WEBVTT

00:00:00.000 --> 00:00:03.397
Yeah, yeah, get in here class.
This is good. Okay,

00:00:03.480 --> 00:00:07.357
how is it going? How is ARP con
this time around? What's changed?

00:00:07.440 --> 00:00:10.317
>> Um, well, we're in a phase.
All each one

00:00:10.400 --> 00:00:14.037
of these things like marks a time. First
of all, you guys are even more ball or

00:00:14.120 --> 00:00:16.117
more successful. >> Thank you.

00:00:16.200 --> 00:00:17.197
>> Some tendies in your pocket.

00:00:17.280 --> 00:00:21.077
>> I think it might have been part two
you got to say thank you. You blew us up.

00:00:21.160 --> 00:00:24.117
You were the biggest guest.
>> You're looking bigger and stronger

00:00:24.200 --> 00:00:26.397
somehow. [laughter] Hey,
are you more attractive in your

00:00:26.480 --> 00:00:29.237
personal life now randomly? >> Well,
here's what we're actually focused

00:00:29.320 --> 00:00:32.197
on dead hangs. >> Yeah,
so you came on last time you said

00:00:32.280 --> 00:00:33.837
your dead hangs around like five.

00:00:33.920 --> 00:00:37.437
>> Oh, no. It's it's well, it's plateaued
in the last couple months at 5:30.

00:00:37.520 --> 00:00:41.237
>> 5:30. Okay, so we the thing
is like people are going to

00:00:41.320 --> 00:00:44.717
hear that. They're going to think
hanging on a bar how long could it be?

00:00:44.800 --> 00:00:47.557
You got to go and do it. The audience
has to go try to do it. We've started

00:00:47.640 --> 00:00:50.397
doing it. We're still
in the >> Under two minutes.

00:00:50.480 --> 00:00:54.917
>> in one between yeah, between somewhere
around a minute 30 you feel like your

00:00:55.000 --> 00:00:59.517
tendons are going to rip.
>> 30 1:30 dead hang is respectable.

00:00:59.600 --> 00:01:01.517
>> Okay. >> Two minutes is super elite.

00:01:01.600 --> 00:01:04.797
>> It doesn't feel respectable when
you have the five minute number you're

00:01:04.880 --> 00:01:08.197
looking at the time. >> Strength
strength strength matters. Now,

00:01:08.280 --> 00:01:11.797
the the the thing is I don't want to go
into rabbit hole in training. The single

00:01:11.880 --> 00:01:15.797
biggest mistake people make is they
try to hang every day. You need recovery

00:01:15.880 --> 00:01:20.197
second thing else. So, if you want to
mimic and get progress, you just do what

00:01:20.280 --> 00:01:23.277
I do which is once a week you hang
as long as you can. Doesn't have to be

00:01:23.360 --> 00:01:25.837
super macho. And then that's your day.

00:01:25.920 --> 00:01:27.677
So, let's just say you can do 1:30.

00:01:27.760 --> 00:01:30.557
>> sets? >> No, no, you know,
one day a week you do your maximum.

00:01:30.640 --> 00:01:32.677
>> Maximum. Wow. >> So,
like let's say you could do two

00:01:32.760 --> 00:01:34.997
minutes. You try to do at least 1:30.

00:01:35.080 --> 00:01:37.397
You fight to get to 1:30, but you
don't fight to get to two minutes.

00:01:37.480 --> 00:01:39.237
>> Got it. >> That's your
dead hang day. And then you

00:01:39.320 --> 00:01:42.237
can basically [__] around the next day
and do whatever you want. Don't overdo

00:01:42.320 --> 00:01:47.077
it, but you could do two times one
minute with a long break. And then can't

00:01:47.160 --> 00:01:50.797
you just screw around and do less
and less and less. Two days before if you

00:01:50.880 --> 00:01:55.237
two minute dead hang you do like four
times 15 seconds. The day before before

00:01:55.320 --> 00:01:58.637
off. And you do that.
just keep doing that,

00:01:58.720 --> 00:02:02.407
and your day What The mistake people
make is they hear my ball or time.

00:02:02.490 --> 00:02:05.557
>> [laughter] >> [__] that guy. I'm going

00:02:05.640 --> 00:02:08.597
>> I mean, the mistake you're making is
not doing a course. This could be a whole

00:02:08.680 --> 00:02:12.117
new revenue Wait,
you guys I mean, you guys have a

00:02:12.200 --> 00:02:15.557
war for you guys.
You could be like, "Hey, call in."

00:02:15.640 --> 00:02:19.717
>> So, um yeah, I mean, the dead hang
is that it's like And also, some of it's

00:02:19.800 --> 00:02:24.117
just genetic. Like, my other metrics
are elite, but that this is somehow alien

00:02:24.200 --> 00:02:28.717
territory. God-given gift. >> What
about What about breath hold under water?

00:02:28.800 --> 00:02:32.317
>> I don't do that, and I'm not sure I
have a Like, I grew up swimming. I I think

00:02:32.400 --> 00:02:36.437
I'm weaker at that. Like, I bet you
I would be in your guys' >> [laughter]

00:02:36.520 --> 00:02:39.717
>> I I I'm a I'm a dive master.
I can hold my breath for 3 minutes.

00:02:39.800 --> 00:02:42.037
>> I'm just say I'm just say So,
I think you'd be like

00:02:42.120 --> 00:02:44.797
I think you'd be crushing me on that,
honestly. But, you have like the lung

00:02:44.880 --> 00:02:46.237
capacity of a whale.
>> That's true. That's true.

