Media Roundtable with Dawn Meyerriecks Asst. Deputy Director of National Intelligence for Acquisition, Technology & Facilities

Media Roundtable with Dawn Meyerriecks Asst. Deputy Director of National Intelligence for Acquisition, Technology & Facilities

Wednesday, 29 May 2013 10:25

.Media Roundtable with Dawn Meyerriecks
Asst. Deputy Director of National Intelligence for Acquisition, Technology & Facilities
Location: Liberty Crossing, McLean, VA
Date: Wednesday, May 22, 2013
Time: 1630

MR. MICHAEL BIRMINGHAM, ODNI PUBLIC AFFAIRS: I’d like to thank everybody for joining us this afternoon. Sorry we’re starting a little later here. It’s been a farewell day for Dawn around the building today. As you know, she’s departing. She’s heading down to CIA to be deputy director for S&T shortly.

I wanted to talk a little bit about ground rules before we get started. As I said in the email, this is mostly on the record, but you can expect from time to time, if necessary, just to discuss a little bit more information maybe we’ll have to go on background. I don’t think we’ll need to go off the record. We’ll just say, sorry, I can’t answer that one.

Q: I’m sorry; if it’s on background, then how is it attributed?

MR. BIRMINGHAM: It will be attributed as a senior U.S. intelligence official, OK, if necessary. But for the most part – and I’ll let Dawn state her title and spell her name for you here in a second – It will be attributed directly to her and her position.

The briefing is focused on IC-wide programs, and –

(Cross talk)
Holy cow.
(Cross talk, laughter.)

MR. BIRMINGHAM: And we’re still on the record.

DIRECTOR JAMES CLAPPER (Enters): We’ve been getting all these – what are you getting up for?

DAWN MEYERRIECKS: I was going to let you sit. (Laughter.)

DIRECTOR CLAPPER: I’m going to stand.

So you’re getting the lowdown on what life is really like here, huh? Is that the deal?


Well, I just wanted to stop by.

First, thanks for coming out here. And your time is well spent talking to Dawn. And, you know, big loss for us but big gain for where she’s going. And she’ll continue to do great things for the community.

Are you writing something down?

MS. MEYERRIECKS: They asked me to tell them what my title was and I was trying to figure out what it was. (Laughter.)

DIRECTOR CLAPPER: Oh. Over there or here?


DIRECTOR CLAPPER: Oh, it’s a long title. (Laughter.)

Anyway, Dawn is – I’ll just tell you, just as a testimonial – has done a superb job with, you know, organizing and improving the health of acquisition across the whole IC. You know, we have 38, 39 maybe, systems acquisition programs, and she’s done marvelous work with the agencies who actually do these things – set up a great tracking system for me so I can sort of keep my pulse on – and as well – and I’ve found, in my experience, a rare person who understands the technology, can think about the future and also sort of using the now and next, after next construct.

You know how to serve a boss, I’ll tell you – but that time frame and, you know, managing day to day, what’s next in the queue, what’s going to be after next – the next, and then the after-next with R&D. And I just want to say Dawn has done a great job here and we’re going to miss her a lot.

MS. MEYERRIECKS: And I have a great boss. Mutual admiration society.

DIRECTOR CLAPPER: So what else? (Laughter.)


DIRECTOR CLAPPER: He’s going to pull me out here.

MR. TURNER: Yeah. Yeah, I am. (Laughter.)

DIRECTOR CLAPPER: Anyway, continue.

MS. MEYERRIECKS: Thank you, sir.

MR. BIRMINGHAM: Yes, thanks.

MR. TURNER: We’ll go ahead and go back up. I just wanted to bring the boss down to say nice things about Dawn since she’s leaving.

DIRECTOR CLAPPER: Which is easy to do.

MR. BIRMINGHAM: All right, thanks.

MS. MEYERRIECKS: Thanks. Thank you.

MR. BIRMINGHAM: Continuing here, as I was saying, this is – generally we’re talking about IC-wide programmatic issues. We’re not going to be doing any deep dives in any particular agencies. Those aren’t our programs specifically. And as I mentioned, it’s already been reported that Dawn is going to CIA. And to the extent that you know that she’s going there and you know what her job title is, we’re not really going to discuss anything in her future assignment, OK?

So with that I’ll turn it over to Dawn for about, I don’t know, five, eight minutes or so.

MS. MEYERRIECKS: I get a head start, that’s all.

MR. BIRMINGHAM: That’s right, you get a head start, and then we’ll get to catch up with the Q&A. We’ll have plenty of time for Q&A, about 35 minutes or so.
So, Dawn, go ahead.

MS. MEYERRIECKS: OK, so Dawn Meyerriecks – M-E-Y-E-R-R-I-E-C-K-S. I do follow directions well. And I think the “A” is assistant – associate. I’m looking. Assistant? See, none of us are –

MR. BIRMINGHAM: It’s assistant.


MS. MEYERRIECKS: Oh, thank you. Assistant director of national intelligence for acquisition technology and facilities. So you see how seriously I take titles. I probably should be more aware of that. OK, so that’s the homework I was given.

So I just want to talk about S&T for a second, or maybe a couple of minutes, and where we are from an overall IC perspective. And then I’ll take questions. And to the extent that we can stay on the record and I can answer them, I’m happy to do so. Those of you that have worked with me for years know that I’m pretty upfront.

So from a community perspective, even in face of sequestration and budget reductions, the three priorities that the leadership established, the first one was people. And you can see that in some of the recent discussions about furloughs and things like that, and the boss’ put his money where his mouth is.

The second is cyber, of course. And number three, every year that I’ve been here has been preserving our R&D investment. And that doesn’t mean that we preserve absolute dollars, but it does mean that as a community we have preserved the percentages invested in relation to top line. So I think that’s a pretty strong statement if you think about that, because our sister organizations, other departments for example, have not taken that position. But I’m very proud of the leadership team here because they get that fundamentally we are a technology company, and that is what enables mission.

And I’ve been places where that was not well understood and we paid the price. We had seed corn and then, you know, got caught short in various ways. And of course anytime the boss talks about how great we’re doing on all our SMAs, I kind of, you know, want to say, yeah, so now something will implode in a horrible way. But right now we’ve been executing the heck out of our science and technology portfolio in the time we’ve been here. And that’s not serendipitous. We work at that really, really hard, and we can talk about that if you want.

With respect to AT&F – and I talk fast, so this is my head start and then we’ll go wherever you want to go – if you think about our role in the DNI but also in the IC, we are responsible for oversight of materiel procurement across the board.

