so i look forward to -- >> i didn't quite say that. >> i know, yes. fair enough. fair enough. there we go again, the administration putting words in your mouth. thank you. >> thank you, david. martin? your thoughts. >> well, i'm from 35 years of trench warfare. and i see these problems and what's been going on lately. obviously a tremendously different perspective. i grew up, as i say, in the trenches. i worked with amazing members who are oversighters like john morse, john dingell, ben

rosenthal, jack brooks, henry waxman, carl levin, and chuck grassley. and all of them, you know, brought -- and staffs that were loyal, longstanding. and they came to me, you know, to -- for perspectives on how do we get things done, did we ever do this before kind of questions. and over the years, i learned a lot, i think. i think i got a reputation as a zealot for congressional oversight and it was well learned. i learned three things. first, congress' oversight power is virtually plenary. the investigative authority is

irrefutable. courts have consistently recognized that in order to perform its core constitutional responsibilities that congress can and must be able to acquire information from the president, the departments and the agencies of the executive branch. the structure of the checks and balances rests on the principle that congress has the right to know everything that the executive is doing, including all the policy choices and all the successes and all the failures in the implementation of those policies. the supreme court has made it absolutely clear that article i presupposes congress has meaningful access to information so that it can responsibly exercise its obligations to make laws requiring or limiting

executive conduct, to fund programs supporting executive policies to which it approves, to deny funds for those policies for which it disapproves, and to pursue investigations of executive behaviors that raise concerns. i found in those years that committees wishing to engage in successful oversight had to establish their credibility with the white house and with the executive branch departments and agencies that they oversee early, often, and consistently, and in a matter invoking respect if not fear. although standing and special committees have been vested with a vast array of formidable tools and rules to support their inquiry, including, including supreme court and appellate court approvals of practices and

processes, including that congress has adopted for the conduct of its oversight and hearings, that do not accord witnesses with the full panoply of procedural rights enjoyed by witnesses in adjudicatory proceedings. it is absolutely critical to the success of investigative power that there be a credible threat of a meaningful consequence for refusals to provide necessary information in a timely manner. in 1795, that threat has been the possibility for citation of criminal contempt of congress or trial at the bar of the house, either of which could result in imprisonment or fine. and there could be little doubt that such threats were effective in the past, at least until 2002. in particular i would point out that between 1975 and 1998 there were ten votes to hold cabinet

level officials in contempt. all of those resulted in complete or substantial compliance with the information demands in question before the necessity of a criminal trial. it was my sense that those instances established such a credible threat that a contempt was possible, at least until 2002. even the threat of a subpoena was sufficient to move an agency to an accommodation with respect to document disclosures and the testimony of agency officials and the white house to allow executive office officials to testify without subpoena. the last such instance was the failed presidential claim of privilege during chairman dan burton's 2002 investigation of two decades of informant corruption in fbi's boston regional office.

i would add that it was a bipartisan effort, which was unusual for chairman burton, in which a contempt vote in a bipartisan manner was a certainty if the president didn't cave in. and he did. i'm kind of surprised that, you know, the session relates to simply these two cases, the holder -- the meyers case and the holder case. it's as if everybody is thinking that it came out of the blue. but that -- those situations represent and underlie that congress is presently under literal siege by the executive. it has not suddenly come out of the blue.

it is a calculated offensive. the last decade has seen, and i catalogued this, among other significant challenges, an unlawful fbi raid on a congressional office, the department of justice criminal prosecutions of members congress that have successfully denied them speech or debate protections, a presidential co-option of legislative oversight of agency rulemaking, presidential refusals to ensure the faithful execution of enacted statutory discretion, the directions, an unsuccessful attempt at usurpation of the senate's appointment power, and with respect to investigative oversight of the actions of executive branch officials, the adoption of a stance that was first enunciated by the office

of legal counsel of doj in 1984, that the historic congressional processes of criminal contempt designed to ensure compliance with its information gathering prerogatives are unconstitutional and unavailable to a committee if the president unilaterally determines that executive officials need not comply. in such an instance the department of justice will not present allegations to a grand jury as required by law. that's where we are today. these two cases that have come up, you know, are a reflection of a concerted effort to undermine congressional oversight. and the only thing that congress can do is to step back -- i agree with the utilization or attempted utilization of

confirmation powers, the appropriation powers, et cetera, et cetera. but they're not targeted. they're not going to frighten anybody. one of the panelists talked about transactional, you know, methods of settling, you know, disputes over information. what that is, when translated correctly, is an ability to negotiate and stymie over a period of time. congress has to look at what powers it has and get back to finding a credible threat that will bring the executive to the table and to negotiate. and if not, issue, you know, either a contempt of congress criminal action or revive the

inherent congressional contempt. inherent congressional contempt stopped in about 1934, 1935, not because it wasn't effective. it was. it just took too much time and the criminal contempt process was thought to be more expeditious. neither of which is true. the inherent contempt process can be made better, more acceptable. there's no reason now -- the supreme court has for over a hundred years in four different cases established the constitutionality of inherent

contempt. it has been deemed unseemly because it requires arrest, incarceration. and that isn't necessary anymore. the supreme court has, you know, made clear in a case involving the senate's, you know, power over impeachment that the trial can be preceded by conduct of investigations that will be able to present to the senate at its trial and cut down on the time it takes, you know, for a trial. the same thing can be done, you know, by internal rulemaking to make the inherent contempt process seemly.

you can provide for, you know, an investigation, a presentation of recommendations to the floor of the house. and the penalty doesn't have to be imprisonment. it can be a fine. there is certainly precedent for that. secondly, with regard to criminal contempt, that is also still necessary. and the olc opinions, you know, misstate the history of criminal contempt. criminal contempt is absolutely necessary as -- you know, to be revived. there is no doubt that there is an analogy to criminal contempt that is issued by courts when there is a contempt of court. the supreme court in 1987 in the louis vuitton case accepted the

right of a court to appoint a prosecutor to criminally prosecute somebody who has, you know, been found to be in contempt of court, to bring a private attorney to bring a prosecution. the next year, in morrison versus olson, that louis vuitton case was cited prominently as a seemly and authorized means to -- you know, for a court to appoint somebody, you know, when there is a criminal prosecution. it should be understood that it is constitutional for congress to have -- congress has the same

self-protection that the courts have, and the analogy is appropriate. and i think that the next time that there is a refusal to bring a contempt of congress to a grand jury, there should be a resolution that authorizes both an inherent contempt at the same time that there is a criminal contempt, and that the supreme court, in morrison versus olson, will come and support it, that there should be an injunction, because there is only -- what the justice department is saying is really that there is a conflict of interest because their client is the president and also the executives.

justice department in such situation has rules which says if we do have a conflict of interest we'll appoint a special prosecutor, either somebody co interest, we'll appoint a special prosecutor, either somebody within the department of justice who is wauld off, or somebody private, like independent counsel. both of those should be looked at, because there needs to be leverage here. this is all about politics. that's what it is. and congress has had it, and needs to revive it, and not go to court. the court process, you know, is an experience, and means delay -- delay for oversight means effective oversight goes away. thank you. >> i'm going to open it up to questions next. and i actually wanted to take my privilege and start with one.

and mark, you started by talking about you served with senator levin, grassley, and we've also heard from david and others on our panel about the contrast between oversight being used well and misused. i'm curious, what, in this -- i'll open the panel, but what do you think enabled that strong, or does enable that strong good use of oversight leadership and the figures who exemplify that in the past, and what is limiting it now? in other words, why aren't we seeing more of the type of -- the leaders in oversight we've seen in years past? what has changed and what can we do to maybe bring some of that leadership back? >> since an institution is missing, and in taking away, there is no sense today that if

we enable effective oversight, and enforce it by contempt of congress, or inherent contempt, that there's not any thought of it. because at the forefront will be the other party will use it. and we want to, you know, be sure that that doesn't happen. there is no sense that -- of the responsibility and the duty of members that's common that is a need for cohesiveness. a need for, you know, underlying, you know -- maintain the integrity of the institution

itself. it's the results. and it's not there today. >> well, i want to go on record as being in favor of effective oversight. so anything that will contribute to that, i'm in favor of. i guess one thing i would say, on this point, is i suspect some of the breakdown, if you will, and -- let me start by saying, i don't agree that there isn't a lot of effective oversight going on. i think there is. but there's a lot of high-profile type matters that suggests it's not doing well. and obviously it's not doing as well as it has in the past. but i think at least in part, that's a function of the more polarized political world that we're living in, that affects any number of things. it bedevils us in any number of ways. we're living in a way of more tribal politics. i think it makes it harder for the minority to trust the

majorities in congress, for the majority to trust the minorities. and i think this at least in part is probably won by a product of that. if we can figure out a wra to, you know, make our politics more nice, then i suspect more bipartisanship oversight will probably flow from that. but i don't have an answer to that. >> it's worth noting that there's -- you know, we have this sort of dpauzy sense that there was this great historical moment that partisanship noticed everywhere. that's the function of moment in which we're living right now, which is to say for much of the 20th century, there were racists everywhere. the reason that bipartisan coalitions are possible for much of the 20th century, is the democratic party was split down the middle, between the northern and southern wings, so the republican could make common cause with usually the northern democrats, occasionally the southern on var issues.

