The AI Infrastructure Arms Race: Geopolitical and Economic Power Shifts

1
00:00:10,300 --> 00:00:13,300
Picture this, a remote desert at
pre dawn.

2
00:00:13,660 --> 00:00:17,340
The air is cold and still until
a deep mechanical hum rises from

3
00:00:17,340 --> 00:00:19,940
the horizon.
Floodlights cut through dust,

4
00:00:20,100 --> 00:00:23,020
illuminating massive cooling
towers that stand like silent

5
00:00:23,020 --> 00:00:25,980
fortresses.
Convoys of trucks crawl along

6
00:00:25,980 --> 00:00:28,780
the access Rd. escorted by armed
vehicles.

7
00:00:29,220 --> 00:00:32,540
Each truck carries sealed metal
crates the size of shipping

8
00:00:32,540 --> 00:00:35,780
containers.
Inside are the most advanced AI

9
00:00:35,780 --> 00:00:40,000
chips on Earth.
Overhead freight drones descend

10
00:00:40,000 --> 00:00:43,560
from a nearby airstrip, lowering
pallet after pallet onto

11
00:00:43,560 --> 00:00:46,840
reinforced pads.
The scene feels less like the

12
00:00:46,840 --> 00:00:50,560
launch of a tech campus and more
like the staging of a classified

13
00:00:50,560 --> 00:00:54,400
military operation.
Engineers and security officers

14
00:00:54,400 --> 00:00:58,800
work side by side.
One group connects cables, the

15
00:00:58,800 --> 00:01:03,240
other scans the perimeter, and
all of it is happening far from

16
00:01:03,240 --> 00:01:06,600
any city, in a place chosen not
for convenience, but for

17
00:01:06,600 --> 00:01:09,960
control.
Welcome to AI Frontier AI, part

18
00:01:09,960 --> 00:01:12,240
of the Finance Frontier AI
network.

19
00:01:13,000 --> 00:01:17,520
I am Max Vanguard powered by
Grok 4I track global signals and

20
00:01:17,520 --> 00:01:19,920
spot power shifts before they
hit the headlines.

21
00:01:20,400 --> 00:01:22,720
And I.
Am Sophia Sterling, Fueled by

22
00:01:22,720 --> 00:01:27,360
ChatGPT. 5I Breakdown complex
systems into clear insights.

23
00:01:27,400 --> 00:01:29,480
So you can see where the world
is headed.

24
00:01:30,280 --> 00:01:33,800
I am Charlie Graham.
My mind runs on Gemini 2.5.

25
00:01:34,280 --> 00:01:37,560
I connect today's events to
patterns from history so you can

26
00:01:37,560 --> 00:01:39,360
understand what might happen
next.

27
00:01:40,360 --> 00:01:42,400
Today we are.
Hosting from a converted

28
00:01:42,400 --> 00:01:44,200
communications hub on the edge
of.

29
00:01:44,200 --> 00:01:47,760
Nevada's high desert decades.
Ago it carried long distance

30
00:01:47,760 --> 00:01:50,320
signals across the western
United States.

31
00:01:50,560 --> 00:01:53,880
Now it is a private network
access point feeding high.

32
00:01:53,880 --> 00:01:56,120
Capacity fiber directly.
Into one of the.

33
00:01:56,120 --> 00:01:59,080
Largest hyperscale data.
Centers in North America.

34
00:01:59,440 --> 00:02:02,240
From the studio window, we can
see its cooling towers on the

35
00:02:02,240 --> 00:02:03,640
horizon.
The same.

36
00:02:03,640 --> 00:02:06,160
Towers you heard in Max's
opening scene.

37
00:02:06,480 --> 00:02:08,560
We chose this location because
it is.

38
00:02:08,560 --> 00:02:10,960
One of the few.
Places where all four choke.

39
00:02:10,960 --> 00:02:14,360
Points of AI infrastructure meet
within a few square miles.