So when the boss first talked to me about facilities, I was like, uh, wall sockets. I don’t know anything about facilities per se. But it made a lot of sense as we got into it. For example, when we started there were over a hundred facilities, either improvements or acquisitions going on, and there were a large percentage of those that had no provision for IT infrastructure. We don’t even do loading docks anymore that don’t have some sort of capability.

So when he first told me about it I was bemused, and then as we worked through it, it started to make sense to me, about why you would actually put those things together. So if you think about it as all the materiel solutions for the intelligence community, we don’t build things generally. There’s an exception and I’ll talk to that, but we do oversight and we try to figure out how best to leverage from a community perspective so that we’re not – I don’t mind duplication in R&D because if we knew what the answer was, we wouldn’t call it R&D.

But to the extent that we can keep people apprised of what they’re doing, you know, what common communities are doing so that we make new mistakes – I like to think that R&D is about making different mistakes than the last time – that’s really what we try to facilitate, and then figure out how we get innovation into mainstream acquisitions – because generally the rub about those is that you can’t be innovative in doing a major system acquisition. So we worked that really hard. And I am a technologist first. And actually I’ll just tell a funny little story. I think it’s funny anyway.

When Director Blair called me, I laughed, and I said I left the Department of Defense because I hated acquisition. I got to first milestone A, and CES (cost element structure) – for those of you who have been covering it for a while – but it was the first attempt to do enterprise services. It was based on an SOA (statement of agreement). And it took about 30 months. And it was – I likened it to the Flemish panel of judges. Nobody really understood what I was doing but everybody had a better plan.

And so I got to milestone A and I thought, well, we need the program but it will never get killed because we’ve got a milestone A. And that’s when I got the call from AOL to come run product strategy for them. So I laughed when he called me about it and I said, I really don’t like the oversight process. And he said, well, I know you know how to deliver, and I figure that the fact that you don’t like the oversight process is probably a good indication of why you’re qualified for the job. And I said, well, OK, on that basis we can actually have a conversation. So I brought that sensibility to an oversight role.

The one exception to oversight is that IARPA is probably actually part of the organization and I can’t talk about the top lines in general for that. If you think about it as kind of the kid sister to DARPA in terms of scale and volume – and that’s about right. And we do very little basic research here. We do very much – it’s very focused on applied because that’s where our mission generally is. There are some exceptions to that, and we might get into that because if you want to talk about that, but we are definitely interested in leveraging what other folks are doing.

My personal philosophy is that we, the U.S., win because we are great at complex systems engineering. We can put things together in ways that nobody else can on the planet. We also have pretty good science and technology too, but there’s a great quote that Marc Andreessen said, is that you can’t steal your way to being innovative.

So our adversaries who think they can steal our IP (Intellectual Property) and beat us, they replicate the same mistakes we do, and they don’t actually know it well enough in general to improve on it in substantive ways or to understand the mistakes that we made. And, you know, my engineering principle – because I’m a working engineer – is that you do it – you know, the first time it’s not great, and the second time it’s better, and the third time it’s actually right.

So stealing our IP, sometimes that’s the biggest setback they’ll ever have, because if we’re early on, that’s actually great. We don’t want them to do that but I think that actually we win because we are really good at thinking outside the box and re-examining first principles and figuring out how to do things.

I think IARPA is a great bellwether of what it is that we’re investing in as a community. And 34 of our 38 programs are unclassified. We try to do unclassified. We believe that the talent is global these days, and that’s just intrinsic. We have one of our major initiatives unclassified; 70 principal investigators, only 17 are U.S. citizens. That tells you the scale of the talent that’s out there, and we recognize that.

And the way – you know, we’ve got to be all over that, and we’ve got to leverage the global marketplace in order to best figure out, you know, what the next steps are. So that’s a big deal for us. And that is countercultural here, and I’m very proud of IARPA, that we continue that tradition. And we’ve done some amazing things there. And we can talk about some of those if you’d like.

One of the things that I think we did very effectively for the community in the 3 ½-plus years that I’ve been here is we actually got people to come together on where we were going to make tech investments, and we came up with kind of a simple little model. Unfortunately, the lowest I could get it classified was Secret-NOFORN, but there are parts of it that I can talk to that aren’t classified, so I will.

We divided things up into things that we either need to lead or influence, and then adapt or adopt. So adapt or adopt is write checks and generally make work in our environment. The lead/influence is we actually need to own the IP or understand it very thoroughly.

And the influence piece – and I’ll give you an example of what I mean by “influence.” So human language translation is one of the things I can talk to, obviously. We don’t have to build all the human language translation capability on the planet. There’s lots of people that are doing that. It turns out, though, based on some really good research – and IARPA has done some in this space as well, but not the only – not the only ones – you can inflect, based on the inflections but also the kinds of nouns and verbs that are used, who is actually the leader in a group as opposed to who is at the top of the work chart.

So we don’t have to do the actual tool that does the human language translation, but we have to ensure that we understand the IP well enough to know that we’re getting the inflections that are important to understand those sorts of things, to suss out, OK, so got it; this person shows up at the top of their tree, but this individual over here is actually the spiritual leader, or the technical leader or whatever, based on the way they’re addressed or the inflection of the voice during conversations. So that’s an example of we don’t have to invent the technology but we have to actually understand the intellectual property well enough so that we know that we’re getting the nuances that we need.

So I have a list and I’ll rattle it off, and we can come back to it. But lead influence, the unclassified part of that list that I can talk to is exploitation, video and motion imagery, big data, human language translation, trust in software and platforms, networks, and then high-performance computing.

The adapt/adopt is the cloud, visualization, cognitive systems and knowledge management. So that’s kind of just – and the good news is, is that we got the community to look at that and created basically communities of interest and computing use of practice. Again, I don’t mind duplication in R&D. I actually think that’s very healthy. Competition is a good thing. It’s a very American thing. What I like to do is have those people talk to each other so that we’re leveraging each other to the extent possible.

And I hit on the role of industry. We can’t do this ourselves. This is like – you know, the days of us going off in a back room and making something up and getting – you know, figuring it out perfectly and getting it all right, we win because we can do complex systems engineering. That’s going to happen when the basic science is done and when large parts of the advanced science – we know how to manufacture and those sorts of things. That’s going to happen after that. And so our trick is to figure out when we can bring things – when we bring things “black” I’ll call it. You know, and that’s a hard call, figuring out where those tipping points are.

