 Good morning. Thank you for all coming out today. Can you all hear me in the cheap seats? Okay I'm Tom ricks. I'm with new america's future of war project and I write books for a living Our guest today is Eric Schmidt You all know as mr. Google also mr. Former chairman of new america He's here because he's been running a panel looking into innovation in the military For the Pentagon and we're going to focus on that today. I'm going to begin by asking a few questions and Then we're going to go to the audience and you can throw your heart balls at him First though, I want to ask a poll question Our polling question has to do with the subject. We're about to discuss it is this Which innovation area should the US focus most on for future? military advantage a Hardware robots be software artificial intelligence cyber see Waveware that is energy weapons D wet where that is Biosciences and human performance enhancements and it looks like software is the Iran of this one How long are we supposed to give this? Okay, we're cutting it off Eric you've been running this panel looking at the Pentagon and defense and innovation What are your two or three big takeaways from that? Well first, thank you for for having me. Thank you for supporting this. Thank you all for coming I want to thank the sort of new america and ASU folks for putting this thing together As I've gotten more and more involved trying to understand what we're up to I think it could not be a better use of everybody's time And Tom, you've got a new book coming out on Churchill and Orwell So so let me do give a shameless plug to to your new book having read many of your earlier ones I do write big deal, right? When is it published? May 23rd, okay? I'm sure you can pre-order it. You can pre-order it on Amazon And by the way that really helps with the first week numbers on for the best seller list By the way, mine is being introduced tomorrow morning sign ahead of you So I Also need I want to give read a disclaimer though just to make sure we're all proper here you I put together at the request of the DoD a group that visited up to 20 ish sites around the world and But so I have a lot of opinions now, but I'm not speaking for the for the department or for the Defense Innovation Board Anything I say here will be my personal views as a private citizen and I've studied it for a long time So with that your question was what are the two or three most important takeaways? I think the biggest takeaway is that you guys got it right and the single overwhelming conclusion you come to speaking as a person is sort of wandered around is that the Defense Department Broadly speaking is run in many ways like a 1980s corporation There's not very much software being used and it feels that way so We were asked to sort of come in and look at what would how our external companies run technology innovation And how's the DoD and the lack of software and in particular the lack of commuter science is a real problem to be addressed And that recommendation has been well heard and and is being listened to today by the military We actually made 10 or 11 recommendations which are public and we have a public process They were released in January and you can just type Defense Innovation Board and you'll read all about them with a lot of justification One of your recommendations was to have a chief innovation officer. The Pentagon already has a chief technology officer What's the difference? Because the DoD is run in this sort of classic old way The internal procedures that it has that are not I don't have to say that they're very long in the in planning right so they plan over a decade and And the culture does not have a culture around rapid prototyping at sort of any level And so our idea was that you could take a chief innovation officer and put that person in To basically try to get these systems to be more responsive to Innovation quicker product cycles and so forth and this is true at every level of the military whether it's the HR systems The transportation systems the weapon systems everything that they do it sounds to me like you're describing a problem That's more cultural than technological. Is that innovation is always a Cultural problem and it's a management problem. It's a cultural problem. It's a technological problem. It's a money problem So there are for example groups in the DoD that are doing exactly the right things and they need more support, right? More political support more funding. It's over just a big bureaucracy But what happens in large having now run a number of very large corporations the culture tends to determine the outcome And so when you go and talk in detail with the military What they'll talk about is their their own view of how the culture works Which in the military is largely determined by their promotion capabilities that where they're reviewed Risk intolerance and risk tolerance risk seeking who's in charge that kind of stuff. It's obviously very hierarchical How do you? make the military bureaucracy less risk averse Well, one of our recommendations is to make it Sorry, it's a bit of background. There's this process generally known as AT&L and people in the audience who have been in the military will know a great deal about this and It's rumored to be a couple thousand pages. No single person seems to understand it So if people spend all day complaining about the process, right? And this is how the weapon systems are purchased is how everything is purchased It drives the planning processes and the organizations that implement it are highly centralized and very very rule bound So that's an area. That's a target rich environment From from our perspective. And so we wandered around and we said well, this is clearly a problem so the AT&L people call back and say well They're a gazillion waivers. You just have to ask for them. I said what? You can wave this stuff Yeah, it's in the document, you know page 697 so we found the the the document we found the paragraph and indeed it's true that Using good judgments the military can Actually do the right thing right from our perspective and so one of the recommendations that we have made is to have much more extensive use of waivers and it's interesting that again military leadership is very sensitive to this