1
00:00:00,090 --> 00:00:01,807
- We have to build systems like Palantir

2
00:00:01,890 --> 00:00:04,417
and we have to explain to people
with working class skills,

3
00:00:04,500 --> 00:00:07,537
people with normal jobs, that their
labor's gonna become more important

4
00:00:07,620 --> 00:00:10,237
and more valuable over time
with products like ours.

5
00:00:10,320 --> 00:00:11,767
If you run around saying that, you know,

6
00:00:11,850 --> 00:00:14,917
oh, the economy is gonna
completely shift to value creation,

7
00:00:15,000 --> 00:00:19,417
only owned by 10,000 people, people on
the left and the right are gonna go nuts.

8
00:00:19,500 --> 00:00:21,787
- I'm Steven Levy,
Editor at Large at "Wired."

9
00:00:21,870 --> 00:00:24,967
I sat down with Alex
Karp, CEO of Palantir,

10
00:00:25,050 --> 00:00:27,637
a controversial company
that orchestrates data

11
00:00:27,720 --> 00:00:30,337
and solves problems
for government and business.

12
00:00:30,420 --> 00:00:33,127
We discussed Palantir's
work with ICE and Israel

13
00:00:33,210 --> 00:00:35,797
and why he feels the company
has no competitors.

14
00:00:35,880 --> 00:00:37,837
Welcome to "The Big Interview."

15
00:00:37,920 --> 00:00:40,590
[intense music]

16
00:00:44,070 --> 00:00:46,657
Well, first of all, lemme say
welcome to "The Big Interview."

17
00:00:46,740 --> 00:00:51,667
- Okay. - Dr. Alex Karp.
You said that Palantir is at odds

18
00:00:51,750 --> 00:00:54,337
with Silicon Valley
from the very beginning.

19
00:00:54,420 --> 00:00:57,337
You have kind of a harsh
critique of Silicon Valley.

20
00:00:57,420 --> 00:01:00,607
You want Palantir to be different.

21
00:01:00,690 --> 00:01:02,947
What do you think the biggest
problem of Silicon Valley was?

22
00:01:03,030 --> 00:01:07,537
- Well, we have a long and managing
relationship with Silicon Valley.

23
00:01:07,620 --> 00:01:11,377
In the beginning we were at odds
because we were pro-American, pro-West,

24
00:01:11,460 --> 00:01:13,087
and pro-making the government functional,

25
00:01:13,170 --> 00:01:15,037
and that was very
controversial in Silicon Valley

26
00:01:15,120 --> 00:01:18,327
because it equated to not making
any money and being a loser.

27
00:01:18,410 --> 00:01:22,057
We won that battle. And I think
Silicon Valley actually has become very,

28
00:01:22,140 --> 00:01:25,177
at least officially and even
behind closed doors, patriotic

29
00:01:25,260 --> 00:01:28,597
and Silicon Valley has
always been pro-meritocracy

30
00:01:28,680 --> 00:01:30,517
and we're very aligned with that.

31
00:01:30,600 --> 00:01:35,077
Where we're currently at like
a misalignment-alignment apex again,

32
00:01:35,160 --> 00:01:38,287
is we believe in using
large language models

33
00:01:38,370 --> 00:01:40,747
in a way that creates
actual empirical value,

34
00:01:40,830 --> 00:01:44,557
but also, is very strong
for workers, technical people,

35
00:01:44,640 --> 00:01:46,717
whether you're a plumber,
carpenter, or electrician,

36
00:01:46,800 --> 00:01:48,697
or you are a high school
grad working in a plant

37
00:01:48,780 --> 00:01:51,207
where you'd normally be doing
engineering things or couldn't do them.

38
00:01:51,290 --> 00:01:54,607
We see our products as being
able to enhance your market value.

39
00:01:54,690 --> 00:01:57,397
And so, we're kind
of pro-labor and pro-military.

40
00:01:57,480 --> 00:02:00,637
So we're much closer to Silicon
Valley on the patriotism thing.

41
00:02:00,720 --> 00:02:04,147
But for decades we fought Silicon Valley

42
00:02:04,230 --> 00:02:07,267
about, you know, government
is something we should support.

43
00:02:07,350 --> 00:02:10,267
We walked towards the government
when everyone walked away.

44
00:02:10,350 --> 00:02:12,637
That's how we ended up powering Maven,

45
00:02:12,720 --> 00:02:14,937
which is the, you know,
US government's military.

46
00:02:15,020 --> 00:02:17,537
- Yeah, you took advantage
when the Google workforce said,

47
00:02:17,620 --> 00:02:19,987
"We don't want Google to work on Maven."

48
00:02:20,070 --> 00:02:22,057
- And then we built it, yeah, but.

49
00:02:22,140 --> 00:02:23,587
- Explain what Project Maven is.

50
00:02:23,670 --> 00:02:26,767
- Well, so to find
an adversary on a battlefield

51
00:02:26,850 --> 00:02:31,597
from any sensor including space, is a
massive integration in analytics project.

52
00:02:31,680 --> 00:02:34,177
So, how do you find basically
a needle in a haystack?

