The Largest Untapped Market Inside U.S. Defense | Emil Michael
00:00:00The military buildup in China is thebiggest military buildup in worldhistory. And so there's a real urgencyon our side to ensure that we are ahead,but we stay ahead. And that's going totake a different level of investment anddifferent type of thinking than we'vehad in the last 20 years. In the 80s,there were 50 defense contractors andthey got merged. So there are only aboutfive. There's a lot of room for newentrance. It's crazy to me that SpaceXand Androl and Palanteer all had to suethe Department of War for their firstcontract. So the idea is you don't haveto sue anymore. come through the frontdoor [laughter]because no one's people are not going tofight you. We're now excited about lowercost, faster, more sophisticatedoptions.>> Hi listeners, welcome back to No Priors.We're here today with Emil Michael, theformer chief business officer of Uber,White [music] House fellow, andcurrently CTO of the Department of War.>> Emil, thanks for joining us. Welcome toNew Prior.>> Good to see you guys. It's been a longtime.>> Congratulations on the new role. Um,very exciting news. Could you describe alittle bit more about what that role isand what's changed at the Department ofWar to sort of create this new war thisnew um role this new momentum newinitiatives that you all are focused on?>> Yeah. So um for a long time at thedepartment of war there was oneorganization
00:01:15calleduh acquisition technology and logisticsand it was all bunched up into one thingand then about 8 years ago they saidwell tech is moving faster on new kindsof weaponry and defense systems than onthe old system. So we're going to splitout acquisition from research andengineering they call it. So I'm theunder secretary for research andengineering which is cool right becauseI get to work on the stuff that I usedto work on when I was in Silicon Valleylike with work with entrepreneurs workwith new companies. I'm now responsiblefor DARPA which is obviously super coolbecause it's most some of the mostadvanced research that happens in thecountry if not the world. uh in the lastfew months I took over as chief AI thechief AI office in the department of warwhich would 3 million employees thebiggest organization with the biggestbudget in the world you know is is not asmall thing to to think about how to doAI right for and then the defenseinnovation unit which is actually basedin Mountain View and it's supposedly theit's supposed to be the link betweendefense industry and uh you know thestartup community that's buildingcommercial products may have dual use.And then last is this strategiccapability office which takes kind of
00:02:30existingum systems and tries to modify them instrategic ways or supplement them to getyou know for strategic surprise theycall it. So all that all has technologyunderneath it and the idea is to unifythat across the department because wespend you know $150 billion a year ontech in one way or the other. So youwant to avoid duplication, you want tobring things to market faster. That thatwas a big announcement yesterday by bythe secretary at uh at Starbase.>> You mentioned DIU has been involved withproviding early funding to a variety ofpromising new defense tech and hardwarecompanies. Could you tell us a littlebit more about what they've done and howyou've approached that so far?>> Yeah, DIU's had some good success withSeronic and and uh and a bunch of othercompanies and now it's you know we havenow it's time to do more of it. So, it'smore of a focus and yeah, DARPA is just,you know, they invented the internet.You go to their gift shop and they havelike a a napkin that where someone drewout the internet uh sort of architectureand they sell that in the gift shop. Incase you guys come to DC, we can get youget you some of those. [snorts]>> This is like a I mean, it's an amazingunification of like a bunch of stuffthat's clearly important to America andto, you know, global security.
00:03:46uh it's avery different environment than a lotand I have seen you in uh in the pastright you've worked with technologycompanies and you know even within themas a leader had this reputation forbeing you know high-speed aggressivedealmaker uh if if I'm allowed to saythat um how does that work with thedepartment of war which like amazingimpact importance and scale not thestrongest reputation in previous decadesand administrations for speed.>> Yeah. Well, I I mean hopefully part ofthe reason I got chosen for the job andcertainly the reason I took it is toinject some of that urgency speed, youknow, being a little bit impervious tobarriers and trying to run them over asopposed to being stopped by them. And itkind it works with leadership in inorganizations is not is somewhattransferable, right? If you can showthat kind of urgency and leadership andcreate a culture in a company, you couldcertainly make some progress in abureaucracy or a government agency, itmight not be as fast, but I also haveway more resources, right? And I have aa way more, you know, in some way isimportant mission to humanity. So thosetwo