when you look at the votes, you will have some democrats and some republicans. but that's an anomaly in american history. if you look at debates leading up to the civil war, i talked about the mid-19th century, the wigs started their life called the anti-jacksonians. we go to the 1800, puts the current election in some per speck shon. you can look at the investigations arising out of the 1890s, which are entirely done along partisan lines. there are very few moments in american history in which bipartisan, this sort of kumbayah happens. it turns out it's not for reasons we want to emulate today. i want to push back on the idea that bipartisanship is something that we should see as necessarily an indication of good, healthy politics. sometimes one party controls a lot of the levers of power and sometimes that's because they're engaging more effectively with

the public. in that situation, i think it's not entirely inappropriate for them to press those claims, to sort of get more of what they want to use the mechanisms of power. sometimes we have divided government. and that's because the government hasn't trusted either party with power. and in the -- it's the manifestation of the fights playing out in the public at large. look around you. we have a divided policy at the moment. why should our institutions not reflect the friction that we have out in the world. that would be actually sort of eliminating some sort of our political diversity. that doesn't seem desirable to me. >> since we're all guessing here, i'll put my own guess out there. two of the three examples i mentioned were oversight initiatives from our authorizing committee, the house resources committee. the reality is, there's not a

lot of legislation going on. the complete absence -- or near complete absence of a dialogue with the administration and across the aisle on legislation i think provides a bit of a vacuum that promotes this kind of thing. i served in the clinton administration as well as the obama administration. i think it's getting worse, not better in that regard. >> my question for the panel is to get a sense of your advice to congressional staff on carrying out effective investigations on a day-to-day basis. i've seen this as an investigator on the senate side, working for a private law firm, seeing this as an executive agency in the white house. and in a very practical way, some of the lessons that i

learned in the senate are not necessarily carried through, things like keeping investigations confidential until they are ready to be final, making them confidential, treating materials that are sensitive as sensitive. and in doing other things in a way to -- for congressional staff to establish credibility. i may be using that in a different than you used it, morton. but in a way so that there is, despite differing interests on either side of the congressional investigations, there is a balance of power. and how both parties can establish credibility. as a lawyer, yes, we do have different interests, but with edo believe that facts should carry the day and the truth should be the truth. and that's what investigations are meant to get after. so if you were giving advice to congressional staff, keeping in mind that many of them are young and under 30 and not necessarily lawyers, what sort of institutional advice would you give them?

>> i would advise them, first, to pay attention to the more senior folks in their ranks rather than going off half cocked. i mean, i'm of the school that the executive branch takes inquiries like this very seriously. and in my experience, it does. you know, a predecessor of mine in the bush administration went to jail because he lied to congress. so there are terrific incentives to be careful about how administration officials work with congress. and i just -- i think it's like any other, you know, potentially challenging situation, professionalism really pays off. when both sides recognize that they have institutional

interests, but ultimately they have the same interests in government working well. and ensuring that, you know, we're serving the american people. but obviously, you know, completely one-sided reports that, you know, that having -- you know, since chairman burton's taking it on the chin, i mean, the opening remarks of chairman burton in one of his oversight hearings was, you know, amazing to listen to. because untethered by facts. but on the other hand, chairman dingle and chairman grassley and others, you know, not that way at all. you know, let's start with the facts, let's talk about the facts, and let's develop the facts. so i do think that maybe part of the problem is, there has been a little bit of a turn. obviously on both sides of the

equation. and some of the real pros in terms of staff, congressional investigators, you know, have moved on. and you do find some of the chairman of -- for example, house resources committee bringing in new groups of investigators who don't have the experience, and who are thinking of this as, i would guess, more of a political exercise than anything else. >> i'll take a brief crack at that from the standpoint as someone who worked in the general counsel's office as opposed to a committee oversight investigator, or oversight lawyer. two things, i guess i would say. number one, i think because of my background as a litigator, my advice would be litigation centric. you need to be focused. you need to be precise. and i'm talking about subpoenas and requests for information. you need to leave as little room

as possible for the department's -- the executive branch departments to concoct objections. all right? so you come at it from that standpoint. the other thing i think i would advise is, given -- notwithstanding my advice that litigation is probably not a great option for the congress, i suspect it will continue to be an option that will be used, so i would be giving advice on how to shape information requests, and how to conduct the investigation in a way that makes it more saleable in the judicial context when we get there down the road. >> one thing, which is maybe sort of the flip side to what you're saying. which is, it probably makes sense to think about it from a sort of lawyerly litigation-like perspective some of the time, but it also makes sense to think about it in terms of public politics. if you think about some of the

most successful in the, like broad scope congressional investigations in american history, so think about the 1920s munitions investigations, which almost certainly delayed american entry into world war ii by creating a large sort of public peace movement, and a movement that was skeptical of the war-making capability of the state. these are committees that were careful, but they were also highly cognizant aboutfyqñ the that their work just didn't face toward the executive branch, just wasn't inward facing to the congress but faced outward to the public. their reports were written, and their hearings were structured so as to convince members of the public to adopt a certain perspective. it's a reminder that facts aren't things -- facts are to some -- in some sense found, but in another sense they're assembled. there isn't a sort of situation out in the world that your job is just to find, your job is to

construct a narrative about the world and convince people of that narrative. that would be my advice insofar as you want the investigations to have real public punch. >> i've always experienced working with those people who were, you know, legends in the oversight, that they all viewed it as a staged process. you start with a problem, and try to identify it. of and construction relationships with the agencies that you're dealing with. at the same time. that's the importance of having long-lasting staff, staff that's still there, and going from case to case, and being credible as a

long-standing group, that information comes in, there's some attachment to the agencies that are being overseen. and before those kinds of relationships develop, it resulted in calls back and fo h forth, trying to avoid being in a particular situation suddenly showing up in the new york times, with the post, or anything like that. and working through the kinds of problems that agencies do have, in order to fulfill their objectives, and, you know, and the objectives of the president, as opposed to, you know, the sense that congress has vesting this power out. the stage process usually went from, you know, one level of

pressure to another. never starting out with a subpoena. a subpoena was, you know, for a long period of time that was there, it was a big event. and a subpoena, you know, scheduling a subpoena conference, to vote for one triggered reactions, triggered some of the negotiation necessary. and if a somebody was issued, that was a big deal. this is no longer there. subpoenas are more committees in the 114th congress have the authority to issue subpoenas on their own that was never there before. and it results in trigger-happy kinds of actions. and going forward without having the full, you know, facts before

th them. you know, cogo's organization, you know, it brings that forward. you know, and it's part of the public panoply there, giving information to make that kind of oversight process really work. helping with whistleblowers, providing -- you know, the background information, you know, that's necessary for going from one stage to the next. and i think that's important. i try to teach the people who are calling me, you know, read this, here's -- you know, it's been done before, and this is why it was done before. and we don't have that institutional memory anymore.

