1
00:00:00,160 --> 00:00:03,560
Speaker 1: Welcome to thrilling threads. So today we are pulling on

2
00:00:03,640 --> 00:00:05,519
a thread that I mean, it feels like it was

3
00:00:05,599 --> 00:00:07,719
ripped straight out of a cyberpunk novel or.

4
00:00:07,679 --> 00:00:09,679
Speaker 2: A high stake spy thriller exactly.

5
00:00:10,000 --> 00:00:11,679
Speaker 1: And I have to say this right at the top,

6
00:00:12,199 --> 00:00:15,320
every single bit of this is real. We're talking about

7
00:00:15,320 --> 00:00:19,039
a company that was born, specifically from the ashes of

8
00:00:19,160 --> 00:00:23,440
nine to eleven, a company funded by the CIA, designed

9
00:00:23,480 --> 00:00:29,920
with one singular, terrifying and well frankly onspiring purpose to

10
00:00:29,920 --> 00:00:31,239
see everything, to see everything.

11
00:00:31,280 --> 00:00:34,759
Speaker 2: And when we say everything, we really aren't exaggerating for effect. Here.

12
00:00:34,799 --> 00:00:38,560
We're talking about an artificial intelligence ecosystem that tracks wars

13
00:00:38,600 --> 00:00:43,479
in real time, It monitors national borders, manages global pandemics,

14
00:00:43,520 --> 00:00:46,520
and it predicts corporate profits before they even happen. It's

15
00:00:46,560 --> 00:00:49,799
a tool that governments rely on for national security, right,

16
00:00:49,840 --> 00:00:53,079
and that massive corporations, you know, the big ones, depend

17
00:00:53,159 --> 00:00:55,159
on just to survive the fiscal year.

18
00:00:55,359 --> 00:00:57,520
Speaker 1: The crazy part is that most people just walking down

19
00:00:57,520 --> 00:01:00,000
the street have absolutely no idea this entity even exist,

20
00:01:00,359 --> 00:01:02,679
none at all. But if you've been on a plane,

21
00:01:02,880 --> 00:01:06,519
or opened a bank account, or received a vaccine, I mean,

22
00:01:06,599 --> 00:01:08,599
even if you've just lived through the last few years

23
00:01:08,640 --> 00:01:12,959
of geopolitical chaos. Your life has been touched by their algorithms.

24
00:01:14,280 --> 00:01:18,079
We are, of course talking about pal and Teer Technologies.

25
00:01:17,760 --> 00:01:24,760
Speaker 2: Arguably the most controversial, secretive and significant AI company operating today.

26
00:01:24,799 --> 00:01:27,480
And to be honest, just calling it a tech story

27
00:01:28,040 --> 00:01:28,959
kind of sells it short.

28
00:01:29,040 --> 00:01:29,799
Speaker 1: It really does.

29
00:01:30,000 --> 00:01:33,239
Speaker 2: This is a story about the tradeoff between security and privacy.

30
00:01:33,400 --> 00:01:36,400
It's a case study on the ethics of what happens

31
00:01:36,519 --> 00:01:39,560
when you hand over the god view of the world

32
00:01:39,599 --> 00:01:40,719
to a private entity.

33
00:01:40,799 --> 00:01:42,840
Speaker 1: So I hope you have your coffee ready or maybe

34
00:01:42,840 --> 00:01:44,640
a piece of tape over your webcam. I don't know,

35
00:01:44,879 --> 00:01:46,799
because this is going to be a wild ride. It is.

36
00:01:46,879 --> 00:01:49,280
We've got a stack of sources here, everything from Peter

37
00:01:49,359 --> 00:01:54,000
Thiel's biography, financial disclosures, investigative reports from New York magazine,

38
00:01:54,200 --> 00:01:57,680
military case studies, you name it. A lot to unpack

39
00:01:57,760 --> 00:01:59,959
a ton. So the mission for this episode of Thrill

40
00:02:00,159 --> 00:02:02,319
Threads is to figure out exactly what this company is,

41
00:02:02,640 --> 00:02:04,879
where it came from, and why it matters to you.

42
00:02:05,079 --> 00:02:08,479
Speaker 2: It's the billion dollar question, isn't it? And to answer it,

43
00:02:08,639 --> 00:02:10,479
we can't start with the tech. We have to start

44
00:02:10,520 --> 00:02:12,360
with the trauma. Right, we have to go back to

45
00:02:12,479 --> 00:02:14,039
September eleventh, two thousand and.

46
00:02:14,039 --> 00:02:17,479
Speaker 1: One, right, because the narrative we usually hear about nine

47
00:02:17,520 --> 00:02:20,719
to eleven is all about the ideology or the geopolitics.

48
00:02:21,479 --> 00:02:26,039
But inside the intelligence community, the post mortem was it

49
00:02:26,120 --> 00:02:26,919
was very.

50
00:02:26,719 --> 00:02:28,560
Speaker 2: Different, completely different.

51
00:02:28,680 --> 00:02:31,080
Speaker 1: It wasn't just seen as an attack, it was seen

52
00:02:31,560 --> 00:02:33,639
as a catastrophic data failure.

53
00:02:33,800 --> 00:02:36,879
Speaker 2: Precisely. The nine to eleven Commission report is I mean,

54
00:02:36,879 --> 00:02:39,800
it's essentially a six hundred page document just screaming about

55
00:02:39,840 --> 00:02:43,360
information silos. The United States actually had the data to

56
00:02:43,400 --> 00:02:44,360
stock the attacks.

57
00:02:44,439 --> 00:02:46,240
Speaker 1: That's the part that's so hard to believe.

58
00:02:46,360 --> 00:02:49,319
Speaker 2: It is the CIA, the FBI, the NSA, they all

59
00:02:49,319 --> 00:02:52,280
held different pieces of the puzzle. But the systems were

60
00:02:52,319 --> 00:02:55,159
so archaic. We're talking about an era where agents were

61
00:02:55,159 --> 00:02:58,680
and this is not a joke, physically walking floppy disks

62
00:02:58,719 --> 00:03:02,000
between offices because the secure networks couldn't talk to each other.

63
00:03:02,120 --> 00:03:04,639
Speaker 1: That is just it's mind blowing to think about. Now,

64
00:03:04,719 --> 00:03:07,080
you just assume that government sees everything, but back then

65
00:03:07,120 --> 00:03:08,439
they were completely separate.

66
00:03:08,199 --> 00:03:08,840
Speaker 2: Totally separate.

67
00:03:09,080 --> 00:03:11,280
Speaker 1: Like one agency knew a guy was taking flight lessons

68
00:03:11,280 --> 00:03:15,439
in Florida, another agency knew that exact same guy was

69
00:03:15,479 --> 00:03:18,719
on a watch list in Malaysia, right, But those two

70
00:03:18,759 --> 00:03:21,960
little sticky notes never made it onto the same whiteboard exactly.

