00:00:00
Trey Stevens, Sean Sankar, welcome to
00:00:02
the All-In podcast at the Hill and
00:00:04
Valley Forum. Thank you guys for being
00:00:05
here. What's up? How are you guys doing?
00:00:06
>> Thanks for having us. Doing great. Good
00:00:08
to be here, too.
00:00:08
>> You guys are friends, right? You guys go
00:00:10
back like a long time.
00:00:12
>> A really long time.
00:00:14
>> Okay. Tell us how you guys know each
00:00:15
other. Palunteer Ander. What's the
00:00:18
connection and the history?
00:00:20
>> Well, I'll start. You can fill in all
00:00:21
the gaps.
00:00:22
>> I feel like I know the story you're
00:00:23
going to tell and it's going to be very
00:00:24
uncomfortable.
00:00:29
Every few years, a new ad channel opens
00:00:32
before the market catches on. That's
00:00:33
axon.ai right now. The AI ad platform
00:00:36
behind one of the biggest runs in tech
00:00:38
with access to over a billion daily
00:00:40
active users. Full screen video ads in
00:00:42
mobile games watched for a median of 35
00:00:44
seconds. Businesses are profitably
00:00:46
spending hundreds of thousands of
00:00:47
dollars a day on it, and most
00:00:48
advertisers don't even know it exists
00:00:50
yet. The window is open at
00:00:51
axon.ai/allin.
00:00:55
So Trey, I think you know in the early
00:00:57
days of Palunteer, I was roaming around
00:00:59
giving demos to anyone who could
00:01:01
possibly want to see them. Trey was
00:01:03
working in an intel agency and happened
00:01:05
to see one of these demos and he should
00:01:08
tell you his version of it, his side of
00:01:09
it, his frustration with bureaucracy,
00:01:11
but I think he he realized like, hey,
00:01:13
this this might be really cool. Maybe I
00:01:14
should uh leave the hell hole I'm in in
00:01:17
the basement of this building getting
00:01:18
nothing done, talking about sports with
00:01:20
other people to uh to go join this
00:01:22
crusade. So Trey reached out and
00:01:25
applied. Um he made a big faux paw. He
00:01:28
came all the way to Palto and he he wore
00:01:30
a fullon suit tie, cufflings, CIA
00:01:34
cufflings, mind you.
00:01:36
>> Coming to interview with I don't know,
00:01:38
we were probably 20 people, you know,
00:01:40
who wore t-shirts and, you know, second
00:01:42
handme-down clothes. Um and I, you know,
00:01:45
he was intercepted in the lobby
00:01:47
receptionist who really cared about him
00:01:48
and told him to get ditch the tie and
00:01:50
try to dress down and don't screw it up
00:01:52
too bad. But we loved him immediately
00:01:54
and he helped us build the uh the
00:01:56
government business.
00:01:57
>> So this was at Palunteer. You were their
00:01:59
employee 13 2006, right? So this was
00:02:02
pretty early on.
00:02:03
>> Yeah. I came in in early 2008, but there
00:02:06
were still, you know, 25 30 people at
00:02:09
that point.
00:02:10
>> Pretty small.
00:02:10
>> And Peter kind of incubated it, was
00:02:13
involved at the beginning, right? Peter
00:02:14
Teal. Maybe you could just recap because
00:02:16
I know a lot of folks know the history
00:02:17
of Palunteer, but just kind of like the
00:02:19
early standup of Palunteer and how
00:02:20
things got going during that era with
00:02:22
that small group, how you guys kind of
00:02:24
figured out how to build the business.
00:02:25
>> Yeah, it was it was one of these things
00:02:26
that was kind of a slow start. It was
00:02:28
there was a a real idea amongst the uh
00:02:30
the five co-founders, including Peter,
00:02:32
that um you know, it's kind of insane to
00:02:34
live in a world post 911 where people
00:02:36
are arguing about what's more important,
00:02:38
privacy or security. Like, aren't they
00:02:40
both really important? And who is
00:02:42
actually spending time pushing out the
00:02:44
efficient frontier? For any amount of
00:02:46
given security, you should have more
00:02:47
privacy than you had before or any
00:02:49
amount of given privacy, you should have
00:02:50
more security. And and this sort of
00:02:53
changing the dialectic there was really
00:02:55
the the entire impetus of what we
00:02:57
started with. And now there's a
00:02:58
technical approach that follows from
00:03:00
that. There's an approach to privacy and
00:03:01
civil liberties that are around that. Uh
00:03:03
but really we started as a business that
00:03:05
was pretty myopically focused on solving
00:03:07
a handful of problems and
00:03:08
counterterrorism for a handful of
00:03:09
institutions in the world. Let me start
00:03:10
by asking a question that I think is
00:03:12
important to ask particularly for a
00:03:14
broad audience and for the two of you to
00:03:17
frame your personal philosophical views.
00:03:20
Is war good? There's a lot of
00:03:23
conversation about there is a
00:03:24
military-industrial complex that has an
00:03:26
incentive for war. What's your view,
00:03:29
your philosophical motivation for why
00:03:31
you do what you do? What your view is of
00:03:33
war and defense and and the work that
00:03:36
your businesses kind of pursue? Well,
00:03:38
anyone who's been to war would tell you
00:03:39
that war is awful. War is bad.
00:03:41
Categorically bad. That doesn't mean
00:03:43
it's always avoidable and that um you
00:03:45
know there will there are people who
00:03:48
will want to use might to make right to
00:03:50
define a set of rules. Um and you have
00:03:53
to be in a position to protect your
00:03:56
people and your interests accordingly.
00:03:58
>> At the end of the day, it's all about
00:03:59
deterrence. you don't want to go to war,
00:04:01
but you want to be prepared so that if
00:04:03
you do have to go to war, that you will
00:04:05
win decisively and quickly. Um, I've
00:04:08
never met a general that has said, "You
00:04:10
know what I really want to do today? I
00:04:12
want to make phone calls letting parents
00:04:13
know that their children have died in
00:04:15
combat." Nobody wants to do that. Um,
00:04:17
and and I think that that's really the
00:04:19
goal of everything we were working on at
00:04:20
Palanteer. It's what we're working on at
00:04:22
Ander, which is make it unthinkable to
00:04:24
your adversaries that they should ever
00:04:26
challenge you. Why do you think it
00:04:27
became taboo and became so negative
00:04:30
particularly in Silicon Valley that
00:04:32
building a defense company or company
00:04:33
building technology to service the
00:04:35
defense sector was viewed with with such
00:04:38
disdain and was an untouchable sector
00:04:41
for so long.
00:04:41
>> I think it's a beautiful consequence and
00:04:43
I mean I mean it unironically of kind of
00:04:45
the peaceful world that we lived in. The
00:04:47
origin story of Silicon Valley is
00:04:49
actually defense. You know Lockheed was
00:04:51
the largest employer in in Silicon
00:04:53
Valley in the 1950s. The Corona spy
00:04:55
satellites were built there. So in a
00:04:56
world that was uh gripped by the threat
00:04:59
of the Soviet Union, you know, you had a
00:05:02
very different posture in Silicon
00:05:03
Valley. But in a world post the end of
00:05:06
the Cold War, the end of history, um
00:05:09
kind of a view of globalism that it's
00:05:11
all going to be great that we exist
00:05:12
beyond the confines of our country. Um
00:05:15
these things were viewed much more
00:05:16
cynically and the threats didn't seem
00:05:17
very real. And I think you could even
00:05:19
back test this through the moment where
00:05:21
Silicon Valley kind of really woke up
00:05:23
and realized, hey, maybe these things
00:05:24
are real or when the Russian tanks
00:05:25
rolled across the Ukraine border, you
00:05:27
know, is when they realized and there
00:05:29
are a lot of Ukrainian, Eastern
00:05:30
Europeans, a lot of people who are
00:05:32
affected by that in Silicon Valley who
00:05:34
started to realize that um maybe there's
00:05:36
a more nuanced issue here than just
00:05:38
simple the simple Eisenhower quote of
00:05:41
the military-industrial complex. I think
00:05:42
many of the people who were protesting
00:05:45
Silicon Valley's involvement in working
00:05:47
on national security priorities were the
00:05:49
exact same people that had Ukraine flag
00:05:51
in bio. Um so there's clearly just like
00:05:54
a policy mismatch or an understanding
00:05:56
mismatch. The other thing I would
00:05:58
mention is that these large tech
00:06:00
companies uh unlike many epics prior uh
00:06:03
were global technology companies. They
00:06:05
didn't necessarily view themselves as
00:06:07
American. Uh and so if you go and you
00:06:09
look at, you know, where these protests
00:06:10
came from, you have to keep in mind that
00:06:12
a lot of the signatures that they had on
00:06:14
these, you know, protests, uh were not
00:06:16
coming from US citizens. And so there
00:06:18
there's like a global character to those
00:06:20
companies that probably was not the case
00:06:22
during the Cold War.
00:06:23
>> Do you think that's changed recently?
00:06:25
>> I mean, to Sean's point, I think that
00:06:26
there's just an increased awareness of
00:06:28
the complexity of the geopolitical
00:06:30
situation. Um, and I I think it's
00:06:33
certainly a lot less controversial work
00:06:35
on these topics than it was in 2017,
00:06:37
2018.
00:06:38
>> You know, it's a 20-year overnight
00:06:39
success, right? And Ander's now
00:06:42
reportedly raising money at a $60
00:06:44
billion valuation, just landed a $20
00:06:46
billion army contract. Palanteer today
00:06:49
is worth $400 billion.
00:06:51
The flip side of what I just said is
00:06:54
that some people are saying that Silicon
00:06:55
Valley is taking over defense and
00:06:58
Silicon Valley is the next story of the
00:07:02
American war machine. This has become
00:07:04
sort of a popular narrative. Can you
00:07:07
just respond a little bit to kind of
00:07:08
where we were coming out of World War II
00:07:11
from a defense industry perspective,
00:07:13
where we find ourselves, where Silicon
00:07:15
Valley seems to be at the center of this
00:07:17
today. You know, the industrial base
00:07:18
that won World War II and the early Cold
00:07:20
War was not a defense industrial base.
00:07:22
It was an American industrial base. You
00:07:24
know, Chrysler made the Minute Man ICBM,
00:07:27
they were the prime contract on it, and
00:07:28
they make minivans, so missiles and
00:07:30
minivans. General Mills, the serial
00:07:31
company, had a mechanics division.
00:07:33
Everything they learned doing R&D to
00:07:35
process grains. They actually used to
00:07:36
build torpedoes and inertial guidance
00:07:38
systems. Ford built satellites until
00:07:40
1990. So the entire economy was invested
00:07:43
not only in pro economic prosperity but
00:07:45
also underwriting the freedom that
00:07:47
allowed us to have economic prosperity.
00:07:48
It's really a consequence of the end of
00:07:50
the cold war. So when the Berlin Wall
00:07:52
still stood in 1989, only 6% of spending
00:07:55
on major weapon systems went to pure
00:07:57
play defense specialists. 94% of it went
00:08:00
to what I call as dualpurpose companies.
00:08:02
You know, yeah, a missile is single use.
00:08:04
It's not a dual use product. You're not
00:08:05
going to buy it at Walmart. But that
00:08:06
actually these companies were invested
00:08:08
in both parts of this. That figure today
00:08:10
is 86%
00:08:12
goes to defense specialists. So we have
00:08:14
a very different structure of the US
00:08:16
economy as a result. And I I think it
00:08:18
leads to very perverse narratives of
00:08:21
what would it be like to mobilize if
00:08:22
things got really bad. We'll just flip a
00:08:24
switch and our auto factories will
00:08:26
magically turn into enabling us provide
00:08:29
for our defense and security. And and I
00:08:31
think a a coldeyed look at history is it
00:08:33
took 18 months to do that. we actually
00:08:36
started ramping up to provide this stuff
00:08:38
for our allies, for the Brits and the
00:08:39
Soviets in the context of World War II.