And we’ve made some mistakes. I won’t talk about those because that would tell you where to look and we don’t want people to know. But we have kept things unclass sometimes in some cases for too long, and there are other cases where I think, because we were worried about it, we brought it in too early and had to figure out how to expose pieces of it again.

So I think with that I’m going to stop and let you all ask questions, because you probably have more interesting questions than my canned spiel. So is that OK?

MR. BIRMINGHAM: It’s a small group. I don’t think we need a moderator. I think there’s plenty of time to get everybody’s questions in. So if anybody has one, just go ahead and start.

Ben, go ahead.

MR. BEN IANNOTTA, DEEP DIVE INTEL: Sure. Could you talk more about the human language translation? I mean, what’s the state of that technology now, and why is it so important? Just maybe put it in context a little more.

MS. MEYERRIECKS: Sure. I mean, there’s lots of – you know, there’s lots of translators out on the Net. And this is back to the where we make investments is really the crux of the discussion. That’s going to happen with or without us. I mean, there’s a real desire to provide products and services that are global in nature.

And so people are going to make their own investments. When I was in product technology at AOL we had, you know, French versions because we had an AOL France at the time. It was just the market drove you there. I mean, Apple, I heard a quote from I think (Tim) Cook today that Apple has more cash in their foreign holdings than they have in the U.S. because that’s where their volume sales are right now. So we just have to ride the market wave.

And so the thing for us is that what people are not so interested in is – from a marketing perspective we’re not really to the nuanced communications yet. It’s like kind of brute force keywords. And I’m not saying that people aren’t doing that. I think it will get there. But where we win from a community perspective is if we’re slightly ahead.

So what we’ve done is we’ve funded some research in IARPA but we also leveraged other people’s research who are doing sentiment analysis. So what you say to me and what you actually mean when you say it as a human, if I have IQ I’m pretty good at figuring that out. Computer, not so much. So we’ve got to actually figure out how to take all of the keys and the indicators and we’ll get what’s incorporated in the algorithms, which is why we need to understand – we need to understand the IP. We don’t need to own it but we need to understand it, and understand how much nuance is being picked up and what that says in terms of how we can apply the information that is translated for us.

And everybody has heard the example, 300 versions of snow – different, you know, for Eskimos. That’s kind of the one that everybody has heard about. And we translate it in English to “snow,” right, or maybe “icy snow.” But this is more about how humans interact.

MR. IANNOTTA: But, I mean, is this an indication of what, it’s just not possible to hire more people who can do communications intelligence or –

MS. MEYERRIECKS: No. Well, I would say more that we have to be more efficient with what we’ve got. The budget is not going to go up. I was with OMB this morning in the Sit Room, and flat on our best day. Flat is the new up, as we say. (Laughter.)

So we’re not going to hire more analysts. And what we have to do is do a better job. And that’s a blanket statement. And, you know, if we go up by 78 next week, you know, “Meyerriecks said,” OK. But in general, we’re planning on better leveraging the analysts that we have and using technology to get move value out of the processing.

And that’s how we win. I think that’s how we’ve won in the past, and that’s how we’ll continue to win. And that’s why it’s important to us to figure out how to wrest that last little bit of meaning, the nonspoken, sometimes the nonverbal, but also the nuanced. The way that I address you in this setting is going to be different than if I meet you in a bar.

Q: Fair enough. (Laughter.)

MR. ROBERT ACKERMAN, SIGNAL MAGAZINE: How has sequestration forced you to alter your plans?

MS. MEYERRIECKS: So I’m really, again – I know this sounds like a love fest – I’m really proud of the leadership here. We made some hard decisions, and we have over the last few years. I’m part of a small group. You know, basically we submit the budget, which is one of the big differences between the IC and the DOD. We actually do the final budget submittal. So we actually do have controls to a certain extent – influence largely, but we also have direct controls.

As a leadership team we’ve met regularly. And I’m part of a small group – I call it the “death-eaters squad” – where we actually go through and say, here’s the priorities based on the last senior offsite, and that means that these programs should be preserved regardless, and these programs should be looked at for reduction, and we’re going to disinvest.

The boss (Director Clapper) doesn’t believe in salami slicing, and I actually think that’s a really good idea, not salami slicing when you’re talking about the kinds of cuts that we’re talking about here. We have to decide what missions we’re getting out of. And form follows function, so then we make decisions about R&D. And it’s not just a straight stick. So for example, if we decide we’re going to get out of a particular technical area, and I won’t give you examples here, because that’s probably – I can’t think of any that are unclassified. What we may do is actually increase the R&D investment in that area so that we preserve a vestigial capability so that if we decide we need to get back into it, we can without restarting from scratch. So it doesn’t necessarily mean that the R&D goes down as a function of, we’re out of this business area. In fact, the R&D may go up a little bit to say we need to preserve an engineering and science core that is still broad enough that if we need to get back into that business, we haven’t completely stopped the manufacturing line.

In other cases, when we decide that we’re going to disinvest, we made kind of a more intuitive decision, which is, OK, so if we’re going to get out of this area in a big way – and we see, for example, that industry is going to pick it up for us, then we may in fact, really, really disinvest from an R&D perspective as well. So it’s not a mindless, you know, sort of thing; it’s actually a thoughtful – and one of the other things – this little framework that we put together – the lead/influence, adapt/adopt, – the other thing that I made a big point of with the Hill was, I would get questions about, well, what percentage should R&D be of your overall budget? And that always bothered me, because it takes three years to get on a space manifest, and it takes six years to build a satellite. So the duty cycle on doing something innovative there compared to IT and the investment that we get from industry – I mean, that’s like apples and oranges.

So if I tell you 3 percent across the board – and I’m making that up – that’s not the number, because I can’t talk about them – but if I tell you that, do you feel better? I mean, it’s kind of a – and I would tell them – I would tell Congress this, that it’s kind of a bad question. And then we do other things like pinhole cameras and batteries and things like that. So we’ve actually been kind of life cycle and severity of loss. So you lose an asset in space – there’s a lot of money with that, and it’s hard to replace.

You lose a cell phone, you have a kind of – slightly different reaction. And so, we actually came up with kind of big trends that looked at the severity of the loss, the life cycle, what it takes to produce, and we bend those out, and then we looked at industry investments, and we said, OK, so from an IT perspective, based on National Science Foundation numbers, the industry does on the order 8 (percent) to 9 percent per year of R&D, if you just, generalize the numbers. And those are 2009 numbers, because that’s the last time that they did it. But regardless, space is on the order of 5 and a half percent, and this other stuff – this special purpose is about 2 and a half percent. So that also tells us what our investment profile should look like.