in the committees, or even in some of the support organizations that are there, like crs, and gao, which had been cut by the appropriations process, and can't keep the steady equals there. we've lost a kind of sense of how oversight is -- should be conducted in a way that is supportive, back and forth. >> another question? we'll get you a microphone. >> thank you. thank you very much. i work for the department of labor. i am an economist by trade, so not a lawyer. and just listening to this fascinating conversation, it seems to me like the executive branch, federal agencies are being asked to do more and more

with less and less. i don't know how much of this is a function of the elections, where more programs are coming online, but where are the resources to fund these programs? particularly in terms of human resources. it seems that civil servants are being asked to do a lot. hundreds of millions of dollars projects, et cetera. i know in the heat of getting this stuff well managed, it's hard to look at congress and oversight as something positive instead of seeing it something as detracting from what i really need to do. i was also very pleased in the last session to hear that resources from professor wright may be a problem affecting congress also. and maybe a lot of these things talking about institutional memory. anyone who's been on congress notices an army of unpaid staff. but any member i think would say that is the staff that is the backbone of how this institution runs. so there seems to be resources lacking as well, to keep people, to maybe do more meaningful

oversight over long term. so my question is, how much is resources at the end of the day, maybe something as simple as that, causing this score, where lack of resources has been a big problem and causes people to fight. and if so, how are you having these discussions to try to increase salaries, to try to hire more people, to maybe even promise less to the american people of what the government will do, so you can get to a better place where you can have programs that are effective, have the proper oversight and bring this together in a way that is less -- seems more on the legal side subpoenas and things, i'm just hoping never to experience. [ laughter ] thank you. >> steve, send him a subpoena. >> i'll be real quick. you're obviously raising big questions about the overall funding for congress and for the agencies. those are really big issues.

and beyond the scope of this conference i suggest. i'll just say that resources are an issue, and when there are investigations that are, you know, very broad ranging, and with lots and lots of document requests, and it appears to be clear from an administration perspective that this is really partisan driven, and not seriously driven by government to get a result, it adds to the resentment, and it makes the accommodation process more difficult. and i think probably elongates the process, and makes it difficult to have the professionals like you would have in a lawyer situation where you've got a tough negotiation, just get to the bottom line more quickly. so i think it's an exacerbating issue. the reality is, the administrations find the

resources that they need to get these things done. but when there's not a lot of respect on either side of the aisle -- not the aisle, the administration, the avenue, then the lack -- the resources issue exacerbates the problem. >> i think i agree it probably contributes. i don't really know, i don't have any empirical data. i suppose you are right there are resource issues on either end of this. i doubt there were adequate resources on both ends of it and everybody was happy with the resources. i don't think the problem goes away. i think it's a more, at the moment, at least, a more philosophically driven between the branches. >> specifically, having adequate congressional staff to do oversight, was one of the main driving factors of the 1947 congressional reorganization act, which did significantly increase the staff resources.

but one of the more sort of robust public polling findings is people across both parties think members of congress have too many staffers and think members of congress have an order of magnitude more staff than they do. people think there's this giant army of congressional staff. everybody in this room knows that's not true. and they want to see it cut. so at least since the '80s, it's been hard to put for the more congressional staff because of those intertwined misconceptions. >> there were more staff over the last two decades. it's shriveled, actually. pay has been cut. people don't stay enough. and that's important. incentive, you know, for having the institutional memory that

allows for, you know, more effective oversight, and, you know, and comfortableness between the branches, and on the hill itself. >> we have time for one more question. yes, brendon? >> brendon sawyer. i had a question. we've talked a lot about committees and subpoena power. if you had any insight into the role of individual members who aren't chairmen of the committee, how can they exercise the oversight? one thing i thought was in the senate, they used polls on nominees to get information wholly unrelied to that individual's -- the merits of their nomination. is it sort of going nuclear

category? >> it's not going nuclear, it's just conventional warfare. [ laughter ] and it's very effective. having been held up twice for many months, and both times i was confirmed. but don't tell anybody how effective it is, please. >> rand paul has figured this out, effectively, right? he did it with david barron, dci, he's done it with a lot of people. he's gotten in many cases a lot of what he wanted out of the administration. i think it's -- okay. i should qualify. when i say i think it's great, what i mean is i think it's great as sort of a way for members of congress to get what they want. i don't necessarily agree with paul's goals in those particular cases, but it's -- what i started out talking about, congress has all these different levers. houses have levers they can pull, individual members have individual levers they can pull.

morton brought this up earlier, leaking state secrets is one way members have influence public policy in really helpful ways, actually. so for example -- so, for example, senator gra vel leaked the pentagon papers to the press. in the '80s. there were all -- actually, in the '70s. a lot of what led to the creation of the church committee was leaks that came out of congress. in the '80s, leaks about the cia activities came again, out of congress, in ways that have been tremendously helpful. as long as the members are doing it on the floor, which in many case it has been, like henry gonzalez in the early '90s, up to the first iraq war, that is one of the constitutional tools they can use. >> yes, and you understand that the executive branch's argument with respect to the production of anything that might be sensitive is, we can't give it to you, because you will leak it.

>> but that's their argument no matter what happens. if you look at all the major leaks recently, whether they leak it or not, that's the argument the executive branch relies on. and you look at who's leaking the most harmful stuff is the executive branch. but they're going to use that argument whether it's true or not. >> next up, we have scott ream ream, the vice president of the project with a few remarks. before we get to scott, i want to ask you to please welcome me in thanking this group [ applause ] >> thanks very much. i have to say, i didn't think that was the discussion i was going to follow. in any event, i want to take two or three minutes to offer a few concluding observations on the basis of what i think were two terrific panels today. i want to start by taking a step back to comments senator levin made right at the outset.

an effective and well-functioning oversight system is critical to our democracy. i think we've heard several panelists echo, it's important both to congress and the executive branch. it should be important to all of us, not just immediate stakeholders in the process. and i think we've heard different folks on both panels sort of touch on some of what makes high quality oversight, and the kind of obstacles that stand in its way. i think it's worth highlighting a few of those things on our way out. particularly given the moment in time that ron white flagged on the first panel, which everybody's aware, congressional and presidential elections coming up with uncertain results, particularly a good time to reflect on all of these issues. so some of the characters of quality oversight, i noted that folks identified. probably the number one is fact based and not plitic di driven, that it's not partisan driven. it has sort of an objective

legitimacy to it. it's bipartisan, though josh might disagree that's an important condition. that it's in-depth, so that there's a mechanism for, if there's an investigation, for oversight to be ongoing. there can be follow-up. there can be regular monitoring beyond whatever the initial sort of investigation is. that executive branch folks, who are the subject of oversight, feel like they're being treated fairly, even if the process is adversarial. some of the challenges, or obstacles that i noted that folks raised. low quality oversight might have been the most repeated one. right? when it -- the executive branch feels, or there's some objective indicia, that the oversight is politically driven, or it's unduly burdensome in some other way, that it isn't bipartisan. again, josh excepting you out.

insufficient resources or capacity for committees and staff, for the members. is there just an inherent philosophical difference between the branches with respect to oversight. i think both of those sort of characteristics of quality oversight and obstacles that stand in their way suggests some potential conditions for facilitating better oversight going forward. and i think some of them are going to be obvious from what i flagged already. avoiding low quality partisan oversight, right? in part, i think senator levin flagged this at the outset, so that courts don't have to step in, and fix rights and responsibilities on both sides in ways that neither branch may find acceptable going forward. making sure committees have experienced professional staff with appropriate resources and training.

ways to build relationships between executive branch personnel and the committees who oversee them. maybe there need to be more opportunities for that particular staff turnover, if it's happening at a rate that it wasn't previously. and then i think one that i find really interesting that was brought up numerous times is that this sort of 80/20 problem. if 20% of the oversight is what the public sees, and a lot of the really sort of problematic, or what the public feel is the 20%, they don't know about the 80% that is working. is there a way to raise up the 80%. both so that people see that as more of a functioning democracy, and so that there can be lessons learned from the 80%. so again, flag this all because i think it's an important time to be thinking about it. as david said, these are not issues in which -- that we sort of dove into in-depth today, and

it's sort of beyond the scope of this conference. but i do think it's grist for the mill for the next one of these which i hope we can have. this one was wonderful. i hope the various stakeholders in the room, and tuned in remotely, you will think seriously about this stuff and the ways in which they can help facilitate more effective oversight in the next congress and administration to come. and with that, i want to, again, thank the levin senator, senator levin for hosting us. please join me in thanking all of them. [ applause ]

defense and intelligence leaders discuss defense innovation and outline the pentagon's third offset strategy, an initiative that seeks to identify next-generation technologies and capabilities in the military. later, remarks on defense secretary ashton carter. >> good morning, everyone. welcome to csis and to the third offset conference that we are holding today to talk about what the third offset is, assessing