00:08:41
And that we've actually missed some of
00:08:42
those signals today. When when Ukraine
00:08:44
went through 10 years of production in
00:08:46
10 weeks of fighting, that probably
00:08:48
should have been a five alarm fire that
00:08:49
we got the fundamental calculus on
00:08:51
deterrence wrong. We thought the
00:08:53
stockpile was going to deter our
00:08:55
adversaries. It was always the factory.
00:08:56
It was the ability to generate and
00:08:58
regenerate the stockpile.
00:08:59
>> Well, let's look at where we are today.
00:09:01
So in terms of defense readiness
00:09:04
relative to our adversaries 10,000 to1
00:09:06
drone production gap versus China 223x
00:09:10
ship building capacity disadvantage and
00:09:13
I think you've said a 2027 Taiwan window
00:09:16
of danger high level we spend more than
00:09:19
any other nation on defense. How ready
00:09:22
are we and how long will it take us?
00:09:25
Well, our joint force is the best in the
00:09:27
world. And so, you know, I think we have
00:09:29
to couch the alarmism and commentary in
00:09:31
that context here. But I think if you
00:09:33
look at like the the rate of change, our
00:09:36
adversaries are moving very quickly. If
00:09:37
you just look at the empirical loss of
00:09:39
deterrence, you had the annexation of
00:09:41
Crimea in 2014, the militarization of
00:09:43
the Spratley Islands in 15, Iran with
00:09:45
breakout capability to get the bomb in
00:09:46
17, you've had a pogram in Israel, you
00:09:48
have the Houthis holding trade hostage
00:09:50
in the Red Sea, not to mention the
00:09:52
present conflict. So when you look at
00:09:54
that you how can you not feel that
00:09:56
deterrence is eroding to to Trey's
00:09:58
earlier point of having capabilities
00:09:59
that are so scary that picking a fight
00:10:01
isn't worth it and people are investing
00:10:03
significantly on the low end of the mix.
00:10:05
Our high end is still unquestionably
00:10:07
amazing. But you know you can't keep
00:10:10
shooting $2 million interceptors at
00:10:11
$20,000 drones and have that math work
00:10:14
very long. I think the second part like
00:10:16
why does it matter that we lost the
00:10:17
American industrial base and have a
00:10:19
defense industrial base is you lose
00:10:20
volume. you lose the R&D stimulus to
00:10:23
come up with creative ideas like using
00:10:24
the methodology of building a bathtub to
00:10:26
build the next generation lowcost cruise
00:10:28
missile, right? And and you get stuck in
00:10:30
these platforms that are absolutely
00:10:32
eyewateringly amazing but are $2 million
00:10:34
a pop and you just can't produce them at
00:10:36
the scale or with the speed that you
00:10:38
really need to.
00:10:39
>> Right. And Trey, you've talked about
00:10:40
this. You wrote an article, no solvency,
00:10:42
no security. just share with us your
00:10:44
view on the importance of this
00:10:46
industrial base and what it takes for us
00:10:48
to have strong defense requisite in a
00:10:51
strong manufacturing industrial base. I
00:10:53
>> I mean we really kind of sent most of
00:10:55
these capabilities away um during the
00:10:58
last 30 years of globalization um and
00:11:00
that gutted entire communities in the
00:11:02
United States. You know, I mentioned in
00:11:03
this article that just in my immediate
00:11:06
family, like grandparents, aunts,
00:11:07
uncles, uh I have family members that
00:11:10
worked at GM, at Ford, at Frigid Air, at
00:11:13
National Cash Register, at Armco Steel.
00:11:16
Every single one of these factories in
00:11:18
Ohio closed. Not a single one still
00:11:21
exists. And they all have different
00:11:22
places they've gone around the world. Um
00:11:24
but you know, one of the things we're
00:11:26
doing with Andrew right now is we're
00:11:27
building out a 5 million foot factory
00:11:29
campus in Columbus, Ohio. And we have
00:11:32
the benefit of going back and tapping
00:11:34
into that knowledge base that exists
00:11:36
that's undermployed at the moment
00:11:38
because those the factories that kept
00:11:40
them busy for many many years are are
00:11:43
closed. Um and I think this is the story
00:11:45
all over America. Like if you think
00:11:47
about the atscale manufacturing for a
00:11:50
new company started in the this century
00:11:53
uh in the 2000s um there's really only
00:11:56
one that comes to mind just Tesla.
00:11:58
That's it. And so we haven't really
00:12:01
built any muscle around uh new
00:12:03
manufacturing capacity. Um and and I
00:12:06
think that as you spin things up to
00:12:08
Sham's point, we don't have the ability
00:12:09
to do that quickly because we've
00:12:11
atrophied massively by by turning all
00:12:13
this away globally. So um I think
00:12:16
readiness is a real problem and uh we
00:12:19
have to start investing ahead of the the
00:12:21
you know conflict being being there
00:12:24
because you can't just start at the
00:12:26
moment that it's absolutely needed. Tell
00:12:28
us about the Arsenal one and then the
00:12:29
manufacturing behind it. And then I want
00:12:31
to just understand a little bit about is
00:12:33
this productled or production capacity.
00:12:36
Do we need to make selections and have
00:12:38
capital committed to bringing product to
00:12:40
market before we start to build out the
00:12:42
base or is there an alternative way to
00:12:44
get ready?
00:12:45
>> Arsenal one is our the factory campus
00:12:47
that we're building in Columbus. Um the
00:12:49
operating system for that the Arsenal
00:12:51
platform um is really intended to reduce
00:12:54
the uh the cost the overhead that's
00:12:57
required to uh automate and you know
00:13:01
handle these processes in the most
00:13:03
efficient way possible. So there's like
00:13:04
kind of the software layer that sits
00:13:05
behind that. We're actually working with
00:13:07
Palunteer uh the foundry platform on
00:13:09
some of these programs as well. And uh
00:13:11
the idea behind this is that you want it
00:13:13
to be as modular as possible. If we
00:13:14
build out the factory campus and we said
00:13:16
we're just going to build furies here or
00:13:18
we're just going to build road runners
00:13:20
here, we're just going to build
00:13:21
barracudas here, that provides a
00:13:23
tremendous limitation in a in the moment
00:13:26
of a conflict to be able to be
00:13:27
responsive to the demand for specific
00:13:30
things that are relevant to that
00:13:31
conflict. And so we're kind of thinking
00:13:33
about this more like contract
00:13:35
manufacturers think about building
00:13:36
assembly capacity where they say, "Yep,
00:13:39
we make, you know, this VR headset for
00:13:42
Facebook. We make this VR headset for
00:13:44
Samsung. We make this VR headset for
00:13:46
Apple. Um, we build a skill set around
00:13:48
building optics systems. Um, but we're
00:13:51
going to build a bunch of them and we're
00:13:53
going we're going to be much more
00:13:54
effective at doing that by hitting the
00:13:57
network of scale. Uh, you know, get
00:13:59
lower costs on all the individual parts.
00:14:01
You're going to build out a supply chain
00:14:02
to to work through that. That's roughly
00:14:04
what we're doing in this case is we're
00:14:05
saying like a contract manufacturer, we
00:14:07
want to be able to pivot on a dime into
00:14:09
ramping up production of roadrunners if
00:14:11
we need road runners or ramping up
00:14:12
production of barracudas if we need
00:14:14
barracudas. Um the counter example to
00:14:17
this of course is what we saw in the
00:14:18
early days of Ukraine where what they
00:14:20
really wanted was more stingers and
00:14:22
javelins. The problem is is that we once
00:14:25
we burned through our inventory in the
00:14:26
warehouses the stingers and javelins the
00:14:28
assembly line to build stingers and
00:14:29
javelins didn't exist. All the people
00:14:32
that worked on those assembly lines were
00:14:34
retired. And so the primes were
00:14:36
literally calling people out of
00:14:38
retirement to come back and teach them
00:14:40
how to build stingers and javelins
00:14:42
again. Um, and so these are problems
00:14:43
that we're trying to avoid by the design
00:14:45
of the factory as a as an initial
00:14:47
concept.
00:14:48
>> Who funds it? So is there a government
00:14:50
contract that supports the production of
00:14:52
these facilities? And are multiple
00:14:54
companies going to be able to to stand
00:14:56
these up to make these investments or is
00:14:58
this just going to end up being an
00:14:59
anderal because you've got most of the
00:15:01
capital and you guys are going to roll
00:15:02
these facilities out and how do we make
00:15:04
this kind of a national interest?
00:15:05
>> Well, our business model is definitely
00:15:07
very different from the primes uh who
00:15:09
are basically responding to things as
00:15:11
requirements directly from their
00:15:12
customer. So they're not really
00:15:14
investing a whole lot forward of the
00:15:16
capability. Um, and on the other hand,
00:15:18
we're doing all of this as private R&D
00:15:20
investment and then we're selling the
00:15:22
outcome, the output of that as a
00:15:24
product. So, this is fundamentally a
00:15:26
different business model. Um, but I I do
00:15:28
think that there's a tremendous amount
00:15:29
of capital that's required to pull this
00:15:31
off as a new entrant to the space. Um, I
00:15:34
don't think that the market is going to
00:15:36
support 100 new primes or something like
00:15:38
that. Uh, I'm hopeful that it's not just
00:15:40
Anderil. I think it would be healthy if
00:15:42
there was more competition in the space.
00:15:44
Um, but you know, being able to raise
00:15:46
that amount of capital, hire the people
00:15:48
that you need to actually pull it off.
00:15:49
Uh, this is not it's not like building a
00:15:52
a a normal tech company. It's not just
00:15:54
the development of a product. It's, you
00:15:56
know, there's a lot that goes into this.
00:15:58
Trey's point might seem kind of obvious
00:16:00
or simple, particularly to an audience
00:16:01
that's largely coming from tech or
00:16:02
Silicon Valley, but that's not how
00:16:04
defense works. People don't build
00:16:06
products. People build to a spec that
00:16:08
the government says, "This is what I
00:16:09
want to buy." And then you kind of say,
00:16:11
"Yes." You know, you have to think about
00:16:13
defense as a monoponyy where there's a
00:16:14
single buyer for the thing and that
00:16:16
concentrates an enormous amount of power
00:16:18
in the buyer whether they're right or
00:16:20
wrong. And if you look at the history of
00:16:21
defense innovation, typically the
00:16:23
monopsin is wrong. You know, it was
00:16:25
Churchill as the head of the Royal Navy
00:16:27
who built the tank because the British
00:16:28
army was not smart enough to realize
00:16:30
that in the next battle having horses
00:16:32
was not going to work. So, every one of
00:16:33
these innovations is kind of an act of
00:16:35
heresy. There there's a founder-like
00:16:37
figure who is so committed
00:16:38
pathologically to a different heretical
00:16:41
concept. They see it through and only in
00:16:44
combat or only when it meets the moment
00:16:46
is it actually validated. Now a grave
00:16:48
threat like the Soviet Union is a sort
00:16:50
of forcing function that allow allows
00:16:52
you to innovate. When we were building
00:16:53
ICBMs we had eight competing programs.
00:16:56
Today in peace time that would be viewed
00:16:59
fastly as wasteful. Why should you have
00:17:01
eight competing things? Why can't you
00:17:02
just pick the right one and put, you
00:17:04
know, spend less or put all your wood
00:17:06
behind one arrow? And of course, that
00:17:07
kind of turns its back on our
00:17:08
fundamental belief in America on the
00:17:10
free market that there's a competition
00:17:11
of ideas. There's fundamental
00:17:12
uncertainty. There's quality like
00:17:14
variance in the quality of the execution
00:17:16
and that you need this competition. And
00:17:18
it's not just principally competition
00:17:20
amongst the primes or new entrance. It's
00:17:22
also competition inside of government.