00:02:46.320 --> 00:02:47.757
>> I mean, like >> Yeah.
>> You got like you got

00:02:47.840 --> 00:02:50.197
>> Yeah, if I'm not moving,
I'm not using any oxygen.

00:02:50.280 --> 00:02:51.757
>> It's like you're
like you're like you're

00:02:51.840 --> 00:02:55.557
like a like a like a whale floating
out there [laughter] under the

00:02:55.640 --> 00:02:57.557
waiting to surface. >> That's true.

00:02:57.640 --> 00:02:59.117
>> So, you know,
I'll say I'll tell you the

00:02:59.200 --> 00:03:02.957
difference. Uh it is just I got they're
always minding me out there. But, like

00:03:03.040 --> 00:03:05.037
okay, when we first met, it was like,

00:03:05.120 --> 00:03:08.957
"AI maybe real." Okay. Then, I would say

00:03:09.040 --> 00:03:14.477
somehow until about 2 weeks ago, there
was like a holy [__] this is real, but

00:03:14.560 --> 00:03:17.237
somehow it's not working, but we're
not allowed to say it publicly cuz we'll

00:03:17.320 --> 00:03:20.237
look stupid. And then there's
a lot of investor hype.

00:03:20.320 --> 00:03:22.477
>> Yep. >> Um there
still is like investors

00:03:22.560 --> 00:03:26.117
printing tendies. So, you have
the investors on one side. I think people

00:03:26.200 --> 00:03:31.317
realize it's real, but you know, it's like
you know, you have the whole token maxing.

00:03:31.400 --> 00:03:33.957
>> Yep. >> And people
are on onto that. And then

00:03:34.040 --> 00:03:35.997
so, there's a whole value lecture there.

00:03:36.080 --> 00:03:40.277
Um uh you have a political situation

00:03:40.360 --> 00:03:44.837
um where, you know, people who
do not understand basic economics

00:03:44.920 --> 00:03:48.317
are winning the political argument.
So, you can you could we could talk about

00:03:48.400 --> 00:03:50.157
where AI >> Yeah, there's
a lot here. Let's break it

00:03:50.240 --> 00:03:53.557
Let's break it up. Let's start with
the token maxing thing. Let's start with

00:03:53.640 --> 00:03:57.797
what's what's real.
How are you actually thinking about

00:03:57.880 --> 00:04:00.797
deploying AI >> First of all,
what what is your what is

00:04:00.880 --> 00:04:04.117
what is Palantir's philosophy
around token consumption?

00:04:04.200 --> 00:04:07.477
>> Well, like we okay, we have a product
that will allow you to be able I mean,

00:04:07.560 --> 00:04:11.917
internally it's called something,
but externally, but really we

00:04:12.000 --> 00:04:16.397
call it the de-masturbatory like get
off masturbation thing internally.

00:04:16.480 --> 00:04:19.477
>> It's like >> People are
just like print like sitting

00:04:19.560 --> 00:04:23.757
there all day kind of like a porn
addiction, and enterprises are like,

00:04:23.840 --> 00:04:28.637
"Okay, we knew this we believe this will
create value, but we cannot have people

00:04:28.720 --> 00:04:31.557
just like some people >> checking
the weather with it, just like

00:04:31.640 --> 00:04:35.157
and just re-re-rearranging deck
chairs on their personal Titanic.

00:04:35.240 --> 00:04:40.717
>> literally like porn. >> Like people
are like full on >> It feels it feels so

00:04:40.800 --> 00:04:43.197
>> objects, right?
>> Tool-shaped objects you're looking at

00:04:43.280 --> 00:04:46.477
more than you want. You hope no one
notices. You're kind of before dinner

00:04:46.560 --> 00:04:49.957
after dinner. >> Yeah,
and every email classified with tags.

00:04:50.040 --> 00:04:55.357
>> It comes down to like business problems
can never be I mean, sometimes they can

00:04:55.440 --> 00:04:58.797
be solved purely with money and just
spending more, but very often

00:04:58.880 --> 00:05:01.237
>> Actually, I think it's the opposite.
So, just to give you a weird analogy

00:05:01.320 --> 00:05:04.117
>> No, and I was going to say very very
often it's more the opposite where it's

00:05:04.200 --> 00:05:08.037
about figuring out the right way to do
something, and then you can use capital

00:05:08.120 --> 00:05:09.757
to fuel that process, but

00:05:09.840 --> 00:05:15.117
>> Let me give you a thing that's sui
generis for you guys. Okay. It it it's

00:05:15.200 --> 00:05:20.357
taste plus money. >> Okay.
>> And there is no the like AI Like if you

00:05:20.440 --> 00:05:23.597
look at like to pick any issue we
want to talk about. Token maxing is a

00:05:23.680 --> 00:05:26.557
maxing uh what's going on with deploy

00:05:26.640 --> 00:05:28.717
codes. Are other people going to build

00:05:28.800 --> 00:05:32.117
ontologies? Why are why why does the our

00:05:32.200 --> 00:05:36.437
political class not understand AI,
especially in Europe? It's like, "Yes,

00:05:36.520 --> 00:05:39.917
because all of these things can be
scaled in a very valuable, but largely

00:05:40.000 --> 00:05:45.317
going to commoditized way. But you can't
scale the taste of like, "What is the