53
00:02:34,260 --> 00:02:38,977
And where that needle is also complex
because it may look like your adversary,

54
00:02:39,060 --> 00:02:41,187
but maybe that person's
actually working for you.

55
00:02:41,270 --> 00:02:44,527
Maybe they're in your service,
maybe their phone has been compromised.

56
00:02:44,610 --> 00:02:48,187
Finding that needle in a
haystack, so find a person

57
00:02:48,270 --> 00:02:51,577
that is your adversary, you wanna
take out across a massive landscape

58
00:02:51,660 --> 00:02:54,937
where you have a human controlled
in the loop, like targeting workbench.

59
00:02:55,020 --> 00:02:57,007
So you say, you know,
this target's been approved.

60
00:02:57,090 --> 00:03:00,517
Why? Under what conditions?
The data protections that are necessary,

61
00:03:00,600 --> 00:03:02,977
the civil liberty protections,
interestingly, that are necessary.

62
00:03:03,060 --> 00:03:06,637
And doing that at scale on the
battlefield was not possible.

63
00:03:06,720 --> 00:03:10,357
And now, it's been used in every
operational context across the world.

64
00:03:10,440 --> 00:03:12,457
And you know, there are
many things that have led

65
00:03:12,540 --> 00:03:15,757
to America's ability
to reestablish its deterrence,

66
00:03:15,840 --> 00:03:18,067
I would say, after
the disaster in Afghanistan.

67
00:03:18,150 --> 00:03:21,817
But one of them was actually
one of many corrosicism

68
00:03:21,900 --> 00:03:25,877
of our soldiers may, I think one of
the other things is just the superiority

69
00:03:25,960 --> 00:03:27,517
of the targeting capabilities
the US government has.

70
00:03:27,600 --> 00:03:30,007
And you know, there are other
people in other companies,

71
00:03:30,090 --> 00:03:31,147
but we are the backbone of that.

72
00:03:31,230 --> 00:03:34,717
- A lot of your government
connections, they go back early.

73
00:03:34,800 --> 00:03:38,437
You know, you've worked for ICE, you've
had long-term contract with them, right?

74
00:03:38,520 --> 00:03:39,847
- Well, that's actually later.

75
00:03:39,930 --> 00:03:43,027
I mean, our first contracts
were at US Intelligence,

76
00:03:43,110 --> 00:03:47,647
so CIA, Special Operations, and FBI

77
00:03:47,730 --> 00:03:50,617
and our biggest contracts
ended up being in the DOD.

78
00:03:50,700 --> 00:03:54,157
And then later we started working
in Homeland Security under Democrats.

79
00:03:54,240 --> 00:03:56,827
Now it's very controversial
because it's Trump in my view.

80
00:03:56,910 --> 00:03:58,207
- Well, it's controversial

81
00:03:58,290 --> 00:04:01,237
because he's doing things
that we haven't seen done before.

82
00:04:01,320 --> 00:04:05,047
I want to ask you,
do you keep an eye on that?

83
00:04:05,130 --> 00:04:10,267
Are you making sure that your products
are being used according to your code,

84
00:04:10,350 --> 00:04:11,947
which supports democracy?

85
00:04:12,030 --> 00:04:15,937
And you know, it says people
can't be discriminated against.

86
00:04:16,020 --> 00:04:19,327
I'm wondering if there's a line
for you where you would say-

87
00:04:19,410 --> 00:04:20,827
- Okay. - To a president, "No."

88
00:04:20,910 --> 00:04:22,687
- We were the, I mean this is like,

89
00:04:22,770 --> 00:04:24,757
there's all these things
that no one believes are true.

90
00:04:24,840 --> 00:04:28,327
I was the first, and I
may be one of the few CEOs

91
00:04:28,410 --> 00:04:30,427
to say we would not
build a Muslim database.

92
00:04:30,510 --> 00:04:33,997
Now, everyone thinks I'm like,
and I am, you know, I've defended Israel.

93
00:04:34,080 --> 00:04:37,597
I think, like, people who
wake up in the morning thinking

94
00:04:37,680 --> 00:04:38,947
about how Jews are running the world,

95
00:04:39,030 --> 00:04:41,287
and actually, you believe
whatever you wanna believe.

96
00:04:41,370 --> 00:04:44,497
I'm not so concerned about people
who are skeptical about Jews

97
00:04:44,580 --> 00:04:45,817
or even their antisemitic,

98
00:04:45,900 --> 00:04:47,947
but like people who wake up
in the morning, in the evening,

99
00:04:48,030 --> 00:04:49,297
and night thinking about this.

100
00:04:49,380 --> 00:04:53,407
I think like, you know, Jewish derangement
syndrome is a little bit of a carcinogen.

101
00:04:53,490 --> 00:04:55,537
So, I think I'm the last
person in the world

102
00:04:55,620 --> 00:04:58,387
people kind of expect in the kind of rude

103
00:04:58,470 --> 00:05:01,327
and crude way people think
of this to refuse revenue,

104
00:05:01,410 --> 00:05:03,787
to build, you know, a Muslim database.