00:17:25
>> And there were originally, I think, or
00:17:26
at one point 51 major defense
00:17:29
contractors. And just for the audience
00:17:31
to understand, they reduce down to
00:17:34
roughly five or six primes. That's
00:17:35
that's the term prime is one of these
00:17:37
prime contractors and then make
00:17:39
subcontract out to other companies to
00:17:41
develop componentry parts. My
00:17:42
understanding in Anderil was to do
00:17:44
exactly this build a product, show up
00:17:46
with something that's cheaper, better,
00:17:48
faster, and actually competes on the
00:17:50
merits rather than be caught in this
00:17:52
kind of prime landscape. And that worked
00:17:54
clearly. But over time there seems to be
00:17:57
a lot of capital concentrating into the
00:17:59
new primes. You guys, you guys, SpaceX,
00:18:03
OpenAI, maybe one or two others. Is the
00:18:06
landscape going to emerge that we're
00:18:08
going to just look like the same old way
00:18:10
of operating where we have a handful of
00:18:12
primes that all have these trusted
00:18:14
relationships with the government
00:18:15
agencies or how do we kind of create
00:18:17
that competitive environment to continue
00:18:19
to drive innovation and make things
00:18:21
affordable for the US government and the
00:18:22
taxpayer going forward?
00:18:24
>> Well, just like uh and I interviewed
00:18:26
Trey in my new book on this like just
00:18:28
like venture, there's going to be a
00:18:30
power law here, you know. So, so one of
00:18:31
the grave mistakes we've made as we
00:18:33
think about innovation which has result
00:18:34
in essentially innovation theater is
00:18:36
this idea that you're going to just
00:18:38
peanut butter spread around the capital
00:18:39
you have for innovation that you know
00:18:41
every company is going to get roughly
00:18:42
the same amount. It's not enough to hit
00:18:44
scale. It doesn't reflect actually the
00:18:46
relative performance differences because
00:18:48
there's an authentic power law curve
00:18:49
like any venture capitalist would would
00:18:51
realize that you know your your biggest
00:18:52
winners are going to return your whole
00:18:54
fund. There's there's something
00:18:55
authentic here as well, which is like,
00:18:56
yeah, you may have 10 bets, but at some
00:18:58
point you're going to have to get smart
00:18:59
about concentrating down on on the
00:19:01
things that are actually working.
00:19:02
>> Yeah. I mean, this is a a kind of a
00:19:04
great point about any category. It's
00:19:06
like if you're a space tech investor and
00:19:07
you didn't invest in SpaceX, you
00:19:09
probably lost money. If you're a crypto
00:19:11
infrastructure investor and you didn't
00:19:12
invest in Coinbase, you probably lost
00:19:14
money. If you're a social media investor
00:19:15
and you didn't invest in Facebook, you
00:19:17
probably lost money. And yet, for some
00:19:19
reason, capital allocators have a very
00:19:21
short memory. they don't have the
00:19:23
ability to go back to the the prior boom
00:19:27
cycle and say, "Oh, wait. I remember
00:19:29
what happened that last time we had to
00:19:31
concentrate capital down to the winner."
00:19:32
Um, and and I think that this is this
00:19:34
isn't any different.
00:19:35
>> So, what was the motivator then for
00:19:36
change in the construction of the
00:19:38
landscape of defense contracting? Was it
00:19:40
software that takes us from the old
00:19:42
primes to the new primes? Is that really
00:19:44
what kind of triggered this
00:19:46
>> consolidation? So in '93 having won the
00:19:49
cold war in '91 or I think more
00:19:51
accurately the Soviets lost the cold war
00:19:53
you know by 93 we expected as a nation a
00:19:56
kind of peace dividend we don't have an
00:19:58
adversary now we should be able to spend
00:20:00
less on defense and uh the department
00:20:02
had this famous dinner where they
00:20:04
brought 15 of the 51 primes together and
00:20:07
said this is going to happen the
00:20:08
budget's going to get slashed we're not
00:20:10
going to save you guys you have our
00:20:11
permission to consolidate some of you
00:20:12
are going to go out of business some of
00:20:14
you should try to make a commercial
00:20:15
business which didn't really work uh And
00:20:17
that's what led down to five. It was
00:20:19
actually there was going to be we were
00:20:20
going to go from five to four in 99. The
00:20:22
justice department put their foot down
00:20:23
and said we're not going to let Lockheed
00:20:24
and Northrup consolidate.
00:20:26
>> Right. And then today, so the change
00:20:28
over, what gave you guys the window?
00:20:30
What gave you guys the window to build
00:20:32
the business that you've built? I mean,
00:20:34
what did you see early on 20 years ago
00:20:36
and continued forward in the hardware
00:20:38
side that made you say this is this is
00:20:41
the moment? Is it software that enabled
00:20:42
this?
00:20:43
>> Well, I'm going to throw you under the
00:20:44
bus a little bit on this. It was not
00:20:46
obvious from the beginning that this is
00:20:47
going to work. Um I can't tell you how
00:20:51
many meetings even we had together where
00:20:53
we walk out and be like
00:20:55
>> this is not going to work. This is very
00:20:57
bad.
00:20:58
>> We were not welcome with open arms,
00:21:00
>> right? You know,
00:21:00
>> early days of Palunteer.
00:21:01
>> Yeah. Early days, mid days, maybe even
00:21:04
late days. You know, famously we had to
00:21:06
sue our customer just for the right to
00:21:08
compete. Uh that that's the strength of
00:21:10
the monopsin, right? And and really I'd
00:21:13
say our our entire business was
00:21:15
validated from the field backwards,
00:21:17
right? It was in DC that we the doors
00:21:20
were closed to us. No one wanted to
00:21:21
interact with us. Like the where the
00:21:22
monopsin is strongest at the margin in
00:21:24
the field where people are saying like,
00:21:25
well, I'm on this deployment. I'm on
00:21:27
this rotation and I I there is more free
00:21:29
market system like what I'm being given
00:21:31
doesn't work. I'd like to come home. How
00:21:32
do I like bend the rules, figure out how
00:21:34
to get the software I need? That allowed
00:21:36
us to empirically show, you know, create
00:21:38
facts on the ground that this stuff
00:21:39
worked. It was a long road to hoe. I
00:21:41
think this is part of why it took 20
00:21:43
years to kind of get to this sort of
00:21:44
point here.
00:21:45
>> And yeah, I think that a lot of those
00:21:47
lessons that were learned were not only
00:21:49
like educational for and it also created
00:21:52
a precedence that already existed by the
00:21:54
time that we showed up where we didn't
00:21:56
have to go through all of those same
00:21:57
growing pains. So you know what took
00:21:59
Palunteer I guess probably about 5 years
00:22:01
to get to 10 million annually in
00:22:03
revenue. We did that in 22 months. But I
00:22:06
don't think that was a credit to Andrew.
00:22:07
I think it was mostly a credit to me,
00:22:08
Matt, and Brian lived the the 5 years of
00:22:12
balance here.
00:22:12
>> 5 years plus 22 months.
00:22:14
>> Exactly. Exactly. And so we kind of had
00:22:16
a bit of a cheat code, but I don't think
00:22:18
it's like culturally shifted to the
00:22:20
point where it's like actually just easy
00:22:21
to do this now. It's not like the
00:22:23
government has fixed all of their
00:22:24
problems. I think it's just that there's
00:22:26
enough people out there who have seen it
00:22:28
that understand how to make it work. Um
00:22:30
that you can work within the system
00:22:32
rather than, as Sham said, trying to
00:22:33
constantly fight from the outside in.
00:22:35
Was there a view at that time in these
00:22:37
years building Palunteer where you saw
00:22:39
an opportunity in hardware? Is that kind
00:22:41
of the motivation for Anderoll that
00:22:42
we've got software but there's more to
00:22:45
do and hardware could be reinvented and
00:22:47
we could build systems. I mean help us
00:22:48
understand the connection.
00:22:49
>> Yeah, in the very early days actually I
00:22:52
came by to pitch Sham uh before we
00:22:55
officially started the company. Uh and I
00:22:57
think part of that pitch was we said uh
00:22:59
we know how hard software is. We don't
00:23:01
want to do that again. We believe
00:23:03
hardware would be less hard. Um, and I
00:23:05
think we ended up being right from like
00:23:07
a getting going perspective. Um, but
00:23:09
yeah, it's just, you know, the
00:23:11
government doesn't know how to think
00:23:12
about the value that's created by
00:23:13
software. Maybe you feel like they're a
00:23:15
little bit better now, but it was really
00:23:17
hard in the early days of Palunteer to
00:23:19
convince them to think about it as
00:23:20
anything other than lines of code, for
00:23:22
example, or some like metrics that
00:23:24
anyone that works in tech knows doesn't
00:23:26
matter. But for hardware, they know
00:23:28
there's a bill of materials. In fact,
00:23:30
the government made stuff for a long
00:23:32
time. During the Cold War, there were
00:23:33
literally ammunition factories that were
00:23:35
owned and operated by the United States
00:23:37
government. Um, so they can kind of go
00:23:39
through a spreadsheet and say, "Okay, I
00:23:41
know how much it costs for you to build
00:23:42
this. I know how much margin I'm
00:23:44
comfortable paying you. Uh, and it's a
00:23:47
much easier, you know, crossing of the
00:23:49
Rubicon. With software, it never felt
00:23:51
like that.
00:23:52
>> Yeah, it's still very hard. I mean,
00:23:53
people want to pay for software and
00:23:55
government on a cost plus basis, which
00:23:57
doesn't make any sense if you think
00:23:59
about the R&D that you put into
00:24:00
software. Like, the marginal price
00:24:02
you're paying is a small fraction of the
00:24:04
R&D we're advertising across commercial
00:24:05
and government customers. And I think
00:24:08
maybe Trey, it's worth telling the story
00:24:09
of how when you got to Founders Fund,
00:24:11
you were looking for defense companies
00:24:12
to invest in and hitting so many dry
00:24:15
holes is is part of what led to the
00:24:17
idea.
00:24:18
>> Yeah. I mean I I have no I had maybe
00:24:21
still don't have any interest in venture
00:24:22
capital. Um and so when Peter asked me
00:24:24
to come over to Founders Fund, the only
00:24:26
thing I knew to do was just to look at C
00:24:28
the category of things I felt like I
00:24:30
understood pretty well, which was GVEK.
00:24:32
And so I uh incidentally, you can look
00:24:34
up who bids on federal contracts. And so
00:24:37
I pulled that list and then at the time
00:24:39
I was using Crunchb. And then I just
00:24:41
started outbounding to companies that
00:24:43
were bidding on federal contracts. And I
00:24:45
met with hundreds of companies in the
00:24:46
first three years that I was at Founders
00:24:49
Fund. We made ended up making one
00:24:50
investment in KDM which was renamed
00:24:52
Expanse and acquired by Palo Alto
00:24:54
Networks. Um but otherwise that was it.
00:24:56
There was nothing else. And even like
00:24:58
looking in retrospect like we didn't
00:25:00
miss anything. There was just nothing
00:25:01
worth investing in. It turns out Gnome
00:25:03
was building in this category. Um, and
00:25:06
so I went back to the founders fund
00:25:07
investment team and said, "Man, I feel
00:25:09
like someone should be building a next-g
00:25:11
prime that builds hardware, but it's
00:25:14
software defined and hardware enabled
00:25:15
rather than being hardware defined and
00:25:17
software enabled." Um, and uh, to my
00:25:20
surprise, the team was like, "Yeah,
00:25:21
sounds like you're probably the guy that
00:25:23
would know that that's the gap. And I
00:25:24
don't know, maybe you should start a
00:25:26
company." And I'm like, "Oh, no. You you
00:25:27
guys hired me to work on the investment
00:25:30
team at Founders Fund." They're like,
00:25:31
"Yeah, that's fine. You can you can do
00:25:33
that, too. It's all good." So yeah, it
00:25:36
was kind of an accidental founding in
00:25:37
many ways.
00:25:38
>> The capital requirements for hardware
00:25:40
business are notably higher if you
00:25:42
didn't have the success you had building
00:25:43
the manufacturing facility you're
00:25:45
building. And I'm assuming all the R&D
00:25:47
cycles are longer and require more
00:25:48
capital to get to a point of product.