If I’m trying to compete with industry on IT, I’m crazy, right? My crystal ball is – if – I can’t go five years out. I’m not that good from an IT perspective? In space, I better be that good, because it takes that long to, get it manufactured and get it into orbit. So it was a much more – we tried to bring a much more thoughtful sort of look at that, and say OK, so from an IT perspective, we want to largely supplement what industry is going to do naturally anyway. From a space perspective, we probably need to have a bigger role and then we ganged up with NASA and DOD and NOAA and some others and revitalized those relationships and said OK, so collectively, we could probably move the needle, from a space perspective with our investments. And then there’s these other kind of – and they’re not cats and dogs, they’re critical to mission, but their investment profile is fairly small. What does that say about ours that in order to supplement, and so we had very, very thoughtful conversations about what, kind of big muscle movements in this areas. So all of that is context to – we still try to get the community to kind of act like a community, but it’s not – it’s a nuanced conversation. It’s not a one-size-fits-all sort of approach, and what we take mission inputs, and we take leadership inputs on the priorities and then we come back to them and say OK, so what we think that translates to is the following investment profile. And they’ve been supportive – completely supportive.

MR. ACKERMAN: How long has this approach been in practice?

MS. MEYERRIECKS: Let’s see. So the first week I was in the job, I had 60 staffers ask me this question about R&D percentage. And they actually see our budgets, so they know what the numbers were. And it just really, really bothered me. So we started working on it with the NSTC which is the National Intelligence Science and Technology Committee then, which is the heads of all of the S&T organizations across the community. We started working on it, and it took us about a year to get together on it, just because we argued a lot, which was all healthy. You know, as an engineer and a scientist, that’s – peer review is a good thing, and we came out with this.

And now it’s the basis for what we do, and the next document – and Dr. Dave Honey is our director of S&T right now – the next document that we’ve done is, so what are the seven or eight really important intelligence questions we need to answer that we don’t have good collection or analytical capabilities on? So now you take the list of technologies, and then you marry that with, what are the questions we’d like to answer but can’t? And now we have really good context, from a mission perspective that also helps advise where we make investments. That helps us focus on certain things.

MR. ACKERMAN: So you’re talking about two years?

MS. MEYERRIECKS: Yeah, two years, probably, for the framework – the initial framework, and then the top eight, nine questions really have come out of the intelligence integration since the DNI – Clapper, got here, and the UISs (unified intelligence strategies), which are the NIMs (national intelligence managers) statements of what it is that they’re looking for. So yeah. So it’s – we’ve been practicing this for a little while. So – yes sir.

MR. GOPAL RATNAM, BLOOMBERG NEWS: So of the 38-something major programs you talked about, what kind of money are we talking?

Can’t answer that.

Is it billions? Is it any scope or size?

It’s big – it’s big.

And how does it break down between the lead/influence programs versus adapt and adopt? Is it kind of equally divided or is it – how does it –

No. From an R&D perspective, what we try to do is put most of our money in the lead/influence in terms of R&D dollars. Otherwise it’s procurement dollars, which is – some parts of the community don’t differentiate between colors of money, but the defense agencies certainly have to.

So I would say it’s more like a 75-25 that we try to do, because even in the adopt – and I’ll give you an example – we needed fine-grained access controls for cloud much earlier than industry did. And actually, we worked with Google, and Vince Cerf came in and did an architecture review for us; he’s a big fan of the community. And he said, yeah, you guys are really ahead of us. And it’s based on an open-source stack, and we just needed it because you can imagine why we might be interested in fine-grained access controls before anybody else is; it wouldn’t take a rocket scientist to figure that out. And then we community sourced it. So that’s where we spend – I’ll call it R&D dollars. And they are – they were. It was NSAR that did a lot of that work. And so we spent R&D dollars there.
But the good news was, because we had floated this framework, we had very robust – and I – this is totally geeky, but we had this amazing conversation at the EXCOM with Keith Alexander kind of taking water from his peers about, why are you spending R&D dollars when they’ve said that this is an adopt thing. And he actually had good answers. And I was – I actually supported him, because it was good answers, because – and he explained – basically, he gave the same explanation.

So we do try to spend most of our money in the lead/influence, because we think those are the tiebreakers in terms of capability, but we do spend money in the adapt, adopt, where it’s kind of nuanced development that we do there, and we try, to the extent possible, put either – put the capability back into the product line, that – from whence we got it, or make it available from a community or open source perspective.

(Off mic.)

AMY BUTLER, AVIATION WEEK: You spoke about systems engineering being sort of a special secret sauce that the U.S. has. And I understand completely where you’re coming from, but having covered defense programs primarily for almost 15 years, it seems like – and maybe I just came in at a really rough patch, but it seems like systems engineering took a beating over the last decade, especially in the space community, at least on the white space side, and we know FIA (Future Imagery Architecture), so there was a little bit of that congenital problem in the dark side, too.

So can you talk to us about where you think that is today in space, but also across the board, because we see this problem in other defense programs. Obviously, we can’t really peer into your portfolio very closely, but –

MS. MEYERRIECKS: Yeah. So – let’s see – one of our initiatives in our organization, I’ll tell you, is how we recruit, retain and grow systems engineers, because I really do believe that’s the secret to us. And this is part of my mantra about, we’ve got to team with industry. Say what you like about AOL; I would tell you that we had the best systems engineering organization or world-class systems engineering, and I – for reasons of employment agreements, I won’t speculate any more about what happened with AOL there, but – (laughter) – we all have those clauses.

Pound for pound we still have – I believe – the U.S. – still have the best systems engineering capability on the planet. I’ll say that distribution is uneven, and it’s a function of – in terms of the DOD, the IC and commercial. That takes nothing away from our national capability; what it does say – and this is why I’m so big on it is – we have to – we have to pair with wherever we find that engineering talent. And so I think, again, that when there was a big push to outsource, and when we told 21-year-old engineers that they were going to write contracts as opposed to write code, that we made a pretty bad decision. And I will say this as a young engineer – there was no way I would have come to government and written a contract to watch somebody else write code. I wanted to go exercise my technical muscles. I had worked hard to acquire those; I wanted to use them. I still – I say I hack HTML for my gardening club – I write HTML. I still like writing in code. I mean, it’s simple code these days, but I still like doing it.