105
00:05:03,870 --> 00:05:07,597
And I would say more like at a high level,
'cause I can't go into the details,

106
00:05:07,680 --> 00:05:11,017
people internally know I pulled things

107
00:05:11,100 --> 00:05:14,677
from places where I thought there
was anything like that going on in the US.

108
00:05:14,760 --> 00:05:16,267
Discrimination, yes, I'm against that.

109
00:05:16,350 --> 00:05:18,337
By the way, even if I was in favor of it,

110
00:05:18,420 --> 00:05:21,547
our products make
that nearly as impossible

111
00:05:21,630 --> 00:05:23,317
as anyone's ever seen in a product.

112
00:05:23,400 --> 00:05:25,267
Because the way you do these things,

113
00:05:25,350 --> 00:05:27,997
the way you discriminate
against people politically

114
00:05:28,080 --> 00:05:31,537
or based on their gender or race,
you do not have what are called ackles.

115
00:05:31,620 --> 00:05:33,907
So, you can't see where
the data's flowing.

116
00:05:33,990 --> 00:05:36,367
What you can immediately see,
not as a technical person,

117
00:05:36,450 --> 00:05:40,147
but as you, your cameraman,
everyone watching this can go look

118
00:05:40,230 --> 00:05:42,187
and see how is the data merged?

119
00:05:42,270 --> 00:05:43,627
What kind of pipelines did you create?

120
00:05:43,710 --> 00:05:45,667
What were the assumptions
that went into it?

121
00:05:45,750 --> 00:05:49,807
Where did the data flow?
But where you and I disagree with maybe

122
00:05:49,890 --> 00:05:52,687
is I spent half my life in Europe.

123
00:05:52,770 --> 00:05:58,927
I am an immigration skeptic. Like,
and I personally think that US people,

124
00:05:59,010 --> 00:06:03,667
meaning citizens,
have to decide by their vote

125
00:06:03,750 --> 00:06:05,407
what our immigration policy is gonna be.

126
00:06:05,490 --> 00:06:07,507
And I'll tell you one
of the reasons I'm very skeptical

127
00:06:07,590 --> 00:06:13,717
of my own party is that open border was
not only a disaster, but it was tolerated.

128
00:06:13,800 --> 00:06:17,317
And that is massively corrosive
and it's corrosive to democracy

129
00:06:17,400 --> 00:06:20,767
and it's corrosive to people. And leaving
all the theory aside, like I was,

130
00:06:20,850 --> 00:06:24,907
I grew up in a progressive family,
classic progressives my age and older

131
00:06:24,990 --> 00:06:27,337
were massive immigration skeptics.

132
00:06:27,420 --> 00:06:30,157
Open borders is not a progressive policy.

133
00:06:30,240 --> 00:06:31,837
It is not a progressive policy.

134
00:06:31,920 --> 00:06:35,437
It never was. And then
the fruits of this, you know,

135
00:06:35,520 --> 00:06:38,107
I spent most of my life,
adult life in Germany.

136
00:06:38,190 --> 00:06:41,827
Look at what a version of open borders
a little different has done to Germany.

137
00:06:41,910 --> 00:06:46,657
On any vector: rights of women, rights
of gay people, future of society. - Well.

138
00:06:46,740 --> 00:06:49,327
- I just wanna make it clear. - Yeah,
yeah. - I am against discrimination.

139
00:06:49,410 --> 00:06:50,527
I'm not saying you're saying this,

140
00:06:50,610 --> 00:06:53,467
but people tend to merge
being against discrimination

141
00:06:53,550 --> 00:06:57,457
with being, to be against discrimination,
I have to accept that there's nothing

142
00:06:57,540 --> 00:07:00,277
that can be done about illegal
immigration in the West.

143
00:07:00,360 --> 00:07:01,747
And I just, I disagree with that.

144
00:07:01,830 --> 00:07:03,997
But more importantly than
me disagreeing with it,

145
00:07:04,080 --> 00:07:05,557
the American people disagree with it.

146
00:07:05,640 --> 00:07:07,447
And I, quite frankly,
you're not asking this,

147
00:07:07,530 --> 00:07:09,337
I think one of the reasons
why the Democratic Party

148
00:07:09,420 --> 00:07:11,497
continues to fail and continues to fail

149
00:07:11,580 --> 00:07:14,197
and continues to fail
is their talent managed.

150
00:07:14,280 --> 00:07:16,747
Which like in businesses,
like you have this problem

151
00:07:16,830 --> 00:07:20,297
where the people doing the work
sometimes need to be told,

152
00:07:20,380 --> 00:07:21,787
"No, you're wrong
and we're not doing that."

153
00:07:21,870 --> 00:07:25,147
And the Democratic party and people
like it, the Green Party in Germany,

154
00:07:25,230 --> 00:07:26,647
I don't view these
policies as progressive.

155
00:07:26,730 --> 00:07:31,277
- Well, I guess what I'm saying is are
you monitoring sort of what's happening

156
00:07:31,360 --> 00:07:34,477
with this government
and democracy in this country

157
00:07:34,560 --> 00:07:38,827
and saying, "Maybe at some
point I have to look at-"

158
00:07:38,910 --> 00:07:40,177
- I would say the more important question

159
00:07:40,260 --> 00:07:42,907
is have I ever worked against
our commercial interest

160
00:07:42,990 --> 00:07:44,407
because it violated our norms?

