Evening! Welcome to the Center for Security, Innovation and New Technologies CSINT Conversations. We're very excited because we have Professor Jon Lindsay visiting with us today and we're going to be talking about his new book which I will show you - Information Technology and Military Power. Have a good look - great book. I'd like to introduce Jon Lindsay first. He is an Assistant Professor of Digital Media and Global
Affairs at the Munk School of Global Affairs and Public Policy and also has an appointment in the Department of Political Science at the University of Toronto in Toronto Canada. Again, today we'll be talking about his most recent book - Information Technology and Military Power which was published by Cornell University press in 2020, but Jon has two earlier edited books beginning with China and Cyber Security: Espionage Strategy and Politics in the Digital Domain which came out in 2015 and Cross-Domain Deterrence Strategy in an Era of Complexity which came out in 2019 and both of them were published with Oxford University Press. I should probably also tell you that I am Audrey Kurth Cronin - I left that out - the Director of the Center. Jon Lindsay has a PhD in Political Science from MIT, a master's in Computer Science and a bachelor's in Symbolic Systems from Stanford University. He served in the US Navy assigned to Europe, Latin America and the Middle East - assignments that he may be willing to tell us a little bit about in a few minutes and I should tell you Jon, that that makes us siblings because I come from a US Navy family and my father always said that the Navy is one big family. So I'm glad to know that we're related.
Currently, Jon is working on a new book project called Age of Deception: Technology Intelligence and Control in International Relations which I think sounds really interesting and I can't wait to see the new one. Our plan this evening is for Jon to talk for a short while about the argument in his book and then he and I will converse a little bit about it and then we'll open the floor for questions from the audience ending the program at about seven o'clock. So I would encourage you as you are listening to Professor Lindsay to enter questions into the QA function because I'll be watching those questions go by and I'll be fielding them when the time comes.
So thank you so much for coming Jon, we're delighted to have you as our guest. We feel truly honored and very excited about your book and over to you. Well, great. Thank you so much Audrey, it is really a privilege to be here. Thanks to you and the Center for Security, Innovation and New Technology, which I think has a great acronym CSINT - since I'll maybe be talking a little bit about intelligence collection and analysis in this talk and it's particularly a pleasure because you know as we'll talk about this book focuses on what war fighters do with open technologies that are available in the commercial commercial market of how they adapt and reconfigure it to try and improve operations as they understand it. Sometimes it goes well and sometimes it goes
not so well and that really overlaps nicely with work Audrey that you've done and has just been published in your own book which is just terrific. So thanks very much. Now this book has a title which is both very short and maybe overly broad - Information Technology and Military Power - maybe that's just because I'm not creative enough to come up with something else but also because it was the simplest way that I could think of to encapsulate the tremendous amount of complexity that the book deals with and in many ways the complexity of modern information systems and their interaction with military organizations in the operational environment is the theme that the book is grappling with. So the book tries to impose some simplicity - hopefully not simplistic - concepts on this tremendous complexity to understand how people are are dealing with this in real time. Now I'll try and hit on three themes or three contributions that I hope the book makes. One is a negative contribution and this is a critique of what I'll call the technology theory of victory. Then there's a more positive aim which is to try and provide a few comments on the micro foundations of military power if you will and then thirdly maybe we can talk a little bit about the methodological approach of the book which uses ethnographic methods as a way of building understanding in this space. So let's
go ahead and start with the negative argument. This is a critique of what I call the technology theory of victory. Now this is my own name for a general set of ideas that I've been kicking around out there and been kicking around for decades that hold that better intelligence - more sensors, more precise weapons and more sophisticated digital networks that tie them all together - will make military forces faster, more efficient, and more effective. Now this general set of claims
is most associated with a set of ideas that really emerge in the 1980s, but really are popularized in the 1990s known as the revolution in military affairs and it was sort of associated with the American victory in the first Gulf War, the dot-com bubble, and just sort of this heady millennial excitement that we are in a new information economy and therefore should also be in a new military reality. So for example Admiral William Owens who did a lot of work to kind of push the US military in this direction wrote this infamous book called Lifting the Fog of War in which he argued that technology gives us the ability to see a battlefield as large as Iraq with unprecedented fidelity comprehension and timeliness by night or day in any kind of weather all the time, but I think we can all agree that US experience in Iraq over the last two decades in Afghanistan have been anything but. These have been wars marked by perennial confusion and frustration. They've been long and protracted and have been anything but decisive so these
RMA ideas generated lots and lots of criticism and scholarship most of it well deserved - the kind of bottom line a lot of this work was that it's not the technology, but social factors like doctrine and organization and civil military relations that make the difference and yet this idea has a perennial attraction, right? So here's the Chinese in 2017 - they're talking about artificial intelligence, talking about in the future intelligent battlefield, networks of robots on land, sea, air will be supported by big data and cloud computing that will allow us to acquire data in a useful orderly prompt and accurate manner, which will enable the striking of targets and the evaluation of losses more effectively than ever before. This is echoed as well in kind of US claims. There's a positive version of this that says new technology will make our military stronger. There's a pessimistic version that says the same technology will empower hackers to at very low cost strike our digital systems and create all kinds of havoc and destruction - so there are many different varieties of the technology theory of victory and the historical context changes and the technological vintage changes, but these claims and these great expectations tend to be evergreen even though you have sort of enduring experiences of fog, friction, right? The way that clouds describe the battlefield - friendly fire incidents, targeting errors that kill civilians and cause other tragedies, right ? These are also perennial experiences of war not just in the information age, but particularly in the information age, and it's a little mysterious and that these technologies are often adopted to improve perception and reduce friction and yet those same technologies often get in the way of seeing they often create new sources of uncertainty. So part of the book is wanting to to highlight
this technology theory of victory which is sort of the gift that keeps on giving. To problematize it a little bit and ask why is it so enduring, right? Is it just a matter that practitioners are naïve? They haven't read enough history? Okay, maybe they need to spend more time at the War College. Are they just cynical, right? Are they trying to sell an organizational theory that will give them more resources and manpower? That's possible, but it may also be possible that there really is something fundamental happening with with military power - certain transformations that are going on that we need to pay attention to. Then I think brings us to the second more positive object of the book, which is to think about the micro foundations and military power.