71
00:03:22,000 --> 00:03:25,919
Speaker 2: It was a format problem, a database problem. The FBI

72
00:03:26,000 --> 00:03:29,960
files were literally in paper boxes or these antiquated mainframes

73
00:03:30,199 --> 00:03:33,599
that couldn't talk to the CIA's databases. There was no

74
00:03:33,800 --> 00:03:36,960
single pane of glass to use the industry.

75
00:03:36,639 --> 00:03:38,960
Speaker 1: Terms single pane of glass, yeah.

76
00:03:38,680 --> 00:03:40,520
Speaker 2: Where an analyst could just look at the world and

77
00:03:40,560 --> 00:03:43,199
see all the connections at once. The question that just

78
00:03:43,280 --> 00:03:46,159
burned in every intelligence report for the next decade was

79
00:03:46,639 --> 00:03:48,560
how the hell did we not connect the dots?

80
00:03:48,759 --> 00:03:52,000
Speaker 1: And this is where Silicon Valley steps in, and specifically

81
00:03:52,240 --> 00:03:54,639
a group of people who had been solving a very

82
00:03:54,960 --> 00:03:57,919
very different kind of problem. The solution, I mean, it

83
00:03:57,960 --> 00:04:01,560
sounds almost too simple in retrospect. The idea was, what

84
00:04:01,639 --> 00:04:04,120
if we built software that could just sift through these

85
00:04:04,280 --> 00:04:09,319
oceans of data, emails, surveillance footage, passport records, bank transfers,

86
00:04:09,560 --> 00:04:12,319
and find the next Osama bin Laden before he even

87
00:04:12,319 --> 00:04:13,240
booked a plane ticket.

88
00:04:13,520 --> 00:04:17,879
Speaker 2: And the person driving this idea was Peter Teal.

89
00:04:17,720 --> 00:04:20,399
Speaker 1: The PayPal guy. Which always feels like such a weird pivot.

90
00:04:20,480 --> 00:04:22,439
It does you know, from I want to make it

91
00:04:22,480 --> 00:04:25,160
easier to email money to my friend to I want

92
00:04:25,160 --> 00:04:28,959
to hunt global terrorists. It feels like two completely different universes.

93
00:04:29,040 --> 00:04:30,839
Speaker 2: It really seems like a leap, but if you look

94
00:04:30,839 --> 00:04:34,360
at the underlying math, it's it's actually the same problem.

95
00:04:34,800 --> 00:04:37,920
Teal had just made millions of dollars by stopping credit

96
00:04:37,959 --> 00:04:41,319
card fraud at PayPal. Okay, now think about what fraud

97
00:04:41,360 --> 00:04:44,720
detection really is. You are looking for a tiny, tiny

98
00:04:44,800 --> 00:04:48,720
signal in a massive amount of noise. You have millions

99
00:04:48,720 --> 00:04:52,959
of legitimate transactions, people buying coffee, paying their rent, and

100
00:04:53,000 --> 00:04:55,720
you need to find the one transaction that is, you know,

101
00:04:55,959 --> 00:04:58,560
a Russian mobster testing a stolen card number.

102
00:04:58,759 --> 00:05:01,959
Speaker 1: Okay, I see the parallel that at PayPal they use

103
00:05:02,000 --> 00:05:05,199
pattern matching. They looks for networks like this IP address

104
00:05:05,279 --> 00:05:07,439
is linked to this bank account, which is linked to

105
00:05:07,480 --> 00:05:09,319
this other suspicious purchase.

106
00:05:09,399 --> 00:05:12,519
Speaker 2: That's it. It's graph theory, nodes and edges. At PayPal,

107
00:05:12,800 --> 00:05:16,439
the nodes were bank accounts and computers. In counter terrorism,

108
00:05:16,600 --> 00:05:19,920
the nodes are people and phone numbers and flight manifests.

109
00:05:20,160 --> 00:05:23,600
So Thial's logic was His logic was, if our anti

110
00:05:23,680 --> 00:05:26,920
fraud tools can detect a stolen credit card. By mapping

111
00:05:26,959 --> 00:05:30,759
these networks, why can't we build tools to detect a

112
00:05:30,879 --> 00:05:31,800
terrorist cell.

113
00:05:32,000 --> 00:05:36,279
Speaker 1: Why can't we apply that same algorithmic logic to national security.

114
00:05:36,360 --> 00:05:37,920
Speaker 2: That was the genesis of the whole thing.

115
00:05:37,959 --> 00:05:40,439
Speaker 1: Because he sets out to build this thing, he recruits

116
00:05:40,480 --> 00:05:43,480
some of the smartest engineers from Stanford and PayPal, including

117
00:05:43,480 --> 00:05:45,800
Alex KRP and we definitely get to him in a minute,

118
00:05:45,879 --> 00:05:49,680
and they start Pollantier And for all the fantasy nerds listening,

119
00:05:49,759 --> 00:05:52,639
and I say that with love as one of them. Yes,

120
00:05:53,079 --> 00:05:54,959
it is named after the Palantary from the.

121
00:05:54,879 --> 00:05:57,399
Speaker 2: Lord of the Rings, which is such a telling choice

122
00:05:57,439 --> 00:05:59,720
of name, isn't it. And honestly a bit of a

123
00:05:59,720 --> 00:06:00,920
red flag, if you know your.

124
00:06:00,839 --> 00:06:02,439
Speaker 1: Tolkien, a huge red flag.

125
00:06:02,519 --> 00:06:05,399
Speaker 2: And the lore the Palenteer are the all seeing orbs.

126
00:06:05,639 --> 00:06:08,879
They're these powerful magical stones that let you see across

127
00:06:09,000 --> 00:06:13,120
vast distances and uncover secrets. But and this is the

128
00:06:13,120 --> 00:06:17,040
crucial part, they are dangerous, very dangerous. They're tools used

129
00:06:17,040 --> 00:06:20,319
by wizards and villains alike. And when you look into

130
00:06:20,319 --> 00:06:23,879
the Pallenteer, the dark Lord Souron can look back at you.

131
00:06:24,439 --> 00:06:25,959
It corrupts the user.

132
00:06:26,319 --> 00:06:30,040
Speaker 1: It's an incredibly bold move to name your company after

133
00:06:30,120 --> 00:06:33,079
a cursed object used by the bad guys in a

134
00:06:33,120 --> 00:06:38,639
fantasy epic. It just signals this obsession with visibility, secrecy,

135
00:06:38,680 --> 00:06:40,519
and power right from day one.

136
00:06:40,639 --> 00:06:42,519
Speaker 2: It's like naming your security company the.

137
00:06:42,519 --> 00:06:45,240
Speaker 1: Death Star exactly. It's like, we know what this is.

138
00:06:45,560 --> 00:06:48,800
But here's where the origin story gets really spy novel esque.