161
00:07:44,490 --> 00:07:46,237
Yes. Have I done this in governments?

162
00:07:46,320 --> 00:07:50,097
Yes. Have I refused? We get no credit for
this, but we almost went outta business,

163
00:07:50,180 --> 00:07:53,527
'cause we were not working in Russia,
China, or anywhere else, so yes.

164
00:07:53,610 --> 00:07:55,747
Have we refused give our product

165
00:07:55,830 --> 00:07:58,297
to foreign governments because
we didn't agree with them?

166
00:07:58,380 --> 00:08:01,657
Yes. Do I agree with
your implicit assertion

167
00:08:01,740 --> 00:08:05,497
that what's going on in immigration as you
formulate it has never been done before?

168
00:08:05,580 --> 00:08:07,837
No, I think actually
that's completely crazy.

169
00:08:07,920 --> 00:08:09,517
You know, when you're in Japan,

170
00:08:09,600 --> 00:08:13,747
if you miss your visa,
do you know what happens to you?

171
00:08:13,830 --> 00:08:16,137
- I didn't overstay my visa.
- Okay, great.

172
00:08:16,220 --> 00:08:18,547
I'll tell you what happens.
You politely are put on a plane.

173
00:08:18,630 --> 00:08:21,907
Is Japan not a democracy? Do you
know what happens in like Singapore?

174
00:08:21,990 --> 00:08:26,017
- Well, I think like- - You're,
but I just, no, I think this

175
00:08:26,100 --> 00:08:27,287
is a very valuable line of questioning.

176
00:08:27,370 --> 00:08:33,337
You're asking does our product allow for
civil rights abuses and will I intervene?

177
00:08:33,420 --> 00:08:36,547
- Yeah. - Our product is the
hardest in the world to violate.

178
00:08:36,630 --> 00:08:40,657
Will I intervene? Yes. Do I agree
with some of the assertions you just made?

179
00:08:40,740 --> 00:08:45,307
No. The legitimate criticism could be,
hey, these are two very rational people.

180
00:08:45,390 --> 00:08:48,397
You know, you're bringing up
exactly the core and right question.

181
00:08:48,480 --> 00:08:52,027
And I'm telling you that I
have done this, I will do it.

182
00:08:52,110 --> 00:08:55,207
And I've continued to do it.
And by the way, not that people our age

183
00:08:55,290 --> 00:08:56,887
like the term constructive engagement,

184
00:08:56,970 --> 00:09:00,937
I am as obnoxious in private as I am
with you, with people I agree with.

185
00:09:01,020 --> 00:09:03,457
So, like, when they're
asking me to do stupid shit,

186
00:09:03,540 --> 00:09:06,767
I can assure you, I'm telling- - Wait,
is this stupid shit or something?

187
00:09:06,850 --> 00:09:07,607
- No, no, I'm not saying
you're asking me to do it,

188
00:09:07,690 --> 00:09:09,097
but I can assure you one of the things

189
00:09:09,180 --> 00:09:13,057
you can be certain of 100%
is when somebody calls me

190
00:09:13,140 --> 00:09:15,397
and asking me to do some
completely illogical,

191
00:09:15,480 --> 00:09:17,257
mostly it's illogical before it's immoral.

192
00:09:17,340 --> 00:09:18,457
That's the thing people understand.

193
00:09:18,540 --> 00:09:21,247
It's illogical. It will not
get you the result you think.

194
00:09:21,330 --> 00:09:24,687
I am telling them, "This is illogical
and we're not gonna do it."

195
00:09:24,770 --> 00:09:26,917
By the way, there's
variants of this commercial.

196
00:09:27,000 --> 00:09:29,287
I have commercial clients,
this is not even moral,

197
00:09:29,370 --> 00:09:31,657
who are like, "Hey,
could you give me 50 engineers

198
00:09:31,740 --> 00:09:33,877
to do some crazy thing that will
not help their business?"

199
00:09:33,960 --> 00:09:35,497
And you know what I'm telling them?

200
00:09:35,580 --> 00:09:38,257
No. One of the things that happens
when people work with me

201
00:09:38,340 --> 00:09:39,847
is they see I'm meeting clients

202
00:09:39,930 --> 00:09:42,067
and they'll ask me to do
something that makes no sense

203
00:09:42,150 --> 00:09:44,557
and I'll tell them,
"I can't go to my engineers

204
00:09:44,640 --> 00:09:45,857
and tell them to do something
that makes no sense."

205
00:09:45,940 --> 00:09:48,847
- Yeah. - And this is, I think a lot
of these questions come down to that.

206
00:09:48,930 --> 00:09:55,537
Now, where we have to expand the prism of
dialogue is I think in complete honesty,

207
00:09:55,620 --> 00:09:58,327
there is a place where you
and I, forget you and I,

208
00:09:58,410 --> 00:10:02,137
you step in, you're just honestly talking
to you's just like talking to my family.

209
00:10:02,220 --> 00:10:05,107
It's like, where me
and my family disagree.

210
00:10:05,190 --> 00:10:07,057
And I'll tell you where those places are.