00:25:51
Does it make sense for venture capital
00:25:53
to be the funer of the next gen of
00:25:56
hardware defense tech companies? How
00:25:58
does the economics and the capital
00:25:59
markets
00:26:00
>> I'll give a take here and then so I
00:26:02
think um if you back up a little bit if
00:26:04
you look at Fairchild if you look at
00:26:06
integrated circuits in 1968 96% of all
00:26:10
integrated circuits sold were sold to
00:26:12
the Apollo program there was effectively
00:26:14
a monopsy there's one buyer for this
00:26:15
thing but Bob who was at Fairchild at
00:26:18
the time and co-inventor of the
00:26:19
transistor he was so maniacally
00:26:21
committed to a future that
00:26:22
semiconductors integrated circuits would
00:26:24
be in everything that he never let more
00:26:26
than 4% of his R&D be paid for by the
00:26:28
government. he was going to be like why
00:26:30
should I have some PM tell me what my
00:26:31
R&D road map I invented the thing so he
00:26:34
kept executing that and that's he so he
00:26:36
chased Moore's law mercilessly and that
00:26:39
meant in the 80s when the government
00:26:40
needed precision guided munitions for
00:26:42
assault breaker to compete against the
00:26:44
Soviets we had the price performance we
00:26:45
needed uh so this idea that you know
00:26:48
because when you get stuck inside the
00:26:50
government loop of what they want you
00:26:51
never hit the price performance
00:26:52
>> by the way there's a good modern analogy
00:26:54
which is Jensen and Nvidia right 93
00:26:57
graphics chips sold them as an OE almost
00:26:59
like an OEM type solution until here we
00:27:01
are.
00:27:02
>> The the one that speaks to my heart as a
00:27:03
child who grew up in the shadow of the
00:27:04
space coast is shuttle. You know,
00:27:06
shuttle is is a it's beautiful, but it's
00:27:09
$50,000 a kilogram to get to orbit.
00:27:11
Starship heavy reuse will be under 20
00:27:13
bucks a kilogram, you know, and there's
00:27:15
no way that you will achieve that vision
00:27:18
if you're in a cost plus world because
00:27:20
every day you're actually deleting your
00:27:21
cost, which means you'd be deleting your
00:27:22
profit instead of deleting your cost,
00:27:24
turning that into margin that you can
00:27:26
provide a better price performance on
00:27:27
and keep writing that curve down,
00:27:29
>> right?
00:27:29
>> Although to be fair, SpaceX sued their
00:27:31
customer in the Air Force on the basis
00:27:34
of the exact same federal procurement
00:27:35
law um that Palanteer did. And so, you
00:27:38
know, even with that massive cost uh
00:27:41
costing down, they were in the same
00:27:43
position as Palunteer was in the early
00:27:45
days.
00:27:45
>> Yeah. Let's go back to the venture
00:27:46
capital at this conference that we're
00:27:48
at. There's dozens of VCs now. This is
00:27:50
clearly become a hot space as defined by
00:27:54
your successes. What's going to happen?
00:27:56
So, these VCs are plowing money into
00:27:58
hundreds of defense tech companies from
00:28:00
drones to satellites to weapons systems
00:28:02
to software tools. How is this going to
00:28:05
play out over the next couple of years
00:28:06
as a venture investor with your
00:28:07
knowledge about Androll?
00:28:09
>> I remember when I first got my offer
00:28:11
letter from from Palunteer. Um there
00:28:14
there we had this kind of high low mix
00:28:16
of salary versus equity compensation and
00:28:19
it gave three examples of how much your
00:28:22
equity at your that was on your offer
00:28:24
would be worth at different valuations.
00:28:25
It was I think it was 15 and 10 maybe or
00:28:29
15 and 20 something like that. And I
00:28:31
remember talking to people around the
00:28:32
company, they're like, "20? That's
00:28:35
ridiculous. We're never going to be
00:28:36
worth $20 billion." Um, and uh, you
00:28:40
know, I feel like now it's like people
00:28:41
are going out and they're like, "Our
00:28:42
seed round will be priced at $20
00:28:44
billion." Um, so I think there's there's
00:28:46
been a big shift in just like the way
00:28:48
that people think about these things.
00:28:50
Um, some of that is good, some of it's
00:28:52
bad. I think on the good side they're,
00:28:54
you know, they're able to look at
00:28:55
companies like Palunteer, like SpaceX,
00:28:57
like Andreal and say there is actually a
00:28:59
path to success. This isn't like an
00:29:01
impossible market to innovate in. But at
00:29:03
the same time, um, you know, there's
00:29:05
there's a real tension around amount of
00:29:07
capital raised and the valuations that
00:29:09
are being applied to these rounds that
00:29:12
add a tremendous amount of risk to the
00:29:14
company that doesn't have to exist. You
00:29:16
know, there's this favorite there's this
00:29:17
famous scene from HBO's Silicon Valley
00:29:20
where a CEO gets fired. um for uh kind
00:29:23
of underperformance on the plan that
00:29:25
they had presented. Um and Richard
00:29:27
Hendricks, the main character, says,
00:29:28
"You know, you could have raised less at
00:29:30
a lower price." And he said, "What do
00:29:32
you mean? No one ever told me I could
00:29:34
have raised less at a lower price." And
00:29:36
I think that's ultimately what my advice
00:29:38
to these companies is is like, "Look,
00:29:40
your product might be awesome. Seems
00:29:41
like you're building a good team. I like
00:29:43
the vision for what you're doing and the
00:29:44
mission, but you can also raise less at
00:29:47
a lower price." And I think that avoids,
00:29:50
you know, playing chicken ultimately
00:29:52
with trying to hit numbers that frankly
00:29:54
don't even make sense. Um, one of the
00:29:57
things that we've been focused on at
00:29:58
Andrew that I think maps back to our
00:29:59
experience at Palenteer actually is
00:30:02
climbing down the multiples tree uh with
00:30:05
every round. We never want to go and
00:30:07
raise the next round at a higher revenue
00:30:09
multiple than the prior round. Um, and
00:30:12
even the series H that you mentioned
00:30:13
before, uh, is is down pretty
00:30:16
significantly from the series G. And
00:30:17
we're not doing that because investors
00:30:19
wouldn't be willing to pay higher
00:30:20
multiples. We're doing that because I
00:30:22
believe the discipline is really, really
00:30:24
important, especially heading into a a,
00:30:26
you know, medium-term IPO.
00:30:27
>> Emil Michael talked about this $200
00:30:30
billion of capital he wants to deploy.
00:30:32
It's been reported he's hiring bankers
00:30:34
to help him deploy it. To counter your
00:30:36
point, if there's a lot of capital
00:30:37
flowing from the monopsin, if there's a,
00:30:40
you know, a big market that's growing
00:30:41
aggressively, you've got a $200 billion
00:30:43
TAM to go after now. Shouldn't that
00:30:45
justify the venture capital coming in?
00:30:47
Maybe even justify going at a higher
00:30:48
price. Is this Department of War
00:30:51
investment activity really going to kind
00:30:52
of change the landscape and increase
00:30:54
more venture capital flow?
00:30:56
>> I think it's it's going to help out
00:30:57
tremendously. First of all, I love the
00:30:58
nickname that this team has, which is
00:30:59
Deal Team Six. So, uh you know, I'd be
00:31:02
honored to be part of that team. uh but
00:31:03
I I think if even if you look back at
00:31:05
the titanium supply chain in the 50s60s
00:31:07
it was it was bootstrapped by the air
00:31:09
force in a similar way where they
00:31:10
strategically injected capital down the
00:31:12
supply chain to enable the aerospace
00:31:13
industry to be created. So I think
00:31:15
there's a tremendous opportunity with
00:31:16
the office of strategic capital to think
00:31:18
about what are the structural
00:31:19
bottlenecks in the production. You can
00:31:21
have a lot of drone companies and
00:31:22
they're all going to bottleneck on
00:31:23
brushless motors. There's going to be
00:31:25
certain key parts of the supply chain
00:31:27
that we don't have enough capacity to
00:31:28
produce here in the US and they're going
00:31:30
to need a bolus of investment and just
00:31:32
like integrated circuits, the first
00:31:34
customer is going to be less economical
00:31:36
than the end customer. So I think that's
00:31:38
one part then the other story I love is
00:31:39
where we really screwed this up. The
00:31:41
drone is an American birthright. Abe
00:31:43
Kareem and the Predator, we we built it
00:31:45
General Atomics. But of course then the
00:31:47
government got in the way. The
00:31:48
government said, "Hey, a drone is a
00:31:49
flying missile. This thing needs to be
00:31:51
ITAR controlled." Also, the FAA got in
00:31:53
the way and said no beyond line of sight
00:31:55
operations. Right? So, you basically
00:31:56
killed the domestic drone market. There
00:31:58
was a counterfactual world where General
00:32:01
Atomics had a consumer subsidiary called
00:32:04
DJI and the consumer drone market was
00:32:08
entirely owned by the US and it provided
00:32:09
for economic prosperity for America and
00:32:12
wrote us down the price production curve
00:32:14
that we could be using these things in
00:32:15
innovation for national security as
00:32:17
well.
00:32:18
>> What's the next market where you're
00:32:19
worried that's going to happen?
00:32:20
>> I have a lot of markets. So, I mean,
00:32:22
we're talking a lot about um weapons of
00:32:24
war here. I'm worried about things that
00:32:26
go beyond weapons of war that affect our
00:32:28
will to fight. Uh pharmaceuticals is one
00:32:30
that's close to my heart. My father was
00:32:31
a pharmacist, and it's one of the things
00:32:32
he always wanted me to work on was
00:32:33
bringing pharmaceutical production back
00:32:35
home. 80% of APIs for generic drugs are
00:32:38
produced by China. What do you think the
00:32:40
American people are going to think when
00:32:41
they have to choose between defending
00:32:43
the free world, defending American
00:32:44
sovereignty, and their 5-year-old dying
00:32:46
of an ear infection that we would have
00:32:48
thought of as trivially curable? So I
00:32:50
think there are these these things where
00:32:51
we've just embibed the globalist vision
00:32:54
that hey we'll do the innovation they'll
00:32:56
do the production without realizing that
00:32:58
innovation is a consequence of
00:33:00
productivity probably to put it in terms
00:33:01
that the that the tech community would
00:33:03
really understand is like what motivated
00:33:04
Google to do the research in 2017 be
00:33:07
behind the paper attention is all you
00:33:08
need a desire for a 3% incremental
00:33:12
improvement in Google translate you
00:33:14
cannot think of something more bal
00:33:16
leading to something more revolutionary
00:33:18
and we have seated all of those
00:33:19
opportunities to realize and harness
00:33:21
that innovation. Now, if you we stick
00:33:22
with pharmaceuticals, there's a reason
00:33:24
50% of all clinical trials, new clinical
00:33:26
trials are happening in China and not in
00:33:28
the US.
00:33:28
>> Do you have a view on a technology or
00:33:30
product that's developed in the US that
00:33:32
we're at risk of getting regulated out
00:33:34
and we're or we kind of going in the
00:33:36
right direction at this moment?
00:33:37
>> I would say that semiconductors is
00:33:39
actually a really interesting one. back
00:33:40
to your your point about um Fairchild is
00:33:43
that you know we were the the home of
00:33:45
the semiconductor industry for many many
00:33:47
years and um you know once TSMC got got
00:33:50
you know ahead of steam in Taiwan uh
00:33:53
they kind of ran away with it and we
00:33:55
didn't rather than investing in a
00:33:57
domestic competitor uh there are all
00:33:58
sorts of reasons for this that we could
00:34:00
spend hours talking about um we
00:34:01
basically just allowed that to happen
00:34:03
somewhere else and now we're in a
00:34:05
position where almost no amount of money
00:34:07
is going to fix this problem for us
00:34:08
certainly on a timeline of relev as it
00:34:10
pertains to the risk that Taiwan has um
00:34:13
in 2027. So, I would say that's like
00:34:15
another really big one.