00:05:01things are different in some wayssort of give me the same abilities tosort of drive fast and people respond. Imean, I'm getting a lot of people whowant to join my team or other teams atDO um and across the government andyeah, it's harder, but uh in some waysit's more impactful. So, I'm just asmotivated.>> Yeah, it's been interesting to watch allthe entrepreneurs kind of gathering inDC in a way that, you know, I reallyhaven't seen in a while. So, it's it'sbeen impressive in terms of the types oftalent that you all have been able torecruit. One thing that I think isrelated to moving fast isprioritization. And I think in anotherum interview in the past, you saidsomething like 14 priorities mean nopriorities at all. And you've kind ofreally honed down into a handful of keyareas. Do you mind walking through sortof those key areas from an overallbroader innovation focus? And then maybewe can touch on one or two of them as wego.>> Yeah, I mean so just to analogize toUber or any company that you you know weworked on together, imagine you went toa management team and you're like what'syour how many product you know how'syour product line going? You're a seriesB company. and you're like, "Well, I'vegot 14 product lines." You say, "Well,wait a minute. Sort of uninvestable,right?" So, um, when I got here, therewere 14 critical technology areas. So,critical critical to our nationalsecurity. So,
00:06:16I I looked at them. Isaid, "Well, number one, that if ifthere's that many, they can't becritical." But number two, hard forpeople to hold in their head that manypriorities and wake up every morningknowing that they have to get up andexecute against those priorities andmake progress because you getdistracted. did you get too much uhdisperate you know uh separation andtherefore not enough resource to any onething. So I I cut them down to six aafter studying them and I made it moreactionoriented. So we put sprints behindthem like you would in an engineeringteam. So uh you know off the top of myhead the the number one is applied AIbecause we're not building a foundationmodel at the DO because there's beinghundreds of billions of dollars beingspent by the private sector on this. So,how do I adapt or use what's beingdeveloped in the in the private sectorand apply it to the Department of Waruse cases? So, that's number one. Um,number two is I don't know if you'veheard of hypersonic missiles, butthey're sort of the newest variation ofum we used to have nuclear weapons andballistic missiles. Now, thesehypersonics, Elon talks about this allthe time, that could go five times thespeed of sound, the minimum Mach five.And so they pose somewhat of a uniquethreat and they're maneuverable and sothey're you know it's the new sort ofthreat
00:07:31environment um weapon that'sentered the picture. Few countries canbuild them the Chinese, Russia, US butmy uh critical technology area is to dothat at scale and at reasonable pricebecause they're all very expensive now.They're exquisite. So how do you makethem like in the way and thinks aboutthem much more producible, much lessexquisite, much cheaper, more mass? So Icall it scaled hypersonics. Another oneis scaled directed energy. So directedenergy um is you know high poweredmicrowave or lasers to take downmissiles like iron beam if you thinkabout what the Israelis did. And nowwith all the drone warfare you see, it'smuch cheaper to take out drones with anenergy zap than it is to send a missileat another at a drone, right? A $5million missile at a $50,000 drone orsomething. So directed energy again isis the technology exists, but how do Ido that at scale?>> I I'm a little bit curious uh on the onthe drone area in particular. One of theum things that I think a lot of peoplebelieve is that we're going to move fromsort of big iron giant systems into moredistributed fleets of autonomous dronesor vehicles.And how do you think about that in thein the context of some of the uhprograms
00:08:46or purchasing behavior of theDepartment of War? You know, we're stillbuying aircraft carriers. We're stillbuying very big sort of systems andthere's all sorts of uses and needs forthose. But what do you view as the shiftto autonomy and the shift to dronesthat's going to happen over time? Imean, what what proportion of hardwaredo you think goes there?>> I think thatstarting with the Russia Ukraine war,you've got this sort of renaissance indrones or robots. Think of it as robotsas the new front line, right? Say in aterritorial battle, if you're fightingover land, you know, less costly inhuman life if you have robots fightingfirst and then before humans come in.Without the drone warfare in in thatarea of the world, you probably wouldhave seen way more human casualties. Itwould it's already devastating. It wouldhave been way more devastating. [snorts]So I think for territorial battles,you'll see that happen real fast wherethe shift the shift mix toward robots ordrones will happen much faster. When youtalk about carriers and sea, there'sstill projection of power, there's stillprotecting sea lanes, there's still uhall those things which will continue forsome time. But the mix shift for isstill also going to move to autonomouswhether it's autonomous submarinesautonomous boats autonomous sort of bigdrones