211
00:10:07,140 --> 00:10:10,237
They're on the border, they're
in Israel, and they're in Ukraine.

212
00:10:10,320 --> 00:10:13,957
And what's super surprising
is I get yelled at all day

213
00:10:14,040 --> 00:10:17,377
about of those three issues. And by
the way, the issue I don't get yelled at

214
00:10:17,460 --> 00:10:19,027
about the elite is about Ukraine.

215
00:10:19,110 --> 00:10:21,247
The issue I get, obviously,
the Israel thing is like,

216
00:10:21,330 --> 00:10:23,437
you know, only thing anyone
ever wants to talk about.

217
00:10:23,520 --> 00:10:25,687
You would not, it's like
somehow people have forgotten

218
00:10:25,770 --> 00:10:28,087
it's a country with a GDP
smaller than Switzerland,

219
00:10:28,170 --> 00:10:31,837
but you know, and then my progressive
friends, it's all ICE all day.

220
00:10:31,920 --> 00:10:35,017
- Some people might actually
not even know what Palantir does.

221
00:10:35,100 --> 00:10:39,727
In the shortest possible way, could
you explain to people who just know it

222
00:10:39,810 --> 00:10:42,577
as sort of something they
see on a page and don't like

223
00:10:42,660 --> 00:10:44,527
and saying, "Hey,
this is what it really is?"

224
00:10:44,610 --> 00:10:48,637
- Or love. We have a lot of people who
love us and some people don't, but yeah,

225
00:10:48,720 --> 00:10:51,037
we tend to elicit strong
reactions on both sides.

226
00:10:51,120 --> 00:10:55,087
So, that's definitely fair. If you're
internal or external intelligence,

227
00:10:55,170 --> 00:10:57,337
you're using us to find terrorists

228
00:10:57,420 --> 00:11:00,967
and find organized criminals
while maintaining security

229
00:11:01,050 --> 00:11:03,637
and maintaining the data
protection of your country.

230
00:11:03,720 --> 00:11:06,157
So, like the highest data
protection environments

231
00:11:06,240 --> 00:11:08,767
in the world, Western Europe
would love to get rid of us.

232
00:11:08,850 --> 00:11:10,417
They try every day to get rid of us.

233
00:11:10,500 --> 00:11:14,317
But building a data protection
environment that actually allows you

234
00:11:14,400 --> 00:11:16,567
to do anti-organized
criminality environment

235
00:11:16,650 --> 00:11:20,047
with that also does data protection
at any level of granularity,

236
00:11:20,130 --> 00:11:23,287
at any level of complexity
is very, very hard.

237
00:11:23,370 --> 00:11:26,797
Ask the German government. They would
love to find another alternative.

238
00:11:26,880 --> 00:11:30,137
And honestly, I'd have no problem
with them finding it, but it's very hard.

239
00:11:30,220 --> 00:11:31,807
So, that's one thing we do.
Then you have the battlefield,

240
00:11:31,890 --> 00:11:34,447
the Project Maven, and also
organizing the battlefield,

241
00:11:34,530 --> 00:11:35,857
what the Special Forces use it for.

242
00:11:35,940 --> 00:11:37,237
How do you know where your troops are?

243
00:11:37,320 --> 00:11:38,947
How do you bring them
home with all the data

244
00:11:39,030 --> 00:11:40,807
that you'd have from any
source environment?

245
00:11:40,890 --> 00:11:42,907
How do you manage assets while doing that?

246
00:11:42,990 --> 00:11:46,717
Meaning assets from the other side,
assets meaning spies, basically.

247
00:11:46,800 --> 00:11:50,377
How do you get the human,
how do you prepare your egress?

248
00:11:50,460 --> 00:11:54,847
How do you get in and out of the
battlefield safely, as safely as possible?

249
00:11:54,930 --> 00:11:57,337
So, avoiding mines, avoiding enemies,

250
00:11:57,420 --> 00:11:59,227
those things change
rapidly on the battlefield.

251
00:11:59,310 --> 00:12:00,757
That's a software problem that's powered

252
00:12:00,840 --> 00:12:04,717
in great part by us across
America and with our allies.

253
00:12:04,800 --> 00:12:08,797
How do you supply targeting
capabilities to allies,

254
00:12:08,880 --> 00:12:11,467
Ukraine, Israel,
so that they get the product,

255
00:12:11,550 --> 00:12:13,897
but obviously, we don't have
access to the full data they use,

256
00:12:13,980 --> 00:12:16,207
'cause we're not part of those countries.

257
00:12:16,290 --> 00:12:21,217
That's a Palantir thing. On the commercial
side, it's, you know, how do you to make,

258
00:12:21,300 --> 00:12:24,127
I mean the current version
would be if you want AI to work,

259
00:12:24,210 --> 00:12:25,897
you're gonna have
to have precursor things.

260
00:12:25,980 --> 00:12:27,757
You can think of it as like
the precursor to make anything.

261
00:12:27,840 --> 00:12:29,227
What are those things gonna be?

262
00:12:29,310 --> 00:12:30,847
You're gonna need high fidelity data.