So there’s no way. And so what I think happened was we went through this period of time and then the bathtub that everybody talks about – we went through this period of time where it was really hard to be a practicing engineer with good sensibilities in the government. And no amount – and I will just say this, and I’ve told Frank Kendall the same thing – a certification as a systems engineer does not a systems engineer make. You have to actually build stuff and get the technical chops and understand that.

And I’ll use a key example: When I got my master’s in computer science, relational databases, the answer was third normal form. Well, the first time I did third normal form in an Oracle database, my performance went in the toilet. Doesn’t actually work because nobody actually builds a database, you know, pristinely. And your data’s never that clean and I could go on and on. But you have to learn that. And the way that you learn that is because you screw up. The things we learn the best are the things we’ve made personal mistakes in our lives.

So I got to grow up as a working engineer, which brings a different sensibility to systems engineering writ large. So I think what we went through and what you’re still seeing is this decision to outsource. I think the healthiest organizations I see in my portfolio today are the ones that didn’t buy into that lock, stock and barrel. I will say that I think NSA and CIA have done fabulous jobs with that. NSA does co-teams in terms of development; they are a software powerhouse. CIA, I think – and I can’t talk about there – but I think they also are very, very good at what they do. And some of our others did the whole outsourcing thing and we’re recovering. That’s my assessments.

MS. MEYERRIECKS: So now we’ve got to grow those kids. And we’ve got to let them actually build stuff, because you got to learn the hard way that that isn’t what you learned in the textbook doesn’t translate precisely into what actually – what actually works.

Q: OK, so fair enough. But I’m curious on the letters that you have to pull, say, with industry or even within your oversight with the agencies, the kinds of contracts that are let, the kinds of milestones that are set for different projects. Have you guys changed your approach to any of that to try to tap into the goodness that’s out there, as opposed to the – maybe the dark side of capitalism, which is just, you know, a quarterly, annual kind of report mentality?

MS. MEYERRIECKS: Let’s see, so yes we have. And what we’ve tried to do is to say there’s – I would have to get into specifics, because there’s not one-size-fit-all – fits-all answer here. But I will – so I’ll talk a little bit about at least what – for software-intensive systems – couple of things that we’ve done there. One is when I first got here, I was told NSA with their time-and-materials contracts was a terrible thing and this co-development model was a terrible thing and we got to hold their feet to the fire and you know, all of that. And I went and did an assessment and, man, they were just churning out really good software. And they were hitting their milestones and they were actually using Agile (project management) and it worked. And the people that were doing the Waterfall (project management) were going three years and putting something in front of a user and the user was saying, well, this doesn’t actually suit my needs.

And so one of the lessons to learn from AOL, was users lie. And it’s not – they don’t mean to, but if I asked you all today what did you Google today, you’d do something fluffy. You say, well, I was looking for this and like, yeah. We’re not actually that cognitive about what – how we do things. So, if you get a user and say, here’s two screens, what would you do, there’s absolutely no correlation to what if you put them in front of the screen and measure what they click on. So that Waterfall methodology does not work for software-intensive systems. You should be doing regular drafts and you measure what the user does with it. And now you know what they actually need. That doesn’t lend itself easily to a different process, if you’re a DOD reporter, that whole write down every thread and write down every derive requirement and hand it to a developer and in three years, they’ll show you something that nobody recognizes. That doesn’t work.

So we’ve encouraged that process across the board, and so things like GED, which is a ground system for NRO and RTM, which is the tasking – the new TPED (Tasking, Processing, Exploitation & Dissemination) basically for NGA and others, they came with waterfall models worse than that. I won’t tell you which one of them came with a 15-year-old spec. And I said, if you seriously think that’s still the spec – (laughs) – we’ve got to talk. We’re not doing that. So we have actually been real proponents of that, and then they come with a – I try not to armchair quarterback. I got armchair quarterbacked in my job. You know, that’s one of the reasons I left the department – don’t tell me how to do it. Make an agreement with what it is that you want, how much you’re willing to spend, I’ll go figure it out. That’s what I get paid to do. So I try – I try to bring that same sensibility. It’s like, no, we’re going to talk about methodologies and things like that. You go figure out the contract types that go with that. I’ll let you and then propose it back. We’ll have a conversation. But I feel really strongly about stuff like that. So we have been big proponents of Agilent Scrum, with regular drops to users and then measuring the heck out of what they use. So that’s a different way of approaching that.

Now on the space side, it’s a slightly different thing. You never want to do control-alt-delete with the command and control of a satellite. I mean, they’ll save themselves, but they kind of go into hiding until they decide that the numbskull that sent me this is no longer in charge and so I can come back out and do my thing.

MS. BUTLER: Unless they need to get shot down. (Laughs.)

MS. MEYERRIECKS: That’s right. That’s bad. That’s a bad thing. So we took a slightly different sensibility there, but we’re working it the same way in terms of there are things. Do we have to do NRE (non-recurring engineering) every time we buy a bus? I would claim that that’s dumber than a box of rocks. So we’re thinking differently and we’re actually implementing different methodologies for how we acquire space capabilities so that the payload drives the bus and the bus doesn’t drive the payload. So there are a number of things that I think we’ve done. And again, I’m a working systems engineer who suffered under acquisition, and I’m all about innovation because, you know, I look at people and say really, we’ve been running that play for 50 years. Do you think it’s going to win this time? OK. So maybe we need to think differently about that.

And the other thing good thing is a bus – this is a great quote that I got from somebody else – but “never let a good crisis go to waste.” “We’ve run out of money so it’s time to think.” So we’ve been very overt about that and we’ve actually instituted from an acquisition policy perspective differentiators for MSAs versus we call them ERDs (engineering, research, development), which is tech development platforms, one of the kinds. We don’t know the con ops. We don’t know what problem exactly it is that we’re trying to solve, but we know that this is kind of in the right space, and then I’ll call them platforms, either physical platforms like buses for space or virtual platforms like Google Maps. And those all have different acquisition oversight models. And what I did there for those – the ones that aren’t MSAs (major system acquisitions), is we started with the minimum list I thought we could get away with from an oversight perspective, and then we have to argue about each one, each artifact that we add to that list, as opposed to they always say you can tailor 5,000. Yeah, you can. You can take things out. I wanted to start the other way, which is let’s ask for three things and convince ourselves we need the fourth thing because that – they all cost. So that’s the nice thing about being able to institute those.