00:05:45.400 --> 00:05:48.677
business problem you want to have
to solve and need to solve? At the end of

00:05:48.760 --> 00:05:51.837
the day, whether it's in the four
whether it's the Ukrainians fighting,

00:05:51.920 --> 00:05:56.037
the Israelis, commercial entities,
it's there's somebody sitting there who's

00:05:56.120 --> 00:05:59.957
like, "Okay, but this problem is valuable.
This problem isn't." And once

00:06:00.040 --> 00:06:03.557
that value that problem
is always has that problem

00:06:03.640 --> 00:06:06.717
alway almost always but not always has
attributes. So, there are some problems

00:06:06.800 --> 00:06:10.477
you could solve with this like,
"I want to write a report on GDP growth in

00:06:10.560 --> 00:06:14.117
China, right?" Okay, but if it's
a problem that requires a knowledge store,

00:06:14.200 --> 00:06:17.637
like I want to understand
the specialized way I underwrite. We're

00:06:17.720 --> 00:06:21.677
going to have a guest here. I want
to understand the specialized way I drill

00:06:21.760 --> 00:06:24.797
for oil and gas that's both legal,
ethical, and reduces the cost of

00:06:24.880 --> 00:06:29.877
production. I want to change the the
the supply chain of my industry, whether

00:06:29.960 --> 00:06:33.717
that's military or whether that's
building boxes, or whether that's cars.

00:06:33.800 --> 00:06:38.437
These things require actual
precise, ongoing processes.

00:06:38.520 --> 00:06:41.237
They are enhanced
by large language models.

00:06:41.320 --> 00:06:44.357
They are not replaced by large language
models. And then you get to security

00:06:44.440 --> 00:06:48.077
issues like But for us like the whole
myth of things is it's just a boon

00:06:48.160 --> 00:06:52.357
because like yeah, we can
take any model, their model,

00:06:52.440 --> 00:06:54.957
OpenAI model, open model, we can

00:06:55.040 --> 00:06:57.517
identify we can now identify

00:06:57.600 --> 00:07:02.437
vulnerabilities at like 10 100x. Yeah,
but then who patches them? Yeah. How do

00:07:02.520 --> 00:07:06.757
you patch them on prem? How do you patch
them on prem so that your specialized

00:07:06.840 --> 00:07:10.237
knowledge stays on prem? Like if you're
any business or intel service, there's a

00:07:10.320 --> 00:07:11.717
lot of these things are very similar.

00:07:11.800 --> 00:07:16.277
Like you're not putting your classified
data in a public cloud. Yeah. Same thing

00:07:16.360 --> 00:07:20.517
if you're like, you have a special way
of farming soybeans. Well, you're not

00:07:20.600 --> 00:07:24.957
So, it's like how do you have how do you
get So, all these problems are exposed,

00:07:25.040 --> 00:07:28.477
identified, and then you always
have a thing of where's the

00:07:28.560 --> 00:07:32.637
charisma, which people really
underestimate. And it's it's not global.

00:07:32.720 --> 00:07:36.357
There's no global charisma now. Like So,
right now the large language models are

00:07:36.440 --> 00:07:40.437
very Frontier companies are
super charismatic with investors.

00:07:40.520 --> 00:07:43.677
>> Mhm. >> I'll give you
some news. They're super

00:07:43.760 --> 00:07:46.637
not un-charismatic with
enterprises and the people.

00:07:46.720 --> 00:07:48.637
>> Yeah. >> In a way even with enterprise.

00:07:48.720 --> 00:07:50.477
>> No, no. I mean it's >> Cuz
I understand with the people.

00:07:50.560 --> 00:07:53.757
>> no. The enterprise people I have
a secret I have a secret like every

00:07:53.840 --> 00:07:57.197
company has a secret way of selling.
You know what my secret way of selling is?

00:07:57.280 --> 00:08:01.517
Don't even call it don't don't come talk
to us. There's a frontier company. Go

00:08:01.600 --> 00:08:05.357
spend 2 days with them. And if you're
lucky after you're done, I'll let you in

00:08:05.440 --> 00:08:07.197
my door. They're like
[laughter] clamoring.

00:08:07.280 --> 00:08:09.917
>> Yeah. >> They're like hey,
I'll take your bad brand.

00:08:10.000 --> 00:08:11.717
>> Yeah. >> We have a great
brand in enterprise. But

00:08:11.800 --> 00:08:15.117
like it it's like it's like secret
knowledge cuz the investors love this.

00:08:15.200 --> 00:08:17.997
They're like hey, my stocks are all up
everything's up. I mean Palantir's done

00:08:18.080 --> 00:08:21.437
very well. We're like it but it's like
and you know you guys are doing very

00:08:21.520 --> 00:08:23.077
well I imagine, right? And it's okay.

00:08:23.160 --> 00:08:26.797
We're we're you know we can but I'll
tell you what you go down the street you

00:08:26.880 --> 00:08:30.717
talk to a marine, you talk to a bus
driver, you talk to the person who owns

00:08:30.800 --> 00:08:33.237
the bus driving company. >> Mhm.

00:08:33.320 --> 00:08:37.157
>> They are not happy. They do not like
these people. They're tired of people

00:08:37.240 --> 00:08:38.477
token maxing. >> Yeah.

00:08:38.560 --> 00:08:44.237
>> It it looks like masturbation at
their that's cost them money. They they're

00:08:44.320 --> 00:08:46.437
like and honestly then you have
something we're not allowed to talk

00:08:46.520 --> 00:08:47.997
about in this country likeability.