00:34:16
>> Getting back to defense, how does the
00:34:18
hardware and software fit together in
00:34:20
this, you know, kind of emerging war
00:34:23
technology landscape? You guys are
00:34:25
building software, you're building
00:34:26
hardware, like help us understand a
00:34:28
little bit about what the systems of war
00:34:29
look like in the decades ahead.
00:34:31
>> Well, I think the the starting premise
00:34:33
would be where does it give us an
00:34:34
advantage? And uh you know, we had the
00:34:36
first offset which was nuclear weapons.
00:34:38
The second offset, precision guided
00:34:39
munitions and stealth. Really, the the
00:34:41
third offset is decision advantage. How
00:34:43
can we outthink and out execute the
00:34:44
adversary? That's where these things fit
00:34:46
together. That's like the thesis of even
00:34:48
having them to begin with. And I think
00:34:50
the reality is that there's kind of a
00:34:52
messy overlap of these things. Like all
00:34:54
innovation is messy and chaotic and
00:34:56
maybe the department suffers at times
00:34:58
because it tries to have a
00:35:00
framework-based approach like we're
00:35:01
going to have MOSA modular open
00:35:03
standards architecture something like
00:35:05
that. you know, it's like we're going to
00:35:07
avoid all of the pain and messiness by
00:35:09
having some sort of process, but in
00:35:11
reality, the process always destroys all
00:35:13
the innovation. And so there there's
00:35:15
obviously a very tight thesis where
00:35:17
these things need to coexist and build
00:35:19
off one one another. They need to be
00:35:21
interoperable, but you have to earn
00:35:23
that. You have to earn that opinion in
00:35:25
the exercises, in the tests, and the
00:35:26
evaluation in combat would be my my
00:35:29
humble suggestion. Do you think that
00:35:31
outside of what Anderl's building,
00:35:34
there's actions that this administration
00:35:38
should be taking or that the private
00:35:40
market should be taking to solve this
00:35:43
hardware manufacturing and capacity gap
00:35:45
that we have right now?
00:35:46
>> My view is they're actually taking a lot
00:35:48
of actions. Not all of them have come to
00:35:49
see the day of light here, but you know,
00:35:51
even something as simple as reimagining
00:35:53
munitions and drones as consumables.
00:35:55
They're not things that you build and
00:35:56
then stock on a shelf. They're things
00:35:58
that you at the time of ordering them,
00:36:01
you already have an exercise test plan
00:36:02
where you plan to expend them, which
00:36:04
means that you're going to have to
00:36:05
replenish them, which means there's a
00:36:06
demand signal to industry to keep going
00:36:08
and and a buying cycle that means that
00:36:10
you can buy the next generation rather
00:36:12
than having the old generation. Uh, and
00:36:14
I think this department sees that in a
00:36:15
very cleareyed way and is doing the
00:36:17
yman's work of like stitching that
00:36:19
through through all the services, all
00:36:21
the portfolio acquisition executives to
00:36:23
the point of not believing in process.
00:36:24
They've moved from a world where you
00:36:26
have a very rigid, hey, I said I was
00:36:29
going to buy X of Y and I can't change
00:36:30
my mind to a world where there's more
00:36:32
autonomy and authority for the people
00:36:34
who are buying things to say, I told you
00:36:36
I was going to buy something that
00:36:37
accomplished this goal. I can change my
00:36:39
mind about how I'm going to accomplish
00:36:40
this goal, which you know, can you
00:36:42
imagine trying to run any private sector
00:36:43
business without that degree of autonomy
00:36:46
and and flexibility in your decision-m?
00:36:48
>> To echo Sham's point about what the
00:36:50
administration is doing, there are a
00:36:51
bunch of swings that they're taking. It
00:36:53
wasn't that long ago that uh in the
00:36:55
Obama administration, there was kind of
00:36:56
the Celindra failure where they invested
00:36:58
in this solar panel company. It ended up
00:37:01
being a bad investment for the US
00:37:02
taxpayer. Um but now the Office of
00:37:04
Strategic Capital is taking a really
00:37:06
hard look at critical minerals. They're
00:37:09
looking at, you know, refining of of
00:37:11
minerals and uh they're engaging
00:37:13
directly with the private sector to come
00:37:14
alongside them in doing deals to get
00:37:16
offtake agreements on these things. So I
00:37:18
think there is some really clever
00:37:20
thinking going on. Um, and then in
00:37:22
addition to that, I think the
00:37:23
procurement process is sort of up for
00:37:25
revision constantly. Um, and Ash Carter
00:37:28
started pulling a lot of these threads
00:37:30
during the Obama administration. That
00:37:31
was the third offset initiative. Um, but
00:37:34
you know, we're in a in a world today
00:37:35
where there's kind of a renewed vigor
00:37:37
around being asked private industry
00:37:39
being asked what do you need like tell
00:37:41
us what you need to be changed about the
00:37:43
way that we do business in order to
00:37:45
streamline this and make it go faster.
00:37:46
So I I think there's a lot of positive
00:37:48
momentum. Do you guys think these
00:37:49
changes have been institutionalized or
00:37:52
are they political party dependent? If
00:37:55
the Republicans lose the midterms and
00:37:58
there's a Democrat president in 2028, do
00:38:01
you think that there's a reversion to
00:38:03
the old way of operating that's going to
00:38:05
have some political influence and change
00:38:07
things or do you think we've really
00:38:08
changed how things are are running in
00:38:09
the government here?
00:38:10
>> I I don't think it's political, but I
00:38:11
think it's people. So you can go back to
00:38:14
again Ash Carter, Democratic
00:38:16
administration, highly focused on fixing
00:38:18
these problems. The current Trump
00:38:20
administration highly focused on fixing
00:38:22
these problems. And you know, there are
00:38:25
a lot of people that live that were in
00:38:27
the middle of that that didn't
00:38:29
prioritize this. It wasn't, you know, it
00:38:31
didn't rise to the top of the stack. And
00:38:33
so I think we always think of like the
00:38:35
bureaucracy as being a political
00:38:37
infrastructure or an institution or, you
00:38:39
know, bureaucracy. I actually just think
00:38:41
at the end of the day it's about
00:38:42
leadership. It's just do we have the
00:38:44
right people that understand the set of
00:38:45
the problems that we're we're facing.
00:38:47
>> I couldn't agree more. I mean I I call
00:38:48
them heretics and heroes. Like you you
00:38:50
know the entropy of the bureaucracy is
00:38:52
always towards some sort of sclerosis.
00:38:54
It doesn't it's not a political
00:38:55
statement. It just is what bureaucracy
00:38:56
does. when David Packard, who was
00:38:58
probably the last major technology
00:39:00
co-founder who served in the Department
00:39:02
of Defense, when he came up with the
00:39:03
5,000 series on acquisition, which today
00:39:05
we view as just sclerotic BS that's
00:39:09
that's tied us down. Well, he his
00:39:11
document was seven pages long in the in
00:39:13
the years between when he wrote that and
00:39:14
now it's 2,000 pages. So, did he really
00:39:17
screw us or did just the entropy take
00:39:20
hold and there was no strong leadership
00:39:21
in between to go do the bushwhacking and
00:39:23
mow the lawn and and and make the right
00:39:25
decisions? Now, you know, this this
00:39:26
administration's blown up jids, which is
00:39:28
kind of one of these insane bureaucratic
00:39:31
processes. Like, can you just take a
00:39:33
forget a scalpel, take a machete, and
00:39:35
start clearing the jungle, and reearn
00:39:37
some of these lessons as you go through
00:39:38
it? If you look at Kelly Johnson, who
00:39:40
was the founder of Skunk Works, he built
00:39:41
41 airframes in his career, including
00:39:43
the SR71, still the fastest flying
00:39:45
manned aircraft, and uh and the U2,
00:39:47
which we still fly. Like one, if you
00:39:49
look at his rules, one of his rules was
00:39:51
he had to play defense to keep the
00:39:53
government bureaucrats out of his
00:39:55
program. And I think you could, you
00:39:57
know, okay, it's not it's not a critique
00:39:58
of government. You could think of big
00:40:00
corporations. When the big corporations
00:40:01
bureaucracy gets into the innovation
00:40:03
folks, the innovation stops.
00:40:04
>> And the heretics and heroes were really
00:40:06
founders. It's like from a tech
00:40:08
perspective, that's exactly what they
00:40:10
were. like to the Kelly Johnson point,
00:40:12
the U2 did not start as a US military
00:40:16
aircraft because they didn't want it. Uh
00:40:18
he ended up going through the
00:40:19
intelligence community to get his start
00:40:21
there. Uh and and it was kind of finding
00:40:24
the right person that was willing to
00:40:25
take the risk to do the thing rather
00:40:27
than relying on the system or the
00:40:29
institution to do that. This is the same
00:40:30
with Benny Shriber with ICBMs, Admiral
00:40:33
Rickover with the nuclear navy. Um, it
00:40:35
used to be about people, but today if
00:40:37
you were to go to the the Pentagon and
00:40:40
say, "Who's responsible for the F-35?"
00:40:42
I'm not sure they even know who is
00:40:44
responsible for the F-35. We're we're
00:40:45
building systems by by committee rather
00:40:48
than trusting the founders and people.
00:40:50
>> There's like a proclamation. It's so
00:40:51
great. It's this X Y or Z project is
00:40:53
built in all 50 states. Like the
00:40:55
objective is it's it's being the money
00:40:57
is being spread around versus the F35.
00:40:59
How do you achieve the objective? The
00:41:00
F-35 has components manufactured in 400
00:41:03
congressional districts,
00:41:04
>> right?
00:41:04
>> I mean, that's like that's a political
00:41:07
project.
00:41:07
>> We had this AB test with SLS and
00:41:09
Starship, right? Like SLS had to have
00:41:12
subs in all 50 states, but this AB test
00:41:14
has been played out.
00:41:15
>> I do think there's something about the
00:41:16
kind of Calvinist spirit of America that
00:41:19
sometimes gets us to misunderstand. You
00:41:20
know, we call it the Apollo program, but
00:41:22
it was probably it's it's really the
00:41:24
Jean CR program. You know, uh it's the
00:41:26
F-16, but it's really John Boyd's plane.
00:41:28
And and there's like a humility where
00:41:30
we'd never want to call it, you know,
00:41:32
you wouldn't call it although maybe the
00:41:34
nuclear navy we kind of do call it Rick
00:41:35
Over's Navy, but there's an element of
00:41:38
like it's obviously both things. It's
00:41:39
bigger than the person, but actually the
00:41:41
starting conditions require the founder
00:41:43
figure.
00:41:43
>> So tell me about where we are in the
00:41:46
state of readiness. You know, we
00:41:48
highlighted some of the statistics, but
00:41:50
if you guys were to think about what we
00:41:53
need to accomplish, the infrastructure
00:41:55
we need to stand up to have production
00:41:57
lines that can be turned on to support
00:41:59
munitions capacity for conflict in, you
00:42:02
know, pick your region, let's say more
00:42:04
than one region around the world, and
00:42:06
the US needs to sustain those conflicts
00:42:08
to defend the United States. How far are
00:42:10
we from being ready based on what's
00:42:13
going on in the investments that are
00:42:14
being made today
00:42:15
>> across the sector? I think the way I
00:42:17
think about it is if you thought about
00:42:19
this as a spear, the tip of the spear is
00:42:21
incredibly sharp. The shaft of the spear
00:42:24
needs a lot of work here. Um, you know,
00:42:26
the Department of Defense such a big
00:42:28
organization. It it's it's structurally
00:42:31
supply and demand is not integrated. You
00:42:33
know, the the demand side is a real
00:42:35
world events that happen in the
00:42:36
combatant commands. The the supply side
00:42:39
are what the services and the industrial
00:42:40
base build. The man, train, and equip
00:42:42
and how you're actually producing that
00:42:44
material. And the ability to drive
00:42:46
consensus, which is really the beating
00:42:47
heart of any private sector company of
00:42:49
like how does supply and demand come
00:42:50
together? What's our review of it? Is is
00:42:52
kind of managed with with great
00:42:55
difficulty in the Pentagon. And your
00:42:56
ability your agility to respond to
00:42:58
scenarios is is weak. That's that's
00:43:01
where we can have a lot more precision
00:43:02
because if you can start changing your
00:43:04
mind and saying actually I thought I
00:43:05
needed X barracudas, but now I need Y
00:43:08
barracudas and Z furies. How quickly
00:43:10
does that take to percolate through the
00:43:12
supply base? you know, instead we get
00:43:14
locked into these like, hey, we've made
00:43:16
a decision. The decision really can't be
00:43:17
revisited. Uh, and then you start over
00:43:20
time, the entropy is like, hey, I need
00:43:22
to I have I have an unexpected bill.