What do I mean by micro foundations? That's just a fancy social science way of talking about what people are actually doing. What motivates people in their interactions with technologies and with one another to generate military power or to undermine it and what conditions leads to one outcome or another. Can we understand that by looking at what practitioners actually do? One goal of the book really is to just kind of call attention to what I call information practice, which is just you know intuitively what people actually do with information technology. If you go into any complex military organization today you'll find that people are looking at and arguing about representations of the world, right? Digital models, symbols, maps, diagrams - right - their perception is mediated by technology. You'll also see that that technology is constantly
breaking down. There are constant glitches, constant operability problems, problems getting data, constant negotiations and outright arguments amongst organizations about access to data and the meaning of data. So there are constant minor breakdowns and frictions, but you also see constant repair and debugging - working through all of that friction. Some of this involves service members leveraging commercially available technologies to come up with creative solutions and band-aids and sometimes even more sophisticated creative solutions for dealing with some of these problems. So you know, one way to to think about this is may be to repackage
the old saw and say that amateurs will talk cyber strategy, but professionals are really debugging information systems and no matter how sophisticated the engineering, no matter how able the engineers and careful the designs, this kind of ongoing breakdown friction and debugging and adaptation is just fundamental to the nature of large-scale complex information systems. Now I'm not even talking at this point about enemy cyber attacks if you will - that certainly could be a major source of friction - I just want to point out that a considerable amount of organizational friction in the information age is self-created and it is critical that we understand what practitioners are dealing with on a daily basis. If you want to go to the next level and understand how cyber operations work and what kind of frictions attackers themselves may be experiencing. Now this is a bit of a perpetual motion machine right? Well-meaning attempts to reduce friction often end up creating more friction. Amateur repairs end up being amateurish - they cause security problems and they're not scalable and they're not interoperable, but the managerial reforms that are supposed to prevent those amateurs mistakes end up being really bureaucratized and slowed down by red tape.
They don't adapt to changes by the enemy or the technological state of the art and there's no way out right? This is one of the big messages here, okay. There is no way out of this problem because continuous breakdown and repair and these cycles of exploitation and reform are just fundamental to the nature of complex digital systems. Now the book makes two arguments about information practice, you know, because it wants to say hey this is constantly going on in organizations, but we would also like to know under what conditions does it actually improve military performance on the battlefield and under what conditions does it make mistakes and fracture size and targeting errors more likely. So the first argument I make is that if the performance of or - excuse me - the effectiveness of information practice and its contribution to military effectiveness is not a function purely of technology, but more importantly it's a function of an interaction between organizations and their environments. You can think about environments as being more or less structured right? So environments can have lots of constraints they can have a set number of things that you are worried about. Organizations
likewise can be very institutionalized and formalized and when you have an organization that fits hand in glove with its environment then you can have very efficient information practices and systems that are really well aligned with the problem you're trying to solve. At the opposite extreme, maybe you have a very dynamic fluid changing problem. In that case it would be best to have an organization that has a more organic bottom-up adaptive way of interacting. That's not going to be the best outcome, but it may be the best you can do under those circumstances - freeing individuals to explore the problem and come up with creative solutions. But then there's these mismatches that are going to be problematic, right? One is going to be when an organization is so set in its ways that it's unable to perceive and adjust to these more dynamic ambiguous changing environments/. And the other case would be when you have people making all kinds of changes that end up interfering and conflicting with other parts of a more stable structured environmental problem.
