressed by the ability of the real world to come up with scenarios and crises far more creative then interagency groups or Pentagon planning staffs. And that too is the story of the Cold War. The Cuban missile crisis wasn’t supposed to happen, It was supposed to be too risky so he wouldn’t do it, but nobody told him.
Peter Hayes [00:38:21] In the moment of crisis it seems to me that a lot of the architecture becomes less relevant precisely because much of the information is overload and can’t even reach the decision makers let alone have them process it in a meaningful time frame.
Peter Hayes [00:38:44] And so they revert to standard procedures and ways of thinking. But they also start to activate the weak links in the networks that get them in direct touch with the commanders on the other side at the highest level. In other words the old human game of personal relationships and trust actually comes into play with high level intermediaries back channels and then eventually direct head of state to head of state contact and communication.
Peter Hayes [00:39:24] How do you see that playing out in the modern context in a crisis or in a pre crisis period. I mean we may have seen a bit of that in the India Pakistan recent headlock for example.
Paul Bracken [00:39:43] I would say that among the major powers my belief and more than my my hold but my belief would be they would rediscover this, what you just described. Because Moscow, Beijing ,and Washington the last thing they want is a conflict that could escalate to nuclear use. In fact I find it almost impossible to write the scenario where that happens. To me the troubles are really on the Korean Peninsula in South Asia and in the Middle East down the road, were Iran to acquire nuclear weapons. I think even there the analogy is often used of nuclear brinksmanship as a game of chicken. But if you look at the Cold War as a model it would be more it’s a game of nuclear chicken where each side has a large beacon on the top of the car signaling which way it’s going to go in order to not have a collision and I could see some of that really happening again. I think it already is happening in U.S. China relations. Each side is extremely cautious in what they do. I know even with the North Korean test the other day. I mean this is a really short range small missile. This is not an ICBM test and so of just sort of nudging the problem. There’s a great deal of caution in the system. I don’t mean to exaggerate the danger. I just would say it’s a fantastically interesting development, this idea that we’re building a 21st century technological version of the multipolar structures that developed in Europe in the 19th century. As I said I see positive developments and potentially positive developments which at the moment are not in the forefront of the news cycle, but that doesn’t bother me very much.
Peter Hayes [00:42:04] If coercive enforcement is made difficult if not impossible against those who acquire the capability or shelter under its shadow from an allied state with nuclear arms, what kinds of commander accountability could emerge from this new penta polar and you could say it’s what it what would be nine poles nana pole system, what kinds of new accountability other than just the senior leaders having sufficient political will to talk to each other in a crisis or use intermediaries to overcome a crisis.
Peter Hayes [00:43:09] What what are the kinds of accountability and crisis avoidance and resolution mechanisms could come as a result of the technology and the integration of a global level.
Paul Bracken [00:43:22] Oh I think there’s a lot that could be done. I mean if any leader in South Asia or North Korea or elsewhere were to read the papers at this conference it would have a sobering effect on any illusions they have about nuclear brinksmanship that’s number one. Number two I think there’s a lot of potential for multipolar major power arms control and I can actually argue that we are seeing this develop already among the major powers. So for example the United States, China, and Russia test ICBMS only under the most control constrained ways with plenty of advance notice. There is another example there’s a de facto no testing norm which has developed the US may not have signed the test ban treaty but put it there is no interest in us in the United States testing and there is a de facto no first use of nuclear weapons has taken hold in the world for the major powers. I’m aware of the technical details of what the Russian subset of this over the years.
Paul Bracken [00:44:54] But don’t kid yourself. They have a no first use policy and these are if you pull these things together I can see the foundations of a multipolar architecture for arms control which replaces the one that usefully governed the Cold War. And more thought and work can go into these things. And I have a couple more examples of what this might look like.
Peter Hayes [00:45:23] Go ahead.
Paul Bracken [00:45:25] Well I just would say no big cyber attacks no big cyber tests. We’ve got to alert U.S. leaders, political leaders on the side, to make sure they’re not authorizing any substantial cyber strikes when in a nuclear crisis unless they know what they’re doing. I can tell you that’s not easy, both in this environment and because the military does not want to display the plans that they have. But you need a civilian to go over and demand to see what these cyber plans are let’s just say they’re bigger than a breadbox. Not surprisingly. So I think there’s a lot of things that could usefully be done which would go into an implicit arms control regime which at the moment is not treaty based, is unlikely to be treaty based in the current environment for obvious political reasons. And the fact that the all these countries are sort of mad at each other, this is a good time for scholars and others who are not in the line of fire to start thinking about these things.