00:43:25
Let's just take a little bit of money
00:43:26
away from these programs. And everything
00:43:27
starts getting down to minimum rate
00:43:29
production, which is this idea of what
00:43:30
can I produce to just keep the line
00:43:32
open, which is not deterrence. That's
00:43:34
how we frog boiled our way here. And
00:43:36
that this is where my my idea of like
00:43:37
you got to tie this into consumption.
00:43:38
Like everything you're building,
00:43:39
particularly on the munition side, needs
00:43:41
to be consumable, such that you know
00:43:43
you're going to replenish it such that
00:43:44
the primes and the neopimes all have the
00:43:48
demand signal they need to keep going
00:43:49
and have a reason to produce them
00:43:51
cheaper in order to make more money. My
00:43:52
most contrarian idea is like the the
00:43:54
cynical view that like you have the
00:43:56
military-industrial complex and they're
00:43:57
just in it to make money. Well, this
00:43:58
it's kind of a crappy business. Like you
00:44:00
know these companies trade at like less
00:44:02
than two times revenue. The problem is
00:44:03
it's not profitable enough actually. But
00:44:06
Trey, Sham said in the past, we have 8
00:44:08
days of munitions on hand for a major
00:44:11
conflict with China versus 800 days
00:44:13
needed. How far away do you think we are
00:44:16
to having the supply chain built and the
00:44:18
production capacity built to meet that
00:44:20
objective?
00:44:21
>> Well, I mean, one of the other things
00:44:22
Sean just said in this conversation is
00:44:24
that there's a hyo mix question as well.
00:44:26
like there are exquisite systems that
00:44:27
are like the multi-million dollar
00:44:29
interceptor missiles and then you have,
00:44:31
you know, dumb munitions, things like
00:44:33
bombs that are being dropped and things
00:44:34
like that. And it's not the same across
00:44:36
that entire stack. You know, there are
00:44:38
some uh some munitions that are woefully
00:44:40
under supplied and we have very low
00:44:42
readiness on and there are others where
00:44:44
I actually feel pretty good. Um, from an
00:44:46
Android perspective, you know, we're
00:44:47
looking at ramping as fast as we can. We
00:44:49
just opened uh Arsenal yesterday uh to
00:44:52
start producing furies. Um, and you
00:44:55
know, looking out over the next 18
00:44:57
months, we're not taking our our foot
00:44:58
off the gas and yet there's still a run
00:45:01
rate capacity that we're running into as
00:45:03
we ramp. Like you it just takes time to
00:45:05
get all these lines stood up. Um, so,
00:45:08
you know, if we were to start today with
00:45:10
unlimited cash, I think over the next 18
00:45:13
months as a country, we could get to the
00:45:15
point where we were on track for having
00:45:17
a sustainable uh industrial base. Uh,
00:45:20
but we're not going to do that. we're
00:45:21
going to trickle it out over time and uh
00:45:24
and I think that unless there's real
00:45:26
political leadership that steps up to
00:45:28
drive this forward, we will likely be in
00:45:30
a similar situation uh for a long time.
00:45:32
Um and so I think we really need to to
00:45:35
take advantage of this specific moment
00:45:37
where there's clearly urgency that is
00:45:39
understood by by leadership across the
00:45:41
administration.
00:45:42
>> How much of the spear is autonomous
00:45:45
systems? Is everything going autonomous?
00:45:47
Is it just drones or is there on the
00:45:49
ground autonomous
00:45:50
>> right now? Well, I I would actually
00:45:53
argue that like the
00:45:54
>> if you were to build for the 18-month
00:45:55
out
00:45:56
>> Yeah. I mean, the wars of today are
00:45:58
fought with the weapons of yesterday.
00:46:00
That's like just a fundamental truth.
00:46:02
And some of our weapons of yesterday are
00:46:04
awesome. Like B2s are incredible. Uh the
00:46:08
bombs that we dropped last summer in
00:46:10
Iran, incredible. Like these are very
00:46:13
high-tech, exquisite systems. Patriot
00:46:15
missiles, incredible. They're very
00:46:17
performant. Uh they hit their targets
00:46:20
almost every time. I mean, these are
00:46:21
great systems. The problem is is that
00:46:23
they're incredibly expensive and we
00:46:24
can't resupply. Like we're just we're
00:46:26
way behind the eight-ball on that. And
00:46:28
so, uh, you know, thinking about the
00:46:30
wars of the future, it means that we
00:46:31
need to start building these attritable
00:46:33
mass systems today so that they're in
00:46:36
inventory to be used for those
00:46:37
conflicts. There are a number of androl
00:46:39
systems that are at a readiness level in
00:46:41
at a level of deployment that they're
00:46:43
active and being used in in combat
00:46:46
today, but it's still a very small
00:46:48
percentage of of the way that the wars
00:46:49
are being waged. Um over, you know, the
00:46:52
next 5 to 10 years, that hyo mix is
00:46:54
going to have to shift massively. I
00:46:56
mean, this is just like clearly evident
00:46:58
in Ukraine. It's clearly evident in
00:47:00
Iran. uh and I think the department is
00:47:02
making steps to ensuring that they have
00:47:06
a better divide of lowcost attra
00:47:10
exquisite systems um but uh you know as
00:47:13
I said it's going to take political
00:47:14
leadership to get there uh but we we
00:47:17
have the ability to do it the reason I
00:47:18
ask is I want to talk a little bit about
00:47:20
the ethics of technology to war the
00:47:23
anthropic runin with the department of
00:47:26
war recently and I think you guys have
00:47:28
both talked publicly about this but
00:47:31
Anthropic refused to let its claude
00:47:33
model be used in Maven without human
00:47:35
oversight constraints. This is what's
00:47:37
been reported what Emil Michael has said
00:47:39
and the Pentagon labeled them a supply
00:47:41
chain risk. Both AI and generally
00:47:44
autonomous systems beg the question what
00:47:47
role should humans have and who should
00:47:49
have the right to play that role of
00:47:51
hitting the kill switch. What's your
00:47:54
view on the role of the technology
00:47:56
vendor to the department of war and
00:47:58
where you guys draw your line on what
00:48:00
your responsibilities are with respect
00:48:02
to ethics?
00:48:04
>> Well, my my view on my responsibility to
00:48:06
ethics is a slightly different question
00:48:08
than my belief in democracy, which is a
00:48:10
different thing. Um, so maybe starting
00:48:12
on on the democracy point. Um, I believe
00:48:14
that the people of America have elected
00:48:16
representatives to make really hard
00:48:19
decisions about how we engage in combat.
00:48:22
Full stop. Um, fully autonomous weapons
00:48:24
are not new. We've had autonomous
00:48:27
systems in operation like SeaWiz, which
00:48:29
is deployed on uh, naval vessels that
00:48:32
shoots down aerial threats fully
00:48:33
autonomously. You don't have time to
00:48:36
make decisions about, you know, incoming
00:48:38
missiles or threats to your ship. You
00:48:39
just have to shoot it down. So, that's
00:48:41
what Sew does. Now, Sewiz has
00:48:43
accountability in the system. There's a
00:48:46
person on that boat that is responsible
00:48:48
for whatever actions that weapon system
00:48:50
takes. And I believe that this is the
00:48:52
future of autonomous systems is that
00:48:54
just like any other system, whether it's
00:48:56
a soldier carrying a gun, they are
00:48:58
accountable for what happens with that
00:48:59
gun or uh you know the captain of a
00:49:02
naval vessel, they are responsible for
00:49:03
what happens with that sea whiz. All of
00:49:05
your autonomous systems are going to
00:49:06
have accountability baked in. Now,
00:49:09
ethically, how do I think about this?
00:49:11
You know, I believe that, you know, what
00:49:13
Sean was saying about the first offset,
00:49:15
the second offset, and the third offset,
00:49:17
um, is, you know, we have we as a
00:49:20
society went from like rocks and sticks
00:49:23
to, you know, knives to guns to bombs.
00:49:27
And then in, you know, World War II, we
00:49:29
kind of plateaued with nuclear weapons.
00:49:31
And we all looked around at each other
00:49:33
and we said, "Wow, that's pretty crazy.
00:49:36
I don't think we want to make more and
00:49:38
more powerful nuclear weapons forever."
00:49:40
And so actually our engagement in combat
00:49:41
has come back down the chain. Uh we're,
00:49:44
you know, precisiong guided weapons.
00:49:46
We're shooting non-exlosive missiles
00:49:48
into windows of apartment buildings,
00:49:51
avoiding casualties, unintended
00:49:53
casualties. Um and I think that's really
00:49:55
the goal. And if you look at AI as the
00:49:58
command center uh for making better
00:50:00
decisions with better precision, with
00:50:02
better discrimination, with uh less less
00:50:06
uh civilian casualties, this is good.
00:50:09
that's actually ethically far improved
00:50:11
from just dropping dumb bombs on areas
00:50:13
of cities to eliminate military
00:50:15
facilities. So, um I don't think
00:50:17
abstension from participating in the
00:50:20
building of technology for national
00:50:21
security is a morally neutral decision.
00:50:24
You are making a moral decision when you
00:50:25
decide to abstain. And I am making a
00:50:27
moral decision as a private citizen
00:50:29
building a company in this space that I
00:50:31
believe that this is ethically just and
00:50:33
uh I trust in our democracy to deploy
00:50:35
those tools with uh the interests of the
00:50:37
American people at heart. There's a lot
00:50:39
that's said about Palunteer enabling a
00:50:42
surveillance state. We had the all-in
00:50:43
summit in September. Alex Karp, your
00:50:45
CEO, spoke and there was a protest group
00:50:47
outside protesting Palunteer powering a
00:50:50
surveillance state. I just want to give
00:50:52
you a chance to respond to that. number
00:50:54
one. And number two, if you saw that
00:50:57
Palunteer's tools were being used in an
00:50:59
illegal way, where's the responsibility
00:51:01
for Palunteer as a technology vendor in
00:51:04
addressing those concerns?
00:51:06
>> It's almost hard to respond to because
00:51:07
it's very unclear what surveillance
00:51:09
people think we're doing. You know,
00:51:10
there's just like a broad almost maybe
00:51:12
an outgrowth of Terminator fear around
00:51:14
technology.
00:51:15
>> And I want to get into that cultural
00:51:16
question next. I think it's like very
00:51:17
important to understand.
00:51:18
>> We don't collect data. We don't have any
00:51:19
data. It'd be like it'd be like accusing
00:51:21
Excel of being a surveillance tool,
00:51:23
right? It's like this is a way of
00:51:25
bringing your your own data that you
00:51:26
have lawful authorities to collect
00:51:28
together to make decisions. Sounds a lot
00:51:30
like Excel. Um, but it's, you know,
00:51:33
because we are unabashedly patriotic and
00:51:36
serve the US military, I think people
00:51:38
have a kind of colored view of these
00:51:40
things.
00:51:40
>> I would argue that it's actually Excel
00:51:42
with cellby cell.