So the second argument then is that because you have these four different outcomes organizations often get caught in a tail chase, right? Whereas they come up with a way to solve one particular problem and they've let out contracts and they built particular systems and then lo and behold the enemy changes right or allies change or politicians decide they want to do another mission and suddenly you have an organization that is adapted for one kind of solution, but the problem is different. In that situation people start defecting from the bureaucratic scheme and they come up with really creative solutions. Unfortunately those creative solutions cause problems for the organization. Managers move in. They start trying to reform the organization. It
becomes more complex. That's good for a little bit, but strategic competition tends to then change the problem yet again. So you go around and around and around where organizations and problems tend to get more and more complex without actually having long-term performance advantages. So that's the - in a nutshell - the general argument of the book and it tries to go into detail with a couple of historical and ethnographic cases which brings me to the third point. I'll make this really really briefly, so we can get onto the more exciting conversation. This book really draws
heavily on ethnographic methods or participant observation methods. Now these are really common in comparative politics or other parts of the sociology of technology, but there may be less widely used in national security studies and the reasons may be obvious - right. We're dealing with classified information in dangerous environments that are hard to access with conflicts that are inherently rare events. Okay, so simply getting into the field to observe
these things could be difficult. But if you buy my argument that information practice - what people actually do with technology in time of war - is important, then you need to have some ability to actually go and see you know what's actually going on. So I was able to leverage my own experience in the United States Navy active duty before defecting to academia and also as a reservist, while I was thinking about this book and the issues that were associated with. And I mobilized to active duty and I got permission from both the Navy and from - I was working on a dissertation at the time at MIT - and my advisors thought this was a little strange, but both sides agreed - okay as long as this doesn't interfere with your actual job you can go ahead and think about this. And what was interesting is that because I was interested
in how information systems worked in complex organizations and because part of my findings slash argument where that practitioners are always having to pay attention to these workings - that the same kinds of questions that i would ask as an academic researcher interested in information practice were very similar kinds of issues that i would deal with as a practitioner right. Actually dealing with the constant breakdown and repair of systems in this particular case systems that were supporting special operations in western Iraq in 2007 and 2008. So that'll kind of just give you a flavor of where this book is coming from, where the ideas came from, and how I tried to really flesh them out and help you to see this I think fundamental problem that exists in all organizations. And I also tried to to pair it with a couple of historical cases and other cases that we could see similar issues arising with very very very different technologies and different interactions between organizations and the problems that they're trying to solve, but still generating familiar patterns of information practice. So I think I'll go ahead and pause there. We can take this in whatever direction you and the audience would like to. Well great - I would like to exercise my prerogative as chairs always
say and engage you in a little conversation first and then I see a bunch of great questions coming in. So I'll be turning to the questions in just a few minutes audience. So one thing I wanted to ask you Jon, because my own training has been in a kind of a broader context, looking not just at the military level, but also at success as a broader societal question - political economic social impact on the outcomes of wars. So is there a tendency - your book says that it's about military, you know military power, so you're specifically looking at that level of analysis - this is not a criticism, I'm just trying to understand - how do you go from that level of analysis which is very oriented toward the organization and what it is doing - and it doesn't really talk a lot about what the enemy is doing either because most of the cases are oriented toward situations where the united states was very dominant - how do you go from that to the question of success in a war? Because a success in a battle is not the same as success in an overall war. And just - you know, managing the practice effectively, adapting effectively, that may
work very well for the military organization, but what does that really have to do with the outcome? Yeah, absolutely. So - you know this this goes back to this question of kind of wrestling with complexity and how really to scope this. You know, I tried to pitch the argument generally enough that information practice and information practice theory could be applicable at a lot of different levels of analysis and it can be applicable outside the military environment as well. And military operations at the operational level have a feature that is very very useful for thinking about this particular problem and that is that it is extreme conflict between extreme cooperators, okay? So you have an organization that ostensibly is working together so that it can defeat or disrupt or degrade - all those wonderful D words - the adversary, okay. So extreme cooperation and extreme competition and these polls - right - pull on all information processes in kind of two opposite directions and it creates these two very very different problems the practitioners are trying to solve at the same time. Trying to say hey I need to understand what's going
on in the environment and what's relevant and what are our effects going to be and what the unintended consequences are going to be, but then I also need to understand how my organization is cooperating and solving collective action problems that are needed to solve that first order problem. So in the military environment those are such extreme contrasts that you really see the fundamental tensions of all information practice which I think manifests in lots of different places really cast in stark relief so that's one answer. Now there is this tendency as you mentioned, kind of in the strategic studies and certainly the way that kind of War Colleges like to teach their students - that you can nicely separate the strategic from the operational from the tactical level and most historians would hotly contest this, right? They would say hey when you're doing operations, right, you're making lots of strategic assumptions. You're designing operations that can work for particular strategies and not others, okay,
and so this is really really a false dichotomy. It works if there is an alignment between strategy and a particular concept of war then this engineering optimization problem can proceed and you can have operational optimization be really really good for political outcomes, but if you have a mismatch that can be hugely problematic. So in many ways this is what the Iraq chapter was about and it focuses on one particular special operations organization that was run by a Navy Seal team and Navy Seals you know - you've watched the movie, you've read the books, right? You know these - this is a very very action-oriented organization with kind of a very proud commando history and it really wants to see the world in terms of targets that can be actioned.
And this was in Anbar Province in 2007, so after the awakening, before the surge, in an environment that is transitioning from intensive combat to stabilization and reconstruction, okay. So counter insurgency and stabilization is arguably the strategic point of being there at that point and yet you have an organization that was very very focused on this particular concept of operations. And what was interesting was that these cycles of friction and breakdown, rather than improving the ability of that unit to do counter insurgency and stabilization, actually tended to further lock in this operational fixation on the fine fix and finish targeting mission. So this was an example where the nature of the problem at the strategic and the political level had changed, right? It had become something else - because you know external problems in competitive environments tend to do that and yet you had an organization that was very very fixed on optimized - optimizing the world as it understood it and so yeah there really really was this disjunction.