its progress to date, the challenges to going forward, and what may be in store for this issue set that the third offset raises into the future. i'm so pleased to have with us opening the conference today the three of the architects, if you will, of the offset and implementation. collectively, they constitute something called the advanced capability and deterrence panel. and so i get to be the moderator of the panel, speaking today about finding the third offset. so let's begin with brief introductions. we'll have each of our panelists speak a little bit about the offset and we'll have a little conversation on how it's defined and how to think about it in terms of how it shapes government activity. all the way to my left is the honorable robert work. he's the deputy secretary of

defense. to his right is general paul selav, the chaerm of the joint chiefs of staff. to my immediate left is stephanie o'sullivan for national intelligence. let me begin with the main architect of the third offset strategy. deputy secretary work, i wonder if you can talk to us a little bit about, when you first conceived of framing this issue set around third offset, what is the issue set you were seeking to define, and the problem and operational challenge, and how do you see the third offset being the frame being helpful in implementing the department's activities? >> well, first of all, i'd like to thank csis for hosting us today. i really appreciate it. i can't really take credit for thinking of the third offset. i traced that thinking back to 2012, when secretary carter, who was then the deputy secretary of defense, established what is

called the strategic capabilities ochs. and it was motivated by the can exact same thing of what i'll talk about. that was followed by a very important presentation that secretary carter gave to the national security council on the growing threats and vulnerabilities to our space constellation. and so when i came in, the job of the deputy secretary, the primary job is to fashion a defense program that is in consonant with the secretary's strategic vision. that's what we do. paul and i chaired what is called a deputy management action group and we tried to make it a cohesive program. let me tell you what the third offset is all about in terms of fashioning a defense program. it is not a unified field theory, it is focused on one thing, and one thing only, conventional deterrence. it is designed to strengthen

u.s. conventional deterrence to hopefully avoid ever any major confrontation with any major state. it's focused on the operational level of war. think of that as the theater level or the campaign level. it's not focused on tactics. it's not focused on specific exchange ratios. it's really focused on having an advantage at the operational level, because from a historical perspective, especially since world war ii, having that advantage is the surest way to underwrite deterrence. everybody always asks us, offset for what? we'll talk about that, i'm sure, in the question-and-answer period. but the first place you start with is your pacing competitors. the competitors who are developing advance capabilities, and you want to start your conventional deterrence approach, focused on those pacing competitors. and the pacing competitors, not

adversaries, not adversaries, the pacing competitors are russia and china. because they're developing an awful lot of advanced capabilities that potentially worries. now, what are we trying to offset? in a case where we are projecting power across the oceans, most of the combat power of the united states is now resident on u.s. territory, either the continental united states, or states outside the continental land mass or in territories. that's different than in the cold war when we had a lot of forces forward, and the theaters they were expected to fight. finished the second thing is, offset strategies always happen when our potential competitors reach parody with us in certain areas. and our potential competitors had reached parity with us, and what we would term battle

networks, theaterwide battle networks. it's just a sensor grid in the theater, an intelligence grid that tries to make sense of what's happening and say what type of effects would we like to achieve. and then you have an effects grid that goes out and tries to achieve the effects and they say let's do it. and a logistics and support grid that keens the whole thing running. so both china and russia now have battle networks, theaterwide battle networks that are approaching parity with us. so to strengthen conventional deterren deterrence, we want to make sure we can extend our advantage in that particular area. the third thing is, both our pacing competitors have put a lot of money in counternetwork operations, because they know how powerful our battle networks are. so they spend a lot of money on

cyber capabilities, on electronic warfare capabilities, because our space constellation is an important part of putting the battle networks together. when you hear those three things, not a lot of forces in the theaters, they might be expected to fight, guided munitions and battle network parity at the operational level of war, and a lot of counternetwork operations, the shorthand for that, we refer to in the pentagon, as a 2 ad, the shorthand for the three objective facts on the ground. and so what is the offset then? that's what we're trying to offset. we're trying to improve conventional deterrence. we believe the best way to go about this on the initial vector, the initial vector, the third offset doesn't have a destination. this is as the vice chairman always says, it's a journey.

and our initial vector is to exploit all of the advances and artificial intelligence in autonomy, and assert them into our battle networks to achieve a step increase in performance that we believe will strengthen conventional deterrence. it's not about technology, however. it's about operational and organizational constructs based on doctrine, based on training, based on exercises that allows the joint force to operate with these type of technologies to achieve an advantage. it's also an institutional strategy. secretary carter's going to talk about that a lot in his presentation, about how we're organizing the entire department of defense to compete in this new dynamic environment. so i think that's where i'll stop, because i want to get to your questions. but again, the third offset is

not a unified field theory. it is focused on one thing, and one thing only. strengthening conventional deterrence to make sure wars don't happen. >> great. thank you, general. i want to get your perspective from the operational and joint combat side in terms of third offset and value as it's framed. >> you bet. i would make a couple of points to expand on what the secretary's already said. first is, the third offset isn't an answer, it's a question. it questions our ability to be able to offset advantages we see emerging in potential competitors forces. i always describe it as a journey, not a destination. if it was a fixed point in space, i would drive those requirements and i would impose them through the chairman on all of the services. except it's not an answer, it's

a question. and by asking the question, and repeatedly asking the question, what are the advantages that our adversaries are accruing over time, what threats do they impose to our field in forces, and can addressing those threats strengthen conventional deterrence, i think we're asking the right question. the way you take technologies and ideas and turn them into tactics, techniques, procedures and doctrines is through experimentation. that begins with designing concepts, testing them in war games and ultimately testing them in exercises. from an operational perspective, the journey we're on has the potential to vastly increase the effectiveness of our conventional forces. but we have to ask the right questions. we have to experiment with the right tactics, techniques and procedures. we have to disseminate those in doctrine to our field forces, to our partners, our allies and our

friends. and figure out how to offset this capability that all of our competitors are bringing to the conventional battle space. which is, in simple terms, long-range precision strike at volume. in space, in cyberspace, in the air, on land, and at sea. and we can sit back and say we invented long-range precision strike, and that would be true. but everyone who wishes to compete with us has read our doctrine. they've watched us in battle. they've analyzed our strengths to find asymmetries, and they're reflecting what we're good at right back on us. so we have to figure out how to offset that in the operational battle space. and i'll stop there, because i'm really interested in your questions. >> okay. miss o'sullivan, if one thing is not like the other on this

panel, you're not a department of defense employee. i would love to get your perspective from the intelligence community side about how you all have come to be partnered with the defense department and how you think about third offset. >> from the moment that the department started talking about third offset, it deeply resonated in the intelligence community. we share a world in which the threats are changing. and in which countermeasures are always advancing as well. eroding collectively our national security advances. so from the time that we deployed a new capability, and increasingly with cyber from the time we conceive of a new capability, our competitors -- i had adversaries, i changed it -- our competitors are working to counter that new capability. so in my view, from the ic, and for the ic, is that if we are not changing, if we're not driving our own offsets, we are

seeding ground. and we are losing our ability to inform policymakers from the oval office to war fighters in oval-shaped foxholes, if you will. so i just don't believe from the intelligence community in support of the department, or any of the policymakers that we support, that not changing is an option that we have. >> great. thanks very much. well, let me -- if i can, i hope to evoke a little bit of a conversation among you on this topic. again, you all see each other probably in various venues that we don't always get to hear the dialogue that happens. so we hope to get a little piece of that today. secretary work, if i could start with you. one of the things some people will say about the third offset is, all well and good to look at russia and china, but there are a lot of other issues going on in the world. the american public is concerned, for example, about isis, homeland security, there are many operational challenges

out there. what's your response to that in terms of the role third offset plays in the broader defense strategy? >> well, i knew this question was going to come. i didn't know whether it was coming from kat or the audience. this is constantly a question we receive across the department. so i actually have a couple of slides. [ laughter ] >> always good to be the straight man. >> okay. so i'd like you to look at the top part first. when we say we're injecting ai, autonomy into the grids, we're looking at five different things. autonomous learning systems. these are learning machines that can crunch big data, and can see patterns that humans simply cannot see. and they can reveal those patterns to the humans, and they affect the ic, i'm sure stephanie can talk about that, but they will increasingly impact operations.