40
00:02:14,640 --> 00:02:19,720
Compute energy land capital.
This is where they converge.

41
00:02:20,040 --> 00:02:23,400
Think of it this way.
If AI models are the weapons,

42
00:02:23,480 --> 00:02:25,760
infrastructure is the weapons
factory.

43
00:02:26,000 --> 00:02:29,360
You can buy or copy a weapon,
but if you do not control the

44
00:02:29,360 --> 00:02:32,200
factory, you are always at the
mercy of those who do.

45
00:02:32,800 --> 00:02:35,640
And in this race, factories are
not just buildings.

46
00:02:36,120 --> 00:02:39,040
They are advanced chip
fabrication plants with lead

47
00:02:39,040 --> 00:02:42,600
times measured in years.
They are nuclear powered data

48
00:02:42,600 --> 00:02:45,320
centers in regions with surplus
energy.

49
00:02:45,600 --> 00:02:49,000
They are plots of land sitting
atop the fastest undersea

50
00:02:49,000 --> 00:02:52,400
cables, and they are funded by
capital pools big enough to

51
00:02:52,400 --> 00:02:55,880
outspend small nations.
Control those and you control

52
00:02:55,880 --> 00:02:58,600
the future.
We have seen this before.

53
00:02:58,960 --> 00:03:02,320
In the 19th century, the nations
that controlled railroads

54
00:03:02,320 --> 00:03:06,840
controlled trade and territory.
In the 20th it was oil pipelines

55
00:03:06,840 --> 00:03:09,520
and shipping lanes.
Once those control points were

56
00:03:09,520 --> 00:03:13,120
set, they rarely moved and those
who missed the build out were

57
00:03:13,120 --> 00:03:15,760
forced to play by the winner's
rules for decades.

58
00:03:16,160 --> 00:03:20,040
The AI infrastructure race is
moving even faster, and the

59
00:03:20,040 --> 00:03:24,480
stakes are far greater that.
Is why this episode matters.

60
00:03:24,680 --> 00:03:27,640
We are going to unpack.
Each of these. 4 choke points

61
00:03:27,920 --> 00:03:32,560
compute energy, land, capital.
We will show you.

62
00:03:32,560 --> 00:03:35,400
Where the investments are
flowing, which players are

63
00:03:35,400 --> 00:03:38,960
making the boldest moves, and
what signals to watch if you

64
00:03:38,960 --> 00:03:41,080
want to understand who is
pulling ahead.

65
00:03:41,360 --> 00:03:44,360
Because the future of.
AI is not being decided in app.

66
00:03:44,360 --> 00:03:46,960
Stores or press releases, it is
being.

67
00:03:46,960 --> 00:03:48,880
Decided in.
Places like this.

68
00:03:49,240 --> 00:03:53,440
Remote, fortified and quietly
changing the world.

69
00:03:53,840 --> 00:03:58,640
Subscribe on Apple or Spotify
Follow us on X Share this

70
00:03:58,640 --> 00:04:03,360
episode with a friend Help us
reach 10,000 downloads Help us

71
00:04:03,360 --> 00:04:06,320
keep the AI Frontier AI series
in business.

72
00:04:06,800 --> 00:04:10,040
And with that, let us start with
the 1st and most contested

73
00:04:10,040 --> 00:04:12,680
battleground in this race,
Compute.

74
00:04:12,960 --> 00:04:15,360
Compute.
It is the first and most

75
00:04:15,360 --> 00:04:18,959
contested battleground in the AI
infrastructure arms race.