And you know a very different case that I look at in the book is the historical case of the Battle of Britain, right? Where Britain was playing defense. The whole goal was to like stop the strategic bombing campaign or actually the goal was to keep the royal air force alive as a fighting force long enough for a German seaborne invasion to not be viable and because that's the strategic goal the operational goal of identifying and intercepting incoming raids and preserving and husbanding your fighter forces, right were really really aligned with the strategic objectives. So in this case that process of information practice tended to be really nicely lined up with the strategic goals. So the key is to be in fear of your survival and then the strategy tends to match the operational and tactical practices. It's hard when you're the dominant party in a way - I mean according to the examples that you give in your book. I would actually like for you to talk a little bit about
FalconView because your discussion about bottom up innovation is very rich - so would you mind telling our audience about that because I'd love to ask you a couple of questions about that too. Yeah absolutely. So this is a case that looks at the 1980s and 1990s practices of automation in primarily, but not exclusively as we'll see in a second. The aviation community - and the aviation community is interesting because it's full of pilots and these pilots tend to have engineering backgrounds - and through you know, most of the history of aviation pilots were responsible for doing their own flight planning so they would get paper maps and they would get out grease pencils and they would kind of draw their routes and they would file their flight plans and they would do their calculations and how much fuel this is going to take and where the targets were going to be and they'd overlap threats and all this kind of stuff like the pilots just did this. So when, you know, electronic calculators and PCs started to emerge - like a
lot of these guys were like hey cool I can do this, you know, a better way right? So they start writing their own routines to do flight planning and this is in the 80s, right? At the same time we're kind of like Microsoft and you know Apple are sort of forming out of garages in Silicon Valley and - no kidding- these guys, you know, were basically getting together with their friends and even had their wives getting kind of the notice to aviators to put in all the information about weather and flight events and things that were going on into this database and they would put the database on floppy disks and as they flew around from airbase to airbase they would distribute these floppy disks around. And so you kind of had this vibrant bottom up kind of like garage startup culture happening in the middle of a military bureaucracy. And we're used to that kind of mythical story in the Silicon Valley context right? Which is often told as this freewheeling kind of alternative to a big government and top down control of which militaries are sort of the apotheosis of and yet even there you had kind of this this surge of kind of bottom-up activity and so FalconView sort of inherited this tradition and because you had a bunch of aviators that wanted to do this - and in fact there were aviators that were in the Air National Guard and the Reserves and so they weren't getting the same funding as the Active Duty Air Force - they got together with a couple of graduate students at Georgia Tech and they put together this basic Windows graphical application. Think of it like Google Maps today - that would run on a PC and it had to run really fast and really efficiently because PCs were low powered in that day and you would never do this on Windows, right? You could only do it on a Unix box - so they created this software that ran fast - it was super popular and it started to outperform the very expensive Unix-based applications that the program offices were funding and then - so we kind of entered this interesting decade of like back and forth where the program offices said hey pilots in the Air Force - and now increasingly in the Navy and the Marine Corps and trickle off into special operations and even the ground services - they said hey we know you like this, but we can't control it, we can't manage it, we can't deal with the cyber security - so we're going to build something better. And they would try and build something better, but meanwhile the users would keep hacking and reconfiguring and extending this. So this was this very
very interesting moment where like the kind of leading graphical planning application that is used kind of, you know, to this day in many places for planning, but certainly, you know, for intelligence analysis, for planning, for special operations planning - is this product of a community of kind of smart hackers that weren't put in that position. They just found a need.... --- TECHNICAL DIFFICULTIES --- Jon your screen has frozen. --- TECHNICAL DIFFICULTIES --- Jess, I think we're having technical difficulties. We are having technical difficulties at the moment and I'm going to try to contact our speaker offline to see if we can get him to at least come in on audio right now. He has unplugged and we're going to get him back on. So, Audrey if you could make some
general chatter - maybe about the topics he's been talking about? Sure, sure - just let me know how things are going. So ladies and gentlemen I'm sorry we have a little bit of a break - you know how these things work. This is almost like a live demonstration of why computers and you know the theory of victory that computational networks will determine success or failure is probably not something we should be depending upon entirely because here we are. I see some wonderful questions in the Q and A, so I thank you for that. I am not um ignoring you by any means, I just would like to have our speaker come back because I'd rather he answer them than - than mine. I see a good
one that's starting out - has the concept of victory changed, and we will talk about that questioner, but let's see if we can find our speaker once again and hope that he's dialing back in. The concept of victory is a complicated question because of course the question of a military outcome is not the same as the question of a victory in war and technological determinism is centuries old and we have to be careful to - not to directly draw a line between effectiveness and winning in a war. Effectiveness can win in a battle, but uh having a good strategy and having it aligned just as Professor Lindsay was saying, having it aligned with what your organization is doing is absolutely crucial. So, I'm sure he'll be right back and talk about this. I have to say that this particular case study on FalconView is my favorite part of his book, so I hope you were enjoying that. I'm hoping he's going to come back and finish this story because Professor Lindsay talks about how FalconView was very insecure and they had to in the end - he will tell you this - but, in the end they had to put in place a different system because they felt that it was too vulnerable to hacking and the other system was not nearly as effective and so what I want to ask him when he comes back and what I'm hoping we'll begin with is - you know, what is the policy solution if you have this wonderful bottom-up innovation? Ah here he comes. Good - it'll just be a few more minutes I hope,