00:10:01essentially autonomous airplanesthat mix will happen that shift mix willhappen over time and you'll see everydefense budget year-over-year more andmore will be allocated to those kinds ofthings as AI gets more prominent andbeing able to control these things andsense the environment around them andand create autonomous action. So youknow in 10 years it wouldn't surprise meif 20 30% of the budget is is is spenton those kinds of systems which are alsocheaper. So for 20 10 20 30% of thebudget you get way more firepower thanyou would for the other systems that uhthat are not designed to be you know forthat case without humans in them withhumans in them. Given the scope of theuh different services, right, and thescope of the organization, like whatdoes the prioritization exercise looklike for you operationally? You know,you go talk to the different services,you have all the people reporting toyou, like how do you decide if it isthese hypersonic missiles or it'selectronic warfare or space or somethingelse that is a is a priority for like,you know, urgent under secretary thisyear versus next. If you're at thePentagon and you're working for theSecretary of War, then your job or myjob is to you as a
00:11:16representative forhim in the technology area is to lookacross the services, Army, Navy, AirForce, Space Force, and say what are thecommon needs, uh, the enablers, the techenablers, if you will, or the enablersthat are useful in a broad way. If it'sjust useful for one service and one usecase, that's probably something thatservice can handle on their own becausethey have their own labs and researchand development dollars and ability toprocure and so on. So I'm looking forcommonalities where you get economies ofscale where there's things that could beuseful for all of them. And so thatthat's why it took me a couple months toget through these 14 technology areas.So to go get a sense of what's reallyimportant to scale them down to six thathave some common thread to them thatcould be more useful across theenterprise.>> So uh genai.millI think was like a surprise and howquickly um it happened. Uh uh can youtell us the story of that and what thelike the goal is? So, ji.m mill becauseif you're on a on a department of warnetwork, you you know it's a even theunclassified networks are secure, right?And and different levels of security asyou go up. So, you still have to figureout how to architect using AI in intothose networks. It's not simple as goingto
00:12:32chatbt.com and and signing up becausethose things are sort of generallyprohibited because you have to haverestrictions on how to use them. So wehad to figure out what are the thepolicies, how do you architect it in thenetwork so nothing goes back into thepool of data that chat GPT or claude orany of these companies have because wecertainly don't want no one in thiscountry wants our data getting out intothe general public right into thesemodels. So you have to architect adifferent uh data flow. So but we movedreally fast. I had some like data bricksengineers former data bricks engineersformer meta engineers former AWS peopletiger team 60 days uh good greatcollaboration from Gemini from Google'sGemini to who already works atdepartment of war has some familiaritywith our architecture and our systemsand got that launched to three millionpeople we've had over a million peopleuniques use it in the last 30 days whichis kind of awesome>> you got onethird of the enterprise onone model That's 2.5, not 3.0. 3.0 iscoming in a couple weeks. Then we'regoing to do Grock uh with XAI and we'llsee where we get to with the other twoand then we do it across classificationlevel. So it's going to it is going tofundamentally change the way we we workhere and we're able to do it fastbecause
00:13:47we got fast people.>> It's amazing. Could you talk about morebroadly what are the set of applied AIinitiatives that you all are focused on?I know it's a major thrust and it's oneof the key areas that you talked about.So yeah, so instead of white papers andtheories and and and committees, we wesaid okay, how are we going to deployAI? So we simplified it and say, well,there's three sort of broad categorieswe're going to deploy it on. One isenterprise use cases like any companywould use it for efficiencies, right?The second is intelligence. We get a lotof intelligence from satellites, fromall the things the United States doesacross the enterprise that uh previouslyused human analysts for. And so if youcould uh improve the theleverage of the human analyst by geingup to them, here's 10 things we saw thatyou should look at, man, you're going tomake that analyst way more efficient,right? And then well if I could takethis type of intelligence and fuse itwith this type of intelligence I'm makethat analyst really capable becauseusually one type of analyst looks at onetype of intelligence. So the more data Ican aggregate from all these sources themore intelligence we can use which is agreat deterrent because we have the bestintelligence collection in the world.And then third for war fighting. So warfighting
00:15:02is a complex sort of thing.You're doing war gaming, you're doing umplanning. You know there's a lot offormalities and how you execute orders,how you simulate what might happen ifyou do different scenarios. So acrossthose three areas, enterpriseintelligence wargaming, we'll have a fewpace setting projects within each ofthem which are designed to sort of bothbreak down barriers to make sure we gettheir data in the system, but to showpeople in the department, holy cow, yourealize this is what AI could do. So wepicked them purposefully so that wecould they could be both demonstrationsor pathfinders we call them in in ourbureaucraties or and and across thesedimensions so we hit all parts of of thedepartment of war. And we think that's away to get people thinking about AI forevery use case possible and thendeciding not to when only when it's likeclearly not useful.>> Yeah. One question that we've askedpeople from some of the major labs uhyou know researchers working on thetechnology and others and they oftendon't want to answer the questionbecause it's hard to project. So if youfeel the same feel free to de mirror butif you think ahead 5 to 10 years and youask what are the big areas where AI isgoing to most transform how thedepartment of war works how you thinkabout autonomous systems how you