00:08:48.080 --> 00:08:50.477
>> Mhm. >> Like Palantir,
we have I think we have

00:08:50.560 --> 00:08:53.997
like 50 100 million global bands.
We have like 5 million people that wake up

00:08:54.080 --> 00:08:56.357
in the morning literally calling me Satan.

00:08:56.440 --> 00:08:59.037
>> Mhm. >> I didn't know
I had that kind of >> Mhm.

00:08:59.120 --> 00:09:02.277
>> warm hand. But uh you know
it's like that's what they believe.

00:09:02.360 --> 00:09:04.717
>> Yeah. >> Yeah. >> And like
and they really believe it.

00:09:04.800 --> 00:09:09.397
Okay. What people are not allowed
to really address is like we have fans and

00:09:09.480 --> 00:09:11.597
enemies. >> Yeah. >> Yeah.
>> These people are polarizing.

00:09:11.680 --> 00:09:13.637
>> Yeah. >> We're polarizing
which means both sides.

00:09:13.720 --> 00:09:14.837
>> Yep. >> These people have one side.

00:09:14.920 --> 00:09:19.797
>> Yep. >> They're just it is so it's like
you and then it's like it's a really big

00:09:19.880 --> 00:09:21.557
>> Social media companies
too have the same problem.

00:09:21.640 --> 00:09:23.517
>> Yeah, no. >> Everyone uses
them but no one likes them.

00:09:23.600 --> 00:09:27.077
>> Yeah, but but then they also live in
a circle and that circle's printing money.

00:09:27.160 --> 00:09:28.917
>> Yep. >> So it's like
you know when you look in

00:09:29.000 --> 00:09:32.757
the mirror and you just printed
a lot of money you look pretty fresh.

00:09:32.840 --> 00:09:37.757
>> Is part of it is is part of it that
that some element of the technology, let's

00:09:37.840 --> 00:09:41.957
just say LLM's, is so magical that

00:09:42.040 --> 00:09:45.437
the companies involved, that the
companies that are making and selling

00:09:45.520 --> 00:09:48.677
frontier intelligence, can be bad
at a bunch of other things and still grow.

00:09:48.760 --> 00:09:51.677
>> Well, no. No, no. They are magical at a

00:09:51.760 --> 00:09:55.997
certain kind of thing, allowing you to
write, for example, code. Now, that code

00:09:56.080 --> 00:09:59.557
doesn't can't be used as a knowledge
store. So, if you look at code in like

00:09:59.640 --> 00:10:02.357
three different ways, like just using
pound as a model. We have code that's

00:10:02.440 --> 00:10:05.077
basically infrastructure. So, what what
are the Ukrainians using? What is the

00:10:05.160 --> 00:10:08.557
Department of War using? What do a lot
of our enterprises that that we call

00:10:08.640 --> 00:10:11.877
that primitives? So, they're basically
hard-coded things that that understand

00:10:11.960 --> 00:10:16.357
the world. Well, way do you do? It would
take millions of technical hours and an

00:10:16.440 --> 00:10:19.717
understanding of all these enterprises
to do it. So, it's it's much more like

00:10:19.800 --> 00:10:21.557
how do you build a steel beam? Then you

00:10:21.640 --> 00:10:25.597
have like code that is written by FDEs.

00:10:25.680 --> 00:10:30.277
Okay, so that's kind of managed. It's
The reason why FDEs work, the secret is

00:10:30.360 --> 00:10:33.277
it's actually managed in something that we
as a product. So, you're writing to a

00:10:33.360 --> 00:10:36.077
code base for managing that. We're
increasing our product. It's not just

00:10:36.160 --> 00:10:40.557
random people writing. Then you have,
let's call it free code. That free code

00:10:40.640 --> 00:10:44.557
is that's magical. Like you can do
it very quickly. It's almost right. It

00:10:44.640 --> 00:10:46.397
doesn't have to be exact. >> Dashboards.

00:10:46.480 --> 00:10:49.037
>> Dashboards, financial stuff,
>> Little little flow stuff.

00:10:49.120 --> 00:10:51.877
>> probabilistic stuff where you you
just have to get it >> One-off analysis.

00:10:51.960 --> 00:10:55.117
>> it was it magical. >> Yep.
>> By the way, it's magical and not only

00:10:55.200 --> 00:10:58.917
creates it and it's magical in a way I
don't I don't know people like the porn

00:10:59.000 --> 00:11:02.317
thing, but it's also addicting. It's
like you know it's not good for you, but

00:11:02.400 --> 00:11:04.597
you know, it may lead to damage.

00:11:04.680 --> 00:11:06.917
>> One more time. It can't hurt that much.

00:11:07.000 --> 00:11:11.077
I know my doctor says it I shouldn't do
it. But it's like it's like that, right?

00:11:11.160 --> 00:11:14.597
And you just keep going and like and if
you're involved in the that thing,

00:11:14.680 --> 00:11:16.437
you're also making money. >> Yeah.