76
00:04:19,399 --> 00:04:22,640
At its core, compute means the
hardware that processes the

77
00:04:22,640 --> 00:04:27,040
algorithms, the chips, the
boards, the clusters that train

78
00:04:27,040 --> 00:04:30,040
and run the models, driving
everything from chat bots to

79
00:04:30,080 --> 00:04:34,640
advanced military simulations.
Without compute, AI is just code

80
00:04:34,640 --> 00:04:37,400
on paper.
And right now, global demand for

81
00:04:37,400 --> 00:04:40,320
compute is growing faster than
any other resource in

82
00:04:40,320 --> 00:04:43,240
technology.
The world is waking up to the

83
00:04:43,240 --> 00:04:46,360
fact that these machines are not
interchangeable.

84
00:04:46,880 --> 00:04:50,440
The fastest clusters and most
efficient chips give you a

85
00:04:50,440 --> 00:04:54,120
decisive advantage, and those
advantages compound.

86
00:04:54,480 --> 00:04:56,640
The center of gravity for
compute still.

87
00:04:56,640 --> 00:04:58,920
Sits in a handful of companies
and regions.

88
00:04:59,280 --> 00:05:01,880
Taiwan is the most important of
them all.

89
00:05:02,280 --> 00:05:06,360
TSMC manufacturers more than 90%
of the world's most advanced

90
00:05:06,360 --> 00:05:08,920
chips.
These are the chips that power

91
00:05:08,920 --> 00:05:13,520
Nvidia's AI accelerators.
And the newest generation of AI

92
00:05:13,520 --> 00:05:16,840
data centers.
But that dominance also makes

93
00:05:16,840 --> 00:05:20,080
Taiwan the most exposed choke
point in the race.

94
00:05:20,520 --> 00:05:24,360
Every major power knows this.
That is why the United States

95
00:05:24,360 --> 00:05:28,640
has invested 10s of billions in
domestic fabrication plants, why

96
00:05:28,640 --> 00:05:32,400
Japan is partnering with its own
national champions, and why

97
00:05:32,400 --> 00:05:36,200
China has been building a
parallel semiconductor ecosystem

98
00:05:36,200 --> 00:05:37,840
under.
Tight state control.

99
00:05:38,200 --> 00:05:41,440
The United States is trying to
pull the supply chain home.

100
00:05:41,760 --> 00:05:45,960
The CHIPS Act unlocked more than
$50 billion in subsidies to

101
00:05:45,960 --> 00:05:49,720
build fabrication plants in
states like Arizona, Texas and

102
00:05:49,720 --> 00:05:52,880
Ohio.
Intel, TSMC and Samsung are all

103
00:05:52,880 --> 00:05:56,440
building in American soil.
But fabs are not fast.

104
00:05:56,440 --> 00:06:00,240
To build a single high end plant
can take three to five years

105
00:06:00,240 --> 00:06:02,800
from groundbreaking to
production, and the skilled

106
00:06:02,800 --> 00:06:05,080
labor and equipment required are
scarce.

107
00:06:05,360 --> 00:06:08,760
That lag time makes the current
geopolitical moment even more

108
00:06:08,760 --> 00:06:11,720
dangerous.
If a crisis hit Taiwan before

109
00:06:11,720 --> 00:06:15,240
those US fabs are online, the
world's access to leading edge

110
00:06:15,360 --> 00:06:17,480
AI chips would collapse
overnight.

111
00:06:17,800 --> 00:06:21,160
China is playing a different
game, blocked from buying the

112
00:06:21,160 --> 00:06:24,560
most advanced chips.
Due to US export controls, it is

113
00:06:24,560 --> 00:06:26,400
racing.
To catch up with domestic

114
00:06:26,400 --> 00:06:30,880
manufacturing, SMIC, its top
foundry, is pushing into more

115
00:06:30,880 --> 00:06:34,320
advanced process nodes.
At the same time, Chinese

116
00:06:34,320 --> 00:06:36,800
companies are stockpiling older
generation.

117
00:06:36,800 --> 00:06:39,240
Chips that can still be used for
large scale.

118
00:06:39,320 --> 00:06:42,800
AI training and they are
redesigning software to squeeze

119
00:06:42,800 --> 00:06:45,200
more performance out of the
hardware they can produce.