00:16:17thinkabout warfare in general and defense ingeneral what do you think are the bigshifts that are going to happen or whatwhat does that future world look like? Imean, I think that, you know, autonomoussystems is the clearest thing becausewe're already thinking about that withdrones and then now you take that toother things like ceronics building oruh animals building or whatever and andyou apply AI to them and moresophisticated AI over time that could dosensing and change how it behaves. Thoseare absolutely going to be some thingsthat happen uh in the next few yearsthat AI will have a dramatic impact onwhich is why Boston Dynamics didn't makeas much progress as Optimus did and theAI you know the things that startedlater on robotics just work betterbecause they were more AI infused. So Ithink um we're actually in a good momentfor for department of war to starttaking advantage of those things andthat's where we see some some bigimprovements. And then the enterprisepiece people, you know, you're like,it's sort of boring for Silicon Valleypeople. You're like, okay, of course.But but you guys, you guys are immersedin that because you you're living inlike the frontier of AI use cases in theenterprise.>> We're just starting over here in DC. SoI think people are going to go, "Holycow, you don't have to
00:17:33write 20 pagepowerpoints, 100page congressionalreports, you know, in the same way thatyou did with all the manpower you did."And that'll that'll be more exciting forpeople to work at too, right? It'sbecause you're not doing as muchdrudgery. So you'll attract bettertalent uh as a result. So I think thoseare two clear areas in the next fewyears. You'll see a lot of uh a lot ofadvancement.>> You mentioned that one thing you have tohave uh to get these changes to happenis like some talent from Silicon Valleyor just talent that is excited aboutthis pace. Like who are you recruiting?Who do you want to be part of thedepartment of war? one of the big untoldstories about Doge, which is yeah, itdid some some aggressive actions in thebeginning that were um controversial andall that, but the talent they broughtinto every department um could then bedeployed on other projects like the AIaction plan uh that we did with Gen AI.So because they were already techn likeElon chose technically sophisticatedpeople. They weren't all engineers, somewere lawyers, some were had fromdifferent but they all had a mentalityof of a fixer builder mentality. So Iborrowed a lot of those people and thenthey attract more people on the way in.And then we have this US tech force
00:18:48thatthat Scott Kapoor from Andre andHorowitz is deployed and we're hoping toget thousands of people out of collegefor a two-year stint sort of make it umyou know this is your service to thecountry as a technologist rather than asa soldier. Um so now we want both andwe're going to sort of you know make tryto make that a badge of honor so thatyou go back into industry with some credyou know additional credibility and beproud of your service here. So, thoseare the ways I'm attracting talent andI'm dialing for dollars. I call everyonewho's had just left a job and I'm like,"Hey, you have a year to spare doing thecoolest stuff you could possiblyimagine. Recruiting Tuesdays." I callthem and I call my friends like you guysand say, "Do you have anyone who's onthe bench right now and we have so muchinteresting stuff to do that I'm pickinga few people off every month.">> You and Secretary Hexf have been on thisArsenal of Freedom tour. Can you tell usabout it or um anything you learned orthat the American people should hearabout? It's been amazing and I I'll tellyou why. So the secretary's first reformor transform or transformation if youwant to call it was restoring thewarrior ethos. It was bringing objectiveum genderneutral race neutral standardsto military service members. Right?These are fighting units. This isn't,you know, this
00:20:03isn't ping pong. This is,you know, this is war fighters are haveto [clears throat] be capable to docertain things. And he was intent onrestoring those things. So that's sortof a fitness standard if you if you wantto think about it that way like what isa what does a fitness mean for militaryand then there's the fitness of def theindustrial base and what you give thewar fighter to execute what tasksthey're given that's weapons it'sdefense defense systems it's cyber cybereffects and so on we noted in thisconversation defense industrial basesort of wasn't ready to provide what'snecessary for the threat environment andif you think about threat environmentwhat we've done in Iran what we've donein Venezuela, uh what we've done withthe Houthis, like all just in one year,right? Not to mention what might come inthe years ahead. You think about were wepostured for that with with the defenseindustrial base. And the answer was weweren't as postured as we wanted to be.So he said, "Okay, my next fitnessstandard is to go to industry and tellthem that we want them and need them toproduce. We have their backs. We'regoing to bring all the resources tobear. We're going to cut out thebureaucracy. We're going to make it anopen wide front door for new entrance tocome in and help us solve problems.We're going to change the way we buy theway we uh do RFPs. Even like we