And what I’m actually trying to do on the platform side from an IT perspective is how people think about what should be encapsulated – sorry, I know, I’ll be geeky here – what should be extracted, what should be hidden – these are all computer science sorts of things. Do the systems engineering first and sense good systems engineering. And the analogy I use is, you know, IOS (internal operating system) or whatever – nobody asks me to upgrade Windows on my machine because it’s a well-established platform, there’s a well-established way that the APIs are updated; everybody understands that. That’s not an acquisition, that’s a procurement. If I can incent the community to treat a capability as a platform and bring that level of discipline, then I’m willing to treat it like a Windows upgrade. And within certain parameters, if the APIs don’t change by more than 15 percent, you know, and there’s a good governance process, that’s a procurement. Now you’re not in an acquisition pile anymore. And that’s a much different model, all right? So you get out from under acquisition, if you actually do good systems engineering up front and establish a stable platform, like a cloud or whatever. So we’ve done things like that.

MR. JASON MILLER: Can you translate MSA and ER.

Sorry. MSA, major system acquisition.

Major system, OK.

MS. MEYERRIECKS: And ERD is engineering, research, development or something like that. Yeah, it’s a technical – it’s a prototype, basically. We can’t just call things easy stuff. I don’t know why we came up with ERD.

MR. MICHAEL McCARTHY, DEFENSE DAILY: Can I ask you about veterans? It’s playing prominently in Better Buying Power 2.0 over at the Pentagon. Can you talk about how you’ve addressed that, being that it’s a pretty contentious issue with industry, as you know? How have you addressed that here in terms of capturing data for future recompetes to upgrade open architecture systems, stuff like that?

MS. MEYERRIECKS: That one’s hard for us, and we actually were in the throes of talking to Congress about it. We have a number of organizations that would prefer not to talk about who they do business with – gee, for obvious reasons. And so – and you’re talking about kind of performance – performance data is what you’re poking at?

MR. McCARTHY: Intellectual property, data rights that the company’s providing.

: Data rights, OK.

How do you capture – how do you capture data rights from the companies? Have you – have you optimized that, or is that still a major challenge?

OK. I thought you were poking at how we share past performance information. That – (inaudible).

No, I’m talking more about capturing data rights through the acquisition process.

So let’s see. So we happen to have, I think, the best intellectual property attorney on the planet working with us right now. I feel very fortunate in this job. And she started with IARPA so she really, really – she’s a technical attorney. She really, really gets this stuff.

So starting from there, what we try to do is we try to retain what we think is important to us, but we don’t want to maintain it. And the example that I’ll give is we made an investment in PittPatt early on, which Google bought maybe 18 months ago, that does face recognition kinds of things. And what we negotiated there – and I always laugh about this – is we negotiated data rights to the source code.
And so when the Google attorney showed up – because we were trying to just – you know, basically we were funding the development of the startup, and when the Google attorney showed up they were like, oh, we don’t do that. That couldn’t possibly be. You know, it’s like, no; actually read the contract. So we actually manage to retain data rights.

And the way we did it is I looked at him and I said, look, you’re going to put 400 staff on this. We’re going to take your source code that we had maybe 40 staffers total on, and we’re going to keep going with 10. Your code baseline is not going to look anything like ours and it’s going to be way out in front of us. We have some special-purpose things that we’re doing with this, even though IARPA does much unclassified. I mean, it’s a stalking horse for what we’re actually, in many case, going to do with it on the dark side.

And so we actually had a very effective conversation with the Google attorneys there, and they got – it was like, oh, yeah, you’re right. And I was like, in 18 months you’re not going to be able to tell that there’s even a common heritage here anymore. So get over it.

In-Q-Tel we have a very different model, as you know. So we have different models based on where we are. We have had some problems in the MSAs with some of the contractors in terms of – my schedule is proprietary and it’s covered – it’s not covered by government data rights. And we’ve basically – again, we have a wonderful attorney. And as my boss used to say, my attorneys get paid whether they’re in court or not. So generally those get resolved long before we get there because we’re not afraid.

MR. McCARTHY: How do you mitigate that when you do have a company that’s saying what they just told you? I mean –

MS. MEYERRIECKS: So generally we have – I mean, we’ve been doing this for a long time. There’s general verbiage that we put into all of our contracts. And when we get into those kinds of conversations, usually the first thing I do, if it doesn’t get resolved, is I’ll pick up the phone and call the head of the P and L and say: OK, so let me explain. And because I have this information: Here’s how much business we do with you. Here’s how much this contract is worth. Really? Really? You want to have this conversation about our schedule? And generally that’s pretty much the end of the conversation.

Now, some of them have taken a little longer than that and the attorneys have gotten involved, but generally we’ve worked through – we’re not getting – and I think people understand that we’re not compromising their data rights. Who’s ever been sued because the government didn’t have the access to data rights? I laugh about that. Like, who have we taken to court? I mean, it’s kind of – I think it’s a paper dragon.


MR. JASON MILLER, FEDERAL NEWS RADIO: You talked about the changes in acquisition, like the Agile Scrum going away from Waterfall. How has that been kind of pushed through, or dragged through, or accepted by the IC community at large? Let me give you maybe an example.

Initially with the IT infrastructure project IC ITE – who was initially the QUAD and then it became the QUINT and then it moved on, and now it kind of grew on people – have you been able to kind of get that collaboration when it comes to technology, when it comes to acquisition, that some people are buying at least in that same direction? Because there’s a sense sometimes that that – you know, are people going in blocks, at least from an architecture standpoint, from a system – like, are we buying the same – it doesn’t have to be the same system per se, but it’s got to meet the same standards or the same proprietary – or open standards I guess is the question. So are you able to kind of push the collaboration through?

MS. MEYERRIECKS: So I think there’s at least two questions there, so –


MS. MEYERRIECKS: Well, if I get it wrong – yeah, explain it to me again and we’ll give it another whirl.

So from an Agile Scrum perspective, absolutely we’ve been able to push that through. Is it ubiquitous across the community? No, of course not, you know. Let’s see; there is some power in the bully pulpit but there is also some power as the acquisition wonk. I can actually stop programs.

In the three-and-a-half years that I’ve been here I’ve stopped one acquisition. I don’t – because I’ve been on the receiving end – not that mine were stopped but I’ve been on the receiving end of: We’re going to stop your acquisition because there’s this other thing over here that this other organization is doing that’s pissing us off, and we know that that will hurt them if we stop you. And I have refused to play that card. It’s like, no, I’m going to take each one as it comes. And you will succeed or fail on your own merits, not because I had a bad day and I don’t like your boss right now. So we didn’t do that.
So it’s not even – but what we could do is for those two I rattled off – NSA is already there, but for GED and for RTM – and those are the – you know, NRO and NGA, in terms of big acquisitions we basically said: No, you’re not going to do it that way. You’re not going to do Waterfall and you’re not going to bring us an acquisition. And I spent enough time with these folks. We meet at least quarterly and very often, more often. I have very good relationships with the leadership.