Peter Hayes [00:46:53] Well one of the things you’ve written about is that market actors, lots of agile and innovative corporate entities not just American but globally often have better information move faster and have more analytic capacity to interpret that data than state based organizations that are often self disabled by virtue of compartmentalization and bureaucracy and so on.
Peter Hayes [00:47:29] And there are also civil society actors and indeed tens of thousands of highly competent cities that have immense amounts of data close by and at their fingertips.
Peter Hayes [00:47:44] And I wonder if these players who increasingly have global reach might actually in some way create a new information media and a new information regime that both surveillance and monitors what states are doing with nuclear weapons as well as proliferation activity that might need to be picked up just as states are tracking mobile missiles.
Peter Hayes [00:48:13] So cities and civil society players may be particularly well-placed to track what states are doing and then alert other states in ways that have never occurred before except perhaps at an individual level where someone might have reported what they saw when they went to a particular country.
Paul Bracken [00:48:38] Well I think that’s a very interesting avenue to pursue. I’ve talked to a couple of friends of mine who work at various foundations and said here’s what you ought to be funding which is the future of the multinational corporation, as a security actor. I don’t have an answer to what that will be. There was a very impactful sophisticated literature on this in the 1970s with people like Ray Vernon at Harvard Business School, Bob Gilpin at Princeton and it really came up with interesting insights which were very relevant to the 1970s and are largely irrelevant to the world we’re living in today. But that’s only to say we need to really relook at this problem on a more narrow front. The biggest demand on that I get at the moment for Washington and I was down there in March, see I have this course where I look at what does technology leadership and technology strategy, what does it really mean. And my answer is, you can only answer this question if you look at see how big corporations have answered it for themselves. One answer is to go into new markets. Another is to make things more efficient. But the corporate world has pioneered what you’re beginning to see in the Pentagon and I haven’t see anybody write this other right this up other than me. Terribly interesting subject which is they’re changing their innovation ecosystems they’re talking to different people. So if you look at NSA for example in the 2000s you know we have to be careful in choosing our words here, but it was the most innovative part of the United States government. How did it do what it did. And what it did I’m referring to the revelations that Edward Snowden released in 2013.
Paul Bracken [00:51:03] Then you know I’ve been interviewing people and what they did was they concluded after 9/11 that the big defense contractors they relied on Raytheon Lock Martin and others didn’t know squat about this stuff so they were forced to go out and make individual contacts out in the valley, to upgrade their digital information skills, shall we call it. And this introduced a totally different line of information into the intelligence community. If you read Charles Caro’s book this is like the key insight, “organizations will never change. You don’t have to change people to change an organization. You only have to change the information that people see.” It’s a very important point. So this had huge impact. I mean the speed standards of the American government almost revolutionary.
Paul Bracken [00:52:09] The other dimension of the multinational corporation is it’s just a lot easier way to get rich than using military force or threats and I think that the United States and Wahway and China will eventually see this that having business conflicts is just a lot better than having destroyers hug each other in the South China Sea. I mean really the Cold War was ultimately won by economic development. This requires more time here. It was the Japan story when they were growing which others in Asia looked at. It showed that getting rich raising your standard of living you could do it if you were Asian. Japan was doing it. I’m well aware Japan bombed out in the 90s but but for a while it was really had a huge impact certainly on China. That is the hope of the world here. So that’s really so sort of reposition business as a force in society. And the international security order something that that Robert Gilpin didn’t didn’t touch on it. It was Soviet Union just didn’t have any companies. Understandable why and this is not just a source of optimism it’s compared to the alternative of nuclear devastation, just everybody gets this when you say when you hear.
Peter Hayes [00:53:58] Well about a year ago General Mattis who was then secretary of defense transferred responsibility for nuclear command and control communications as a weapons system directly to the combatant commander saying that the attempts to fix it essentially failed over the previous three or four years. And so we needed to do something extra ordinary and General Hyten took over that responsibility. And he seemed to take this very seriously whilst he was at Strat Comm as far as I know yet moved to his new position at the Joint Chiefs.
Peter Hayes [00:54:37] There are three or four major elements underway extract comm to educate a new generation of practitioners in this field. And at the moment at the battle deck at strategic command with a monitoring launch data, potential attack data is there staring at computer screens showing the satellite and radar feeds and of course they have the intelligence flows and right smack in the middle of it is Twitter. We have the president of the United States staring at a TV screen with Fox News and his phone with Twitter on it so that’s the reality of the information flows that they’re sitting in when they move into their professional role. Having gone through this education program at Strat Comm on how does NC3 work, what are the issues.