00:21:18for anRFP, it used to be here's a 300pagerequirement document. Please fill it outand we'll decide if you win or loss. Nowwe say, well, here's our problem. We'retrying to achieve this. Why don't youcome back to us with a solution? And youallow them to to come in with differentideas about how to solve the problem.And there's novel concepts out there. Soit's a different. So the primes willexcel at filling out a 300pagequestionnaire.>> A startup won't. So if you change itinto here's the problem I'm trying tosolve. Well, startup may say, "Oh, Ihave I have ways of solving this." Sothat Arsenal freedom tour is aboutgiving them permission and promoting andmaking sure they understand that we'reserious about the fitness and theexpansion and health of the defenseindustrial base. So that's the ArsenalFreedom Tour. Um, so it's been a lot offun. and been we've been gosh six dayson the road in the last two weeks andvisited about 20 companies and and it'sbeen awesome both the workers and thetech together you know again thesecombined hardware software systems umit's really been inspiring to see themotivation um in in Southern Californiain Texas in North Carolina and a lot ofthe places we've gone>> amazing the other audience uh that um Ithink is going to be inspired by the youknow leadership that they recognize orjust pace
00:22:33of doing business like um isentrepreneurs, right? And you havetalked about the need for more newprimes and like people to solve some ofthe problems that you're looking at andare important to American dominance andcompetitiveness. um what shouldentrepreneurs I I think like it's it'swonderful that um you know you've beenso early in this and a lot as wellinvesting in defense technologies thathas now become much more mainstream interms of and like it that soundsridiculous to say but it is acceptableand exciting to work on American defensenow what should people look at in termsof opportunities and what shouldentrepreneurs know about doing businesswith the department of war now>> I would say that so a few things sonumber one we're more open for businessthan ever before because there's arecognition that our adversarieswho started from a cleaner sheet ofpaper in the last 1015 years like thebuildup the military buildup in China isthe biggest military buildup in in worldhistory and so there's a real uh urgencyon our side to ensure that we are aheadbut we stay ahead and that's going totake a different level of investment anddifferent type of thinking than we'vehad in the last 20 years in the in 80sthere were 50 defense contractors andthey were got
00:23:48merged so they're onlyabout five now. So there's a lot of roomfor new entrance and and it's crazy tome that SpaceX and Andreal and Palanteerall had to sue the Department of War fortheir first contract. [snorts]So the idea is you don't have to sueanymore. Come through the front doorbecause people are not going to fightyou. We're now excited about lower cost,faster, more techly sophisticatedoptions. The caution I giveentrepreneurs is it's still not likeselling to another business. So don'thave illusions that you know this issort of a deal with Uber deal with MX orsomething. This is still a governmentand we have to make sure stuff worksbecause lives are at stake, right? Thisis not just uh just money.>> Uber MX is your example of like the easystuff, [laughter] man. Yeah,>> you know, a lot of entrepreneurs outthere are like, I'm going to sell to mybuddy NYC today. Um, so we're we're[clears throat] several tears up in thehierarchy of pain, but [laughter]>> yeah. Okay, fair enough. Fair enough.That So, yes, I consider that was it waseasy at least in my day. And then theother thing is when you're selling acombined hardware software product,um, you also have you have tomanufacture it. And that's not somethingthat goes back to the hardware is hard
00:25:03in Silicon Valley. Well, hardware ishard also in the defense industry. Andso producing things at scale that areflawlessu means you have to have a differenttype type of capability. Elon's masteredit with satellites and and all that, butthere's not a lot of startups that havemastered it. So mastering theproduction, the scaled manufacturingpiece of it is a thing that startups whoare selling to the defense departmenthave to also prove hire for demonstratecompetence in and so on. And that's theone thing I'm learning in my my firsteight months here is that's the piecethat I'm encouraging them like, hey, ifyou come come with a tech pitch and aprototype, that's great. What's the planto get me 10,000 of these? Right? And sothat's sort of the the advice I'd giveto the community.>> How do you think about flexibility um ofbudget? So, one of the things that oftenhappens is, you know, there there's alot of line items in the congressionalbudget that basically establish certainprograms that are then the ones that areuh implemented and sort of paid for orprocured. And you know, that's somewhatdifferent from how you think aboutinnovation in terms of something newcomes up, you want to rapidly be able toiterate, to launch it, to deploy it. Andso, if how do you think about eitherflexible budget spend or reallocation ofbudget over time? And how does thatrelate