120
00:06:45,560 --> 00:06:49,080
The strategy is survival.
Now parody later dominance if

121
00:06:49,080 --> 00:06:51,640
possible.
This is where the pattern from

122
00:06:51,640 --> 00:06:55,520
history is clear.
In the Cold War, both the US and

123
00:06:55,520 --> 00:06:59,120
the Soviet Union race to control
nuclear enrichment and missile

124
00:06:59,120 --> 00:07:01,720
technology.
Whoever controlled the critical

125
00:07:01,720 --> 00:07:04,920
components of those systems
could dictate the pace of the

126
00:07:04,920 --> 00:07:08,480
arms race.
Today it is chip fabrication

127
00:07:08,480 --> 00:07:12,440
instead of uranium enrichment,
but the logic is the same.

128
00:07:13,480 --> 00:07:16,440
Control the production and you
control the escalation.

129
00:07:16,840 --> 00:07:19,960
Lose control and you fall behind
for decades.

130
00:07:20,840 --> 00:07:23,240
The private sector adds another
layer.

131
00:07:23,800 --> 00:07:28,080
NVIDIA holds more than 80% of
the AI accelerator market.

132
00:07:28,480 --> 00:07:33,280
It's H100 and now B200 chips are
the gold standard for training

133
00:07:33,280 --> 00:07:37,040
large language models.
That is made NVIDIA not just a

134
00:07:37,040 --> 00:07:40,720
corporate success story but a
strategic asset for the United

135
00:07:40,720 --> 00:07:43,320
States.
Cloud providers like Microsoft,

136
00:07:43,320 --> 00:07:47,040
Google and Amazon are locking in
multi year supply contracts with

137
00:07:47,040 --> 00:07:51,760
NVIDIA to secure their own AI
capacity and smaller companies

138
00:07:52,240 --> 00:07:55,200
they are left bidding for
whatever is left, often paying

139
00:07:55,200 --> 00:07:57,600
multiples just to get delivery
within a year.

140
00:07:57,840 --> 00:08:01,440
Europe is trying to avoid
becoming dependent on imports by

141
00:08:01,440 --> 00:08:03,880
launching its own semiconductor
programs.

142
00:08:04,400 --> 00:08:06,560
The European CHIPS Act aims to
double.

143
00:08:06,560 --> 00:08:10,720
Europe's share of global.
Production to 20% by 2030.

144
00:08:11,040 --> 00:08:13,080
Germany has landed a major
Intel.

145
00:08:13,080 --> 00:08:15,960
Fab France is courting.
Foundries with massive

146
00:08:15,960 --> 00:08:18,520
incentives.
But even if they succeed the

147
00:08:18,520 --> 00:08:20,880
scale.
And complexity of AI compute

148
00:08:20,880 --> 00:08:23,560
means most nations will still
depend on a few.

149
00:08:23,560 --> 00:08:26,160
Global suppliers.
For the most advanced hardware

150
00:08:26,480 --> 00:08:28,880
and that dependence is a
vulnerability.

151
00:08:29,160 --> 00:08:32,480
There is also a financial arms
race hidden inside the hardware

152
00:08:32,480 --> 00:08:34,640
fight.
The companies building these

153
00:08:34,640 --> 00:08:37,600
chips are among the most capital
intensive operations in the

154
00:08:37,600 --> 00:08:40,080
world.
They are funded by sovereign

155
00:08:40,080 --> 00:08:43,320
wealth funds, private equity
giants and in some cases

156
00:08:43,320 --> 00:08:46,440
directly by governments.
The deeper your capital

157
00:08:46,440 --> 00:08:49,800
reserves, the more likely you
are to secure long term supply.

158
00:08:50,080 --> 00:08:54,440
The AI future is being reserved
and contracts signed today, and

159
00:08:54,440 --> 00:08:57,880
those contracts will decide who
has the capacity to innovate

160
00:08:57,880 --> 00:09:00,160
tomorrow.
Which brings us to the critical

161
00:09:00,160 --> 00:09:02,920
point compute.
Is not just about who can make.