human machine collaborative decision making is providing diffused information, advanced visualization, coupled with machine-to-machine communications with humans being in it, that allow humans to make more timely relevant decisions. and we refer to that as human machine collaborative decision making, with human at the beginning. it's not machine-human collaboration, it is human-machines. machines using humans to allow them to make better decisions. assisted human operations. this is providing as much information to the individual in the battle network as possible, that they can pull this information to allow them to make better decisions at every level. and it also is some physical assistance, such as xo skeletons, wearable electronics, disposable sensors right off of the soldier.

that's what we mean by assisted human operations. advanced human machine combat teaming, you see this all over the place with manned and unmanned systems working together. and the final are network enabled autonomous weapons, and high-speed weapons, like directed energy, electromagnetic rail guns and hypersonics. all of those things will be injected into the sensor grid, into the c-4-i grid and support grid, allowing a big performance impact. and again, it's not about the technology per se, it's how paul goes from the process saying here's the requirements, to the doctrine developers, who say this is how we will use this, to exercises in the field to train our forces to fight in a new way, that is what we are talking about on the third offset. so people always say, yeah, just

like kat said. why are you focused -- we're fighting isil on a day-to-day basis, why wouldn't you focus there? could i have the yankee slide. this is a white board -- this is a white board writing. in the bottom lower right, you'll see ya-1, code name yankee 1. it came off a mission in 2003. he stacked up his special operators on a specific target, and he said, look, i need to bring the entire power of the battle network to me, at this point in time, to accomplish this effect on the battlefield. and he sketched out, this is what i need. i need forward looking infrared, i need radar, i need sigit, i need emmett, and i need it all connected, and i need it connected to me. the five areas that we are doing

battle networks are scaleable from micro networks, completely transferable across the military operations. let me give you a concrete example. what this is is a learning machine that was trained to look across all social media. i mean, all social media. and that is the story of the image 17 shootdown. on the lower left, there's a twitter shot of mh-17 taking off. the next one comes from paris match.com. it is the picture of the russian sa-11 launcher with a serial number on it, date and time stamped, near the village where the shootdown occurred. then on bellingtap.com, exact same sa-11, exact same serial number, in the very same location. then there's a twitter shot of a

contrail of a missile rising at the time of the shootdown. then a rebel leader takes credit for the shootdown on vk.com. that was immediately taken down, by the way. and finally on youtube, there's a picture of the exact same sa-11 with a missile rail that is now mysteriously empty going back into russia. learning machines did this without any human interaction. if we had had this capability at the time of the mh-11 shootdown, we would have been able to prove, in my view, that this truly was a russian missile provided to the separatists in eastern ukraine, and they were responsible for the shootdown. this type of stuff will allow us new indications and warning in gray zone operations, the little green men. learning machines can say, there's an influence operation going on. we don't know who is doing it, but here are the key themes of the influence operations.

there will be new means of inw, new means of going after terrorists, there will be new means of operating against regional policy. there will be new ways of operating against great state powers. this is totally transferable across the range of operations. >> anyone else want to comment on that before i go on? >> let me give you a simple example, because everybody pushes back on this notion of ai, and algorithms helping us do what we do better. it's not a military example. i'm a horrible guitar player. but i play my guitar every night for an hour. you know there's an app that can take an algorithm and decode a song and give you the chord progressions. and lets you play along with the band. so my wife's new favorite party trick is to pull the app out, hand me my guitar, pick the title of the song i've never heard before, and ask me to play it for dinner guests.

which is a little bit of torture for them. [ laughter ] it's always a surprise for me. but what's interesting is this algorithm is 90% accurate at picking out the chord progressions on any song you put into your computer. it makes me better at what i do. all of you sometime today will interact with a piece of artificial intelligence that makes you do something a little faster, a little better, a little more efficiently. we have not to this point harnessed the capability of that part of our i.t. inventive and innovative community and applied it to broad military problems. we have in very narrow spaces. and part of what we're trying to do with this question that we're asking about the potential for a

third offset is to simply plant that question in the minds of the people who work for us. is there a better way? and if there's a better way, can you assist that operation by taking your intellectual effort and putting it where it is most value added and letting machines do the rudimentary work for you? and i think there's great promise in that space. but we have to be willing to take that step. >> so, as was just pointed out and you just illustrated, we're all swimming in a sea of information. and our analysts in the ic are charged with pulling out from that sea of information, indications and warnings in a world that has -- that pervasive instability becomes the norm. and what these learning machines and capabilities allow us to do is to see patterns, to do

sense-making of movement and imagery, to pull out -- triage documents or things in foreign languages so you can rapidly pull out the key pieces of information, or to see patterns that weren't there before in all sorts of information when you add it all up. above all, it would allow us also to enhance our trade craft, allow us to see the analyses, so we pursue that intelligence holy grail of not having bias in the asse assessments that we present. >> i think one of the other things that often comes up is this idea of the achilles heel, our strengths, our weaknesses, you know, everything in life can be a tragedy, the more we are information enabled, the more we are reliant on the information systems. we've had recent cases where the ic has come together to say, for

instance, in the most recent case that the russians are responsible for hacking. we've had hacking before. obviously for an operator in the battle space, this is a real concern that the military deals with routinely. how do you approach this issue from an intelligence community perspective with the degree you can leverage the artificial intelligence to enable everything that you do, while recognizing the challenges it can impose because it can be exploited? >> well, i have a technology background, and one thing i've known from the beginning, when i first started working in the intelligence community, is all technology is a double-edged sword. it's both an opportunity and a challenge. all of these things that we're talking about can present new opportunities. we've been talking a lot about biology. and the fact that that could

make huge advances for humankind. it also poses problems in inscrupulous hands, and challenges for societies. basically what's really hard for us in the business of warning about threats and what's going to happen in the world is that technology is changing so fast, that it's challenging and stressing governments and our society's ability to adapt to the change it's bringing. that's part of what's driving the pervasive instability that i was talking about. here's how you can use this technology to advance our capabilities, and here's what it's going to mean as a challenge to us, whether it's a direct confrontation by competitors, or just what it means to government's ability to handle the change that it's bringing. >> secretary work? >> from an operational perspective, we worry about this all the time, too. the best example i have is in

war gaming, over a long period of time, kind of two schools of thought developed. one school of thought says, look, your adversary is going to try to break up your network. it's going to try to sever the connections within the network. it's going to try to blind the network. so we take that as a fact. all of our exercises now are starting to inject what happens if you lose global prepositioning system data, gps data. what happens if your communications links are severed. what do you do. we're training on this all the time. but this were two schools of thought that developed. one was, look, you have to fight to keep your network intact. you have to try to keep all the connections there. and, you know, you really have to work on the network. then there was another school that said, fool's game. the network will dissemble under attack. what you have to do is train your force to be able to operate

with thin line communications, and thin connections between the force. and so that second school of thought is the way that animates us, operationally. and that's why we're training the force. force. and i know we'll get there, but that's one of our key advantages. our assumption is that a young man or woman who grows up in a democracy in the i-world will have an inherent advantage over young men and women who grow up in an authoritarian regime. we expect and rely on our people under mission type command to continue to operate. that's how we train the force from top to bottom. so we believe this is a very, very big advantage. >> let me give you a current operational example. two forces use the same kind of data links to do two different things. in western military forces, generally, we use data links to

network the force and allow people to see across that network what everybody else sees. so that broad view of the battle space allows our forces to collaborate and maneuver together. in a country that shall not be named, data links are used to issue orders to individuals. so central leadership believing that they have the best view of the battle space issues individual orders to individual portions of the maneuver element, using a similar system of data links. so in the case of the western network, it actually degrades gracefully. in the case of the competitor network, if you can shut down the network, their forces don't know what do. because they are conditioned to react to orders. so when you talk about the networks that we might build

into the future and the dependencies on vast arrays of data that move through those net works, we have to tend to the resiliency of the network and the resiliency of the force that subscribes to the network. so when we exercise without the precision navigation and timing of gps, or without the connectivity of the network element to element, we're exercising the maneuver elements within that force. typically that does not happen in authoritarian organizations that believe that central command is an absolute. now, can they adapt? of course they can. we better adapt faster than they do. part of the requirement that we see going into the future for these networks is both accessibility for our force and resiliency against attack. and the third part of that discussion is in the

absence of the network that resilient force has to actually disassemble and continue to operate. so we're going to work that. >> general selva, you've hit on something that has come up in secretary work's comments as well, how you institutionalize this. the department has a rich history, some very good examples, some maybe that didn't seem to pan out in terms of trying to drive innovation, particularly from the center. and then examples of course of innovation arising if you will from the bottom up. what is your assessment of how well the department today is organized and acculturated as a challenge for the future? if there are areas where the department could do better, it would be great to hear about that. >> as a general principle, i

would propose to you that we're not organized for innovation. >> yeah. >> now i'm going to refute that proposition a little bit. so where our organization, the entire organization, and that is, every part of the government that reacts to national security threats, is in fact organized to innovate, is when we see a compelling threat. when we see a change in the dynamic between the competitors who might become adversaries, and our current state of play. and we have identified that a threat to our ability to prevail in the battle space that threatens the stability of conventional deterrence. we've already talked about it, long-range precision, strike at volume. it's countries that would counter our ability to project power in an effort to preserve their view of their influence over their near/abroad.