263
00:12:30,930 --> 00:12:34,087
To have high fidelity data, you're gonna
need to be able to integrate your data

264
00:12:34,170 --> 00:12:36,787
with a model you understand. You're gonna
have to build something called pipelining,

265
00:12:36,870 --> 00:12:40,717
meaning you can serialize and deserialize
a technical way of taking apart

266
00:12:40,800 --> 00:12:42,457
and putting together
the assumptions within the data.

267
00:12:42,540 --> 00:12:44,797
You're gonna have to be able
to maintain your security model

268
00:12:44,880 --> 00:12:46,957
and maintain the logic
and actions of your business.

269
00:12:47,040 --> 00:12:49,657
And you're gonna have to be able
to extract that, so ontology.

270
00:12:49,740 --> 00:12:54,697
So shorthand, if you're doing anything
that involves operational intelligence,

271
00:12:54,780 --> 00:13:00,217
whether it's analytics or AI, you're
gonna have to find something like ontology

272
00:13:00,300 --> 00:13:03,217
or foundry and you know,
and FTEs to install it.

273
00:13:03,300 --> 00:13:05,527
That's Palantir.
- It's interesting, 'cause you,

274
00:13:05,610 --> 00:13:10,207
going back a little, you implied that,
you know, the rest of Silicon Valley

275
00:13:10,290 --> 00:13:12,367
hasn't figured out how to do that.

276
00:13:12,450 --> 00:13:15,367
That you're unique in exploiting that.

277
00:13:15,450 --> 00:13:20,317
- Well it's, yeah. What I'm really
saying is we know how to do it.

278
00:13:20,400 --> 00:13:22,447
If you find someone else who can do it

279
00:13:22,530 --> 00:13:24,397
and you don't wanna work
with us, buy it from them.

280
00:13:24,480 --> 00:13:26,647
- I mean, is there anyone
you consider a competitor?

281
00:13:26,730 --> 00:13:28,747
- Our competition is actually political.

282
00:13:28,830 --> 00:13:30,247
The woke left and the woke right

283
00:13:30,330 --> 00:13:32,827
wake up every day figuring
out how they can hurt Palantir

284
00:13:32,910 --> 00:13:35,347
and if they get into power,
they'll hurt Palantir.

285
00:13:35,430 --> 00:13:36,757
How could you build a society

286
00:13:36,840 --> 00:13:40,007
with unfettered migration
in Europe and not have Palantir?

287
00:13:40,090 --> 00:13:42,307
I don't know how you're gonna do that,
but they think they can do that.

288
00:13:42,390 --> 00:13:45,727
If they win, if the communists
in France win, they will pull us out.

289
00:13:45,810 --> 00:13:49,597
If like the Mamdani win
of the Democratic party takes over,

290
00:13:49,680 --> 00:13:52,867
I viewed that as my party.
But if that's the party I'm not in it.

291
00:13:52,950 --> 00:13:57,007
Or if the right woke wing, which
is like everything is a conspiracy.

292
00:13:57,090 --> 00:14:00,427
Any use of technology
is actually gonna only be used

293
00:14:00,510 --> 00:14:05,437
to eviscerate and attack us is like,
our data protection part of Palantir

294
00:14:05,520 --> 00:14:06,787
is built into the software.

295
00:14:06,870 --> 00:14:09,607
It is literally the hardest
software to abuse in the world,

296
00:14:09,690 --> 00:14:10,867
but they don't seem to want it.

297
00:14:10,950 --> 00:14:12,907
But if you like, don't want meritocracy,

298
00:14:12,990 --> 00:14:15,967
either left or right, you don't like
the consequence of it, you hate Palantir.

299
00:14:16,050 --> 00:14:18,337
And that is actually our competition.

300
00:14:18,420 --> 00:14:21,517
And by the way, not that they believe it.

301
00:14:21,600 --> 00:14:23,527
The thing that is protecting those people,

302
00:14:23,610 --> 00:14:26,077
the people that hate us
the most is actually our product.

303
00:14:26,160 --> 00:14:28,477
Because it is the hardest
product to misuse

304
00:14:28,560 --> 00:14:31,927
to take away the rights of human
beings that's ever been built.

305
00:14:32,010 --> 00:14:34,267
And don't believe me, just go look at it,

306
00:14:34,350 --> 00:14:37,537
spend 10 minutes looking at it,
find a technical person

307
00:14:37,620 --> 00:14:39,637
and look at our architecture
or look at it yourself.

308
00:14:39,720 --> 00:14:41,317
And that's actually our competition.

309
00:14:41,400 --> 00:14:43,627
And you know, in a weird way,
we embrace the competition

310
00:14:43,710 --> 00:14:45,607
because they say these
crazy things about us

311
00:14:45,690 --> 00:14:47,827
and then the smart people go look at it.

312
00:14:47,910 --> 00:14:49,957
- My head's spinning a little,
how did Mamdani get in there?

313
00:14:50,040 --> 00:14:52,717
- I would say some of is downstream from,

314
00:14:52,800 --> 00:14:56,707
if you explain to the world
that labor's gonna be valueless,

315
00:14:56,790 --> 00:14:59,557
people are gonna elect
the most ridiculous people ever.