00:11:16.520 --> 00:11:18.237
>> And then, last not least, in certain

00:11:18.320 --> 00:11:20.917
circles, like if you have you want to be

00:11:21.000 --> 00:11:25.717
a research researcher or you believe,
essentially it's a religion. So, like,

00:11:25.800 --> 00:11:28.557
you know, and like one of the things
is very charismatic, especially people

00:11:28.640 --> 00:11:31.757
who've never had a religion, cuz all
of a sudden that hole in your heart that

00:11:31.840 --> 00:11:33.877
was yearning for I don't know. I would

00:11:33.960 --> 00:11:36.597
say, you know, a a established religion,

00:11:36.680 --> 00:11:39.517
Judaism, Christianity, Islam is like

00:11:39.600 --> 00:11:43.917
being filled. And and all the answers
are there. But it is very very

00:11:44.000 --> 00:11:46.237
successful at doing things that a

00:11:46.320 --> 00:11:47.957
company has to do.

00:11:48.040 --> 00:11:51.397
But it is not actually solving the

00:11:51.480 --> 00:11:55.037
problem that enterprises are. It is Now,
it can solve them. That's the trick.

00:11:55.120 --> 00:11:58.477
It's not It's not binary. It's not
like You can't say they're not valuable.

00:11:58.560 --> 00:12:00.237
They're totally putting our business on

00:12:00.320 --> 00:12:03.597
steroids. Like without LLMs, nobody

00:12:03.680 --> 00:12:09.077
would be talking about our ontology,
about Apollo managing secure exploits,

00:12:09.160 --> 00:12:12.677
about our ability to manage
an enterprise essentially, uh turning all

00:12:12.760 --> 00:12:16.117
these companies into FDEs. These
deploy codes, we love them cuz now every

00:12:16.200 --> 00:12:18.757
company wants to deploy code.
You know how you do that?

00:12:18.840 --> 00:12:22.437
You re-platform on Palantir. And like
and it actually works. It's not somebody

00:12:22.520 --> 00:12:25.157
with no taste who's never done enterprise,
who has no earthly clue how

00:12:25.240 --> 00:12:28.237
these things work, who's done something
else and is like just imagining they

00:12:28.320 --> 00:12:29.637
know how to do it, right?

00:12:29.720 --> 00:12:32.397
>> Yeah, is part is part of this moment

00:12:32.480 --> 00:12:35.477
quite entertaining for you because

00:12:35.560 --> 00:12:37.717
you guys have been working on

00:12:37.800 --> 00:12:39.997
understanding businesses at a deep

00:12:40.080 --> 00:12:43.277
fundamental level, creating

00:12:43.360 --> 00:12:47.037
You guys have effectively been
doing the work that it that

00:12:47.120 --> 00:12:49.797
people are promising AI could do for 20

00:12:49.880 --> 00:12:51.837
years now, but actually doing it,

00:12:51.920 --> 00:12:57.277
finding all the really rough finding
all the really rough edges and

00:12:57.360 --> 00:13:00.717
uh and not and and being at a point
where you don't have to oversell the

00:13:00.800 --> 00:13:02.517
technology. You can sell both things.

00:13:02.600 --> 00:13:07.477
But now there's maybe Here we go.
We got it together. Um

00:13:07.560 --> 00:13:11.037
now there's maybe >> Oh,
it's the wrong side. That's why.

00:13:11.120 --> 00:13:14.997
Flip it around. >> There you go.

00:13:15.080 --> 00:13:19.480
There you go. There you go.
Living the dream. Living the dream.

00:13:19.920 --> 00:13:24.437
>> I I trailed off, but uh but is part
of it entertaining to you that it feels

00:13:24.520 --> 00:13:29.117
like you You Palantir
has always been in some

00:13:29.200 --> 00:13:32.197
ways um not had
competitors because there's

00:13:32.280 --> 00:13:35.197
nobody with Alex Karp running a company

00:13:35.280 --> 00:13:39.397
that is does what Palantir does besides
Palantir. But at the same time, there's

00:13:39.480 --> 00:13:42.477
been tens of billions
of dollars deployed now

00:13:42.560 --> 00:13:46.477
to effectively do what Palantir does, but

00:13:46.560 --> 00:13:51.157
just selling the intelligence part,
not selling all the underlying kind of

00:13:51.240 --> 00:13:53.157
infrastructure that you guys >> Well,
they're doing two things. They're

00:13:53.240 --> 00:13:56.837
selling They're trying to sell the
intelligence part and they're trying to

00:13:56.920 --> 00:13:59.597
pretend if you just hire a bunch
of people and let them run around their

00:13:59.680 --> 00:14:02.917
FTEs. Now, the the very cool thing is

00:14:03.000 --> 00:14:04.597
when you've been in your

00:14:04.680 --> 00:14:07.717
basement doing your thing and everyone

00:14:07.800 --> 00:14:12.117
kind of use it as the freak show,
it it's really interesting and and and

00:14:12.200 --> 00:14:14.637
great to have adoption. The pretty uh

00:14:14.720 --> 00:14:17.477
ironic thing is half the people adopting

00:14:17.560 --> 00:14:21.917
now don't even know they're copying.
But now, the copying thing it helps and

00:14:22.000 --> 00:14:27.197
hurts. Where it hurts is in the beginning
it puts clutter in the market.