00:51:47
You know, it' be as Karp says, it's the
00:51:49
most insane platform in the world to try
00:51:51
to do something illegal in because you
00:51:52
are going to be caught. You know, like
00:51:54
that was part of the idea of how do you
00:51:55
enhance privacy and security? It's how
00:51:57
do you build more civil liberties
00:51:58
protections in how do you have a
00:51:59
normative view that enables a democracy
00:52:01
to say these are laws and rules, the
00:52:03
system will enforce it. We're not just
00:52:05
relying on people happening to do the
00:52:07
correct thing. Uh so that's one piece of
00:52:09
it. The other piece of it to come back
00:52:10
to this broader point is the the need
00:52:12
for epistemic humility. like one of the
00:52:15
um to the point of we have elected
00:52:16
officials um they are accountable for
00:52:19
these policy decisions I think at the
00:52:21
limit it's actually kind of indefensible
00:52:23
to have a perspective other than lawful
00:52:26
use because if you are salami slicing
00:52:28
the policy that's actually tyranny by
00:52:31
techb bro you know a small number of
00:52:33
people are constraining the maneuver
00:52:35
space of a democracy with no
00:52:36
accountability to the populace so I I
00:52:40
think that's a pretty challenging
00:52:41
perspective to be and we've been in this
00:52:42
perspective if you go back to How did
00:52:43
the Soviets get the nuclear bomb? You
00:52:45
know, there's really there's two sources
00:52:47
of treason. One was committed communists
00:52:49
like Klaus Fuks who was always a spy.
00:52:52
But the other were people like Theodore
00:52:54
Hall. He was one of the youngest
00:52:55
scientists in the Manhattan Project at
00:52:57
18 years old. Uh his brother Edward Hall
00:52:59
built the Minute Man missile. His
00:53:00
brother was, you know, kind of a heretic
00:53:02
and hero in my terms. But Theodore said,
00:53:04
you know, I'm one of the best physicists
00:53:06
in the world. I'm probably also one of
00:53:07
the best geopolitical strategists in the
00:53:09
world. And I think the only way to have
00:53:10
world peace is if two countries have the
00:53:12
bomb. So in 1944, Theodore Hull walked
00:53:15
into the Soviet trade mission in New
00:53:16
York and gave them critical secrets to
00:53:18
the bomb. Now Theodore thought he was
00:53:20
going to deliver world peace. Instead,
00:53:22
every death from communism since 1949 is
00:53:25
actually on his hands. And there's no
00:53:26
accountability for that. Yeah. I think
00:53:28
one of the other things that comes up in
00:53:30
these uh these conversations about
00:53:33
powering the surveillance state is this
00:53:35
belief that we have policies but we
00:53:37
don't actually want our civil servants
00:53:39
to have the best technology to enforce
00:53:41
those policies. And it's almost like a
00:53:43
lack of belief in the institution of
00:53:44
democracy. It's like you know traffic
00:53:47
cameras for example. It's like man I I'm
00:53:49
sort of libertarian in some ways. I
00:53:51
don't love traffic cameras but if
00:53:52
traffic came say I'm going to send a
00:53:54
ticket to every single person that blows
00:53:56
through this red light. I don't know,
00:53:58
maybe we have to change the policy
00:53:59
around red lights if you don't like
00:54:00
that. Uh, or
00:54:01
>> I've gotten three speeding tickets in
00:54:03
San Francisco from autonomous cameras in
00:54:05
the last month.
00:54:05
>> Well, they did ramp it up significantly,
00:54:07
which is incredible that they're
00:54:08
enforcing laws at all if we're being
00:54:10
honest.
00:54:10
>> I mean, like, here's my ticket. And then
00:54:12
I'm literally watching the guy while I'm
00:54:13
doing three miles over the speed limit,
00:54:15
you know, take heroin needles out of his
00:54:17
arm and put it on the floor next to
00:54:19
elementary school and I'm like, totally
00:54:21
fine. Yeah.
00:54:22
>> But it's the same thing with like, you
00:54:23
know, using t using tech for better
00:54:26
enforcement of tax. uh law or something
00:54:28
like that like do you is what we're
00:54:30
saying by criticizing what Palantry is
00:54:32
doing is what we're saying that we don't
00:54:34
want our civil servants to have the best
00:54:36
tools possible because you can have that
00:54:38
position I tend to think that that
00:54:40
position is pretty morally bankrupt but
00:54:42
I guess you could have that position but
00:54:44
ultimately I think that's what it is
00:54:45
>> so scenario play this for me because
00:54:46
there's a public perception I don't know
00:54:48
how farreaching it is that there's some
00:54:50
tied corruption between government
00:54:51
officials that use technology that's
00:54:53
super advanced that's hard for people to
00:54:55
understand and the technology ology
00:54:56
vendors. If you saw government agencies
00:54:59
using your technology, either of you in
00:55:00
an illegal way that you knew broke the
00:55:02
law, do you report it?
00:55:04
>> 100%.
00:55:05
>> I there's an entire mechanism.
00:55:07
Absolutely.
00:55:07
>> There is an IG in every agency.
00:55:09
>> Right. So can just just explain that for
00:55:11
a second. The inspector general maybe.
00:55:13
>> Yeah. Every agency has an inspector
00:55:15
general who's an independent
00:55:16
organization that you can provide
00:55:19
anonymous or non-anmous complaints to
00:55:21
that then have the ability the statutory
00:55:23
ability to do an investigation in an
00:55:25
unfettered way inside of that
00:55:26
organization whether it's the department
00:55:28
of war or um housing and urban
00:55:30
development like literally every single
00:55:32
agency. Uh, and this mechanism is used.
00:55:34
I actually, you know, in this case it
00:55:36
was weaponized against one of my
00:55:38
favorite heretics and heroes, Colonel
00:55:39
Drew Cukor, who invented Maven, the
00:55:41
founder of Maven really, but you know,
00:55:43
people would file complaints claiming
00:55:44
that he was um hiding illegal immigrants
00:55:48
in his basement, a basement that he
00:55:49
doesn't really have, you know, but but
00:55:50
all of those things were investigated.
00:55:52
Naval Criminal Investigative Services
00:55:53
went out to his house and actually
00:55:54
looked into these things. So, people
00:55:56
take this incredibly seriously. Where
00:55:58
does the culture come from? The
00:56:01
anti-defense tech alignment culture. Is
00:56:04
it because of the peace era that we had
00:56:06
and folks took for granted national
00:56:08
security?
00:56:09
>> I think the first schism was really
00:56:10
during Vietnam where people felt like um
00:56:13
they were lied to about the war. It it
00:56:15
drove a fundamental schism between
00:56:16
academia and defense and that we've
00:56:19
never really healed from that schism.
00:56:20
And so there's this this this kind of
00:56:22
sense this distrust that's that's brewed
00:56:24
there and kind of escalates through
00:56:26
society. The second schism is like the
00:56:28
number of people who are prior service
00:56:30
or are connected to this community who
00:56:32
actually see these people as humans and
00:56:34
have a fully informed mental model of
00:56:36
how diligent they are, how the work
00:56:37
actually gets done, what do these words
00:56:39
actually mean, what is the process like
00:56:41
is is is evaporated. So then their own
00:56:44
fears fill how they think it happens and
00:56:47
it seems like maybe it's happening more
00:56:49
in a cowboy way, maybe it's happening in
00:56:50
a way without oversight that there is no
00:56:52
such thing as doctrine. like there it's
00:56:54
kind of a cartoon version of what's
00:56:55
actually happening that I think you know
00:56:58
you're unable to reconcile.
00:56:59
>> Yeah. I think if you even look back at
00:57:01
what happened with Snowden uh you know
00:57:03
what was that 15 years ago now something
00:57:05
like that
00:57:06
>> um you know that there was almost no
00:57:09
discussion about the investigations that
00:57:10
went into like was the data collection
00:57:12
actually abused and the answer was like
00:57:15
basically not at all. there were like
00:57:16
less than 12 documented cases where
00:57:18
someone got access to data that they
00:57:20
shouldn't have. And it was because of
00:57:21
like technology errors, not because of
00:57:24
the policies that were implemented. Um,
00:57:26
and so, you know, we can have
00:57:28
disagreements about whether or not the
00:57:30
intelligence community should have
00:57:32
collected the data and stored it. Uh,
00:57:34
but that was that policy was renewed
00:57:37
multiple times by multiple
00:57:39
administrations, multiple political
00:57:41
parties that had the majority in
00:57:43
Congress over decades. And so apparently
00:57:46
our elected representatives thought that
00:57:47
it was important enough to keep in the
00:57:49
system. Um so I think it ends up being
00:57:52
this kind of weird policy discussion. Uh
00:57:54
but the second point that you make I
00:57:56
completely agree with about like the the
00:57:58
kind of distrust of the institution uh
00:58:01
in in creating the stories for
00:58:02
themselves. Uh I oftentimes go and do
00:58:06
guest lectures at Stanford and I always
00:58:08
try to ask like raise your hand if you
00:58:10
have an immediate family member that
00:58:11
serves in the military. No one ever
00:58:13
raises their hand. It's crazy. Like at
00:58:15
at Stamford, there are veterans that
00:58:18
come in that go to the GSB and things
00:58:20
like that, but in like the undergraduate
00:58:22
undergraduate student population, it's
00:58:24
incredibly rare for anyone to have any
00:58:26
connection whatsoever in their immediate
00:58:28
family to the military. And I I think
00:58:30
there's like, you know, some of this
00:58:32
goes back to criticisms that people like
00:58:34
JD Vance, uh, the vice president, has
00:58:36
made about, you know, elites in America
00:58:38
and things like that, but there's just
00:58:40
this incredible divide that has
00:58:42
happened. Um, and we're kind of losing
00:58:44
touch with that kind of salt of the
00:58:46
earth, middle of the country, you know,
00:58:47
veteran community that I I feel like was
00:58:50
way more present during the the Cold
00:58:52
War, uh, and and, you know, the World
00:58:54
War II before that.
00:58:55
>> Do you think that there's any external
00:58:56
influence that's driving this culture?
00:58:59
>> Are there influences on social media in
00:59:02
mainstream media? And maybe just talk
00:59:04
about destabilization
00:59:06
and the attack vectors. I've seen it in
00:59:09
other areas of science. I don't want to
00:59:10
spend time on this show talking about
00:59:11
it, but there have been traces that I
00:59:14
found on external folks that want to
00:59:17
destabilize American science and
00:59:19
industry progress and they create fear
00:59:22
and they put out articles and then they
00:59:24
go social and they they become viral and
00:59:26
suddenly everyone believes it even
00:59:27
though it's not true. Do you see that?
00:59:29
And have you actually kind of seen that
00:59:30
in the sense of like attacking the tech
00:59:33
companies that are now supporting
00:59:34
American defense?
00:59:35
>> 100%. Even if you go back to, let's
00:59:37
stick with Vietnam. Uh the Soviets spent
00:59:40
$7 billion in 2026
00:59:43
funding the peace movement, the anti-war
00:59:45
protest. Now, there's obviously some
00:59:47
there is an organic element to it, but
00:59:49
this is just dumping gasoline on the
00:59:51
fire to sew division and discord. Uh and
00:59:53
I think in the present day, you I I
00:59:55
think you know certainly we see it
00:59:56
against Pounder where there's CCP money
00:59:59
flowing to organizations that are
01:00:01
protesting us for various domestic
01:00:02
issues here that it's not isolated. It's
01:00:05
it's broadly a successful strategy for
01:00:07
our adversaries to sew division.
01:00:08
>> No one will believe it because no one
01:00:10
wants to believe that they're being
01:00:11
influenced,
01:00:12
>> right? But but I I think Sean's point is
01:00:14
exactly right. It's like it's actually a
01:00:16
brilliant strategy. Like good for them.
01:00:19
They they are not our friends. They're
01:00:21
our adversaries. What would you do if
01:00:22
you were in their shoes? It makes a ton
01:00:24
of sense.
01:00:24
>> Do we do that?
01:00:25
>> I mean, look, we we have all sorts of
01:00:27
counter intelligence operations
01:00:28
operating around them. The modern art
01:00:30
movement was really funded by the CIA to
01:00:32
undermine the kind of Soviet control of
01:00:35
art, you know, and it was it was broadly
01:00:37
funded and it wasn't it wasn't directed
01:00:39
by CIA, but you you can see how we have
01:00:41
cultural values that we want to
01:00:44
inculcate and spread.