139
00:06:49,600 --> 00:06:52,600
Pallenteer didn't start like a normal startup. They didn't go

140
00:06:52,600 --> 00:06:55,720
to the usual Silicon Valley venture capitalists.

141
00:06:55,040 --> 00:06:57,000
Speaker 2: No Sequoia, no and recent Horowitz.

142
00:06:57,040 --> 00:06:59,519
Speaker 1: They didn't go begging for seed money on sand Hill Road.

143
00:06:59,680 --> 00:07:03,000
Speaker 2: No. Their first major outside investor, the entity that kept

144
00:07:03,000 --> 00:07:05,439
the lights on when nobody else really understood the product,

145
00:07:05,519 --> 00:07:06,879
was a firm called in q tel.

146
00:07:07,279 --> 00:07:08,759
Speaker 1: And for those of you who don't spend your free

147
00:07:08,800 --> 00:07:11,680
time reading government budget reports, what is in qtel.

148
00:07:11,920 --> 00:07:15,120
Speaker 2: In qtel is the CIA's very own venture capital firm.

149
00:07:15,199 --> 00:07:17,079
Speaker 1: I still can't get over that. Yeah, the CIA has

150
00:07:17,079 --> 00:07:18,959
a venture capital firm. It sounds like a joke from

151
00:07:18,959 --> 00:07:19,560
a movie.

152
00:07:19,680 --> 00:07:21,959
Speaker 2: It's very real. It was founded in the late nineties

153
00:07:22,000 --> 00:07:25,800
because the CIA realized they were falling way behind the

154
00:07:25,839 --> 00:07:29,879
government moves so slow, tech moves so fast. If the

155
00:07:29,920 --> 00:07:33,240
CIA wanted the best spyware or the best data analysis tools.

156
00:07:33,519 --> 00:07:35,959
They couldn't wait for a government contractor to build it

157
00:07:36,000 --> 00:07:38,199
over ten years. They needed to buy it from startups.

158
00:07:38,759 --> 00:07:42,399
So in Qtel exists to fund startups that build things

159
00:07:42,480 --> 00:07:43,519
useful to spies.

160
00:07:44,240 --> 00:07:48,160
Speaker 1: So this funding is just it's so critical to the story.

161
00:07:48,279 --> 00:07:52,079
It signaled immediately that Pallenteer wasn't just another tech startup

162
00:07:52,120 --> 00:07:54,879
trying to sell ads or you know, cloud storage. Right

163
00:07:55,279 --> 00:07:57,800
Their client from the very beginning.

164
00:07:57,519 --> 00:07:59,839
Speaker 2: Was the state, and that shaped the entire product. The

165
00:08:00,079 --> 00:08:04,160
goal was to build this godview, to blend all this messy,

166
00:08:04,279 --> 00:08:09,639
siloed government data, passport logs, drone footage, field reports into

167
00:08:09,680 --> 00:08:10,720
one single interface.

168
00:08:10,879 --> 00:08:12,480
Speaker 1: But from what I've read in the early histories of

169
00:08:12,519 --> 00:08:15,399
the company, the first few years weren't exactly smooth sales

170
00:08:15,439 --> 00:08:16,040
so far from it.

171
00:08:16,399 --> 00:08:19,120
Speaker 2: The reports say the early software was buggy and way

172
00:08:19,199 --> 00:08:20,000
over engineered.

173
00:08:20,079 --> 00:08:22,759
Speaker 1: I saw a great quote from a government official who

174
00:08:22,759 --> 00:08:26,879
said using early Palenteer was like trying to fly a

175
00:08:26,959 --> 00:08:29,160
space shuttle just to find your car keys.

176
00:08:29,480 --> 00:08:33,679
Speaker 2: That's perfect. Bureaucrats hated it. The NSA apparently thought it

177
00:08:33,720 --> 00:08:36,759
was cute, which is about the most devastating insult you

178
00:08:36,799 --> 00:08:39,519
can get in the Intel world ouch and the FBI

179
00:08:39,600 --> 00:08:44,720
apparently found it terrifying because it made accessing data too easy.

180
00:08:44,840 --> 00:08:46,200
It broke down their silos.

181
00:08:46,200 --> 00:08:49,080
Speaker 1: But they just kept building. They were obsessive. And the

182
00:08:49,080 --> 00:08:52,799
culture they built, yeah, wow, it wasn't the Google campus

183
00:08:52,919 --> 00:08:55,320
let's ride bikes and eat free sushi kind of vibe.

184
00:08:55,320 --> 00:08:56,879
Speaker 2: Not at all. It was more like a bunker.

185
00:08:57,080 --> 00:08:59,320
Speaker 1: A bunker, that's a great way to put it. Let's

186
00:08:59,320 --> 00:09:02,120
talk about that culture because it all centers around the CEO,

187
00:09:02,279 --> 00:09:05,080
Alex Karp. If you haven't seen a picture of Alex Krp,

188
00:09:05,080 --> 00:09:07,559
you need to google him right now. He does not

189
00:09:07,639 --> 00:09:09,600
look like a tech ceo, not even close.

190
00:09:09,679 --> 00:09:12,480
Speaker 2: He has a PhD in philosophy from Frankfurt University. He

191
00:09:12,519 --> 00:09:16,960
studied under Jurgen Habermas. He has this wild einstein esque hare.

192
00:09:17,039 --> 00:09:20,200
He wears bright orange tracksuits or puffy vests and these

193
00:09:20,320 --> 00:09:21,519
wireframe glasses.

194
00:09:21,840 --> 00:09:24,039
Speaker 1: He looks like he wandered out of a Nietzsche book club,

195
00:09:24,320 --> 00:09:25,120
not a boardrome.

196
00:09:25,240 --> 00:09:26,799
Speaker 2: That's the perfect description.

197
00:09:26,840 --> 00:09:29,399
Speaker 1: And he acts like it too. I read he's often

198
00:09:29,440 --> 00:09:33,279
referred to as the monk inside the company. He's a bachelor,

199
00:09:33,440 --> 00:09:36,919
obsessive about fitness. He practices martial arts in the office.

200
00:09:37,039 --> 00:09:39,240
Speaker 2: But more importantly, he would give these lectures to the

201
00:09:39,279 --> 00:09:42,759
engineers on the ethics of power. His famous line, the

202
00:09:42,759 --> 00:09:45,480
one he always repeats, is we're not in the business

203
00:09:45,480 --> 00:09:48,559
of surveillance. We're in the business of understanding.

204
00:09:49,039 --> 00:09:52,399
Speaker 1: And that distinction is so key to their internal mythology.

205
00:09:52,840 --> 00:09:55,120
They saw themselves as the anti Google.

206
00:09:55,240 --> 00:09:56,720
Speaker 2: Absolutely, if Google was.

207
00:09:56,679 --> 00:10:00,000
Speaker 1: This data mining theme park that sold your attention to advertisers,

208
00:10:00,519 --> 00:10:03,080
Pallenteer was a fortress designed to protect the West.