One of the more interesting things that the senior leadership did is with the B21 They actually used a rapid evaluate a rapid acquisition process rather than a more cumbersome process, which was controversial But that's an example of innovation within the bureaucracy that that I certainly hope will produce a better outcome and last question for me for the moment if you were king of the Pentagon for it a day is There one change you would make and how the US military approaches this problem personal policy education whatever Well, I should say that the people we've met are incredibly impressive I don't know about that you all but when you spend time in there The leadership skills the training the level of education the commitment to serve the mission. It's all very good the systems are just sort of I'll just say horrific and An example is that there's this process called the program of record Basically, it's called the palm and it it's a two-year planning cycle During that two-year planning cycle you have to budget things for two plus years from now and then it's ten years after that and It's very difficult if you're doing rapid acquisition rapid software rapid iteration to figure how to fit in that So I would start by attacking that problem and insert This sort of rapid prototyping model one of our recommendations Is to build a culture of rapid prototyping and experimentation when you try when you talk to the generals and admirals? They say they're frustrated right these are people want to do the right thing and they'll say yeah, I want to do that I want to do that but then I have to adhere myself to that to this process which is sort of turgid would be the best description and So I say well, okay, I'm gonna I do I do this with a little bit more respect I'll say something to the effect of you're a general you're in charge. Why can't you do this and Then they'll look at me and I say this very politely and then give a very polite back answer Which is I tried right? What I did is I cut this budget I made this more flexible and the moment I did that the money that I had saved was taken away from me right That's a disincentive to taking risk. That's a classic bureaucratic problem by the way So creating a space of not just and by the way We're not talking about a lot of money here in the Pentagon numbers Small amounts of money and then what the beauty is about the survey that you all did about software software is relatively inexpensive Compared to the many other things the military does Okay, we're gonna go to audience questions to the people wielding the microphones have them ready While the microphones are getting ready, I just want to mention In World War two George Marshall the army chief of staff Was at a meeting in which the army was represented by a general the navy was represented by an admiral and the Air Force Who was represented by a major? And it was pointed out to him that the this was kind of inappropriate to have a major arguing for the Army Air Force And he said you're right. He promoted him to Brigadier General Immediately Okay audience questions over here, please Hi Ted Johnson and Eric and Wendy Smith fellow here at New America. So thank you for that. Thank you. I'm glad you're here That was not a plant I'm also a retired Navy commander and recently wrote about how military bases can bring smart technology To sort of streamline base operations naturally the initial concern is about security So so as we think about how to incorporate more technology software hardware, etc For military applications, how do you address the real security concern of attaching? thousands of wireless sensors to military infrastructure and you know Protecting them from hackers and other well indeed. Thank you and thank you for working with New America One of our recommendations as a group was to focus on this question of cyber more more thoroughly I'll just use my own language. I am concerned that outside of the secure networks Which have a pretty good? I'll just say generically a pretty good approach to try to be secure The in non-secure networks, which is where the majority of the military activity is these are the unclassified networks How do we know that they are secure and a given example a specific example? Many of the cyber attacks are done on downward of Windows machines and the military has lots of downward of Windows machines So the military has recently announced that it's going to upgrade all those to Windows 10. It's a reasonable decision However, they are not upgrading at the same time the systems the PCs that are inside the program of record Weapon systems and other systems So Here's an example where they're trying to do the right thing, but it's such a massive problem. It creates a vulnerability So one solution here would be there are many solutions one solution would be to try to create and we recommend this Some kind of a facility inside the military or appropriate place that can really go and address the question You asked it's specialized problem. I agree with it. However, it's not an excuse to not automate the basis Right that the standard joke about the nuclear stuff is the nuclear stuff is not hackable because it was all the software was done in 1970 and it's still stored on eight inch floppy disks, which you can't purchase anymore Right. It's not a good answer. By the way, that's sort of the joke Right. I'm not endorsing that as a cyber strategy Um, God with the beard who looks like Andrew Waxham. Let's get some mics over here if you have mics as well If you have multiple mics Thanks, Zach Bakes with James. I wanted to ask you, you know the last administration with Secretary Carter innovation was a huge priority He opened the three branches of DIUX You can look at several different major speeches he made on innovation And some of the travel that you mentioned you've done to bases and facilities were with him He was on those trips. So secretary level issue Does the new administration view it with that same sense of urgency from what you've seen thus far? And is there any difference in the approach you're seeing from the new administration? I grant there are a lot of open jobs 18 L's open etc. But do they view it from a different viewpoint? I haven't seen any Change so far. In other words, it looks