but you have this wonderful bottom-up innovation and you have to have security. Ah, there you are. Can you hear me? I can and I was just doing a little tap dancing talking about how much I love your FalconView case study. So I'm sorry that somehow the system went down, but it sort of is a real-time example of why we can't just depend upon computational systems for victory, right? That's true. The book develops this idea of information friction and so this is the
performative dance example of that. And look how well you've adapted - so adaptive innovation there Jon. So may I ask you if you don't mind if you can manage - I think you must be using your phone. Yeah. Okay. Can you start we lost you when you were getting to the meaty part about
the FalconView system and how it was showing itself to be too vulnerable, so if you could pick up that story there I think we'd we'd love to hear the rest of it. Okay - sure. So FalconView was very popular throughout the Air Force and at this time it had developed acolytes and fans throughout the Navy, but people in the program offices in particular were very very concerned that this effort was ultimately not going to be scalable or supportable and there would be some applications that you really wouldn't want users tinkering around with, right? So safety and flight weapons release, you know, functions all of these sorts of things that you know a large-scale planning system would also interface with. So the idea was okay, we need to shut this down, but we know that it's popular with users and one thing was very interesting is that - so this kind of story there are little versions of this that happen all over the place - usually the bureaucracy wins, right? It's easy to say hey Lieutenant that is a neat little hobby shop project you have, but it doesn't meet these different, you know, requirements or breaks these regulations - we're going to pull it off the network. But several FalconView
proponents actually got themselves working in the program offices and this was interesting in that they were working both sides. They were both kind of keeping this project alive as a stop gap until the larger systems could provide this functionality and trying to create the same systems that were going to replace it. So while it was really easy to kind of have this morality tale of you know wonderful adaptive live user innovation efforts versus the slow ponderous systems bureaucracy - in reality, you had this more complicated symbiosis where there was top cover for user innovation efforts and an attempt to try and build that on a more stable platform that would leave it open to users to come up with new applications, extensions, and customizations without then kind of interfering with its large-scale stability and sustainability. What was interesting is that there were at least two cycles that this went through where there was a large system that was designed to replace FalconView and by the time it came out FalconView had already moved beyond. That system struggled to meet the requirements that had been specified two or three years earlier and already requirements had changed as warfighters went out explored new possibilities and new functions and dealt with new requirements. Because let's face it the US
was kind of a hyperactive superpower in the 1990s and that's tended to put people into contact with you know slightly new data processing problems that they needed to interact with. Wonderful, okay. Well I'm grateful to you for finishing that story. I'm sure that our audience will have questions about it and we have about 50 minutes and a lot of great questions already. So if it's all right with you Jon, I'm going to turn to their questions.
Okay. All righ, let's start with Christopher Basson and he asks has the concept of victory changed as a result of the use of information technology by military organizations? So you know, the concept of victory is really really fraught in, you know, military effectiveness studies and I think this really came up in the conversation you were having about - do you mean victory at the political level where you actually accomplish your objectives - accomplish your objectives within, you know, an acceptable level of cost - whether the human cost or financial costs - or other political opportunities. Or are we going to speak more narrowly about keeping the other military organization from doing what it wants or destroying that other organization or occupying territory - I think even that very concept kind of at the strategic political level really kind of varies depending on the conflict that you're looking at, right? This kind of takes us back to kind of Clausewitz 101 right? Understand the kind of conflict that you're involved in because whether it's a war - a total war of annihilation - or whether it's just an armed reconnaissance you're going to organize strategy differently to deal with it. So I guess I'm going to dodge that question a little bit and I wanted to focus on information practice because I think no matter how you define victory you're going to have to figure out what kinds of things matter in the world. Figure out where your own forces are. Figure out how to get your forces into contact with those or avoid contact - if you're playing defense or trying to be stealthy, right? So they're going to be kind of these operational problems that you're going to need to solve. Those don't add up to victory, but they may add
up to some degree of operational effectiveness. Now maybe your larger question is - a person didn't ask this, but I'll just maybe kind of guess a little bit - if you're talking about different kinds of conflicts that we're now thinking about in terms of cyber security, cyber operations, cyber warfare, right? That looks like a totally different set of interactions than kind of the traditional clash of arms on the battlefield and so if there are no militaries involved at all then what meaning could military victory even have? Certainly the cyber influence interaction between, you know, Russia wants to sway an election and cyber security professionals want to prevent that, right? That is a very very different conversation than might be happening otherwise and that's actually one reason why I decided to pull the entire cyber conflict discussion out of this book altogether and have you know a larger conversation about it. I think no matter what kind of political interaction you're looking at information practice itself is going to either help or or hinder it hopefully in the way that I describe. That's what I mean. Yeah that makes sense - that's a great point. So Jon, did you have a big part of this book that you pulled out of it?