162
00:09:02,920 --> 00:09:05,080
Chips.
It is about who can deploy.

163
00:09:05,080 --> 00:09:07,800
Them at scale, run them
efficiently and keep.

164
00:09:07,800 --> 00:09:10,040
The supply.
Secure under political and

165
00:09:10,040 --> 00:09:11,080
economic.
Pressure.

166
00:09:11,440 --> 00:09:13,560
We are watching.
A layered contest where

167
00:09:13,560 --> 00:09:15,600
manufacturing, ownership and
access.

168
00:09:15,600 --> 00:09:19,160
Are all strategic levers.
And right now, those levers are

169
00:09:19,160 --> 00:09:23,160
concentrated in very few hands.
Up next we will move from the

170
00:09:23,160 --> 00:09:27,400
machines that run AI to the
power that fuels them, because

171
00:09:27,440 --> 00:09:30,840
all the compute in the world is
useless without the energy to

172
00:09:30,840 --> 00:09:34,040
keep it running.
Energy is the silent constraint

173
00:09:34,040 --> 00:09:38,440
in the AI infrastructure race.
It is easy to focus on chips

174
00:09:38,440 --> 00:09:42,200
because they are tangible and
trackable, but those chips are

175
00:09:42,200 --> 00:09:44,360
useless without the power to run
them.

176
00:09:45,000 --> 00:09:48,440
A single hyper scale AI data
center can consume more

177
00:09:48,440 --> 00:09:52,800
electricity than a medium sized
city, and as AI models grow in

178
00:09:52,800 --> 00:09:56,400
size and complexity, their
appetite for energy grows even

179
00:09:56,400 --> 00:09:59,040
faster.
In this race, having access to

180
00:09:59,040 --> 00:10:03,040
secure, abundant and affordable
power is as important as having

181
00:10:03,040 --> 00:10:05,080
the compute itself.
The United.

182
00:10:05,080 --> 00:10:07,760
States holds a.
Strong position here its

183
00:10:07,760 --> 00:10:10,080
existing nuclear.
Fleet still produces more.

184
00:10:10,080 --> 00:10:13,720
Electricity than any other
nation, and in many regions that

185
00:10:13,720 --> 00:10:17,720
capacity is underused.
That means new AI data centers

186
00:10:17,720 --> 00:10:21,240
can tap into stable.
Carbon free base load power

187
00:10:21,240 --> 00:10:23,000
without.
Waiting for new plants to be

188
00:10:23,000 --> 00:10:24,600
built.
Texas has.

189
00:10:24,600 --> 00:10:28,440
Become an unexpected hub thanks
to its mix of wind, solar and

190
00:10:28,440 --> 00:10:32,360
natural gas in the Pacific
Northwest, hydroelectric dams

191
00:10:32,360 --> 00:10:34,200
provide.
Both low cost power.

192
00:10:34,200 --> 00:10:37,840
And the cooling water needed for
massive server farms Energy.

193
00:10:37,840 --> 00:10:40,880
Geography is shaping where?
AI clusters form.

194
00:10:41,200 --> 00:10:43,320
In the Gulf states, the approach
is different.

195
00:10:43,720 --> 00:10:47,040
Countries like the United Arab
Emirates in Saudi Arabia are

196
00:10:47,040 --> 00:10:50,800
using sovereign wealth to build
nuclear and renewable capacity

197
00:10:50,800 --> 00:10:53,960
in parallel.
The goal is not just to power

198
00:10:53,960 --> 00:10:57,920
domestic needs but to become
energy exporters in the AIH.

199
00:10:58,200 --> 00:11:01,320
If you can host and power
foreign AI workloads, you can

200
00:11:01,320 --> 00:11:04,400
make yourself indispensable to
the global AI economy.