209
00:10:03,159 --> 00:10:06,399
Speaker 2: And they didn't hire salespeople for years, not a single one.

210
00:10:06,480 --> 00:10:08,360
They hired forward deployed engineers.

211
00:10:08,600 --> 00:10:12,360
Speaker 1: That term forward deployed is so interesting. It sounds military.

212
00:10:12,480 --> 00:10:15,720
Speaker 2: It is. It's totally military. Unlike other software companies that

213
00:10:15,759 --> 00:10:17,840
send a salesman in a suit to take you to lunch,

214
00:10:18,279 --> 00:10:20,840
Palenteer sends engineers to live with the client.

215
00:10:21,080 --> 00:10:23,240
Speaker 1: So the client is the US Army in Afghanistan.

216
00:10:23,440 --> 00:10:27,159
Speaker 2: The Palenteer engineer goes to Afghanistan. They sit in the

217
00:10:27,200 --> 00:10:30,120
tent with the soldiers and they code the software in

218
00:10:30,200 --> 00:10:33,519
real time to fix problems as they're happening on the battlefield.

219
00:10:33,720 --> 00:10:38,080
Speaker 1: Wow, that level of embedding creates a stickiness that you

220
00:10:38,360 --> 00:10:39,960
just don't get with Microsoft excelf.

221
00:10:40,000 --> 00:10:40,879
Speaker 2: You can't compete with that.

222
00:10:41,000 --> 00:10:42,799
Speaker 1: So let's look at what they were actually building. The

223
00:10:42,799 --> 00:10:45,519
product that made their name. It was called Gotham.

224
00:10:45,639 --> 00:10:49,279
Speaker 2: Gotham again with the superhero branding, right. Gotham is the

225
00:10:49,279 --> 00:10:51,679
platform designed for intelligence and defense.

226
00:10:52,200 --> 00:10:54,960
Speaker 1: So what makes Gotham different from just a really really

227
00:10:54,960 --> 00:10:55,919
good search engine.

228
00:10:56,039 --> 00:10:58,679
Speaker 2: It's about ontology. That's a word you hear a lot

229
00:10:58,720 --> 00:11:01,840
with Pellenteer. It' the philosophy. Okay, most databases are just

230
00:11:02,000 --> 00:11:05,399
rows and columns of text and numbers. Gotham maps the

231
00:11:05,399 --> 00:11:09,159
world as objects and relationships. So an object could be

232
00:11:09,200 --> 00:11:12,639
a person, a car, a building, or a bank account.

233
00:11:13,039 --> 00:11:16,639
Gotham switches together all these disparate data streams to show

234
00:11:16,639 --> 00:11:18,240
you how those objects interact.

235
00:11:18,559 --> 00:11:21,480
Speaker 1: So instead of having one spreadsheet of phone calls and

236
00:11:21,559 --> 00:11:25,840
a separate map of bomb explosions, Gotham just layers them

237
00:11:25,840 --> 00:11:27,480
on top of each other exactly.

238
00:11:27,559 --> 00:11:30,600
Speaker 2: Let's take a real world example and ied a roadside

239
00:11:30,639 --> 00:11:33,960
bomb goes off in Bagdad. That's an event. Okay, Gotham

240
00:11:34,000 --> 00:11:37,799
pulls in the report, but then it automatically overlays the

241
00:11:38,000 --> 00:11:40,879
cell phone tower logs from that exact time and location.

242
00:11:41,519 --> 00:11:45,720
It notices a specific burner phone pinged tower right before

243
00:11:45,720 --> 00:11:49,480
the blast. Then it checks that phone number against a

244
00:11:49,559 --> 00:11:53,480
database of captured pocket litters, you know, scraps of paper

245
00:11:53,759 --> 00:11:55,519
found in a raid three months ago.

246
00:11:55,840 --> 00:11:57,279
Speaker 1: And suddenly you have a name.

247
00:11:57,399 --> 00:11:59,200
Speaker 2: You have a name, you have a location, you have

248
00:11:59,240 --> 00:12:01,879
the guy who paid him, you have the entire kill chain.

249
00:12:01,879 --> 00:12:04,320
It wasn't just here's where a bomb went off. It

250
00:12:04,399 --> 00:12:06,559
was here is the network that made it happen.

251
00:12:06,639 --> 00:12:08,960
Speaker 1: And the soldiers on the ground loved it. I mean

252
00:12:09,080 --> 00:12:11,120
from all the accounts, they absolutely loved it. For the

253
00:12:11,159 --> 00:12:13,879
first time, they weren't drowning in data, they were surfing it.

254
00:12:14,080 --> 00:12:17,080
Speaker 2: But there was a major hurdle. Palenteer was an outsider

255
00:12:17,080 --> 00:12:20,919
in Washington, DC. Defense contracting is a club, it's the

256
00:12:20,960 --> 00:12:24,320
Iron Triangle. You need lobbyists, you need golf retreats, you

257
00:12:24,360 --> 00:12:27,240
need retired generals on your board of directors. Pollenteer had

258
00:12:27,399 --> 00:12:27,799
none of that.

259
00:12:28,000 --> 00:12:29,799
Speaker 1: They just had better software, and.

260
00:12:29,960 --> 00:12:32,919
Speaker 2: Usually that's not enough. In DC. The army actually tried

261
00:12:32,919 --> 00:12:35,840
to block Palenteer. They did, oh yeah, they wanted to

262
00:12:35,840 --> 00:12:40,279
build their own system called DCGSA. It cost billions of

263
00:12:40,320 --> 00:12:44,320
dollars and by all accounts, it barely worked. Soldiers were

264
00:12:44,399 --> 00:12:48,600
literally leaving the official Army computers in their boxes and

265
00:12:48,679 --> 00:12:51,960
buying Palenteer laptops with their units petty cash because they

266
00:12:52,039 --> 00:12:53,159
knew it would keep them alive.

267
00:12:53,480 --> 00:12:56,360
Speaker 1: So when the Army tried to give this massive, like

268
00:12:56,919 --> 00:13:01,120
two hundred million dollars intelligence contract to a legacy contractor

269
00:13:01,600 --> 00:13:05,679
without even holding a competition, Pallenteer did something completely unthinkable.

270
00:13:05,799 --> 00:13:06,200
Speaker 2: Mm hmm.

271
00:13:06,480 --> 00:13:08,559
Speaker 1: They sued the United States Army, a.

272
00:13:08,559 --> 00:13:12,039
Speaker 2: Tiny Silicon Valley startup suing the Army. It's just unheard of.

273
00:13:12,080 --> 00:13:13,320
Its career suicide.

274
00:13:13,360 --> 00:13:13,840
Speaker 1: It should be.