to me like the same problems same actors same recommendations I would probably say that that our recommendation our recommendations were designed to be Independent of who the leadership are. I think these are generic They're not specific to any particular leadership approach. So I haven't seen haven't seen much of a change Same focus on this. I think at some point There'll be some kind of crisis That will sort of bring this into form. Maybe it's a budget crisis or a military crisis or something. That's usually how it happens speaking of the new administration if President Trump Called you into the Oval Office and asked you what the next war would look like. What would you tell him? You know the There's a there's a sort of The military doctrine on strategy is pretty well set and it's fairly well set from a long historic reasoning So the nuclear triad All of those things which are horrific and terrifying by the way, so we're clear And the need for an army Navy and so on What is missing is what radical innovation? Could occur that would change war Conflict and also defensive positions Bob work who's the Departing Deputy Secretary has a thing called the third offset where he talks about autonomy in particular and artificial intelligence and That my guess is that's where the real problems would come from that if we don't invest in those areas and The reason is that our competitors I'll just use the word competitor for lack of a better term are clearly investing in these areas Whether it's cyber autonomy. Who's the larger? Who's the largest manufacturer of autonomy? All right, the Chinese with DJI So we need to really think about We have a whole recommendation about this and I want to stick on recommendations It's important that the military make AI and machine learning a component of their work All right, we looked pretty hard to find such groups And there's a few people here and a few people there who were valiantly trying to do this But it's not part of the core competency Whereas you can imagine if you were to let's go do it One of the things you do when you look at change in an organization is imagine You didn't have the organization at all it. No people at all. Just do the thought experiment Of course, it was not really going to happen. What are the things that you would build right? How what would you do first and I think one of the first things you would do is try to use Artificial intelligence and machine learning to change the way people think about deterrents An example which I've used publicly in other forums Is the question of who has advantage in cyber war? you we in traditional conflict the The attacker has advantage because the advantage of surprise Right, and you know, you don't know what's coming. You have all these people monitoring You know, oh my god, you know, this thing's coming. We have to react to it And then you react. That's typically how it works in any form of human human contest But in the in the AI world training matters It may very well be that well-designed cyber defense systems are stronger than the attacker systems Because the attack the defense systems are constantly being trained by the attacks that are coming so anyway, we focused a great deal in our report on the importance of becoming Expert at this indeed. This is one of the recommendations that has been adopted so far We'll see how far the military gets in this on artificial intelligence. I'd like to ask you a Rosa Brooks question Not that she placed this but that I think she would ask it and I would too if I were smarter she is So can humans hope? To keep to ever fight against autonomous weapons Well the general again speaking generically it just seems like I should say that the military has told me that their doctrine is that the human is always in the loop That's that's what they've told me. Is that a recipe for disaster? I'm gonna let I'm not gonna comment on that I'm gonna that's that's they've told me That's their position, right? If you have an autonomy autonomous attacker, you're gonna want to have an autonomous defensive system Mm-hmm, and if you think about autonomy again, would the defender have more advantage than the attacker? It may be because of training, right? And that's a concept that it's for background for those of you who don't know about it Today computer vision is better than human vision and Machine learning allows you to recognize patterns more accurately and more depth than humans can and We all operate as humans with sort of reasoning and common sense all the things that got us to where we are today but in fact the revolution that's going on in my industry is because The algorithms technically deep neural networks are so good at recognizing these patterns so for example the ability to have Automatic translation between languages something we do very well now Recognition of images much of what governments do is in fact pattern matching, right? So what does a security guard do? What does a military guard with the gun do? They're all about what does the submariner who's looking at video of underwater footage looking for mines or other things underwater They could threaten the ship because they're all vision problems when you go to the doctor What does the what does the dermatologist do? It's a vision problem, right? Well, we have technology that can do a better job equal or better job than almost every dermatologist Using vision now it doesn't replace the dermatologist. He just makes the dermatologist better So I think the general theme here is if since this is about about the military if the military were to adopt these things The humans would be more accurate in terms of watching monitoring You know being ready which is part of readiness is a big thing within the military This hadn't occurred to me before but listening to you for the last 60 or so years the offense has predominated in warfare The last time we had a war in which the defense predominated was World War one It sounds to me like you're describing a terrific war of attrition in which victory goes to the side with the ability to sustain itself Economically and demographically is that the war you're describing to mr. Trump Again, I I want to be specific on where the technology can take things And I think the biggest innovations in the