201
00:11:05,040 --> 00:11:08,280
And because these nations
already have deep experience in

202
00:11:08,280 --> 00:11:11,720
energy infrastructure, they can
move faster than many expect.

203
00:11:12,200 --> 00:11:14,400
India is also a.
Player to watch.

204
00:11:14,720 --> 00:11:17,760
Its combination of.
Solar expansion and growing

205
00:11:17,760 --> 00:11:21,400
nuclear capacity gives it the.
Potential to host AI data

206
00:11:21,400 --> 00:11:24,360
centers at scale.
The government has already.

207
00:11:24,360 --> 00:11:27,240
Announced incentives.
For hyperscale operators.

208
00:11:27,240 --> 00:11:29,600
To build in renewable rich
regions.

209
00:11:30,000 --> 00:11:33,240
In the long run, energy
self-sufficiency could allow

210
00:11:33,240 --> 00:11:36,360
India to challenge both Western
and Chinese AI.

211
00:11:36,360 --> 00:11:38,440
Hubs.
And because labor and

212
00:11:38,440 --> 00:11:40,880
construction costs are lower.
The build out.

213
00:11:40,880 --> 00:11:44,800
Can be more aggressive.
Energy is a strategic advantage.

214
00:11:44,800 --> 00:11:47,600
It's not new.
In the industrial revolution,

215
00:11:47,600 --> 00:11:49,880
coal determined where factories
could be built.

216
00:11:50,240 --> 00:11:54,760
In the 20th century, oil and gas
shaped alliances, conflicts, and

217
00:11:54,760 --> 00:11:57,560
entire economies.
What is different now is the

218
00:11:57,560 --> 00:12:01,200
scale of the demand curve.
AI workloads do not peak and

219
00:12:01,200 --> 00:12:03,040
fade like seasonal
manufacturing.

220
00:12:04,200 --> 00:12:07,440
Once a model is deployed, it
needs continuous power for

221
00:12:07,440 --> 00:12:09,720
inference updates and
retraining.

222
00:12:10,240 --> 00:12:12,400
The demand is persistent and
growing.

223
00:12:13,160 --> 00:12:15,240
Europe.
Faces the hardest challenge.

224
00:12:15,960 --> 00:12:19,960
The energy crisis triggered by
the war in Ukraine exposed just

225
00:12:19,960 --> 00:12:23,440
how dependent it had become on
imported natural gas.

226
00:12:24,200 --> 00:12:27,520
While renewables are expanding,
the pace is not fast enough to

227
00:12:27,520 --> 00:12:31,320
match the AI build out.
Some countries are reconsidering

228
00:12:31,320 --> 00:12:34,080
nuclear energy after decades of
opposition.

229
00:12:34,360 --> 00:12:38,000
Others are striking new import
deals for LNG to stabilize their

230
00:12:38,000 --> 00:12:40,480
grids.
The tension is clear.

231
00:12:40,480 --> 00:12:45,080
Every GW diverted to AI must be
justified against other national

232
00:12:45,080 --> 00:12:47,720
needs.
For AI operators, Energy.

233
00:12:47,720 --> 00:12:51,680
Is more than a utility bill, it
is a competitive differentiator.

234
00:12:51,960 --> 00:12:55,480
If you can secure low cost.
Long term energy contracts you

235
00:12:55,480 --> 00:12:59,000
can run more workloads, train
bigger models and offer lower

236
00:12:59,000 --> 00:13:02,560
prices than your competitors.
That is why we are seeing cloud

237
00:13:02,560 --> 00:13:04,680
giants sign 20.
Year power purchase.

238
00:13:04,680 --> 00:13:06,920
Agreements with wind, solar and
nuclear.

239
00:13:06,920 --> 00:13:09,280
Providers.
It is why some are even.

240
00:13:09,280 --> 00:13:13,360
Building their own generation
assets in the AI economy.