275
00:13:14,200 --> 00:13:17,039
Speaker 2: But they had a secret weapon. It was this neglected

276
00:13:17,080 --> 00:13:20,639
law from nineteen ninety four called the Federal Acquisition Streamlining Act.

277
00:13:20,720 --> 00:13:21,559
Speaker 1: Okay, what did that do?

278
00:13:21,799 --> 00:13:24,879
Speaker 2: This law basically says, if a commercial product exists that

279
00:13:24,919 --> 00:13:27,759
can do the job, the government must consider buying it

280
00:13:27,799 --> 00:13:30,639
before it goes and builds its own custom version from scratch.

281
00:13:30,679 --> 00:13:31,080
Speaker 1: And they won.

282
00:13:31,360 --> 00:13:34,159
Speaker 2: They won. A federal judge ruled that the Army had

283
00:13:34,159 --> 00:13:37,879
broken the law by refusing to even consider Palenteer's innovation.

284
00:13:38,600 --> 00:13:40,480
And this was a watershed.

285
00:13:39,919 --> 00:13:41,320
Speaker 1: Moment, a huge moment.

286
00:13:41,399 --> 00:13:44,240
Speaker 2: It didn't just help Palenteer, it forced the door open

287
00:13:44,360 --> 00:13:47,320
for all of Silicon Valley and defense tech. It sent

288
00:13:47,360 --> 00:13:50,320
a message that the Pentagon couldn't just keep feeding the

289
00:13:50,360 --> 00:13:53,720
old defense giants if better tech existed somewhere else.

290
00:13:53,919 --> 00:13:57,639
Speaker 1: So they conquered the battlefield. They proved Gotham worked. They

291
00:13:57,679 --> 00:14:01,639
won the legal battle. But then well, the war on

292
00:14:01,759 --> 00:14:04,759
Terror started to wind down, least in Iraq and Afghanistan.

293
00:14:05,320 --> 00:14:08,039
Speaker 2: Troops came home, and Pallenteer came home with them.

294
00:14:07,919 --> 00:14:10,799
Speaker 1: And this is where the story gets well a lot stickier.

295
00:14:10,879 --> 00:14:13,600
Speaker 2: This is the pivot that caused the most controversy for sure.

296
00:14:14,039 --> 00:14:17,919
Pallenteer began working with domestic agencies like IC, like IC

297
00:14:18,720 --> 00:14:22,080
Immigration and Customs Enforcement and the Department of Homeland Security.

298
00:14:22,240 --> 00:14:25,840
Speaker 1: So the exact same software those designed to track terrorists

299
00:14:26,000 --> 00:14:29,080
in the mountains of Afghanistan was now being used to

300
00:14:29,120 --> 00:14:31,799
track undocumented immigrants in the United States.

301
00:14:32,000 --> 00:14:35,759
Speaker 2: Correct, they utilized a system called Falcon, and it allowed

302
00:14:36,200 --> 00:14:41,000
IC agents to access all these disparate databases, employment records,

303
00:14:41,120 --> 00:14:45,799
address histories, family connections, all in that same single pane

304
00:14:45,799 --> 00:14:46,399
of glass.

305
00:14:46,639 --> 00:14:50,759
Speaker 1: And we have specific examples of how this changed enforcement. Right.

306
00:14:50,799 --> 00:14:52,559
It wasn't just theoretical, we do.

307
00:14:52,879 --> 00:14:56,320
Speaker 2: There are reports detailing how the software could identify, say

308
00:14:56,440 --> 00:15:00,279
a suspect's cousin simply because they rented a tr rock

309
00:15:00,360 --> 00:15:02,919
near a known safe house or shared an address five

310
00:15:03,000 --> 00:15:06,039
years ago. The system would just map these social networks

311
00:15:06,039 --> 00:15:09,679
and effectively flag individuals for investigation or detention.

312
00:15:09,960 --> 00:15:12,759
Speaker 1: It makes the whole needle in a haystack problem easy.

313
00:15:13,039 --> 00:15:15,600
But when the needle is a family living in the suburbs,

314
00:15:16,000 --> 00:15:18,879
the moral implications just feel very, very different than when

315
00:15:18,879 --> 00:15:20,440
the needle is an insurgent.

316
00:15:20,080 --> 00:15:23,000
Speaker 2: Sell That's the absolute crux of the backlash. Civil rights

317
00:15:23,039 --> 00:15:26,600
groups like the ACLU were incredibly alarmed. They argued that

318
00:15:26,639 --> 00:15:29,519
Pollenteer turned deportation into an assembly.

319
00:15:29,039 --> 00:15:31,399
Speaker 1: Line did remove the friction, exactly.

320
00:15:31,039 --> 00:15:33,559
Speaker 2: Remove the friction. When you make it incredibly easy to

321
00:15:33,559 --> 00:15:36,440
find people, you inevitably increase the number of people who

322
00:15:36,480 --> 00:15:37,000
get found.

323
00:15:37,200 --> 00:15:42,039
Speaker 1: And it wasn't just external critics pointing this out inside Polantatier,

324
00:15:42,480 --> 00:15:44,360
the employees started revolting.

325
00:15:44,679 --> 00:15:47,639
Speaker 2: Yeah, this was during the Trump administration, when immigration policy

326
00:15:47,759 --> 00:15:50,799
was I mean it was incredibly heated. You had engineers

327
00:15:50,799 --> 00:15:54,360
at Pallenteer writing letters saying I signed up to stop terrorists,

328
00:15:54,519 --> 00:15:58,320
not to deport families. It really highlights the problem of

329
00:15:58,480 --> 00:16:00,480
dual use technology right.

330
00:16:00,600 --> 00:16:03,200
Speaker 1: Dual use a hammer can build a house, or it

331
00:16:03,240 --> 00:16:04,320
can crack a skull.

332
00:16:04,240 --> 00:16:08,360
Speaker 2: Exactly the underlying mechanism. The math of connecting data points

333
00:16:08,480 --> 00:16:12,360
is neutral, but when you apply that math to domestic

334
00:16:12,399 --> 00:16:16,559
immigration enforcement, the moral calculus changes significantly.

335
00:16:16,720 --> 00:16:18,159
Speaker 1: So how to Pollaneer respond?

336
00:16:18,279 --> 00:16:21,840
Speaker 2: Cooluntier held the line. Alex Karp argued that tech companies

337
00:16:21,840 --> 00:16:25,480
shouldn't be the ones deciding policy. He basically said, if

338
00:16:25,559 --> 00:16:28,600
the democratic process alects a government that wants to enforce

339
00:16:28,639 --> 00:16:31,039
immigration laws, we don't have the right to say no.

340
00:16:31,240 --> 00:16:33,679
Speaker 1: Which is a pretty convenient stance when the contracts are

341
00:16:33,720 --> 00:16:36,279
worth millions and millions of dollars, of course, but despite

342
00:16:36,360 --> 00:16:40,240
the protests and the resignations, the contracts just kept coming.