that means holding us out of the pacific, it means moving us to the far reaches of western europe. so that we can't defend our nato allies. and those are conditions we cannot allow to come to pass. so when we see a common threat, we all move to figure out how to defeat it. and so in that respect, we are in fact organized to innovative. what we don't do is innovation on a micro scale. innovation on a micro scale happens in industry everyday. it is healthy. it is how industry makes step function changes in the services and the capabilities that they bring to us in the commercial market. i've used this analogy. it's like a brush fire in nature. it burns out all the underbrush. it fertilizes the ground and good things sprout and grow. it's something important as national defense and national security. innovation is treated like a forest fire. we bring out the fire brigades and we put it out because

innovation on a micro level tends to threaten institutions. part of what we've been able to do, the three of us, the advanced concepts of deterrence panel, is reach very deep down into the services and find common cause in this particular thread. and we have all of the doctrine development centers and all of the war gaming laboratories in the services spending a portion of their energy trying to decompose this problem and come up with new ideas. and we've even put money behind it, so we have a war gaming incentive fund and we have a war fighting laboratory innovation incentive fund and we provide grants to the services when they're ready to do experiments and when they're ready to do innovative war games. i think we have organized to innovate, but i will continue with my pitch that at a micro

level innovation can be a little bit unsettling for us, but at a macro level, at a strategy level when we're looking out into the future, our military has shown, our defense enterprise has shown, we're incredibly innovative. and we just have to reach for it. >> i want to give a shout-out to my uniformed battle buddy here. because a lot of time when we talk about innovation, we're talking about technology and how fast we can develop or exploit technology. but the third offset is all about operational organization. in 1975, the u.s. army said we are going to own the night. we are going to complete the entire transition of our force to operate 24/7, 365 to maintain a tie operational tempo and press the enemy 24 hours a day. 1975 anybody could have bought night vision goggles. they were kind of clunky, heavy. they weren't all that great.

it wasn't the night vision goggles. that was just enabling the army to own the night. it was all the tactics, techniques, and procedures to allow the squad leaders to control a squad quietly at night. it was the company commanders and battalion commanders, and in ten years the army owned the night. so the innovation i look for in the third offset is not how fast again someone can get a specific innovation into the fight. that's easy. but having the operators trust that innovation and knowing how to employ it is the key, and that is where we really do, i think, are pretty darn good at operational and organizational innovation. >> so i completely agree. it's not just technology. it's also how you operate things. so you could argue that for the intelligence community one of the first offsets was the overhead constellation and the

tremendous insight that that brought to our policymakers and war fighters having that initial capability in constellation set up. we evolved a way of tasking the at sights that was very much based on how precious they were, and we competed and we had much debated targeting decks. but what technology is giving us today and the new tactic we need to embrace is how we can do tipping and queueing via automated talking between -- across the constellation, and we can get much more capacity out of that which we already have than we do today, but we need to embrace the change and the way we allocate those resources which is a cultural and a bit of a challenging change for us. >> what i'd like to do now is open it up to the audience for some questions. let me give you some rules of the road. there'll be a microphone that comes around.

when we call on you, we'd like to hear your name and affiliation if you have one. the most important rule always involves the question itself. the first is, it should be a question, not a statement. and the second is it should be on topic. we purposely pulled this day together to have a focus on third offset and the issues around it. and i'm just going to move to the next question if it is on a different topic. with that, please raise your hands if you have a question you'd like to ask. i have one right here. >> jeff vialos, i'm a partner at a law firm. first of all, let me congratulate you all on the dii and the focus on particularly bringing technology to the war fighter, which i think has been well overdue at the department. secretary work, one of the key premises is technology has proliferated and essentially eroded our prior offsets. and allowed or adversaries to

take advantage of innovations they've brought. in that context, what is the basis to assert that autonomy, a technology that looks to me like it is going to be pretty ubiquitous could conceivably be the basis of our next offset? all you have to do is drive the next tesla and you'll see autonomy. i don't mean to be glib, but i understand how autonomy would be one element of a future force. but what's the basis when giving a sustained overmatch based on that? >> well, first i'll start with your last point. we are not arguing that we will have a sustainable overmatch. this is a very dynamic environment. two examples. the rifle, telegraph, railroad revolution, and war. ai and autonomy are very much like railroads and telegraphs. they were being driven by the commercial sector. it changed our society, and as a result it inevitably changed the character, not the nature of war.

ai and autonomy is changing all of our lives. as the vice chairman said, at an astounding rate. it's happening so fast that sometimes we don't even notice it. this is going to be a world of fast followers. what we have said, jeff, again is it's not so much ai and autonomy. it's injecting it into our battle networks and allowing our battle nets to work better than our potential competitors. and we believe we do have an advantage in ai and autonomy on the operational level of war at this time. we think we might be able to have an advantage for some time if we move to operational or organizational constructs, but we're organizing it ourselves temporarily. we may only have an advantage for five years. so you better be thinking about the advantage you want to create in the next five years and the years after that.

we're thinking like a competitive business where the market is constantly changing and you're having to adapt. again, we know that ai, and autonomy will vastly improve the power of our battle networks. we know that our competitors will probably conclude the same thing. in a world of fast followers, as long as you're a fast leader you have an advantage, but you always have to be thinking about what happens when we achieve parity and what's the next step. >> i also think we need a broader view of what autonomy is, just to pick on your point of cars that can drive themselves. that's interesting, but it barely scratches the surface of what could happen if every car was part of a network. if every vehicle on the road subscribed to a network that optimized our traffic, my guys wouldn't have had to go to lights and sirens to get here on time. >> i will say washington traffic

is the best anti-access area denial of a network in the work. >> the reason i pick on that point a little bit is, having spent a lot of time with people who think critically about artificial intelligence and autonomous automation, every one of them would tell you that we are barely scratching the surface. we are now only beginning to learn the promise of these potential technologies for the future, and that's why i say this is a question, not an answer. if we locked into systems that are not adaptive, that can't continue to bring in new information and new networks, if we don't truly build resilient open architectures, what we'll doom ourselves to is a fleet of singular autonomous things. and we won't be able to adapt. so at the heart of your question is of course autonomy is going to evolve over time and we have

to evolve with it. the current advantage we have in the battle space is born of technologies that are 40 years old, bonn of ops concepts, tactics, and procedures that while they have evolved are largely 40 years old. we're not going to be able to sustain any kind of advantage over potential competitors in an environment where software applications and machines allow organizations to adapt quickly, so we really have to get to the heart of your question and sort out how deeply we're going to look into this problem, how far we're going to predict our capabilities are going to be, and how adaptive we're going to make the architecture. for those in industry in this room, i can't tell you how many times i've asked the following question. will your widget subscribe to an open architecture? answer is always, yes, sir, of course. because the j-rock says we have

an open architecture, so we'll have an open architecture. i say, wonderful. here's an application i would like you to put into your widget to make it more useful to the military, to the service that ordered it. oh, sir. we can't do that. it is an open architecture, but only inside of our company. or only inside of our proprietary ip that's in the system. we really have to unlock the potential, and the only we can do that for ai and autonomy is to find a resilient open architecture to which all of our systems can subscribe, and we've only scratched the surface on that as well. >> i'd just reinforce that by saying that asking that question is like asking the question the way that machine, when arpanet first came out. this is kind of interesting, but what is this going to do for us?