01:00:45
>> Do you think we have a shot at recasting
01:00:48
the defense industry, the tech industry
01:00:50
that's addressing defense and aligning
01:00:53
it with a notion of patriotism? And what
01:00:56
is it going to take to make that happen?
01:00:58
I don't know, maybe writing a book, home
01:01:00
mobilize.
01:01:01
>> Oh, yeah. You have a book.
01:01:02
>> That's right. That's right. I'm trying
01:01:03
to do exactly that. You know, look, if
01:01:05
if we look at a cleareyed sense of the
01:01:07
world and and how much deterrence we've
01:01:09
lost, you know, you could you could
01:01:10
really say like maybe World War II has
01:01:12
already started and 10 years from now,
01:01:13
we'll look back and and and be able to
01:01:14
perceive that, right? And I I I think
01:01:16
you don't need literally everyone to
01:01:19
view this, but if you can if you can
01:01:20
create a more cleareyed view of what's
01:01:22
at stake here, not only for you, but for
01:01:23
your children and their future, I I
01:01:25
think you can you can get people to show
01:01:26
up and participate. And this is this has
01:01:28
been America's story all the time, you
01:01:30
know, and usually when we start these
01:01:32
things, we are the underdog. All periods
01:01:35
of American greatness have started when
01:01:37
we realize that we're the underdog. We
01:01:39
were the 17th largest army in the world
01:01:40
at the beginning of World War II, you
01:01:42
know, and what did a rag tag bunch of
01:01:44
farmers and random tradesmen have taking
01:01:48
on the world's largest army in the
01:01:49
British during the Revolutionary War,
01:01:51
you know, and and I I think um
01:01:54
having some clarity about what's at
01:01:56
stake, what the counterfactual is. It's
01:01:57
so easy when you're successful to kind
01:01:59
of let the nealism grow and say like
01:02:01
look how imperfect we are. There's like
01:02:03
the self-loathing creeps in. And this is
01:02:04
where I take coming back to your point
01:02:06
on the legitimacy of our institutions.
01:02:08
like should these institutions work? Do
01:02:09
they deserve the best software? I mean,
01:02:11
look, if they're public or private, you
01:02:12
can't have doors falling off planes, you
01:02:14
you know, are are government
01:02:15
organizations need to provide the basic
01:02:17
services they've signed up to do without
01:02:19
fraud, without corruption in an
01:02:20
efficient way. And the reason is not
01:02:22
just an aesthetic. It's like in the
01:02:23
absence of that, it breeds nealism. And
01:02:25
the younger generations look at that and
01:02:27
say, you know, we should just tear all
01:02:28
of this down. And things will absolutely
01:02:30
get worse in a world that looks like
01:02:32
that. So, it's incumbent on us to wake
01:02:34
up every day and fight for the
01:02:36
legitimacy of these institutions to make
01:02:38
them more functioning.
01:02:39
>> America's story is never written. It's
01:02:40
every chapter seems to be a whole new
01:02:42
arc and we're in one right now. There's
01:02:45
a rising socialist movement in the
01:02:47
United States. Can you guys just comment
01:02:49
on how much you think that that
01:02:51
socialist movement cast with whatever
01:02:53
term they want to use is going to affect
01:02:55
our capacity for defense and resiliency
01:02:59
going into the next decade particularly
01:03:01
as our adversaries are rising?
01:03:03
>> Well, the argument I've always made is I
01:03:05
think our our greatest threat as a
01:03:06
nation is not homicide, it's suicide.
01:03:09
And it's it's in this vein. It's the
01:03:10
internal discord. It's the division. Uh
01:03:13
it's the it's the self-loathing. It's um
01:03:16
you know things like the socialist
01:03:17
movement which I think are symptomatic
01:03:19
of this internal discord and ensuring
01:03:21
opportunity for our people to the point
01:03:23
of a functioning elite it if you were to
01:03:25
go backwards and think about the root
01:03:27
cause like maybe we don't have an elite
01:03:28
that cares enough about the prosperity
01:03:30
of the American people and we've made
01:03:32
decisions through globalization we were
01:03:34
told NAFTA was going to be a great thing
01:03:36
that actually if you lost your job in
01:03:37
manufacturing why don't you just learn
01:03:39
how to code you know there there's a
01:03:40
certain sort of callousness in that and
01:03:42
you know I'm not an unabashed free
01:03:44
market and I'm clearly not a communist,
01:03:46
but there's a sense of where the
01:03:47
decision on the margin actually really
01:03:49
does matter and that comes down to
01:03:51
leadership.
01:03:51
>> I think one thing that makes me a little
01:03:53
hopeful is that socialism literally
01:03:55
doesn't work. Um, and so, you know, you
01:03:58
look at, you know, what happened in San
01:04:00
Francisco even where, you know, we
01:04:01
elected as the district attorney and
01:04:04
then we're like, "Oh, wow. This is not
01:04:06
going well. Recall him." We have the
01:04:08
board of education
01:04:09
>> rooted in a selling point of empathy,
01:04:10
right?
01:04:12
Yeah. Uh what happened with the board of
01:04:13
education getting recalled, you know,
01:04:15
Daniel Lur coming in as the mayor. Uh
01:04:18
you know, we felt like we kind of went
01:04:19
into this valley and sort of hit rock
01:04:21
bottom. Um but I do think that people
01:04:23
eventually realize that it doesn't it
01:04:25
just fundamentally doesn't work.
01:04:27
>> Seattle and Washington State are likely
01:04:29
going to lose a large number of their
01:04:31
biggest employers. And as that happens,
01:04:33
they'll come through on the other side.
01:04:35
Probably might take them 7 to 10 years
01:04:36
to get there. Yeah, it might be painful
01:04:38
and it might take them a long time, but
01:04:40
I think they'll eventually come to their
01:04:41
senses.
01:04:41
>> What do you two disagree on?
01:04:44
>> That's a great question.
01:04:45
>> Um, we actually, I'll be honest, we
01:04:47
bickered with each other quite a bit at
01:04:49
Palunteer. Um, and usually it was Sean
01:04:52
being right about something and me
01:04:54
taking a long time to come come along to
01:04:56
his point of view. Um, but uh, yeah, I
01:04:59
think you know, any of these cultures
01:05:01
that are like well functioning are
01:05:02
rooted in debate. Um, and you know,
01:05:05
eventually,
01:05:06
>> but you're avoiding my question.
01:05:08
>> Well, whether it's whether it's actions
01:05:11
or outlooks,
01:05:12
>> whether you should show up to your
01:05:14
interview at a tech company in a suit, I
01:05:16
don't know.
01:05:16
>> Well, no, I I will say this because it
01:05:18
actually is part of the story of
01:05:19
Androll. I thought that it was a bad
01:05:22
decision at Palunteer to be as quiet as
01:05:25
we were. I thought that we needed to get
01:05:27
out there and tell the story so that
01:05:29
there would be data that says things
01:05:31
like what Sham said about like we don't
01:05:33
have any data. We are Excel. Um and we
01:05:36
had a very kind of quiet reserved comm
01:05:38
strategy. Um and we went back and forth
01:05:41
on that a lot. And uh when I started
01:05:42
Andrew, I was like, you know, all the
01:05:44
positive things we learned about uh
01:05:46
doing business with the government from
01:05:48
Palunteer, the one lesson that I learned
01:05:50
that we didn't implement at Palanteer is
01:05:51
we're going to go out there and tell our
01:05:52
story. Um and I think that's worked for
01:05:54
us incredibly well. And I'll be honest,
01:05:56
I think Palanteer has come along to to
01:05:57
my my side of the of that debate.
01:05:59
>> And last question, if we don't do things
01:06:01
right, what does 2040 look like? And if
01:06:04
we do do things right, what does 2040
01:06:06
look like? Economic and defense.
01:06:08
>> Well, the the economic part is, I think,
01:06:10
the critical one because national
01:06:11
security is not an end unto itself. It's
01:06:13
a means to an end. And that that that
01:06:15
end is economic prosperity, the
01:06:16
prosperity of the American people. Um,
01:06:18
you know, I think no country has done
01:06:19
more to develop the world. Like how did
01:06:21
integrated circuits and micro
01:06:22
electronics get to Southeast Asia? We
01:06:24
sent them there. Yeah, we benefited in
01:06:26
terms of trade there, but you know, like
01:06:28
which other winner in a war spent their
01:06:31
own capital to rebuild the conqueror
01:06:33
conquered, you know, in Japan and
01:06:35
Germany and and now you have the
01:06:36
stability as a result of it. So when I
01:06:38
think about what could go wrong is you
01:06:39
actually have a Chinese century that we
01:06:42
never recover from that literally
01:06:43
everyone in the world is a vassal state
01:06:45
to China and might makes right in their
01:06:48
sort of world you know and we we
01:06:49
shouldn't forget it's like it's very
01:06:50
clear even in the present moment that
01:06:52
for the CCP it's not enough for China to
01:06:54
prosper America must fall that's an
01:06:56
explicit part of the strategy look it is
01:06:57
a business decision if you want to buy
01:06:59
American or Brazilian soybeans I
01:07:01
actually don't begrudge you one iota
01:07:02
which decision you make there it's an
01:07:04
entirely different decision when you're
01:07:05
smuggling agricultural funguses into the
01:07:07
US so that we can't grow soybeans.
01:07:09
That's the sort of zero sum frame that I
01:07:12
think 2040 will look like if we get this
01:07:14
wrong. And if we get it right, I think
01:07:15
what we actually see is a massive
01:07:17
re-industrialization of America followed
01:07:19
by the West. Um, we see a thriving
01:07:21
middle class, which I define
01:07:23
qualitatively as a middle class that
01:07:24
believes their children's future will be
01:07:26
better than their future, which is
01:07:28
something that I feel is is a
01:07:29
fundamental promise that's broken down
01:07:30
over the last 30 years or so. Uh, and a
01:07:33
belief in our institutions once again.
01:07:35
>> Yeah. No, completely agree with all
01:07:36
those points. Uh I I think there's a big
01:07:39
component here around education that we
01:07:40
haven't talked about as well where we
01:07:42
figured out a way to educate and
01:07:45
successfully enter our young people into
01:07:47
a marketplace that uh needs them and
01:07:50
that benefits from their services. Um
01:07:52
and I I think that we we haven't quite
01:07:54
nailed that. Um, but I I do think that
01:07:57
the re-industrialization point of this
01:07:58
is going to be kind of the the central
01:08:01
point of uh making sure that China's
01:08:04
many many decade belt and road strategy
01:08:06
is not going to put us in a position
01:08:08
where we literally just can't do
01:08:09
anything.
01:08:10
>> Do we need military and industrial
01:08:12
primacy? or can we operate in a
01:08:15
multipolar world where the US can share
01:08:18
influence and economic prosperity with
01:08:21
China, perhaps Russia, perhaps one or
01:08:23
two other nation states?
01:08:25
>> Well, the the challenge with not being
01:08:26
the leader is that you don't get to set
01:08:28
the terms of engagement. And so I think
01:08:30
the the benefit that we have had uh
01:08:32
since really the end of World War II is
01:08:34
that we've had the primary seat at the
01:08:36
table to say this is how we're going to
01:08:38
do uh you know semiconductors. This is
01:08:40
how we're going to do supply chain. this
01:08:42
is how we're going to protect trade
01:08:44
lanes. And I think the moment that you
01:08:46
step back from that, uh, and someone
01:08:48
else has all those incentives, uh, you
01:08:50
start playing the rules of their game
01:08:51
and it doesn't stay multipolar for long.
01:08:54
>> Well, I really appreciate the two of you
01:08:56
being here, Trey, Sean, thank you guys.
01:08:58
This has been great.
01:08:59
>> Thank you.
01:08:59
>> Thank you guys.
01:09:16
I'm going all in.