00:26:19to the legislative process andhow does that impact entrepreneursultimately? Yeah, it's been it's been aproblem such a problem that they'vegiven like an awful name to it calledthe valley of death which [laughter] Iwhich Evo you'll hear eight ways toSunday here and the ways you know we'vecome at it and I've come at it is wehave this defense innovation unit whichis rapid contracting has a billiondollars has like a reasonable amount ofmoney to do things really fast to getstartups off the ground and get themthrough. We have several other programsthat have that same capability. One'scalled AFIT that takes companies thathave developed a product but now need toscale their manufacturing. So it's adifferent line in the pro you knowdifferent point in the process. I havethe office of uh strategic capital whichhas $200 billion in lending authority umlowcost loans and that both sends ademand signal to these companies thatother private capital crowds around itequity capital often and it it's lowcostloans right it's treasuries plus 100bips and in the last I don't know fouror five months we've done five criticalminerals deals like really fast becausethat's that's a target area and we'llhave other target area. So
00:27:34if you're acompany that's in one of the the supercritical areas, um we have this hugelending authority there, too. So I'mtrying to collapse the value of death bycrowding capital around in differentparts of the capital cycle. Um and thenalso we're reserving a bit of the budgetevery year so that we can make in-earbudget decisions because what what whatthe real problem is that Congressdoesn't pass funding bills until halfwayinto the next year for the year ofexecution. And so if you're not planninga year and a half ahead, you're notyou're going to miss. So all thesefunding cycle mechanisms are a way toget at that valley of death. And I thinkwe've got we're really close to sort ofpinching pinching that into to beingvery minimal if if it exists at all. Andand by the way, helps that you guys areinvesting in defense tech also, you andother your your peers, because>> um equity capital has a lot of has a lotof uh value here too. when you seepromising companies and we see promisingcompanies, you're going to say, "Oh,this is a promising company. They've gottheir DIU deal. I'm happy to fund themnow and that, you know, to help solveit, too.">> You've also, I think, advocated for dualuse companies. They're doing things fordefense as well as the broader world.Are there any specific areas that youmost wish a company existed today forwhere, you know, you think it's a sparselandscape or other areas
00:28:49that if thereare founders watching this, you thinkthey should really focus from aninnovation perspective? Yeah, I meanthere's there you know this hypersonicweapons thing is is kind of a big deal.There is a startup um or you know verysmall number of startups that areinnovating here but former SpaceX peoplewho are running around who understandthe physics of of that kind of thing. SoI wish there were more. I think there'sgoing to be when you think about dronesnot just the small drones get the biggerand bigger drones that really look likelong range missiles if you will beinnovation there. There's innovation inthe technology that goes into a dronethat seeks things in a GPS deniedenvironment. Um, you know, it makes themmore precise. For Golden Dome, whichyou've heard about, we're looking at howto take things out from space, which isa compelling, really complex butinteresting proposition.>> So, there's those are kind of someareas. And then from an AI standpoint,you know, I feel like we're generallyokay with the investment. I mean there'sso much money going in there that that'slike a well-funded area both thefoundation models and the derivativesthat are trying to do specific apps youknow uh on those models. So, you know,that's some direction for folks>> when you think about the pace of
00:30:04uhtechnology adop. I'm just thinking aboutyou as like the CTO of a 3 millionperson organization, you know, whetheror not those people are war fighters,right? Um and and you know, selling ifyou have portfolio companies that sellto very large organizations, you see alarge span of adoption of those toolsover time. is amazing that genai.millhas gone a million users recently. Um,how do you think about like the theworkforce, right? You said it's thebiggest single organization in the worldthat's consuming this technology, butyou have to do a bunch of changemanagement. Is that something that youyou think through?>> Yeah, I mean, I try not to think throughit so hard so that I'm like, "Oh mygosh, this is going to be too hard todo." Um, sort of the again the Uberentrepreneur mentality. It's like let'slike move and see how far we can getthings into the enterprise until someonesays stop. Um so it's a different way ofthinking about it.>> Just make forward progress as fast asyou can. That's a good strategy too.>> It's you know it tends to work. Does itwork as fast as in as in privateindustry? No. But it still works.There's still human desire to do betterin in every case. And I do think uhbureaucrats get a bad rap wherebureaucracies get
00:31:20deserve the bad rap.Right? It's the collection of rules,regulations, people, culture that end upbeing hard to change. Whereas any oneindividual person, they want to thrive.Especially if you're a department ofwar, you kind of you don't you don'tcome to Department of War if you're notmissionoriented.You don't have sort of an ethic thatsomething you're working for somethingbigger than yourself.You're certainly not coming here to makea buck, right?