00:14:27.280 --> 00:14:29.197
>> Yeah. >> Um and there's
there's no doubt about

00:14:29.280 --> 00:14:33.877
it. Where it helps and then we saw this
with defense tech, honestly. So like in

00:14:33.960 --> 00:14:37.437
defense tech, we were the only people
we're the first people despite what it's

00:14:37.520 --> 00:14:41.877
I I love these honestly other podcasters.
They're interviewing people

00:14:41.960 --> 00:14:44.877
who are parroting things I said 20 years
ago they don't know it. And it's like,

00:14:44.960 --> 00:14:47.637
"Oh, that's so insightful." It's like,
"Yeah, of course it's insightful. Karp

00:14:47.720 --> 00:14:51.237
said it 25 years ago." >> [laughter]
>> Uh and like but it's it's but so that

00:14:51.320 --> 00:14:54.037
kind of that part is super weird. But uh

00:14:54.120 --> 00:14:57.597
and like but um but it's but what really

00:14:57.680 --> 00:15:01.677
happens we see is like it expands
the market. So like in defense tech, we

00:15:01.760 --> 00:15:05.917
would not be doing this well
in just purely in government unless

00:15:06.000 --> 00:15:09.957
there weren't 50 companies that were
doing similar things because then the

00:15:10.040 --> 00:15:13.757
people are like, "Okay, first of all,
>> You view it as like off-balance sheet

00:15:13.840 --> 00:15:17.517
sales resources where people are
basically doing it themselves.

00:15:17.600 --> 00:15:19.557
>> now that that's
the largely it's They do

00:15:19.640 --> 00:15:24.717
two things. They increase the size of
the market because de facto nobody wants

00:15:24.800 --> 00:15:27.997
to find it under in a market where
there's only one person. Sure. So like

00:15:28.080 --> 00:15:31.837
if you're the one person, the percentage
of the defense budget you can get is

00:15:31.920 --> 00:15:36.397
much smaller. And two, they set up
a comparator. It's like, you know, you may

00:15:36.480 --> 00:15:40.277
not like the freak show. Okay.
If like but have you noticed the

00:15:40.360 --> 00:15:44.077
people who are serious buy it? And then
then it it change and then three it

00:15:44.160 --> 00:15:48.957
changes the standard. Now, what you're
seeing now is like that times a 100x.

00:15:49.040 --> 00:15:51.117
>> Yeah. >> And it does
change like recruiting,

00:15:51.200 --> 00:15:55.277
retention, and like how you build
a company and we're always think you have

00:15:55.360 --> 00:15:58.197
to think about how to being dyslexic
huge advantage there cuz like you don't

00:15:58.280 --> 00:16:01.637
have a playbook and now that we need you
need things to shift and we're doing that.

00:16:01.720 --> 00:16:02.597
>> Um

00:16:02.680 --> 00:16:05.437
>> The the the central thing though that

00:16:05.520 --> 00:16:10.317
it's just cannot be developed it even
if you understood the playbook a lot of

00:16:10.400 --> 00:16:14.037
these things are like appear like it's
like, you know, LM code appears like

00:16:14.120 --> 00:16:17.597
Palantir code but isn't for a deployed
thing appears like Palantir it isn't.

00:16:17.680 --> 00:16:21.997
Ontology you could a theoretically
copy parts of it but they're essentially

00:16:22.080 --> 00:16:25.837
structures that are built deep into
organizations that we own. And by the

00:16:25.920 --> 00:16:29.557
way, take you 3 years. And then 3 years
we're in a completely different world.

00:16:29.640 --> 00:16:34.917
But there is this magical thing called
taste. Like it in the end of the day the

00:16:35.000 --> 00:16:36.477
reason why you guys have done so well.

00:16:36.560 --> 00:16:39.277
It's of course there's
aptitude and diligence

00:16:39.360 --> 00:16:41.437
and showing up and all those things.

00:16:41.520 --> 00:16:45.237
Yeah, but you have to be able to
differentiate between two people who are

00:16:45.320 --> 00:16:48.517
in business one of whom is saying
something that sounds weird that is

00:16:48.600 --> 00:16:52.357
insightful. One of whom is parroting
something that sounds weird and that's

00:16:52.440 --> 00:16:56.797
all they're doing. And a lot of people
very few people can do that. And it's

00:16:56.880 --> 00:16:59.277
the same thing like the enterprises that

00:16:59.360 --> 00:17:02.757
succeed there is a taste arbiter. And at

00:17:02.840 --> 00:17:05.517
Palantir we have taste we have taste in

00:17:05.600 --> 00:17:09.797
every product, taste in every deployment,
taste in every casting. Who

00:17:09.880 --> 00:17:12.557
puts the people there? How do you put
them there? How do you organize the

00:17:12.640 --> 00:17:15.997
thing? Our ontology then does that
technically. How do you manage the whole

00:17:16.080 --> 00:17:18.077
org with taste? Who should be in charge?

00:17:18.160 --> 00:17:21.277
What data sets should come in? What
What are What are the ways in which you

00:17:21.360 --> 00:17:25.157
protect What is What should you push
into the public crowd? What should be on

00:17:25.240 --> 00:17:27.197
prem? What What should I mean, leaving

00:17:27.280 --> 00:17:32.197
aside the law and like wars war ethics,
what do you want to protect? What should

00:17:32.280 --> 00:17:35.757
you protect? What should you not protect
because quite frankly, you want that to

00:17:35.840 --> 00:17:37.557
be out there so you can get more data.