next 20 years will be in this area of machine learning AI autonomy right and America is in a very strong position to be a leader here, right? We have all the best people We have this huge industry, and I think that part of why we were asked to to work with the DoD was try to get some of those ideas in Okay, the right wing has some questions over here. I Think the guy and the standing against with his back against the wall Thank you I'd like to follow up on the human capital question When you look at my name is Mark Hager. I'm a cyber fellow here But I also served on the Defense Science Board study of autonomy long service in the military now. I'm an academic When we had the industrial revolution we had to create entirely new educational institutions West Point Thomas Jefferson the Navy Responded to mechanical engineering you look at the companies that you led Apple led they weren't to knock off on Kodak They weren't a knock off on Honeywell computers. They were something entirely new. Is it time to build a AI? Cyber force with the institutions from the 18-age year old all the way a new culture to augment West Point Air Force the Navy the Army but time for a new force that gets to those rapid culture adaptation and right now Probably is just too slow So again, that would be something for the military to decide how to do it but we made a strong recommendation to Essentially invest in both computer science and AI as a core competency for precisely that reason There are a number of initiatives within the military already that need to be to get a lot more support Was earlier mentioned something called DIUX which is a group that essentially takes military money and Advances companies in partnership that will help the military reach its mission And they're essentially if you will providing additional capital to the private sector to jump start the ability for the military to purchase those Activities so that's a big deal There's another one called SCO or strategic capabilities office. It's run by a guy named will roper a true hero in my view and what they do is they take the existing systems of the military and they Lengthen their lifespan by adding software autonomy and things like that So again solving a very tactical problem, but one which is you know sort of serious So those are two that the military has already I think any form of your idea right whether it's a center a building and so forth makes sense Another idea that we had was just have every major command have a software group. That's under command of the commander Right, so the military is full of the following problem You have screen a and screen B and you have human and the human it spends all day a very well-trained soldier Typically copying the data from screen a to screen B right because the systems don't coordinate each other So in in traditional systems, you know the first thing a new in a turnaround CEO The first thing that CEO would do is go and raise some capital replace the systems automate everything and get huge efficiencies in the system So from my perspective any form of adding software people into the commands and into some part of the university Make sense. I'll give you an analogy the military Nuclear work is done not by the DOD, but by the DOE labs so we visited them as part of our normal visiting and I knew very little about this area and I was struck by the level and scale of what our nation built and Invested in way back when right now. We're not at the same level and I'm not suggesting We have to have Sandia and you know Kirkland and all that kind of stuff, but my point is here. You've got a clear problem a clear opportunity It's pervasive across our nation's things and not just military but everywhere Right, you would want to have a set of buildings a set of programs analogous to what you described And also you didn't mention it But one of your recommendations was that ROTC have a specific computer track of computer sciences again, we go on to There's plenty of Technical work going on in the DOD and there's plenty of people who want to come in as young technical people who want to serve the nation You know and so forth. There's something called the Defense Digital Service Which is another one of these early programs where again software and people will come in I think if the DOD identifies this as a major priority the the capitalist system around it Which is basically the contractors and the and the consultants and so forth will respond We'll get themselves educated up to what you're describing. I like your idea And I think we should propose that one as well Of course It does amaze me that to my knowledge There is still not a reserve unit in Silicon Valley for information warfare operations You'd seen any service chief who wanted to pick up some cheap talent. I mean my college roommate was nuts But he became the executive Apple and if you told him he could he could shoot a 50 caliber machine gun once a year and Do a couple of hours a month of helping out? He'd be happy to but nobody's ever asked him. I Saw Aaron Simpson had a question Thanks Tom Aaron Simpson until recently I was the CEO of Keras Associates Which was a research and consulting firm did a lot of work with the Defense Department including DARPA and as a business person I'd like your thoughts on why both small and large companies should want to do business with the Pentagon There are incredible compliance costs. Yes. I'm margins. It's not the easiest place to work Well, in fact, we have a recommendation on that how how perfectly asked I've I omitted to say that that my career and my world Is largely due to DARPA, right? So you go back to Bob Taylor as a leader of DARPA in 1965 6667 and the funding that he did that ultimately created the industry that we that we live in I mean I worked for Bob So I know this is all true These are extraordinary moments in American history. So let's thank DARPA for among other things helping create the computer industry One of our recommendations as a bit of background working with the military as a procurement matter is very difficult and One of our recommendations is to move to the cloud All right, but we don't mean the military cloud. We mean use commercial cloud Phenomena so you say seems perfectly reasonable after all for non for non