00:31:51And that gives us a lot of really goodmotivation from people. And if I couldjust get some tools in their hands andtalk to them and give them someleadership and some uh you stuff, yousurprising how receptive you'd find evenin this large of a department. in adifferent direction because youmentioned like I was just thinking aboutsome of the hardest problems that youhave to go work on. Um one is likerebuild the defense industrial base,right? Uh and how do you like reignitethat ecosystem when um a lot of peoplehave given up on American manufacturingin different domains? What do you thinkare like the the necessary missingcomponents if you know you're saying thedemand is there? Because I think a lotof the arguments has just been like it'stoo expensive. we no longer have thecapability or the knowledge in theUnited States. Like how do you how doyou think about that?>> I mean, I think in some cases it's amatterof national survival, right? In in sofar as like I'm I'm as free market aperson as as anyone. But I think if youlook back on the last 20 years and sayshould we have outsourced all ourpharmaceuticals, all our criticalminerals, all our semiconductors,all of it,
00:33:06all of our nuclear powerplant building capabilities.So that the vi like if there's ascenario where you're ad the adversaryhas control of those things, you don'thave control of your own destiny. Youwould say that well it's worth the costto bring it back. And bringing it backis not a snap your fingers thing, butwith American ingenuity andentrepreneurship and capital markets, Ithink we can do it and we have to do itand we are doing it and will do it. Soyou pick, you know, it's a little bit ofa Maslo's hierarchy of needs. Like whatare the most important things we have togive back? Well, this deal with Chinasays that in a year we have to, youknow, the critical minerals exportcontrol thing. You know, you know who ifassuming that stays there. So, we'relike operating like we have to beself-sufficient in a year. And then yougo down the list of of things and youtry to tackle them. And the good news isthe enthusiasm about doing it is againwidespread. And it's some it's I Iattribute it to President Trump. Iattribute it secretary secretary Hexat.I attribute it to just smart peoplelooking at the playing field and sayingah you know we have to be proficient insome of these things and we have to doit ourselves. you know, there's 200pharmaceuticals on the critical shortagelist because we're dependent
00:34:22on othercountries. So, that's not that's not aviable thing. We don't want that for ourchildren our children, right? So,whatever it takes, we have to do it. AndI'm relying on folks like you and thepeople you invest in to to help on that.And I think we're going to we just haveto create the market conditions whereit's viable.>> My optimism here amounts to in certainareas there is demand and that was notactually recognized before, right? likewe need to make it here and we need tohave the ability to make it. And like ifif there's a massive regrowth ofAmerican manufacturing, it will not bein exactly the form it was like 40 yearsago, right? It will be because it isdriven by automation, better processes,AI, like being able to make things otherpeople are incapable of making versuscompeting in necessarily a globalmarketplace for cost. Um, but I thinkthat's totally doable. That's exciting.>> Yeah. Yeah, but the one thing I wouldadd there is like who's going to who'sgoing to uh dig and weld and electrifyall the data centers we're building?There's going to be people to a largedegree. There'll be robots doing some ofthis, but there'll be people. And who'sgoing to build all the new housing weneed in these cities because of thehousing shortage? Who, you know, sothere's there's going to be work. It maynot be in an auto plant to the degree wehad workers in auto plant as apercentage of sort of the workforce, butthere's
00:35:37going to be other big jobs. Imean the welding jobs we're trying tobuild ships here>> can't find enough and it's a job thatpays six more than six figures in areasthat are not high cost to live in. Soit's like how do you point people withthe right capital trade do the righttraining and so on. So I think there's alot of opportunity uh for all categoriesof workers. Um but yes you're right insome cases you know on farms are yougoing to have robots pick strawberry?Are you going to are you going to userobots and automation to figure fillholes that are that we can't otherwisefill? Like yes. And we're bestpositioned to do it.>> And I I was really also thinking in thelike effective productivity sense,right? Like American workers with betterautomation and better lines and all thatcan produce things that are bestin-classand cost effective.>> The cost effective thing too issometimes a little bit overstated interms of um there are subsidies forexample for our earth minerals by someof our adversaries etc. And so I don'tthink these have acted as free markets.I think in many cases there's been verysmart strategic action by China andothers to ensure the market isn't a freemarket so that there isn't competition.>> Yeah. And that's and that's a greatpoint because if it was a true freemarket it wouldn't be as it wouldn'thave disrupted but there was a lot ofdumping
00:36:52going on in certain areas tocrush the other industries. I'm somewhatfamiliar with this from the Uber Liftdays, but but the there's a lot of howhow do you inject capital into in acertain way that gives you an an unfairadvantage and and some, you know, peoplecould debate the tariffs, but some ofthe tariff notion is to balance thatdumping uh capability. So if you couldin theory have a free market price forsomething well if we're if we can'tcontrol one side of it we have at leastsome levers on the other side of it totry to create some balance.>> So you know I guess related to that youall did some really interesting dealsaround uh rare earth minerals throughthe office of strategic capital. Couldyou talk about some of the otherinitiatives that you have going on thereright now? Yeah. So those are targetedat what part of the defense industrialbase has a brittle supply chain wherethe technology exists. There's actuallyprobably a company that actually does ithere, but we need them to produce 10times more than they're producing now.And so we've identified, we went throughlike exhaustively these supply chains orwhat are the things that go into this?Like who knew Germanmanium was a thingthree years ago? The American publicwasn't thinking about it. So what we'retrying to get ahead of it is what arethe 10