316
00:14:59,640 --> 00:15:02,827
We have to build systems like Palantir
and we have to explain to people,

317
00:15:02,910 --> 00:15:05,917
but also let them see the results,
the fruit of our labor

318
00:15:06,000 --> 00:15:09,487
that their labor, meaning
people with working class skills,

319
00:15:09,570 --> 00:15:13,837
people with normal jobs, that their
labor's gonna become more important

320
00:15:13,920 --> 00:15:16,747
and more valuable over time
with products like ours.

321
00:15:16,830 --> 00:15:19,507
Which happens to be true.
If you run around saying that, you know,

322
00:15:19,590 --> 00:15:21,577
oh, the economy is gonna completely shift

323
00:15:21,660 --> 00:15:25,567
to value creation only
owned by 10,000 people,

324
00:15:25,650 --> 00:15:27,787
people on the left
and the right are gonna go nuts

325
00:15:27,870 --> 00:15:30,427
and they're gonna vote
for people who are like basically,

326
00:15:30,510 --> 00:15:32,917
yeah, nothing I've said
has ever worked in history.

327
00:15:33,000 --> 00:15:34,357
It has never worked in history.

328
00:15:34,440 --> 00:15:39,247
You cannot find a time in history where
any of that stuff has ever worked, ever.

329
00:15:39,330 --> 00:15:42,577
They still will gravitate toward it
because they're afraid of what could work,

330
00:15:42,660 --> 00:15:46,507
which is like an AI-driven AGI
environment where no one has a job

331
00:15:46,590 --> 00:15:49,327
and the only people who make money
are like people sitting out here.

332
00:15:49,410 --> 00:15:51,967
And so, that's part of the problem.

333
00:15:52,050 --> 00:15:56,347
Now also, honestly, I mean, you know, I
think universities and elite institutions

334
00:15:56,430 --> 00:15:59,137
have played a really corrosive role here.

335
00:15:59,220 --> 00:16:02,407
People are teaching pagan
religion views, again,

336
00:16:02,490 --> 00:16:06,877
pagan religion as it's a new
of religion with sacrifices.

337
00:16:06,960 --> 00:16:09,187
Who's the sacrifice?
Me, I'm the sacrifice.

338
00:16:09,270 --> 00:16:13,507
And that, again, it's also like, it's
a form of thought that has never worked.

339
00:16:13,590 --> 00:16:15,097
- You don't seem constrained to me.

340
00:16:15,180 --> 00:16:17,917
You're you're doing great.
- I mean, I'm not complaining.

341
00:16:18,000 --> 00:16:21,937
- Yeah. - Palantir,
we're not victims of Palantir.

342
00:16:22,020 --> 00:16:23,607
- Yeah. - We're not playing
the victimhood thing.

343
00:16:23,690 --> 00:16:25,327
We don't, I don't think I'm a victim.

344
00:16:25,410 --> 00:16:28,777
I don't actually believe we should,
these things should be framed as victims.

345
00:16:28,860 --> 00:16:32,017
And yes, we're doing very well,
and you know, we might even do better.

346
00:16:32,100 --> 00:16:35,257
- You'd have this bonkers
earnings call in February.

347
00:16:35,340 --> 00:16:39,847
Huge, huge growth and- - Well,
it's like people who are not financial,

348
00:16:39,930 --> 00:16:42,547
which is most people, and honestly,
I wouldn't be paying attention to this

349
00:16:42,630 --> 00:16:43,537
if I wasn't running Palantir.

350
00:16:43,620 --> 00:16:47,647
Those numbers are a little bit
like flying saucer took off numbers.

351
00:16:47,730 --> 00:16:49,507
So, and you know, people are financial,

352
00:16:49,590 --> 00:16:51,877
many of whom doubted us
are like, wait a minute,

353
00:16:51,960 --> 00:16:54,667
the plate dropped,
because what it basically, it's

354
00:16:54,750 --> 00:16:56,797
what non-financial people
might not appreciate

355
00:16:56,880 --> 00:17:00,187
is we're not just growing,
we're growing with high margins.

356
00:17:00,270 --> 00:17:02,907
with, and that's
in at our scale basically.

357
00:17:02,990 --> 00:17:06,487
In my view, never been done before,
not in a similarly situated business.

358
00:17:06,570 --> 00:17:10,927
- You said it was an anomalous like,
you know, some sort of historic moment,

359
00:17:11,010 --> 00:17:12,637
that the company is anomalous.

360
00:17:12,720 --> 00:17:14,707
- You know, in a weird
way we undervalue these

361
00:17:14,790 --> 00:17:17,647
or I undervalue the whole
numeric side of our business,

362
00:17:17,730 --> 00:17:19,747
because it's downstream of value creation.

363
00:17:19,830 --> 00:17:23,767
So, if you have a, no,
if you are doing anything

364
00:17:23,850 --> 00:17:26,497
like what we're doing,
so call it 68% growth in the US,

365
00:17:26,580 --> 00:17:28,987
and again, we're doing
it, you know, our way.

366
00:17:29,070 --> 00:17:33,787
Salesforce is anemic. We're not running
around really trying to sell our wares.