I did. Right, so you know, it may seem strange that the book is called information technology and military power and yet there is very little discussion - in fact no discussion - of cyber conflict, which is probably the greatest manifestation of information technology and military power in the 21st century, right? We have this entire new domain of conflict. We've got new military organizations, a four-star unified command in the United States that can manage cyber warfare, every joint task force now has a cyber component commander - so how the heck could I possibly leave that out, right ? And so you know, I had originally written up something about this but the key point that I think makes cyber conflict different and in many ways more complex, although not necessarily more dangerous - and that's an important point, is that in the classic military case you have two feuding organizations and each of them have a command and control system. They have intelligence sources, they have to integrate it, they have to communicate with all their people - and it may be very very complex and distributed. At the end of the day each side largely controls its own stuff and that's different from the cyber case where you have common tools, common networks, common protocols, right? Both competitors are interacting in the same organization and that's very interesting because now you have a strategic conflict between - in a very real sense - members of the same organization. So while I think that we can think about information practice as an input and a constraint on cyber conflict, that makes it more complex because cyber conflict, right, is both targeting and protecting, exploiting, undermining and reinforcing information practice itself. So I thought it was important first to up and say -
well what's going on inside of organizations to begin with right, in the best of situation when people are ostensibly working together - what kinds of frictions and breakdowns are they encountering and then later on we can add in - oh now what happens when you're actually sharing an organization with your adversary. When your adversary is a source of the frictions that you are dealing with in the information systems itself. Fascinating and I think my Co-Director Josh, as you know well Jon, would probably say that it's all about espionage right? But I'll give Josh his own ability to talk. I think he has a question up here in a minute okay. By the way let me just check technical - I noticed my internet has has recovered. I unplugged my router and put it back in - if the sound and everything is okay we'll just continue this otherwise I could go back to the other one. I think it's okay and okay so if you don't mind. I mean the the visual isn't quite as
good as it was before, but it's good. So we're fine with it if you're fine with it. Okay. All right - so next question from Scott Crawl - wait a second, now I'm having problems. Okay - how much autonomy do you think the DOD will give to AI based management systems in the everyday analysis and decision making at the Pentagon and how will that impact human employment at the Pentagon? You know, so here the big looming question of AI yeah. Okay ,that's a great question. So,
maybe to back up and contextualize I think the AI conversation that we're having today is definitely a new manifestation of this tendency that I was talking about as the technology theory of victory - the idea that will be able to have information technologies that substitute for what militaries are doing and will be able to do it so quickly and efficiently that you have a radical step function change in military performance. So there is a real rhyming familiarity in the way that we talk about AI and I would say I agree. Okay, now the robust finding in the kind of economics and organizational sociology of information technology is that these technologies are never pure substitutes. While they can substitute for some functions they also end up generating all kinds of new tasks and positions or complements that are needed to manage those informational interactions. So the more you add in information technology to try and substitute for what you already understand, what you're doing in order to improve your performance, the more you start creating this overhead of kind of complementary, kind of human interactions. And I think that we are going to see - and we are seeing exactly that happening with AI, right? AI is not a pure substitute for human decision making.
It can do some things well, right? It can provide predictions when you have lots and lots of really good data about fairly well understood problems. And that's not the operational warfight problem. We have terrible partial fragmentary data about, you know, ambiguous evolving problems so AI in that situation is going to create a huge demand for human compliments, for human interactions, for people that can interact with and debug these these systems. So I think that, you know, in all our talk about where do we get AI autonomy and should we be concerned about killer robots and where do we put the man in the loop? It's kind of losing sight of this very very pragmatic fact that organizational complexity is going to increase and that complexity cannot just be you know automated away with AI - because AI is going to be part of what's driving this more intensive interaction. But it will tend to pull people more and more into this information processing debugging and sense-making role that makes judgment really really complicated and you know and the questions that you're asking before Audrey about like how do we, you know, just entangle strategy and politics - like AI is going to amplify exactly that, right? I mean like that's the key compliment is why are you doing what you're doing. And that's going to become more ambiguous more distributed
throughout the force in the society, right. So there'll be kind of this change maybe in the way that you're thinking about conflict precisely because these are not substituting technologies. That's very thought-provoking. Wonderful. Okay, let's see. Katherine Urban - we've heard a
lot lately about the prospect for public-private partnerships to develop cutting-edge military technology - she wants to know about the different organizational cultures between the Pentagon and Silicon Valley and how can that work. Yeah, this is - this is definitely real, right? So this is kind of what I was getting at with the FalconView story. There, you know, on the one hand the stereotype is absolutely true, right? Silicon Valley wants to move fast and break things and it's got, you know, all of these people that are starting things up and attracting venture capital and you know, making things, you know, move at internet speed etc. etc. And that's a little bit of a stereotype because some things in Silicon Valley actually don't move very quickly and some things do get locked in and are a little bit resistance to change, but okay that's there. And then there's this story of military bureaucracies as kind of these ponderous beasts that have to write up these requirements and they have to then you know deconflict them and they have to go through all this committee process everybody likes to make fun of this defense acquisition chart that maybe some of you have seen which is just this incredibly painful, you know, research development test and evaluation process you have to go through to invent weapons. Okay so like how the heck could you possibly put those together? And what I was trying to argue with the FalconView piece is that actually it's more complicated this you see this ferment of innovation already happening within military organizations, but not necessarily formally blessed and sanctioned by military organizations um and you know you don't have to spend much time in a military organization or start reading for this before you start to see example after example of you know somebody that you know came up with a way of using a drone in their iphone in some creative way or a unit that solves some problem that you know their command and control systems administrators weren't able to solve, you know.