00:38:0720 things that are componentswhether it's a brushless motor abrushless motor goes into a drone we'redependent on chi the on China forbrushless motors so uh you know whatkind of semiconductors besid besidesTSMC go into sort of a data center whatare other things we have to make sure wedomesticate and how do I um ensure thatthere that I'm taking the companies thatexist and making sure they have thecapacity to serve the whole nation'sneeds. needs and that's pretty excitinguh too because that's another way of ofgetting at um the supply chain fragilityand our adversary dependency. So youkind of can solve potentially both. Weused to have a very robust defenseindustrial base and an industrial baseperiod. Now you're down in some casesone supplier of a nozzle that goes in 20different you know systems and I wantthat one nozzle maker to succeed but Ialso maybe want two or three so thatwe're innovating while we're doing thatand that's part of the mission of theoffice of strategic capital and we havea lot of money to do it.>> I think one of the things that's umreally cool about studying some of thisuh stuff on the supply chain side isactually like how distributed it is. Umobviously you know you spent a long timein Silicon Valley a lot of innovation isgoing to happen
00:39:23in you know in SouthernCalifornia and Silicon Valley but Iremember like as you said I didn't knowwhat geranium was I remember looking itup when there was this department of warinvestment in uh lattis materials and Iwas like it's in Bosezeman Montana rightthat's where we are making you knowgermanmanium and silicon crystal rightand we need it for defense optics and soI I do think that even just like showingpeople how much demand there is for likethe entire supply chain across differentcommunities is really important.>> Yeah, I think that's right. I mean, youknow, fiber optic cable, we don't makefiber optic cables in this country.Fiber optic, you know, basic fiber,which they're using in the UkraineRussia war to do drones because they'rejamming and all that and we use fortelecom and all that. So, so yeah, it'screating that list, communicating itbroadly,>> um, trying to ensure the marketconditions are such that we're notsubject to the dumping sort of practicesthat could, you know, really kill anindustry very early on. So, it's acomplex set of things, which is why whya research and engineering organizationwould have a strategic capital arm,right? That's why those two things gotogether, which is not apparent from theoutside, is because both things gotogether a bit. Is there a listpublished anywhere where founders whoare listening in can go and see whatthey should be building or areas where
00:40:38there's more demand than either peopleunderstand or is anticipated or wherethe DO is willing to help providecapital or other things to accelerateit.>> So the DIU for defense is like four orfive categories of things that they lookat. So that's that's sort of a publishedand public thing. The office ofstrategic capital is a little moreproactive. They're going to say, "Wehave this whole I'm going to go findcompanies that already exist that dothis." So, it's not for startups becauseit's a lending facility, if you will.So, they're looking for companies thatalready exist and how do I expand theirproduction? Uh, but it's a good pointlike, you know, at some point, I wastalking to the secretary about this theother day. It's like we should have someforum for>> for talking about the things that weneed more proactively so investors andentrepreneurs can think about them.Because if you're expanding capacitythrough loans, you could also expandcapacity through new companies doing thesame thing. And so that's why I'mwondering is there good point>> mechanism to get>> we're just looking for a to-do list.>> Yeah. Yeah. Exactly. [laughter]>> Just to be clear. Yeah.>> Exactly.>> I I just have a last question for you.Um you're you know couple couple monthsin guns blazing in terms of uh youimpact already. what most surprised youor is like a thing you've changed about
00:41:53how you operate or something you knowyou learned>> and and this is your sec second stint ingovernment by the way right and so it'dbe really interesting to contrast firstin versus second stint through that lensin terms of what's what's shifted orwhat's different this time around>> yeah so I was a white house fellow callit a glorified intern in 2009 to 2010and we're in the middle of the globalwar on terror and so what were fightingat that time was, you know, a bunch ofinsurgents who had sort of crude weaponslike improvised explosive devices and soon. Now we have a nearpeer adversarythat has space capabilities,you know, long range strikecapabilities, drones, AI. So thebattlefield has shifted so dramaticallytoward being tech superior than it wasback then. back then we were it was morelike we had the superior tech but thenwe were just vulnerable by umadversaries who had nothing to lose thatsuicide I mean it was a totallydifferent kind of adversary so that'sbeen a dramatic change and I didn't youknow I guess I did realize that and partof
00:43:08why I took this job is that I sawthat change and I wanted to be part ofsort of the solution there um but Ithink what surprised me is that the therecognition broadly speaking speakingthat this a this nearper adversary hadbeen doing so much and so little time ina concentrated way didn't really hit meuntil I got in and started sort ofgetting you know really understandingwhat had happened to some of our marketslike again the Germanian example thefiberglass example the Brussels votersexample uh the importance of Taiwanthese things you know when you're on theoutside you you understand themtheoretically but now I understand themviscerally>> well amazing thank you so much foreverything you're working towards today.I think um obviously you're one of thepeople that's been motivated by showingup and trying to have a big impact forthe country. So, you know, it's uhexciting to to watch what's what'shappening and changing.>> Yeah. Thanks for your for both of yoursupport and I will come back to you withthe to-do list. [snorts]>> We're ready.>> Wonderful. Yeah.>> All right. Thank you so much.>> Good. Good talking to you.
00:44:12>> Find us on Twitter at No Prior Pod.Subscribe to our YouTube channel if youwant to see our faces. Follow the showon Apple Podcasts, Spotify, or whereveryou listen. That way, you get a newepisode every week. And sign up foremails or find transcripts for everyepisode at no-briers.com. [music]