Peter Hayes [00:55:43] What do you think is the one thing they really need to be aware of as they move into more senior command positions as this tumultuous change occurs in the legacy technology. And they have the new AI capabilities at their fingertips and the cyber threats to manage. What do you think they really need to think about as they take command.
Paul Bracken [00:56:07] I think what they need to do is take a longer term 10 year look at where these things are going and not take the immediate perspective of what we would do in North Korea. It doesn’t matter, but we had a major Scholar of North Korea at Yale yesterday and difficult talk. But I was struck about the information sources he uses are satellite pictures and sort of this week’s news cycle. The test of the short range weapon. And looking at North Korea through the straw of a satellite picture it tells you no idea where it’s going to go I would prefer that they look at things like major power arms control systems. Do these make any sense. I don’t mean to avoid the questions of immediate crises because I think they are also not well handled. And I know where I by speak on Strat Comm, because I’ve been met with some of the aforementioned people that we just don’t play cyber space very well. We don’t play limited war very well.
Paul Bracken [00:57:40] There’s two modes. There’s peace, there’s published peace crisis, then unrestrained conventional war and then when you bring up nuclear weapons the answer is well that’s why we paid the president the big bucks. To which I say we’ll look at the president.
Peter Hayes [00:58:01] So really any opportunity to get those intermediate management level commanders who you know learning on the job, cognizant of the longer term perspective both looking backwards and forwards and then getting them into a room with their counterparts from another nuclear weapons state whether they’re allied friendly or adversarial or just you know somewhere out there is probably got huge payoffs because they the people who are going to do the learning and actually manage the crises into the future.
Paul Bracken [00:58:37] I think that’s right. I would also encourage the track two activities of the kind that you guys are among the best in the world on.
Philip Reiner [00:58:46] And if I could add into this the one of the pieces that you alluded to was with the NSA example of how they had to kind of change their innovation procurement if you will, process and who they turn to for ideas. What that resulted in is something that you talk about in your in your paper which was that the technology itself has you know it’s not just raced ahead of diplomacy it’s actually raced ahead of the of the policy and so for those commanders or those who are going to be commanders to be able to develop a longer term understanding or perspective of the technologies that they’re actually dealing with so that they’re the ones that are who are at the table who actually understand what the problem is with AI.
Paul Bracken [00:59:39] That’s really another key point is that junior people today both in universities and the military and certainly mid-level people are not given an education on the big picture defined to be how a purchase of certain technologies will have larger consequences for strategy. So for example very striking to me I wrote this article years ago called net assessment a practical guide. Net assessment does is put what you’re doing into a context of other sort of big think pictures of other actors and that’s very very useful for somebody in middle management given the culture that we live in here where this is in my view a generally under emphasized feature of university education at the Masters and PH-D level which is I don’t know if you agree with me but has tended toward the hyper specialization in almost all fields not just in security studies. You see that and you see that in your conference.
Paul Bracken [01:01:09] Where some of these papers were like really specialized on whatever country they were doing. And the international context is it was I thought under emphasized in what was a tremendous conference.
Philip Reiner [01:01:27] When it happens down into the very operational elements of art large corporations even today right.
Philip Reiner [01:01:39] So 10 15 to 20 years ago someone who was working in information security for example or developing some of these systems think they would know how to build the entire stack themselves from you know from soup to nuts. And now everyone is specialized into one layer of the stack or even elements of said layer. But they have no context into what that thing gets plugged into kind of as an analogy.
Philip Reiner [01:02:07] I think the interesting example it was for us at least was in speaking with folks as part of the workshop from strat comm in terms of how they were conceptualizing the novel 21st century architecture they’re trying to build in light of its potential interactions with global systems.
Philip Reiner [01:02:33] And you could get the sense that they’re having an incredibly tough time wrapping their heads around the US system looks like and what it could look like much less trying to then think about how the complexity of how it integrates or interacts with other NC3 global systems so just the jumping off point is incredibly difficult.
Philip Reiner [01:02:58] Much less trying to put it in a broader system of systems context and I guess to a certain extent that’s part of why we’re trying to do this right we’re trying to make sure that we we play that role from the outset and provide that sort of a forum to help them think about it in that through that lens.
Peter Hayes [01:03:38] Paul the pentapolar concept which says that basically five nuclear great powers sort of now define a global structure of nuclear threat and of territorial control management of inter-state relations.