I have these conversations early: If you bring me something that looks like Waterfall, we’re going to have a really bad conversation, so let’s not have that conversation. And I’ve been around long enough that they take that in the spirit in which it’s meant and not I’m just being a jerk. And you earn your way in, so it took a little while to get there, but fortunately we have had those opportunities as we’ve gone through the cycle.

And so I think as they actually implement, then it speaks for itself. So now they come to me with lessons learned about, oh, we started this thing and we used Waterfall and it was a disaster. I had one of those meetings last week. It was like, yeah, entirely predictable. And they all shake their heads and say, yeah, we won’t do that again. I mean, it’s self-reinforcing. At that point life is good.
Now, you asked about IC ITE.

MR. MILLER: And just using that as an example is IC ITE, but the key is that are people moving in the same lockstep as much as possible within reason within –

MS. MEYERRIECKS: They are. This is a big iceberg to turn, and so – I mean, there’s lots of things – so generally the answer is yes. This is a weekly discussion. We have a weekly VTC (video teleconference) among – with the deputies that the PD (Principal Deputy Director National Intelligence) runs where we actually hold each other accountable for migration planning for IC ITE.
But if you think about all of the complexities that we have to do – so for example, NSA outsourced their desktop support for the, you know, NCR (National Capital Region) area. OK, so even if I had the perfect desktop for them tomorrow, they have a contractual obligation that they have to finish out, you know, the option year, at least the current option year on their contract. And oh, by the way, it’s structured kind of like an NMCI, so they don’t actually own the end-user devices anymore.

So yeah, this is all interesting. Even if I say in a year they have an option they can exercise, do they have the cash to go replenish if the incumbent doesn’t win? And oh, by the way, the incumbent is not our desktop provider, so a pretty good chance we’ll have to buy – recapitalize.

So, I mean, there’s lots of things that – there’s a lot of moving piece parts here. I will say that generally I believe everybody’s all in and doing their best to try – this is a huge – think about taking something the size of this, 120,000 or whatever, you know, person corporation, five major divisions, and saying, OK, put a plan together for how we’re going to collapse to a single infrastructure. You’d get a five-year plan, even in a corporation.

So I think we’re being very – I think everybody’s in. And I think we’re working it hard. And there is – at the end of the day there isn’t going to be a one-size-fits-all for certain classes of things, but for a majority of what we’re doing, there’s no reason for us to pay the same or different vetters over and over again to stamp VMs (virtual machines). It’s not even interesting computer science anymore. So really?

And what we’re trying to do is say, OK, let’s move the value up in the mission value chain. The LockMarts and the SAICs and the Boeings and everybody can really, really help us with big data analytics for our style data. I don’t want them to differentiate because they got a, you know, a different VM than the one next to them. And so that’s – and there’s been some concern about that as well.
But I think we’re actually on the right track and it feels like there is an inflection point that we hit, I’ll say maybe a month ago, maybe six weeks ago, just in terms of the leadership getting it. And that sounds like they’re – they’re certainly not stupid but it feels more like a moral commitment now to me, just in terms of the sense of the discussions that we have. And that’s really – that’s been fantastic.
And it will never be lockstep. You don’t run an enterprise this big with everybody slavishly assigned to, you know, one version of a desktop or one version of a computer file. Even in our cloud we talk about there’s a run time – you know, a virtualized run time, there’s a data cloud, and then there’s the analytics cloud. You optimize those very, very differently, and what we’re just discovering, or learning kind of again, is – and your analytics cloud is optimized for the kinds of things that you’re trying to do. So it’s not even one-size-fits-all there. But that’s a much more interesting problem than figuring out how to configure, you know, VM one more time. And people are getting it.

Yes, sir?

MR. ZACHARY FRYER-BIGGS, DEFENSE NEWS: I wanted to ask – you mentioned turning to industry and certain areas of R&D to pick up the slack in some ways, especially from some of the traditional defense and intel contractors. We increasingly hear, well, given the fiscal pressures, we may not be able to invest in some of those R&D areas.

Do you see some of those pressures might force you to lean more on commercial enterprises outside of the traditional sort of contracting community? And how do you see that mix adjusting? What does that mean?

MS. MEYERRIECKS: Yeah, I think so. I think that – we will lean more on commercial, but I – so maybe I come in prejudiced – I’m happy with that. I mean, I wish we had done that sooner, but as the boss (DNI Clapper) says, “don’t let a good crisis go to waste.” And I think that’s why I was so pleased when I got here about the approach that IARPA has taken. This is countercultural to the community, but so is In-Q-Tel and the things that we are seeing come out of those explorations more than pay for themselves. I mean, there’s a 9-to-1 kind of investment strategy with In-Q-Tel; for every dollar we put in, there’s nine of other peoples’ dollars. That’s phenomenal. You guys should be proud of us.

I mean, I’m proud of us, that we usually have that kind of buying power that we’re thinking that way. Same thing with IARPA – I gave you the stat on the 70 and 17 – 1200 relationships with universities. I mean –we’re kind of – we can’t go back. If we want to do, for example, really good work with identity intelligence, that’s – we’re not going to invent that. I mean, you know, what people are doing in terms of profiling each of us – like the fact that you all turned your cell phones in – you’re probably aberration on somebody’s screen right now, and I’m being a little facetious; they’re not that – you know, they’re not at that level yet, but they’re working on it.

So I think that we don’t have a choice. We absolutely don’t have a choice. How that will play itself out in terms of, you know, percentage investments – that’s – I don’t – my crystal ball isn’t that good, but I think that a lot more – and that’s – we’re going out – we’re taking some of the senior leadership out to the Valley next week, and we’re going to go see – I’ve known Marc Andreessen for a long time, and I’m not a name-dropper – I actually have known him for a long time – I hung out with cool kids when I was a kid – and we’re going to talk to him, and we’re also going to talk to some of the In-Q-Tel investments with the senior leadership in the community, because they get it.

MR McCARTHY: What would you identify as some of the key advantages of going to the commercial industry rather than the traditional contractors?