we're right at the beginning, and we can't see all the possible places that it can take us, but we need to remove the barriers to allow it to be explored. i guess the valley calls it vast failure, but for us it's fast learning, see what this thing could mean, explore it, and then move on if it doesn't -- if it's the brush that should have burned down. >> great. okay. next question. i have one right here in the middle. there's a mike coming, hopefully. >> thank you. dr. ted johnson from the deloitte center for government insights. my question is the first offset basically benefitted from the national security act of '47, the second from goldwater-nichol reform. what's the statutory reform coming that will help realize all the advantages in the third offset? >> i don't think we have to

reorg to do this. i would challenge your premise. i would hate to do that in a group this bright because i'm not the brightest guy in the room, but those offsets were not just the product of changing our national security policy apparatus. in fact the policy apparatus changes may actually have been a result of the things that were happening around us. so i would suggest to you the big part of the first offset is we found a relatively inexpensive way to miniaturize nuclear weapons. and allowed them to be used across the breadth and depth of the battle space. that actually posed some very serious questions for our national security leadership. in the mid-'70s, we realized that precision would get us to a different equation in that same battle space. by the way, some of our competitors say we create nuclear effects with conventional weapons because of our precision. that is so wrong. we create the operational

outcomes with precision munitions that our competitors ascribe to nuclear weapons. there's a huge difference in the way you say that and what it means. that means we don't have to deploy all that fire power in the battle space in that configuration. by the way, it wasn't just precision. it was stealth, precision, and a sensor network that gave us indications and warning to make that capacity useful in the battle space and we reorganized around it. those were the seeds of airline battle. we reorganized concepts around it. we didn't reorganize our services around it. the compelling issue with goldwater-nichols was we had these services that were very, very capable of working together, but we didn't have an organizational construct to encourage that. so goldwater-nichols was born.

i don't think we're making that big a shift in what we're doing that causes us to have to reorganize the national security apparatus of the nation. are there places where we can trim and tweak on the edges? yes, but last night i find myself describing the defense department as a diamond. it's already been cut. if you polish it, it's only going to make it brighter. and what we're doing right now is polishing the diamond. >> okay. i have a question right here. >> i'm from the swedish ministry foreign affairs. first of all, thanks to csis for arranging this discussion. where do allies and partners fit into this equation? and how do we solve long term interoperability without giving away the goodies to the conventional exerts?

competitors? >> i will just free advertise to say we have a whole panel on allies and partners. we welcome your guidance to that panel. >> we have many of our close allies here, norway, united kingdom, japan. there are many others, and this is what we would say. the first advantage we believe we have, the competitive advantage, going back to the ai question in this world is we believe our people within the framework of what we're trying to accomplish provide us an enormous competitive advantage. second thing was talked to by paul. jointness. jointness is hard. it takes all of the western armies and our armed forces to a greater or lesser degree are going more towards joint solutions. why? because your battle network, if you have discrete functional networks, it will never operate as well as a cohesive joint network, so we're very -- 30

years since goldwater-nichols, so we have -- and our competitors are trying to copy us in this regard. the third thing you hit on is our allies. we believe in our national security strategies that alliances are absolutely central to not only our security and our partners' security but to global security. if you compare us with our potential pacing competitors, they don't have a lot of allies. we do. so from the very beginning, we have been talking with our partners and our allies about how we work this together and the way we describe the third offset is very coalition friendly. the second offset the coin to the realm was a mechanized infantry battalion or an armored division or a heavy artillery battalion. now anybody can come up with an application, an ai or autonomy, in any one of the domains and it improves the power of the entire network. so this is very coalition friendly. we're thinking of it from the very beginning as interoperable,

exchanging data from machine to machine, discussion to discussion. so far all of our discussions have been very fruitful. >> intelligence sharing is such an important piece of this. i wonder if you can address that. >> well, clearly, we're in the midst of what director clapper brought to the ic, which is integration. and i would argue that that's our current concept. integrating all the pieces that we have, fully leveraging what we have, and working deep with our partners is foundational to the advantage we've gained, that we can pool together what we already know. as far as the risks involved, sure, but it requires the additional work to think about what we're sharing. it's not an open kimono everything. that's foolish. but it's thoughtfully engaging

in how can we best leverage and help each other. and we're doing all this structural work to make that possible. i won't go into the details, but it's largely in the way we're structuring our i.t. systems. it's tagged and labeled so we can more easily share it. >> okay. other questions. all the way over here. there's a young lady in the second row back, yes. >> good morning. thank you again to the panel. trish martin from the office of naval intelligence, ops intel center. so there are a lot of pieces to the puzzle you're putting together in regard to innovations in the defense department. some of the resources i have observed is how you articulate your return on investment to doing better or to doing faster or an increased level of fidelity with the data points you have. how do you articulate return on investment now and how do you keep kind of the wind at your back when the pentagon doesn't

actually have a competitor where the war fighter can go for a different level of service if they're not getting what they need? >> i think you'll hear from secretary carter in just a few minutes on how he has tackled this problem as far as the entire innovation agenda within the department. but let me just say the way we'll determine return on investment is through the operators. the second offset took life when the army and the air force got together and conceived a battle, which employed the technologies. and within a very short period of time, by 1984, the soviet general staff said this has completely unhinged the way we were thinking about the fight. so what we will do, paul talked about a war gaming incentive fund, where we incentivize to

look at new concepts where we look at technologically operational constructs. then we say it looks like an operational return on investment to do this. then we incentivize the concept developers and the doctrine developers to develop it further. then we do exercises. it will be the feedback from this loop on how it improves the performance of the joint force that will tell us the return on investment to tell us we really should go this way. we are making modest investments, a lot of levers, a lot of tests, a lot of demonstrations. i've stole paul's line. this is a journey. it's not a destination. so the return on investment for us is going to be when the army, the marines, the navy, the air force, our allies and our special operators come back to us and say this is the real deal. that's where we'll really start to pour our money. >> this is about the force taking ownership of the question. i'm very, very, very suspect of

hard objective criteria that we can advertise as success. and i'll just make up an example and you can push back if you want. ddg 1000 is an incredible ship. she's stealthy. she's smart. she's networked. she's resilient. and she's ten times more lethal than any competitor's ship of the same class, to which somebody who just crunches numbers for a living would say you wanted 20, so 2 will do. we have a history in the department in the services of taking the subject of outcomes of war games before the services and our soldiers, sayers -- sailors, marines, have taken ownership of the

tactics, techniques, procedures, and doctrine and advertised them as success and it hasn't been fruitful. so my view is i think the same as the secretary's. when i have soldiers, sailers, airmen, and marines echoing back to me in their work spaces, in their units, in their maneuver elements they're seeing the benefit of this kind of thinking, then we win. and that's where the return on investment comes from because we can experiment and knit things together in different ways. it'll be viewed as a bright idea unless it makes a difference on the deck plates and that's where we have to go. so i would use the same measure when we have articles in our professional journals, when we have commanders responding, when we have the doctrine centers and the war gaming centers putting in the exercises and bringing us the ideas of the young men and women in uniform. that's the return on investment. >> and the best example of this is, we would not have gps today

had bill perry and harold brown had listened to pa&e. they said, the return on investment for gps is hard for us to calculate and therefore you want to kill it. bill perry was in a helicopter and he was in the middle of a kind of really dicey whiteout/brownout situation. that helicopter just happened to have an experimental what was then called -- it wasn't called gps at the time. the pilot landed and perry said -- asked him, and the pilot said, i did it all by gps. what is the return on investment being able to time sync an entire battle? it is hard for us to tell you. when we get into operations like desert storm, and we say, whoa, being able to time sync everybody and being able to do guided munitions attack 24/7 through

the weather, the return on investment is as paul said, when the theater commanders come to us and say this is how we will win in the future. we have to be careful not to say i don't know if you're going to get a six by return on investment on this. it's all about operational and organizational constructs. >> i'm afraid we're going to have to make that the last word. this has been an incredibly interesting and helpful panel. in helping to frame up the rest of the day. we do have secretary carter following this. after i thank the panel formally, i will ask everyone to stay where they are and not get up and get coffee. we will switch over very quickly here. i do want to thank our panelists for taking time out of their very busy calendars today to come here and help frame up for us what we hope will be a very fruitful discussion about the third offset. please join me in a round of applause for our panel. [ applause ]