classified work Military is just a big business in that sense and yet. It's extraordinarily difficult Because of all sorts of strange rules that don't make any sense, right? At least from my perspective speaking as the outsider And so that's an example where if the military were to move to more standard ways of interacting and operating So that they can look like a large corporation for their non classified work That's a big deal to change the way that knowledge be changed the way they would and indeed this is one of our recommendations Okay, we have a centrist question here. We can run a mic up To the guy at the bow tie Will Duffield London School of Economics human machine interfacing and Biological enhancement will alter both how we fight wars, but also how our servicemen will be injured in them How can we best anticipate and prepare to care for the casualties of future wars before we see them coming home? So far it looks well There's a lot of assumptions in that in your series of questions. So one is that It looks like death by conflict are actually declining Right globally, right? It was obviously a good thing However, the advent of more precision means a much more precise outcome, right? And there are just horrific scenarios that you can go through in your mind So I don't know how to answer your question except to say that There are people who were working on human augmentation and so forth But I think the correct answer is to figure out how a way not to have conflict, right? It's just it's always a bad outcome and Based on again speaking as the outsider The and again people it's people get used to these things I guess but when you when you look at the capability of human fighting human We really want to be against it. I just do the math David Wood has written a book that Gets into a lot of that. It's what David. Do you have a comment you want to add to that? I'm not sure what the interface is between Combat stress and combat stress injuries and Cyber war so I'd be interested in your thoughts on that if any well Jared Cohen, and I wrote a book about the sort of the new digital age and and how Wars would get started, right? And as part of this we talked about the fact that it's possible for Countries that are at peace to have cyber conflicts Right because there's relatively little deferred deterrence and so forth. We see that today With the Russian and Chinese scenario where we're clearly not you know quote war unquote and yet both countries have done things to America that we're upset about and Analogously and I don't know this we weren't read in on this But I'm assuming that America did things that annoyed them no end as well and yet we tolerate that so Cyber at least at the level of peacetime if you will Maybe a different kind of animal if you will strategically And I don't think we know there's been lots of discussion about What's it like to be a drone pilot remote so forth and so on I won't comment on that But I think these are new problems to be researched One new problem. I've seen in the field is in the initial steps of a war like the invasion of Iraq in 2003 American ex-soldiers experienced for the first time in their lives extreme information deprivation For the first time in their lives. They have no screens to look at they have no cell phones. They have no computers at night And they they seem to me to go through a form of information starvation They are not stimulated by by The things that normally get their brain going Early in the Iraq invasion of 03 I Would show up at checkpoints with my big heavy Satellite phone. I think the company would bust I came up with their name but it was my big satellite phone and I would offer it to the soldiers I want to call home for 60 seconds to say hi mom And to them it was like it was much more meaningful than just a phone call. He was being reconnected to the real world It's what they called in Vietnam the big px. I think now it's the big database So one of our recommendations is to make computing and bandwidth abundant You know we traveled around and talked to the to the folks who were just doing normal jobs in the DOD and They spend a good deal of their time Emptying their mailbox of mail because the mail system is so primitive right these problems were solved a long time ago I know that's trivial, but if you're a soldier on deployment, this is a really big deal and The technology now allows to have essentially abundant computing and abundant communications around the world in general the military communication systems are Rare and expensive and that's true in general of military things that the that the because of this lack of iteration and sort of the sort of the Building a small number of big beautiful and very expensive things as opposed to a large number of things Which are far more common which again is endemic in the in the approach that the military uses They don't have those kinds of solutions I think your scenario in 2003 can be addressed now by Much better connectivity, but the principle is correct that we recommend very strongly that The the everyone in the DOD is a knowledge worker to write and knowledge workers need access to information knowledge Sharing and so forth and so on the downside is you need extreme information discipline of a sort We've never had especially in American units If I'm the enemy military intelligence commander I want to know what frequencies those cell phones are operating on but those things are easily in general almost all of these problems can be Solved by very strong encryption So again the way you would solve and this is a generic statement about companies All of you should be using two-factor authentication And I can give you lots of examples of people who are unhappy because they didn't All of you should be using chrome by far the best browser. It's also free. Okay, so it's the right price Right, so a simple list of cyber rules of that category would go a very long way But I'm not just worried about secrecy. I'm just saying emissions become a target Especially worries me when embedded reporters don't understand signals and start using unencrypted Devices in the field and get us all killed so again in the in the computer war in my industry We've all moved to fully encrypted HTTPS And so when you are using