343
00:16:40,600 --> 00:16:43,519
Pallenteer wasn't just a counter terrorism company anymore. It had

344
00:16:43,519 --> 00:16:46,600
become a kind of state sponsored crystal ball m HM.

345
00:16:46,840 --> 00:16:49,320
And once they had the government locked down, they looked

346
00:16:49,360 --> 00:16:53,720
at the only battlefield left untouched, corporate America.

347
00:16:54,159 --> 00:16:55,879
Speaker 2: And this was the launch of Foundry.

348
00:16:56,559 --> 00:17:01,679
Speaker 1: Okay, so Gotham is for spies, Foundry is for who business.

349
00:17:02,320 --> 00:17:04,759
Speaker 2: Think of Foundry as Gotham, but wearing a suit.

350
00:17:04,480 --> 00:17:06,759
Speaker 1: And tie Gotham in a suit. I like that.

351
00:17:07,440 --> 00:17:10,599
Speaker 2: The realization Palenteer had was that while governments have basically

352
00:17:10,640 --> 00:17:15,599
infinite money, they move incredibly slowly. CEOs, on the other hand,

353
00:17:15,680 --> 00:17:19,039
sign checks very fast. True, but big corporations have the

354
00:17:19,039 --> 00:17:21,240
exact same problem as the CIA did in two thousand

355
00:17:21,279 --> 00:17:24,960
and one. They have data silos. Sales doesn't talk to logistics,

356
00:17:25,240 --> 00:17:27,200
HR doesn't talk to finance. It's all separate.

357
00:17:27,440 --> 00:17:30,119
Speaker 1: So they pitched Foundry to the Fortune five hundred. But

358
00:17:30,160 --> 00:17:31,960
what does it actually do? I mean, is it just

359
00:17:32,000 --> 00:17:33,960
a bunch of fancy charts because every company has a

360
00:17:34,000 --> 00:17:34,880
dashboard these days.

361
00:17:34,960 --> 00:17:36,799
Speaker 2: No, and that is the key difference. A dashboard shows

362
00:17:36,799 --> 00:17:40,559
you what happened yesterday, sales were down five percent. Foundry

363
00:17:40,559 --> 00:17:43,319
integrates all the enterprise data to create what they call

364
00:17:43,400 --> 00:17:46,920
a digital twin of the company. It simulates the entire

365
00:17:47,279 --> 00:17:49,880
business and then it.

366
00:17:49,880 --> 00:17:52,160
Speaker 1: Tells you what to do, like a digital boss.

367
00:17:52,680 --> 00:17:56,200
Speaker 2: A digital concilier is the term that's often used. Wow,

368
00:17:56,480 --> 00:17:59,519
let's look at a real example. Palenteer works with Airbus,

369
00:17:59,640 --> 00:18:03,079
the air plane manufacturer. They built a system called Skywise.

370
00:18:03,640 --> 00:18:07,119
It pulls data from every single sensor on thousands of

371
00:18:07,160 --> 00:18:10,720
planes in the air, temperature, vibration, pressure, everything, and.

372
00:18:10,680 --> 00:18:12,680
Speaker 1: So it predicts when a part is going to break.

373
00:18:12,759 --> 00:18:15,440
Speaker 2: Before it breaks, it might send an alert to an

374
00:18:15,440 --> 00:18:18,359
airline that says, the fuel pump on plane four H

375
00:18:18,440 --> 00:18:21,240
two is vibrating at a frequency that suggests failure in

376
00:18:21,279 --> 00:18:24,200
the next fifty flight hours. Replace it tonight in London

377
00:18:24,200 --> 00:18:26,240
where you have a spare part, rather than letting it

378
00:18:26,279 --> 00:18:27,559
break down and dubuy where you don't.

379
00:18:27,640 --> 00:18:30,279
Speaker 1: That's incredibly valuable. I mean that says millions in delays

380
00:18:30,279 --> 00:18:31,200
and cancelations.

381
00:18:31,279 --> 00:18:34,279
Speaker 2: It does, but it goes deeper in other industries. It

382
00:18:34,319 --> 00:18:36,559
might tell an oil company to shut down plant three

383
00:18:36,640 --> 00:18:39,680
because market prices are dropping. It might tell a logistics

384
00:18:39,680 --> 00:18:42,640
firm to reroute a container ship to avoid a coming storm.

385
00:18:43,160 --> 00:18:45,599
Speaker 1: Or it might tell a manager to fire Dave.

386
00:18:46,000 --> 00:18:48,000
Speaker 2: It might tell a manager to fire Dave.

387
00:18:48,480 --> 00:18:51,319
Speaker 1: Poor Dave. But seriously, it's making personnel decisions.

388
00:18:51,319 --> 00:18:54,240
Speaker 2: It's providing the data that leads to those decisions, often

389
00:18:54,319 --> 00:18:58,079
with such algorithmic certainty that the humans just they just

390
00:18:58,160 --> 00:19:01,440
follow the instruction. It replaced intuition with data.

391
00:19:01,799 --> 00:19:03,039
Speaker 1: I read the hitge funds use it.

392
00:19:03,160 --> 00:19:06,000
Speaker 2: Oh yeah, they use it to predict markets based on

393
00:19:06,119 --> 00:19:09,319
satellite imagery of Walmart parking lots. If the lots are

394
00:19:09,359 --> 00:19:12,519
full earnings will probably be good if they're empty.

395
00:19:12,400 --> 00:19:15,279
Speaker 1: Short the stock exactly. This leads us straight to the

396
00:19:15,279 --> 00:19:18,680
black box problem, doesn't it. Companies and governments are making

397
00:19:18,799 --> 00:19:22,079
huge decisions based on the output of this software, but

398
00:19:22,119 --> 00:19:25,880
they often don't fully understand how the software reached that conclusion.

399
00:19:26,240 --> 00:19:29,319
Speaker 2: That is the danger. It is a black box with

400
00:19:29,559 --> 00:19:32,680
massive influence. You're just trusting the algorithm, You're trusting the

401
00:19:32,720 --> 00:19:37,119
ontology that the engineers built. And this trust was pushed

402
00:19:37,119 --> 00:19:40,119
to its absolute limit during the COVID nineteen pandemic.

403
00:19:40,240 --> 00:19:43,519
Speaker 1: Oh right, Palenteer was everywhere during COVID, and I feel

404
00:19:43,559 --> 00:19:45,599
like that really flew under the radar for a lot

405
00:19:45,640 --> 00:19:46,000
of people.

406
00:19:46,200 --> 00:19:49,279
Speaker 2: It totally did, but they were the underlying operating system

407
00:19:49,319 --> 00:19:51,839
for the pandemic response in both the US and the UK.

408
00:19:52,200 --> 00:19:54,319
The project in the US was called Tiberius.

409
00:19:54,599 --> 00:19:58,359
Speaker 1: Tiberius another Roman empurn name. They really stick to the.