On and on and on - so this is a very very kind of typical problem so maybe the the distance between these two cultures is not as great as we actually think it is, except that these public-private partnerships I think too rarely are actually really focusing on that firmament of innovation itself. There's still this idea that we're going to tap Silicon Valley for its innovative potential to come up with the new technology in this widget and then we're going to throw it over to the transom and then the military will implement and operationalize it and that's a little bit backwards, right? If we thought of this as less of a technology problem more of an organizational problem then the question would be well how do you make sure that technical expertise is forward, right? That the technical expertise that is already forward is supported, right? That they have the ability to adapt and innovate and those adaptations and innovations are being kind of folded back into existing, you know, command and control systems - that'll be a very very different approach to I think public-private partnership. That would focus on the people less than the technology. Because you know, again like you don't ever get the command and control system working the way that you would like it to it is perpetually in a state of breakdown, right? And so attending to that repair I think is going to be an inherent part of of increasing military capability in this more complex environment. Great. Okay, actually on that point, I'm always arguing with my friends that the military promotion system that keeps people in vertically is so 19th century. It's going to be very difficult to be that agile until
we fix that. All right, let's go to Josh Rovner. He says he's interested in the implications of the argument for civil military relations should civilians take a hands-off approach given that the modern battlefield is information dense and full of fast moving unstructured problems? Yeah - wells okay. So my first answer is is no, definitely not. In fact I would think you know, exactly the opposite, right? The battlefield is sometimes full of, you know, really fast-moving problems. That doesn't necessarily mean they're fast-moving problems - that are problems that strategically you want to really solve. We talked about, you know, militaries that solve these, you know, quickly moving problems but end up solving them in a way that's strategically inimical to what you want to accomplish politically. You could certainly tell yourself stories in a China scenario right, where you freed the military to - innovate in a very very aggressive way that ended up starting to build in some escalatory risks with Chinese and especially with kind of Chinese conventional and military systems that may or may not share command and control systems etc. etc. Those are exactly the situations where you kind of
would want civilians deeply embedded with those interactions to understand what kind of the technical consequences might be. I think the larger question here is that, you know, this larger historical trend that we're starting to discuss - which in the book I call the informational turn in military practice - which is this shift from military practitioners being, you know, on the battlefield - physically fighting in blood, on deck plates washed by the ocean, right? To living in these remote command and control spaces where they're still frustrated dealing with lots and lots of things, but they're physically removed from the pointy end of the sphere - which is increasingly automated, right? And so - these organizations get more and more complex which means that the content of military labor actually is becoming more and more civilianized, right? What do military practitioners do? Well, you know, they work in offices and they go to video teleconferences and they work on reports and build powerpoint briefs and you know - write word documents, right? I mean like it is very very similar to what's happening in the civilian world, but it's connected to these fast moving parts. So you have kind of military practitioners that are more and more involved in civilian type tasks doing things that are still leveraged because of the automation linked to the to the pointy end, which means that they're going to be more and more politically consequential. So I think this speaks to needing to have kind of a robust two-way conversation - if anything, you know - I think one implication would be, you know, a strong endorsement of the Eliot Cohen view of civil military relations where you should have a robust conversation between politicians that are really really sensitive to objectives and competing political tradeoffs and military professionals that understand the tactics. Because those tactics have political consequences and they're going to be different political objectives that they're trying to meet. So. That makes sense. Okay, I'm going to give you two
questions and I'll paraphrase them, but I want to make sure we get - there are a lot of questions here and I want to make sure we get as many in as we want. You can answer whatever aspects you want Jon. The first one - has the United States learned from Iraq and Afghanistan that it requires a shift in information systems or do many of the bad habits of the RMA framework still linger? So that's question A - hold that for a moment. And then from one of our PhD students - ethnographic studies, as you mentioned, aren't all that common with respect to national security and technology - do you think your work opens up more avenues to such research and how do you go about writing up the final draft - balancing the possible classification issues with telling an accurate empirical story? You can tell that he has some connection to the military. Yeah, great. Thank you. Okay.
The first one which was, you know, does military culture - and I think the question was specifically about army culture - does it just sort of persist throughout these conflicts or are there changes? Am I getting that right? Okay. One thing that I think is well - there are many things that have managed to, but one aspect of these wars and the conversation about technology that's been very very interesting is that there was sort of this morality tale that emerged that the RMA was all about network-centric operations and then we found ourselves in this population-centric fight and the attempt to apply network-centric warfare to a population-centric counter insurgency was a terrible idea, right? And you were looking at networks, but the real networks were the social networks and by not understanding that we ended up getting ourselves into a terrible mess. And it took, you know, many many years of reform until we finally, you know, figured out counter insurgency theory and surge...yadda yadda yadda. Okay, that's true to an extent, but what's interesting is that in the
process of learning how to do counter insurgency the US military had a voracious aptitude - appetite - for new technologies and there was massive kind of sustaining low-level innovation where they figured out all kinds of different ways to - yes do the targeting mission - but also to do population surveys and also to do social network analysis and also to figure out, you know, the relationship between, you know, economic performance and small grants and military, you know, violence levels. Okay, so you had the kind of technology being applied across the board so that morality tale really was too simplistic, right? It wasn't that technology is only good in one place. No, technology can be used in a lot of different places even when you know a military is now having to look in a very very different direction. That said I do think that kind of, you know, deep
in their DNA, you know, military organizations do have kind of styles or types of wars that really accord with their, you know, core identity and they're probably embedded deep in their history right? Scholars like Austin Long make this argument - and I certainly found this with kind of a Navy Seal team that I was focusing on, right? That there was just such a strong kind of historically based identity that really shaped the way that they were going to be innovating - and therefore perceiving and actually influencing the battlefield. So I tried to tease this out a little bit by, you know, in Anbar comparing the special operations task force that I was with with Marine Corps units in the same battle space that were also innovating with technology, but doing it in a very very different way. They had a much more kind of population policing mentality that they brought to it rather than you know, the Navy Seal counterterrorism mentality. And I also tried to compare it - even though this is very very hard to do with open sources - compare it with the JSOC mission which is a much more highly resourced version of counterterrorism.