367
00:17:33,870 --> 00:17:36,937
We have a conference here, people
fight to get in, which is pretty cool.

368
00:17:37,020 --> 00:17:39,667
We have the best clients in the world,
especially in America.

369
00:17:39,750 --> 00:17:41,977
Clearly we've tapped into something.

370
00:17:42,060 --> 00:17:43,537
I'll tell you what we've tapped into.

371
00:17:43,620 --> 00:17:45,757
There's essentially a simple thing.

372
00:17:45,840 --> 00:17:50,467
LLMs and software
empowered LLMs create power

373
00:17:50,550 --> 00:17:54,457
per the expectation of preyed
optimality between partner and vendor.

374
00:17:54,540 --> 00:17:58,627
You are going to get paid in the future
a percentage of the value you create.

375
00:17:58,710 --> 00:18:01,277
And the value you create
is going to be, you know,

376
00:18:01,360 --> 00:18:03,277
at some point objectively measured

377
00:18:03,360 --> 00:18:06,307
and it's going to flow into
your financials and vice versa.

378
00:18:06,390 --> 00:18:11,377
- So, you have this unique culture
and you even have called it a cult.

379
00:18:11,460 --> 00:18:12,427
- It's been called a cult.

380
00:18:12,510 --> 00:18:17,437
- I wonder how much you
cultivate that outsider mentality.

381
00:18:17,520 --> 00:18:19,297
It reminds me we're
both from Philadelphia.

382
00:18:19,380 --> 00:18:24,157
Jason Kelce, after the first Super Bowl
saying, "Nobody likes us, we don't care."

383
00:18:24,240 --> 00:18:25,537
- The thing people don't understand

384
00:18:25,620 --> 00:18:29,407
is there's a massive feature side
to being an outsider or even popular.

385
00:18:29,490 --> 00:18:33,547
People forget, people always think
of the bug size of being unpopular.

386
00:18:33,630 --> 00:18:36,697
Like, you know, you're unpopular, it's
not pleasant, you're not gonna like it.

387
00:18:36,780 --> 00:18:40,297
I don't like it actually.
But there are massive feature sides.

388
00:18:40,380 --> 00:18:42,007
Like, you get the best
people in the world.

389
00:18:42,090 --> 00:18:45,037
Like someone says something
ridiculously stupid about Palantir.

390
00:18:45,120 --> 00:18:49,657
Five people look at it and they, the fifth
person who's exactly the kind of person

391
00:18:49,740 --> 00:18:53,527
you want inside your business or will
be the person buying your product someday

392
00:18:53,610 --> 00:18:56,767
or investing in you says, "Hey,
it can't be as simple as that.

393
00:18:56,850 --> 00:19:01,207
It's a conspiracy." It can't be as simple
as they want to eviscerate my rights,

394
00:19:01,290 --> 00:19:03,067
because those rights
would be the same rights

395
00:19:03,150 --> 00:19:06,517
that like an agent would need
protected by the agency or something.

396
00:19:06,600 --> 00:19:08,707
Then those people
investigate and investigate

397
00:19:08,790 --> 00:19:12,367
and investigate and they're like, "Wow,
this is a real interesting company."

398
00:19:12,450 --> 00:19:15,427
And they're thinking of the 10th
derivative of this problem,

399
00:19:15,510 --> 00:19:18,307
not the simplistic bullshit
that somebody online saying

400
00:19:18,390 --> 00:19:19,807
has no idea how our products work.

401
00:19:19,890 --> 00:19:24,757
And then you get the best people like
a really unexpected consequence of this.

402
00:19:24,840 --> 00:19:28,177
There's no country in the world
where our brand is as bad as France.

403
00:19:28,260 --> 00:19:32,887
We have the best French employees
in the world, especially in US commercial.

404
00:19:32,970 --> 00:19:36,337
And why did they join?
Because every time the communist party

405
00:19:36,420 --> 00:19:39,847
says we are doing these crazy
things and they clearly know nothing

406
00:19:39,930 --> 00:19:42,317
about what they're talking about,
the smartest person in France says,

407
00:19:42,400 --> 00:19:44,767
"Wait a minute, if it was that simple,
any idiot would've done it."

408
00:19:44,850 --> 00:19:46,747
By the time they're done
doing the diagnostic,

409
00:19:46,830 --> 00:19:49,137
they're like, "Hey, there's only one
company in the world I wanna work at."

410
00:19:49,220 --> 00:19:51,547
And that's how we have the best
French employees in the world.

411
00:19:51,630 --> 00:19:55,207
The best French employees arguably
in the world are at Palantir in America.

412
00:19:55,290 --> 00:19:56,827
And that's very counterintuitive.

413
00:19:56,910 --> 00:20:00,127
Working at a company that's
apparently a CIA front,

414
00:20:00,210 --> 00:20:02,557
which is like, yeah,
obviously complete bullshit.

415
00:20:02,640 --> 00:20:03,967
- Well, yeah, no, you may not be a front.

416
00:20:04,050 --> 00:20:06,157
Thank you so much, Alex.
- Thank you very much.

417
00:20:06,240 --> 00:20:09,970
Bye. [intense music]