the internet you're on a very encrypted communication as long as you're using modern technology and so forth It's funny that not all the military does that And you're talking about newsrooms are cheap. I remember being laughed at the first time I asked for body armor You know, what would you need that for my editor said? Way in the back here Wave your hand There you go. Thank you. Hi, Eric Jody Shelton, and I'm the president of the Global Semiconductor Alliance I am curious as to whether or not your role with the DoD Has adversely affected Google or Apple in your global business when oftentimes that means dealing with Foreign governments. It certainly means being geographically neutral much of the time and Conversely does the DoD? Think that it's strange to have a global leader as part of their team well I Can't speak for them about how we appear to them. They probably think of us sort of like the Silicon Valley show But you know, who knows we're just trying to do the best we can We've been pretty careful to identify ourselves as trying to bring Technological and business ideas to the military but not telling the military what to do. We're not in charge We're not part of it. That's why I give my disclaimer and so so far that's been fine And and we have representatives on our board from pretty much every every Silicon Valley company Jeff Bezos is on the board Just lots of interesting scientists and so forth. So I think there's It's there's not been an issue so far Guy was standing at the table right behind you. Thank you Zachary Sharma graduate student at the University of Kentucky studying security and intelligence matters a Lot of what you've talked about to this point hits upon sort of broader organizational steps I feel like in the earlier panels, they seem to discuss networking more as a way of Integrating the force or rather Developing sort of a more mobile agile unit in sort of a battlefield context it Do any of the recommendations you have sort of touch on the battle space aspect making that or making the force more effective in terms of So the so the military has a strong view on this which they which I'm just going to label as data fusion Right and so the military way of expressing this is that you have all the sensors You take all this data big data and so forth We listened to that for a while and we said well We don't think you're on your way to building that in other words. That's what you want, right? But so our recommendations are in our view the necessary conditions to create that vision if that's what the military wants So in particular, you need connectivity You need data systems that can carry the data the sensors and so forth generate an enormous amount of data Which they don't do anything with So typical example is you've got all of this information coming in and it gets lost It's stored and forgotten and so forth. Nobody knows what to do with it And there's no programmers to take it and analyze it So again, you don't have the you don't have any of the normal big data scenarios that a company would have One of the recommendations that we are considering making we have a meeting on April 4th Everybody here is invited to it's at the Pentagon to our meeting and we have our meetings in public We're going to have a discussion about should we make a recommendation about what to do about this big data problem Because we keep encountering it over and over again My own view is that that some centers around big data need to get formed You need to get some softer people. You need to start figuring out what the data is useful for And I should say that I think this problem in fact going out on a limb I would say that the majority of our recommendations would apply to the u.s. Government in general In other words, I I suspect I suspect that if you visit kind of any other Department of the government you'd find a lack of cyberspace awareness a lack of computer scientists a lack of proper systems An over reliance on down rev systems Um You know lots of information being leaked to competitors and adversaries and so forth because the systems aren't properly architected I think this is a generic list in that sense. We're down to a precious few final moments One question I want to step back and ask is a broad one. What's the biggest global trend you see out there that is under recognized I sent most of my in my real job I spend most of my time thinking about What happens when you adopt an ai first strategy in the business And ai first meaning taking the systems that we have built so far And using the the technologies that I've been describing here to make the systems more effective and in google's case That would be ads and search and How can you use these systems to hire better people? Can you use these to to promote people better and so forth and so on? So I think that the people are generally aware in this room probably that this is coming But I think most people don't appreciate how profound it can improve Whatever you're doing. So uh, health care is an easy one it looks like Um, the computers are very good now at the following they can have a series of numbers and they can break the next one Right. Well, that's not very interesting, Eric. By the way, that's exactly what happens when you go to a hospital, right? You take your EHR record, right and with enough data and enough training We can predict why you've come to the hospital and more importantly your mortality morbidity Scenario, which is what the doctors really care about So whether it's in health care, which we all care about or information markets, which I care a great deal about Or whether it's the military, which is the conversation we're having I don't think people understand how profoundly transformative in almost all a good way This technology is and so I guess I would leave you with Our core recommendations are fundamentally about this right to change the way the processes work Build this as a core competency Try to understand how it affects. This is a future war conference. How does it affect war? And I think that's a very good thing to do for the next year or two Our time is up And I'm the only thing standing between you and lunch So but before you run for the doors Please give Eric Schmidt a big hand Thank you. Thank you very much time