You could argue that both of those alternatives had a fair degree of success in their local scoped operational mission, whereas the organization that I was looking at kind of did both of them poorly. Now to the second methods question, right. That is a fantastic one - you know, ethnographic methods are very very different than the, you know, kind of KKV causal inference things that we learn about in our normal political science education, right? Where you know, you come up with a deductive theory and then you go and get some data and you pick it on the independent - dependent variables. All these sorts of things. Whereas every single ethnography that you read has this weird idiosyncratic story about how the researcher happened upon, got involved with, developed the trust of this community that they were working with and you know they developed this very very kind of seemingly narrow story about - often kind of a historically insignificant case and yet done with a mind of trying to say something broader about culture, about social practices.
So I think there's kind of two different questions here - one is how is that even possible? Like, how can ethnography - how can you trust anything an ethnographer says given that their access to this group is so circumstantial and so highly personalized, right? In my case, right - you know, I mean I happened to be a reservist that had access to this particular unit copacetic with me, you know, doing this and this that the other thing. I had been steeped in, you know, while I was in active duty -I had been um tinkering with and developing some FalcoView applications myself, so that meant that I was able to develop some trust with developers and program managers, you know that went on. So like a very very personalized particular story. Like, why should you trust anything that I have to say? And I think the answer is that ethnography tends to work in this more sort of Bayesian conversation if you will, with the literature - where you're like, this is what lots of people have said about organizations. This is the experience that I have. Does it make sense? If not, what do we need to update? Do we go back? And you go back and forth between this very very richly described casework and this very very general theory to try and get theory that stabilizes to the point that it can then be applicable for understanding other similarly originally described cases. And so you know that's kind of where you get to the point you're like okay I'm now through this very particular experience that I have recognizing kind of enduring patterns that seem to emerge in lots of other different places. So I think that ethnography is less of a theory testing than
a theory generation mechanism that gives you some confidence that what you're seeing may make sense. And so what I tried to do was use this as a way of kind of laying out this more general theory which then I think would have kind of implications that could - that you could go and test. Now as far as like doing this in the military again - like it worked in this case, but it was so difficult and so frustrating that you know I finished my dissertation, put it aside and worked on cyber conflict and cyber warfare for the next six or seven years because I wasn't sure kind of how to bring this back into the realm of security studies and you know, make it fit better with other theories. So that's kind of a cautionary tale - I'm like if you go down this
road right, it is weird and strange and different and it may not actually fit really well with kind of the the discourse and conversation that people you know are used to having about testing and evaluating theories about kind of security. As far as the classification realms that's a great question. Okay, so - you know part of the deal with this was - hey if you're gonna write about Naval special warfare in a combat environment then yeah your manuscript is going to have to be reviewed and it was and they found you know nothing wrong with it - although they didn't like my conclusions, you know - there was nothing classified about it. But part of that is because I specifically was interested in things that were tacit, that were cultural, that were not being explicitly written down, right? Things that are explicitly written down are the things that are explicitly classified - right? Those are the categories of information that get embargoed for 10 or 25 years, but I was more interested in kind of how people were making sense of the world, right? Exactly the things that didn't get written down. So it was kind of that -
kind of, you know, more general, but slightly abstracted work that I was looking at and I kind of found this sweet spot where I could be in a classified environment and yet describe kind of more general tacit organizational interactions that were general enough that they didn't kind of like run afoul of the classifications. How interesting. Jon, one thing that is happening is that your hand is wiggling and it's making me a little dizzy when you talk. So, set it on something - otherwise we can see you. Yeah that's better. Can I just say - calling this targeting approach counterterrorism has always driven me crazy. Not you, I mean - the way that
is the case in the military, because what the Marines were doing - that kind of practical house-to-house understanding of collecting information and data - that's closer to traditional counterterrorism that integrates a lot of different types of information and capabilities rather than just targeting. Oh there you are. Oh good you're back. Okay, good. Can you hear us? Yes, yes. Perfect. Yeah, that is a little bit better. All right, well thanks Jon. Let's keep going. We have another one of our wonderful PhD students - his name is Ian as well Ian Campbell and let me see - get to the meat of this question - did you find certain organizational doctrinal practices and structures are more effective in dealing with friction associated with competition, either in conventional wartime situations or even in gray zone scenarios, for example do flatter or more hierarchical organizations perform better in different situations, thank you. Yeah. That's a complicated question. Can you generalize on that? I'm not sure I could. Yeah, before - let me - your previous comment, I just
really wanted to pick up on because I think it's so important. You're absolutely right. Right? I mean - when special operations says counterterrorism they mean find, fix, finish, evalu
2021-02-26