410
00:19:58,279 --> 00:20:02,680
Speaker 2: Brand they do. Tiberi's tracked everything. It tracked where the

411
00:20:02,720 --> 00:20:05,599
outbreaks were happening, where hospital beds were full, where the

412
00:20:05,640 --> 00:20:09,279
ppe was needed, and crucially, when the vaccines came out.

413
00:20:09,359 --> 00:20:13,240
Palenteer's code helped determine where every single vial was sent.

414
00:20:13,359 --> 00:20:15,599
Speaker 1: So wait, if you got a vaccine in twenty twenty one,

415
00:20:15,839 --> 00:20:18,799
Palateer likely had a hand in deciding which pharmacy it

416
00:20:18,839 --> 00:20:19,119
went to.

417
00:20:19,519 --> 00:20:23,640
Speaker 2: In many cases, yes, they managed the supply chain logic. Now,

418
00:20:23,680 --> 00:20:27,359
on one hand, this is incredibly impressive. The government's own systems,

419
00:20:27,640 --> 00:20:31,319
you know, faxes and spreadsheets had completely collapsed under the pressure.

420
00:20:31,480 --> 00:20:35,119
Right Palteer mobilized in weeks and almost certainly saved lives

421
00:20:35,160 --> 00:20:37,039
by getting shots in arms more efficiently.

422
00:20:37,240 --> 00:20:40,160
Speaker 1: But on the other hand, we gave a completely unaccountable

423
00:20:40,160 --> 00:20:43,799
private company total control over our national healthcare logistics.

424
00:20:44,079 --> 00:20:46,599
Speaker 2: Critics argued exactly that, They said, we were handing over

425
00:20:46,640 --> 00:20:49,319
the keys to the kingdom. Because the interface looked cool

426
00:20:49,519 --> 00:20:52,599
and you know, it worked. It deepened the reliance of

427
00:20:52,640 --> 00:20:56,799
the state on this one private entity. If Palenteer turned

428
00:20:56,839 --> 00:21:00,440
off the servers, the government would have been completely blind.

429
00:21:00,640 --> 00:21:03,319
Speaker 1: That is a terrifying amount of leverage for one company

430
00:21:03,359 --> 00:21:07,599
to have. And speaking of leverage, eventually Palenteer had to

431
00:21:07,599 --> 00:21:10,400
go public. They couldn't stay in the shadows forever if

432
00:21:10,440 --> 00:21:13,880
they wanted to grow, And in true Palenteer fashion, they

433
00:21:13,920 --> 00:21:15,319
didn't do it normally.

434
00:21:15,160 --> 00:21:18,759
Speaker 2: No traditional IPO roadshow for them. They did a direct listing,

435
00:21:19,119 --> 00:21:21,160
and when they finally opened their books to the public,

436
00:21:21,359 --> 00:21:25,519
investors were really confused. I remember that the company wasn't profitable,

437
00:21:25,640 --> 00:21:29,119
the margins were weird. They spent an astronomical amount on

438
00:21:29,240 --> 00:21:33,359
stock based compensation for their employees. The numbers didn't make sense.

439
00:21:33,200 --> 00:21:35,960
Speaker 1: In a traditional way, but people bought it anyway. Yeah,

440
00:21:36,000 --> 00:21:39,440
the retail investor hype was just insane. Why what were

441
00:21:39,440 --> 00:21:39,880
they buying.

442
00:21:40,160 --> 00:21:42,799
Speaker 2: They weren't buying the financials, they were buying the mythology.

443
00:21:42,839 --> 00:21:45,119
They were buying the mission. They were buying a piece

444
00:21:45,160 --> 00:21:48,960
of the operating system of power. Palenteer had positioned itself

445
00:21:49,000 --> 00:21:51,720
as the company that knows more than the generals and

446
00:21:51,759 --> 00:21:53,279
the central banks, and.

447
00:21:53,200 --> 00:21:57,160
Speaker 1: That reputation has only grown with the war in Ukraine.

448
00:21:57,359 --> 00:21:59,160
This is where we get into the realm of modern

449
00:21:59,200 --> 00:22:01,200
warfare that really feels like science fiction.

450
00:22:01,359 --> 00:22:04,599
Speaker 2: Volunteer has been deeply, deeply involved in Ukraine since the

451
00:22:04,680 --> 00:22:09,400
Russian invasion. Alex Karp was actually the first Western CEO

452
00:22:09,440 --> 00:22:12,599
to meet with Zelenski in Kiev after the war started.

453
00:22:12,319 --> 00:22:17,160
Speaker 1: The first before all the defense contractors, before Boeing and Raytheon.

454
00:22:16,799 --> 00:22:20,000
Speaker 2: Before everyone. He went into an active war zone to

455
00:22:20,039 --> 00:22:23,240
pitch his product and they branded their involvement as AI

456
00:22:23,359 --> 00:22:24,200
for war Fighting.

457
00:22:24,319 --> 00:22:24,559
Speaker 1: Right.

458
00:22:24,599 --> 00:22:27,759
Speaker 2: They aren't just analyzing past data anymore. They are helping

459
00:22:27,920 --> 00:22:32,599
Ukraine map Russian supply lines, optimize artillery targeting, and predict

460
00:22:32,640 --> 00:22:34,359
drone strikes in real time.

461
00:22:34,559 --> 00:22:36,319
Speaker 1: I read a quote with one of the Source articles

462
00:22:36,359 --> 00:22:39,440
that described it perfectly said this software was like algorithms

463
00:22:39,440 --> 00:22:40,960
whispering aim here.

464
00:22:41,359 --> 00:22:43,799
Speaker 2: That is the reality of it. It's a digital war

465
00:22:43,920 --> 00:22:47,759
room that updates faster than CNN. It speaks five languages,

466
00:22:48,400 --> 00:22:52,279
It processes satellite imagery almost instantly. It is helping a

467
00:22:52,400 --> 00:22:56,400
much smaller force punch way way above its weight class.

468
00:22:56,440 --> 00:22:58,160
By using information dominance.

469
00:22:58,279 --> 00:23:00,160
Speaker 1: They're creating something they call a metica.

470
00:23:00,039 --> 00:23:03,640
Speaker 2: Installation, a meta constantly, which doesn't even mean it means

471
00:23:03,839 --> 00:23:07,359
linking space to mud. Satellites in orbit are talking to

472
00:23:07,480 --> 00:23:09,920
drones in the air, which are talking to tablets in

473
00:23:09,960 --> 00:23:13,599
the hands of soldiers in a trench. Paalenteer's software is

474
00:23:13,640 --> 00:23:16,079
the glue that's holding that entire network together.

475
00:23:16,240 --> 00:23:19,160
Speaker 1: So it reduces the sensor to shooter time massively, which

476
00:23:19,200 --> 00:23:21,359
is the time between seeing a tank and actually blowing

477
00:23:21,400 --> 00:23:21,680
it up.