MS. MEYERRIECKS: Let’s see. So there’s good and bad on both sides. They’re less constrained by – I’ll just say security, although they’re getting better about it. They’re less constrained in terms of – they think very differently than we do about problems. Where we have been a very homogenous – and I mean the IC – we’ve been very internally focused for reasons which should be obvious, and they don’t carry all of that baggage. On the other hand, they’re willing to do things like first to market at the expense of security, at times. That never happened in my previous life at any other job that I’ve had – maybe. So there are pros and cons, and I think the folks that are close to us really, really, really do understand our mission kind of intrinsically, and that’s what we’re telling them is, there’s plenty of room at the top.

Help us with mission. Stop with the infrastructure stuff. It’s actually – you’ll be more interesting, and we’ll be more interested, and I think that’s where their big value add is. And I think big data is one of those areas – and that’s why I’m so passionate about IC ITE – every study that I’ve read, you have to make the infrastructure disappear. You can’t continue to stovepipe and talk about big data, because the value is being – and this is my systems engineering thing – part of the big data thing is being good systems engineers about, oh, we can combine this, this and this, and all of a sudden, we have an insight that we wouldn’t have had if we just looked through either one or two of those telescopes by itself.

That’s where our close partners can help us a lot, because they get that. And we can learn a lot about big data toolsets from commercial because they’re solving those problems for the financials and, you know, various other verticals. But populating that so that it’s useful to us is going to be the job of the people that are close to us, and so I think there’s a division of labor here that’s perfect once we get through the inflection point, but everybody’s nervous in the meantime.

MR. ACKERMAN: Dawn, what would you say the technologies on which you will rely most in the industry? I mean, you just mentioned big data.

MS. MEYERRIECKS: All of it. (Laughter.) I mean, there’s none of it that we’re inventing by ourselves anymore. And if we’re – even for batteries – we don’t want to manufacture batteries, we want to introduce the capability back into the manufacturing base and have other people build it for us. So there’s nothing that I would say we weren’t interested in. Big data is definitely one of them.
MR. ACKERMAN: What are the top ones, though?

MS. MEYERRIECKS: New phenomenologies in terms of collection. I used this like, really? We think nobody gets EO (electro-optical) and atmospherics anymore. Who’s not flying microsatellites with cameras on them, or, you know, airborne stuff with cameras on them. So I think we’ve got to talk about new phenomenology and combinatorials that are different, and that’s why I think the intelligence integration is exactly right, not just for mission but also from a technology perspective. We’ve got to think about multiple things to bear, so I think multiple phenomenologies are another area that’s going to be really important to us.

MS. BUTLER: Can you be specific? Like, what are you talking about? HSI or –

MS. MEYERRIECKS: Yes, all of it. I laugh about – we’ve drawn arbitrary lines in the spectrum that say, this is visible, this is IR, this is – you know, this is HF – really? The spectrum doesn’t know that. And so – and our collectors are getting good enough that what we used to consider kind of lobe sorts of things are actually really, really interesting. So it doesn’t – the old paradigm – we’re shooting a lot of holes in kind of the old paradigms in terms of, you know SIGINT versus other things, for example.

It’s like – that’s an arbitrary line in the spectrum. Nobody explained to the spectrum that oh, well, you don’t fit in this – in this spot, so we can’t collect on you, because that’s one of these kinds of things – that – it’s all just physics and math, at the end of the day. And so I think breaking those down has been one of our more interesting challenges. Because then, so what’s the role of the functional manager in that? If you can actually task a sensor to collect in a particular piece of this – areas of spectrum that are independent of where we’ve drawn arbitrary human lines even from a tasking perspective, wow, now you’ve really kicked over the apple cart. But that’s fascinating to me. You know, that’s a whole different – so I didn’t really do a good job answering questions.

MS. MEYERRIECKS: Yes, sir. Yeah, sorry.

MR. JOHN WALCOTT: I don’t want to misunderstand something you said at the outset about the dangers of IP (intellectual property) theft, because it seems to me that the adversary’s a moving target. They’re getting better at innovation.


MR. WALCOTT: And the country that we’re not talking about that everybody talks about is maybe not at the head of the line when it comes to their ability to innovate. There’s some others we could name that are – that we consider allies who are big in the IP theft department. So how important – what – where on the spectrum of interest for you do the security issues play?

MS. MEYERRIECKS: They’re huge. I mean, that’s –

MR. WALCOTT: Yeah, I didn’t want to misunderstand that.

MS. MEYERRIECKS: And that’s – I – no, I’m glad you clarified that, because if anybody walks out and says, oh, Meyerriecks says IP theft isn’t a big deal, no, that’s not what I meant at all. No, I think that’s huge. And again, one of the reasons for ICITE is because let’s at least understand what we have to defend. If we had a countable number of versions, it would still be better. So yeah, that’s important to us. So yes, I’m glad you asked the question because that would have been a horrible implication to walk out of here with. Yes, no, we’re very – we are stewards of the taxpayers’ money but also the very, very critical national mission. And I don’t want to make light of – and, oh, by the way, we are counting on commercial to move the ball forward with us, for us, in specific areas, and we don’t want that given away. I mean, when we are overtly saying we’re going to – we’re going to leverage what’s coming out of Silicon Valley or a LockMart or whoever, depending on the area in order for us to win? Then, yeah, it matters, that that’s – we don’t want that stuff getting in the way.

MR. WALCOTT: So, do you have the tools, finally, to maintain that level of security into the commercial sector all the way through the supply chain?

MS. MEYERRIECKS: So – I’ll give you the same answer I gave him. I think it’s uneven. We are doing much better, I think, from a defense industrial base, but we’ve been working at it for 15 years now. So, I think we’re doing better. I’m not naïve enough to believe that we have this thing down, and there are very different security postures among our supply chain. You know, very interesting and a conversation with a head of a large wireless carrier in the U.S. And let’s say, from a business perspective, he would love to stay U.S. equipment-based. What do I tell him? Who do I tell him to go talk to? Who would you tell him? Huawei the cheapest, and right after that, Ericsson. And who’s third? We probably can’t even name it as a group. Gee, either one of those U.S.-based companies? So I think what we have to do is for those areas where we’re in those positions, we actually have to think through the implications of somebody being in our supply chain, likely, and what do we do with that. How many ASICs (application-specific integrated circuits) are manufactured in the U.S. today? How many FPGAs (field-programmable gate arrays)?

MR. BIRMINGHAM: Well, with that, folks, we can wrap it up.

MS. MEYERRIECKS: Wow, that was kind of a downer. (Laughter.)

MR. McCARTHY: What were you going to say on background today that we didn’t – (Laughter).

MS. MEYERRIECKS: No, no, I fended off the ones I couldn’t, so –