00:17:37.640 --> 00:17:40.917
All those things are arbitrated
by taste and then you have to have the

00:17:41.000 --> 00:17:44.917
credibility of having taste.
That's a real problem for a lot of these

00:17:45.000 --> 00:17:47.517
places cuz they don't have They They're
popular with their friends. They don't

00:17:47.600 --> 00:17:50.857
They really don't understand how
unpopular they are in enterprise. Like

00:17:50.940 --> 00:17:53.157
[laughter] they think it's like, "Oh,
yeah, it's like the way I think I have a

00:17:53.240 --> 00:17:54.757
problem with like professors at

00:17:54.840 --> 00:17:57.277
Columbia." It's like, "No, it's a real

00:17:57.360 --> 00:18:02.197
problem." Like they think I'm Satan.
And you know, it's like I I I I think, you

00:18:02.280 --> 00:18:03.277
know, we grew up in the same community.

00:18:03.360 --> 00:18:06.317
Let's talk about Heidegger. They're like
they don't want to talk about Heidegger.

00:18:06.400 --> 00:18:10.237
So, it's like it's like Yeah, and so
that's just a It's a weird thing. It's

00:18:10.320 --> 00:18:12.277
going to be a super The one thing I

00:18:12.360 --> 00:18:15.477
would say for anyone listening,

00:18:15.560 --> 00:18:19.797
if you're listening to this and you're
chillaxing and not active I'm not saying

00:18:19.880 --> 00:18:21.957
you have to agree with me
politically or anything.

00:18:22.040 --> 00:18:26.157
There the like partly because
of this dynamic and it very self-inflicted

00:18:26.240 --> 00:18:29.157
because I I I tell you
I can't name names. I

00:18:29.240 --> 00:18:33.397
called many of the titans
of this world and and

00:18:33.480 --> 00:18:36.397
like started this 6 months
ago like every couple days.

00:18:36.480 --> 00:18:37.437
We're going to be nationalized.

00:18:37.520 --> 00:18:40.317
>> them every couple days?
>> Uh like some of them I'm like, "Yeah,

00:18:40.400 --> 00:18:42.557
we're going to be I mean, you know,
it's like honestly they're like the batch

00:18:42.640 --> 00:18:44.397
They They They find me very entertaining."
[clears throat] Like I'm

00:18:44.480 --> 00:18:47.797
not sure Like so they call cuz yeah,
it's like it's like Yeah, this is going

00:18:47.880 --> 00:18:48.957
to be entertaining. >> to pick up.

00:18:49.040 --> 00:18:53.397
>> Yeah, so in any case, so I've been
telling them for 6 months 6 We're going

00:18:53.480 --> 00:18:55.677
to be nationalized. >> Yeah.
Yeah, you talked about

00:18:55.760 --> 00:18:58.797
>> We're going to be nationalized.
And they're like, "Why would anyone

00:18:58.880 --> 00:19:00.877
nationalize? Never happened in America.

00:19:00.960 --> 00:19:04.757
It's never Why would anyone nationalize
us? We're so likable. We're creating so

00:19:04.840 --> 00:19:08.957
much value." Like, "Okay, I'm not going
to debate that. I know how likable I am.

00:19:09.040 --> 00:19:12.517
I'm not going to tell you how likable
you are. But I am telling you and you

00:19:12.600 --> 00:19:16.877
know the momentum on this is on the side
of people in national and we don't get

00:19:16.960 --> 00:19:20.277
our act together and figure out ways we
can say, "Hey, look, there are problems

00:19:20.360 --> 00:19:23.157
here. We're going to deal with these
things that are not going to Yes, they

00:19:23.240 --> 00:19:26.477
are going to create opportunities.
You have to talk openly about how these

00:19:26.560 --> 00:19:28.957
things are valuable because we have
adversaries. You can't just say these

00:19:29.040 --> 00:19:33.517
are All that stuff." So, the primary risk,
honestly, to Palantir and to a lot

00:19:33.600 --> 00:19:37.677
of these other countries is And then,
it's going to be nationalized before

00:19:37.760 --> 00:19:40.197
national is going to be regulated
by people who don't understand understand

00:19:40.280 --> 00:19:43.037
this. And now, they'll tell you in
private, "I'm working on this. I'm da da

00:19:43.120 --> 00:19:46.797
da and this and this lobbyist."
It's like, "Not going to work." So, like,

00:19:46.880 --> 00:19:49.077
that's something like if you're
listening to this and you're like,

00:19:49.160 --> 00:19:52.837
"Look, you know, you don't have to agree
with me on all my proclamations. I got a

00:19:52.920 --> 00:19:55.637
lot of By the way, there's some people
who think I'm saying we should have a

00:19:55.720 --> 00:19:59.197
draft." Too lazy to read. I'm just
saying we should Like, in a world where

00:19:59.280 --> 00:20:02.757
everything is changing,
everything is changing.

00:20:02.840 --> 00:20:06.197
Don't we have to find some communal
structure to remember we're American?

00:20:06.280 --> 00:20:09.797
You don't like my You don't like my idea
of like we all do a week in the park.

00:20:09.880 --> 00:20:11.957
Great. Come up with some other idea.

00:20:12.040 --> 00:20:15.277
But, we can't have no idea. You know,
and like and then they're like, "Well,

00:20:15.360 --> 00:20:17.117
I'm saying I I do not want to draft."

00:20:17.200 --> 00:20:20.197
Just to be explicit. They're like, "Oh,
that's pro-war." Now, honestly, you know

00:20:20.280 --> 00:20:23.397
what most of our wars are fought because
no working-class person is making a

00:20:23.480 --> 00:20:26.877
decision. You start making sure everyone
is involved in everything. I'll see you

00:20:26.960 --> 00:20:30.797
how few wars we fight. It's actually
the anti-war position. But, any case,

00:20:30.880 --> 00:20:32.957
disagree with everything. If you're We

00:20:33.040 --> 00:20:38.117
have on the right and on the left people
people who have no earthly clue what

00:20:38.200 --> 00:20:39.677
they're talking about, right and left.