478
00:23:21,799 --> 00:23:24,480
Speaker 2: Exactly in the old days that could take an hour

479
00:23:24,559 --> 00:23:27,680
of radio calls and approvals. Now it can take minutes

480
00:23:27,960 --> 00:23:28,839
or even seconds.

481
00:23:28,920 --> 00:23:30,519
Speaker 1: But the bit brings is right back to the black

482
00:23:30,519 --> 00:23:33,720
box problem. If the algorithm says aim here, and you

483
00:23:33,799 --> 00:23:36,480
aim there and it turns out to be a civilian building,

484
00:23:37,799 --> 00:23:40,519
who salt is that? Is it the soldier who pushed

485
00:23:40,519 --> 00:23:43,200
the button or the engineer in Palo Alto who wrote

486
00:23:43,200 --> 00:23:43,640
the code.

487
00:23:43,839 --> 00:23:46,200
Speaker 2: Volunteer has a defense for this, of course. It's often

488
00:23:46,240 --> 00:23:49,039
called the hammer defense. The hammer defense, they say, we

489
00:23:49,119 --> 00:23:51,720
build the hammer. If someone uses it to build a house,

490
00:23:52,039 --> 00:23:53,960
or if they use it to break a window, that's

491
00:23:53,960 --> 00:23:54,480
not on us.

492
00:23:54,559 --> 00:23:57,680
Speaker 1: We just build the hammer. Yeah, that feels, I don't know,

493
00:23:58,039 --> 00:23:58,960
a little convenient.

494
00:23:59,160 --> 00:24:02,319
Speaker 2: It's a very common defense in the arms industry, which

495
00:24:02,359 --> 00:24:05,799
is essentially what big tech is becoming. They claim the

496
00:24:05,839 --> 00:24:09,279
moral high ground by saying they refuse to work with

497
00:24:09,400 --> 00:24:13,039
China or other authoritarian regimes. They pick a side.

498
00:24:13,079 --> 00:24:15,240
Speaker 1: They are very explicitly Team West.

499
00:24:15,599 --> 00:24:20,119
Speaker 2: Yes, they are in their own words, Western chauvinists.

500
00:24:19,559 --> 00:24:22,440
Speaker 1: Which is such a stark contrast at companies like Apple

501
00:24:22,559 --> 00:24:26,720
or Google who try to be these neutral global entities

502
00:24:26,759 --> 00:24:30,200
at work everywhere. Right Coolunteer says, no, we are Team America.

503
00:24:29,880 --> 00:24:33,039
Speaker 2: And Team NATA. But the line between defense and domination

504
00:24:33,440 --> 00:24:37,240
is it's pretty blurry, and they aren't stopping at war.

505
00:24:37,359 --> 00:24:39,920
They're moving into what they call the meta constellation of

506
00:24:40,039 --> 00:24:45,119
prediction for everything. For everything, they want to simulate economic collapses,

507
00:24:45,559 --> 00:24:49,920
disease outbreaks, civil unrest. The goal isn't just to react

508
00:24:49,960 --> 00:24:52,720
to the future anymore, it's to shape it. They are

509
00:24:52,759 --> 00:24:56,279
positioning themselves as the operating system for the modern state.

510
00:24:56,680 --> 00:24:58,519
Speaker 1: So if we zoom all the way out here, we

511
00:24:58,599 --> 00:25:01,880
have a single company that knows more than the cabinet,

512
00:25:01,960 --> 00:25:04,400
more than the generals, more than the central bank. It

513
00:25:04,519 --> 00:25:09,160
started with this, you know, arguably noble goal of stopping

514
00:25:09,160 --> 00:25:12,640
the next nine to eleven, but it has evolved into

515
00:25:12,680 --> 00:25:16,119
a tool that essentially runs the world's most critical infrastructure.

516
00:25:16,279 --> 00:25:19,240
Speaker 2: It raises what might be the most provocative question of

517
00:25:19,240 --> 00:25:23,079
our time, when we move from human intuition to algorithmic certainty,

518
00:25:23,400 --> 00:25:26,160
who is actually in charge? Is it the elected officials

519
00:25:26,160 --> 00:25:28,680
we vote for, or is it the engineers maintaining the

520
00:25:28,680 --> 00:25:31,119
code that tells those officials exactly what to do?

521
00:25:31,240 --> 00:25:34,200
Speaker 1: Because if the computer screen says detain this person to

522
00:25:34,200 --> 00:25:37,319
stop a terror attack, what politician is brave enough to

523
00:25:37,319 --> 00:25:38,599
say no? I trust my.

524
00:25:38,559 --> 00:25:42,319
Speaker 2: Gut Instead, no one you can't. The data becomes the dictator.

525
00:25:42,079 --> 00:25:44,240
Speaker 1: And that is the chilling thought to leave you with.

526
00:25:45,319 --> 00:25:48,400
We now live in a world where an all seeing

527
00:25:48,440 --> 00:25:51,599
eye actually exists. It might keep us safe, it might

528
00:25:51,640 --> 00:25:54,519
catch the bad guys, but the price of that safety

529
00:25:55,079 --> 00:25:58,880
is a level of surveillance and centralized power that we

530
00:25:58,960 --> 00:26:01,640
have never ever seen before in human history.

531
00:26:01,680 --> 00:26:04,519
Speaker 2: And unlike the Palanteerian Lord of the Rings, we can't

532
00:26:04,559 --> 00:26:06,559
just throw a cloth over this orb and hide it

533
00:26:06,599 --> 00:26:09,759
away in a tower. It's woven into the very fabric

534
00:26:09,799 --> 00:26:13,319
of our society. Now. It's running the trains, the banks,

535
00:26:13,400 --> 00:26:14,200
and the armies.

536
00:26:15,000 --> 00:26:16,839
Speaker 1: So I want to turn this over to you listening

537
00:26:16,880 --> 00:26:19,759
right now. Are you comfortable with an all seeing eye

538
00:26:19,799 --> 00:26:22,160
if it keeps you safe? Is the loss of privacy

539
00:26:22,200 --> 00:26:24,759
a fair trade for security? Or is the idea of

540
00:26:24,799 --> 00:26:28,759
a digital conciliere running the state just too dystopian for you.

541
00:26:29,119 --> 00:26:31,400
Speaker 2: It's a question we all need to start answering, because

542
00:26:31,400 --> 00:26:34,160
the technology isn't waiting for our permission, it's already here.

543
00:26:34,559 --> 00:26:36,160
Speaker 1: Leave a comment let us know what you think. This

544
00:26:36,200 --> 00:26:39,240
has been thrilling threads. Thanks for listening and stay curious.

545
00:26:39,480 --> 00:26:42,079
But maybe you know, keep that webcam covered. Catch you

546
00:26:42,119 --> 00:26:42,559
next time.