we've been a little bit delayed because of the traffic. i would give you a 30-minute introduction or a 30-second introduction. i've known secretary carter for 30 years. i've marvelled at his acts. there's no one better positioned to lead the department right now especially on the question we're exploring today. he's doing wonderful work across the board and just came back from a grueling trip. he's not going to take questions from the floor but i do have questions on your behalf and if

i don't cover them, yell at me later. please welcome secretary of defense ash carter. [ applause ] >> thanks, john, for that introduction. and more importantly, where did john go, for your many years of service, many years of friendship to me. wonderful service to our country over so many years and leadership of this great institution. i also want to thank csis for hosting this important conference. and i'm going to commend my deputy secretary of defense, bob, generally paul selva, guys, thank you for holding the fort down, for their leadership and hard work in leading the technology investment we call the third offset strategy. i'll speak about that. but of course in this speech i also want to speak about an innovation in all its dimensions of which technology innovation is a piece, a very important

piece, because being more innovative in every way we can is critical to the future success of our military and our defense department. today, we have the finest fighting force the world has ever known. there's no other military that's stronger, more capable, more experienced, or frankly more innovative. that's why our military edge is second to none. and it's a fact every american ought to be proud of. but it's also a fact that our military's excellence isn't a birth right. it's not guaranteed. and we can't take it for granted in the 21st century. we have to earn it again and again. that's what this is all about, innovating to stay the best. i want to talk to you today about how we're doing that in some different areas. our technology, our operations, our organization, above all our

people. right now it's imperative that we do so, because we live in a relentlessly changing and fiercely competitive world. there's the faster pace of change which sets up a fierce competition between the present and the future. competition with other nations, not only with us but also with each other. and competition with terrorists and other mal malefactors for w we are the game to beat if they can, even if only at one place and one time. technology is one example of such change and competition that many of us have long been familiar with. when i began my own career in physics decades ago, most technology of consequence originated in america. and much of that was sponsored by government, especially the department of defense. today, we're still major sponsors. but much more technology is

commercial. the technology base is global. and other countries have been trying to catch up with the breakthroughs that for the last several decades made our military more advanced than any other. and much of the frontier innovation is commercial, leading to additional sources of competitive dynamism outside our five walls. against this background, your defense didn't is confronting a world security environment that's also dramatically different from the last generation. and even the generation before that. indeed, the u.s. military is at this moment addressing five major, unique, rapidly evolving challenges. we're countering the prospect of russian aggression and coercion, especially in europe. we're managing historic change in the asia-pacific, the single most consequential region for

america's future. we're continuing to address middle eastern nuclear developments. we're focusing on protecting our friends and allies in the middle east. we're accelerating the certain and lasting defeat of isil, destroying it and its parent tumor in iraq and syria and everywhere else it metastasizes around the world, even as we help protect our homeland and our people. and at the same time as all of this, we're preparing to content with an uncertain future, ensuring that we continue to be ready for challenges we may not anticipate today. we don't have the luxury of choosing between these challenges. we have to do them all. and as the world changes and complexity increases, we'll have to change too. how to invest, how we fight, how

we operate as an organization, and how we attract and nourish talent. as we do, we have to be able to move fast, because the advantageo advantages we expect to derive from each innovative cycle today will not last as long as they used to. all the commercial and global change that's occurred across the technology landscape has made repeated and rapid cycles necessary, and made high end tech a lot more accessible to competitors. think about it. while the cold war arms race was characterized by the in ekex or able improvement in strength, there are additional variables of speed and agility, such that leading the race depends who can out-innovate faster than everyone else and even change the game.

in the area of investment, it's no longer a matter of what we buy. now more than ever it matters how we buy things, how quickly we buy things, whom we buy them from and how rapidly and creatively we can adapt them and use them in different, innovative ways, all this to stay ahead of future threats and future enemies technologically. that's why i've been so intent as secretary of defense not only to plant the seeds for a number of different technologies that we think will be determinative in giving us a war fighting advantage in the future, more on those in a moment, but also to be more innovative and agile in all aspects of dod, in our operations, in our organization, and in the talent management of our all-volunteer forces. in each of these four areas, i along with the chairman and vice chairman of the joint chiefs, the service chiefs, all our excellent combatant commanders and the defense department

civilian leadership have had a lot of help. we've had help from washington think tanks like csis, from our defense labs and industry partners, and also from many innovative americans who understand the innovation imperative and who aren't in our community now, but understand the need for our mission of national security and want to help. and all of us have been pushing the pentagon to think outside our five-sided box and invest aggressively in innovation. and i want to focus on that in the rest of my remarks, the clear strategic imperative we have to innovate in each area, how we have been innovating so far and how we need to innovate going forward. given the topic of this particular conference, i'll stark with technology. the strategic imperative to innovate technologically is well-known, to those who have been paying attention, many of you here at csis. nations like russia and china

are trying to close the technology gap with the united states. and as i noted, high end military technology is diffused, sometimes becoming available to countries like north korea and iran, as well as non-state actors. at the same time, our own reliance on satellites and the internet has grown, creating vulnerabilities that our adversaries are eager to exploit. to stay ahead of these threats and stay the best, we're pushing the envelope with research and development in areas like biotech, electronic warfare, robotics, artificial intelligence, machine learning, and much, much more. and i'll repeat yet again, since it keeps coming up, that when it comes to using autonomy in our weapons systems, we will always have a human being in decisionmaking about the use of

force. now, we're making some serious investments here. just to remind you, the latest budget we've proposed, a budget i strongly encourage congress to pass when they return to washington next month, will invest $72 billion in research and development in the next year alone. that's more than double what apple, intel, and google spent last year combined. this budget marked a strategic turning point for the department of defense. the third offset strategy driving a wide range of new, innovative technological investments in order to advance and sharpen our military edge. we're making these investments because we aren't yet exactly certain what are wear thor wher of offset is going to come from. it could be one area of technology or several.

remember, previous offset strategies were general rangesal successes. reflections of the security environments of their eras, and only recognized as such after the fact. today speed and agility are key. and because the world we live in, the next offset will not look like the previous ones. it may not even end up what we might consider a traditional offset strategy at all. that's where we're seeding these investments in lots of different technologies, so we can see which they germinate, how they can produce, and how to use them most effectively. in addition to these critical investments, it's important to note how dod is innovating technologically how we're innovating tech nonologically, developing technology from within, bringing in technology from without, and repurposing technologies and capabilities we already have, because different entities are focused on each. within the defense department,

we have dozens of dod labs and engineering centers across the country, each one home to great technologically innovators, both civilian and military, who work closely with very innovative defense industry that's long supported us and kept us on the cutting edge. and they're continuing to do so today across a wide range of critical technologies. for example, our navy labs are developing and prototyping undersea drones in multiple sizes with diverse payloads, which is important, since among other reasons, unmanned undersea vehicles can operate in shallow waters where manned submarines cannot. also our army labs are working on gun-based missiles defenses which can help defeat incoming missile raids at much more cost per round than intercepters, imposing higher cost on the attacker. in our air force labs, we're

developing hardware, software, and systems inspired by the working mechanisms of the human brain, which offers the prospect of overcoming limitations of current computer architectures and enabling superiority in air, space, and cyber space. as i said, america's innovative defense industry is a key partner in this. because remember, we don't build anything in the pentagon. that's not the american way. the soviet union tried that, and it didn't work out very well for them. today, with more technological innovation happening in the commercial sector, we need to be able to identify and do business with companies outside our traditional defense orbit as well as those within, and welcome them into our defense technology community. that's why last year i created our defense innovation unit experimental or diux, to help build bridges and startups and

other commercial technology firms located in innovation ecosystems across the united states and help us more quickly adopt technologies that can help our troops accomplish their missions. diux opened its doors last august with a west coast office in silicon valley. and since then, we iterated and launched diux 2.0 in may and opened a diux east coast office in boston and established an outpost in austin, texas. one important area where diux recently solicited proposals was in microsatellites and advanced analytics. leveraging the revolution in commercial space and machine learning to transform how we use space-based tools and advanced data processing to provide critical situational awareness to forces around the world, and also have added resilience, by the way, to our national space architecture. meanwhile, under the guidance of the