00:20:39.760 --> 00:20:43.037
All they're talking about is how much
they hate us. And those of us who are

00:20:43.120 --> 00:20:46.797
sensible in the middle, you know,
too many of us are chillaxing.

00:20:46.880 --> 00:20:51.437
Like, it Like, nationalization, it can't
happen. It's America would never do that.

00:20:51.520 --> 00:20:54.917
Sleepwalking into And you guys have
tendies to protect now. You [laughter]

00:20:55.000 --> 00:20:58.437
guys should be on the front line of this.
Like, you got full Oh, I'm sorry.

00:20:58.520 --> 00:21:02.997
I have a I have a full-on very
impressive corporate leader coming on.

00:21:03.080 --> 00:21:04.917
>> Yes. >> So, I got
to I got to turn it down.

00:21:05.000 --> 00:21:07.757
>> Last Last question if we have time.

00:21:07.840 --> 00:21:10.677
How are your conversations going with

00:21:10.760 --> 00:21:13.917
Fortune 500 CEOs around headcount

00:21:14.000 --> 00:21:17.797
planning? There's been so many layoffs
this last year that people were saying,

00:21:17.880 --> 00:21:19.277
"Hey, we're getting so much out of AI.

00:21:19.360 --> 00:21:21.277
We're able to, you know, cut back here

00:21:21.360 --> 00:21:22.437
or there."

00:21:22.520 --> 00:21:27.037
People inside tech often know like these

00:21:27.120 --> 00:21:31.597
maybe there's just reduction because
there needs to be a reduction. The org

00:21:31.680 --> 00:21:34.997
got bloated. Maybe they do need
to fund some AI initiatives.

00:21:35.080 --> 00:21:37.357
>> business model like they're getting
out competed by someone else. Yeah, yeah.

00:21:37.440 --> 00:21:38.997
Yeah, the business just doesn't
have momentum. But how are those

00:21:39.080 --> 00:21:40.997
conversations going?
>> What does it look like?

00:21:41.080 --> 00:21:44.317
>> Like the like I talk By the way,
I talk to Fortune 500 companies. I talk to

00:21:44.400 --> 00:21:47.837
unions. I talk to soldiers. I talk to fire

00:21:47.920 --> 00:21:52.837
If you upskill somebody,
they're more valuable. And like all

00:21:52.920 --> 00:21:56.197
these whether it's people working
on batteries, people driving trucks,

00:21:56.280 --> 00:22:00.077
people, corporate leaders, and again,
this is where I think we have to be very

00:22:00.160 --> 00:22:03.517
careful to be more disciplined on the
corporate side. Like if you run around

00:22:03.600 --> 00:22:07.477
saying AI allowed you to fire 2/3
of your workforce and you did it because

00:22:07.560 --> 00:22:10.197
maybe your competitor's kicking your ass.

00:22:10.280 --> 00:22:14.517
That could that is a really like you
might as well just go sign up for Bernie

00:22:14.600 --> 00:22:18.637
Bernie Sanders manifest. And part of the
thing is they really believe that can't

00:22:18.720 --> 00:22:22.397
happen. So they're free riding on the
fact that it could. Like we have And

00:22:22.480 --> 00:22:25.797
then it just cannot work anymore.
These things are very, very explosive. The

00:22:25.880 --> 00:22:29.757
American people sense that there
is something dangerous here. And when

00:22:29.840 --> 00:22:33.597
people are playing with that fire, it's
like it's a they assume the fire won't

00:22:33.680 --> 00:22:36.437
burn their hands. Well, that's not
the world we're in. That fire is going to

00:22:36.520 --> 00:22:39.397
consume us. And what we see again, the

00:22:39.480 --> 00:22:41.437
war-fighting example is just the most

00:22:41.520 --> 00:22:46.477
neutral, not for everybody. But like the
soldiers at the bottom have gotten much

00:22:46.560 --> 00:22:49.117
more valuable. Well, and and I don't
even just mean the special operators,

00:22:49.200 --> 00:22:52.997
which obviously they're in a different
league. But like every the people doing

00:22:53.080 --> 00:22:56.317
a lot of the operations now are
doing our product their high school

00:22:56.400 --> 00:23:01.117
vocationally trained. You see this
everywhere. The the modern enterprise is

00:23:01.200 --> 00:23:03.237
going to have like we have a

00:23:03.320 --> 00:23:06.997
a true like very very very smart

00:23:07.080 --> 00:23:10.637
person coming on and it's like you're
going to have a very smart executive.

00:23:10.720 --> 00:23:14.437
He's much better at hiding it than I
would be if I were him, but that's you

00:23:14.520 --> 00:23:16.957
can talk to him about that. But um

00:23:17.040 --> 00:23:19.477
uh um and

00:23:19.560 --> 00:23:22.477
uh and then very talented creative

00:23:22.560 --> 00:23:26.517
people with taste all up and down the
stack. In any case, I think this is time

00:23:26.600 --> 00:23:29.437
for me to >> this is time.
>> We uh >> Thank you so much.

00:23:29.520 --> 00:23:33.240
>> Great to be here. >> Yeah, always fun.