Peter Hayes [01:04:03] Roughly I think if you add them all up there are about 35 states involved in the pentapolar system directly reliant on nuclear threat against each other to keep that system from going to war.
Peter Hayes [01:04:20] And those 35 states that have nuclear weapons or allied with an under the umbrella of or subject to the threat of nuclear weapons constitute about fifty five fifty six percent of the world’s population which leaves one hundred and fifty eight states and about 44 percent of the world’s population essentially outside that system.
Peter Hayes [01:04:46] And as you know not long ago they essentially came together and put together a global nuclear weapons prohibition treaty. The question I have is for those who are outsiders of the system but actually do live on the same planet and share the earth in its support systems including what could be destroyed by nuclear war.
Peter Hayes [01:05:16] What do you think is the shared interest and the overlap and how they could play a role in nuclear command and control in ensuring that the system doesn’t go off the tracks or over the precipice.
Peter Hayes [01:05:30] What’s the role for these small non-nuclear states many of whom are so alienated now from the nuclear weapons states that they would go as far as to declare nuclear weapons prohibition treaty?
Paul Bracken [01:05:44] Yeah it’s a very good question and I think there is much more that they can do and it’s in their interests. I with invited a couple times recently to talk to the Canadian folks on these issues in Ottawa. And pointing at me to them was the they have to understand what’s going on on the nuclear front. Simply to be good allies of the United States to point out really stupid things that we’re doing. I mean Canada has a whole arms control tradition going back to the Cold War where they in a friendly way pointed out really stupid things the United States was doing. Maybe a little more directly involved because of the warning systems which are Canadian base and by the way we should have put it’s probably 19 if you count Canada, Peter I think you came up with 18 countries but Canada should be there.
Paul Bracken [01:06:54] It’s up to them to raise the level of debate. In private meetings with the United States and in multilateral meetings in NATO or other international fora. The United States is and isn’t going to do that invariably. So the world is changing. They have to understand the significance of these changes for them in order to make intelligence suggestions and to ask penetrating questions both in private and in public. I don’t think it’s useful for them to go into a sort of hawkish anti-nuclear stance is like some of my friends in Ireland do. Look it’s their country that can do what they want. And that does serve a role but I think there is a repositioning for them which is going to necessitate the understandings of the changes in the world as reflected in your conference.
Paul Bracken [01:08:11] Now I want to say one thing is that I’ve been talking to many think tanks foundations parts of DOD, ect and you guys are really onto something here and my counsel to you would be to not underestimate either the impact or the importance of what you’re onto.
Paul Bracken [01:08:40] I’m going to say this may sound harsh but I mean it is an incredible compliment. One of the things I try to emphasize in my classes at Yale is this anything worth doing is worth doing poorly. It’s clearly like nobody would know how to put together like a global system of command and control system conference to talk about these things. So there’s gaps in the conference you ran. But you’re looking under the right lamppost. OK. You’re looking at the right issues. It’s not clear to me that other think tanks and groups are looking as nearly at the right issues as you folks are. And anything worth doing worth doing poorly is I offer that as very high praise because so many organizations focus on things that are of only secondary or tertiary importance and you are definitely not one of those.
Peter Hayes [01:09:49] Well Paul before the conference you wrote to us that you felt that dangerous issues including nuclear issues just below the surface and that the skin of civilization you said is exceedingly thin and you said it was a good time to think about these matters when we’re not in a mass panic or open psychoses of rage and revenge.
Peter Hayes [01:10:20] We actually read those words to the group as we began the conference and urged them to really internalize that in their approach. And so you know your input was whilst you were at the conference was really important. And of course your review paper is even more important. So we want to thank you for that.
Paul Bracken [01:10:43] OK. Good. Well you know keep it up. You’re doing a good job
Philip Reiner [01:13:35] And I hate to actually end the conversation there’s so many more pieces of this. That reference you made to the necessity for new institutional arrangements within each country but then putting that in a global context I think for example was a piece of what came up in the in the workshop itself in terms of the discussion of how each country approaches this from the command structure like who gets to make decisions along these lines and how are they empowered by the legal apparatus that supports the implementation of direction, etc. And so if there’s the opportunity to continue a discussion on your thinking on what those institutional arrangements might look like.
Philip Reiner [01:14:22] I think there is definitely a desire in our end to continue the discussion I think. With this though thank you so much for all of your time. I really appreciate you doing this with us.
Paul Bracken [01:14:36] I want to thank you Grace and Philip and Peter let’s just say to be continued.
Philip Reiner [01:15:16] Thank you both. Paul thank you. Thank you.