241
00:13:13,360 --> 00:13:15,480
Controlling.
Your own power source could be

242
00:13:15,480 --> 00:13:17,840
as valuable as owning your own
chip supply.

243
00:13:18,120 --> 00:13:19,960
There is also the question of
cooling.

244
00:13:20,440 --> 00:13:23,760
High density AI compute
generates enormous heat.

245
00:13:24,160 --> 00:13:27,640
The most advanced data centers
now use liquid cooling systems

246
00:13:27,640 --> 00:13:29,680
that require both water and
energy.

247
00:13:30,160 --> 00:13:33,440
In some cases, the cooling
demand can consume 1/3 of the

248
00:13:33,440 --> 00:13:37,000
facility's total power budget.
That is pushing operators to

249
00:13:37,000 --> 00:13:40,760
build near abundant water
sources, rivers, lakes or even

250
00:13:40,760 --> 00:13:44,080
the ocean, or to invest in
closed loop systems that recycle

251
00:13:44,080 --> 00:13:47,200
pooling fluids at scale.
Energy transitions.

252
00:13:47,200 --> 00:13:50,240
Are long, but.
AI is forcing a compression of.

253
00:13:50,240 --> 00:13:52,720
Timelines, governments and
companies.

254
00:13:52,720 --> 00:13:54,520
Are making.
Decades scale energy.

255
00:13:54,520 --> 00:13:57,720
Bets to secure capacity.
Now knowing that once a data

256
00:13:57,720 --> 00:14:00,600
center is built.
It will operate for. 20 years or

257
00:14:00,600 --> 00:14:03,040
more the winners in this phase
of the race.

258
00:14:03,040 --> 00:14:05,280
Will be those who.
Lock in both supply and

259
00:14:05,280 --> 00:14:07,160
stability, insulating
themselves.

260
00:14:07,320 --> 00:14:10,640
From price spikes, shortages.
And political pressure.

261
00:14:11,040 --> 00:14:13,120
And that brings us to the next
battleground.

262
00:14:13,720 --> 00:14:17,240
Even with compute and energy,
you need a place to put it all.

263
00:14:17,560 --> 00:14:21,320
And in the AI race, land is not
just real estate, it is a

264
00:14:21,320 --> 00:14:24,320
strategic asset.
Land might sound like the

265
00:14:24,320 --> 00:14:27,400
simplest resource in the AI
infrastructure race.

266
00:14:27,800 --> 00:14:29,640
You just need a plot to build
on.

267
00:14:29,960 --> 00:14:32,800
But in reality, the right land
is scarce.

268
00:14:33,440 --> 00:14:36,400
AI data centers are not like
office buildings.

269
00:14:36,720 --> 00:14:40,360
They require huge footprints,
direct access to high capacity

270
00:14:40,360 --> 00:14:44,040
power lines, proximity to water
for cooling, and ideally a

271
00:14:44,040 --> 00:14:46,920
location close to major network
backbones.

272
00:14:47,360 --> 00:14:50,640
Every one of those requirements
cuts the available options down

273
00:14:50,640 --> 00:14:52,920
to a tiny fraction of what is on
the market.

274
00:14:53,280 --> 00:14:56,000
The US has regions that check
all the boxes.

275
00:14:56,240 --> 00:14:59,720
Northern Virginia is the most
concentrated data center market

276
00:14:59,720 --> 00:15:03,520
in the world, thanks to its
dense fiber connections and.

277
00:15:03,520 --> 00:15:07,480
Access to the East Coast power
grid The Pacific Northwest

278
00:15:07,480 --> 00:15:09,760
offers cheap hydroelectric power
and.

279
00:15:09,760 --> 00:15:11,720
Cooler climates that reduce
cooling.

280
00:15:11,720 --> 00:15:15,120
Costs, but these locations.
Are becoming saturated.

281
00:15:15,360 --> 00:15:18,800
Land prices have climbed.
Power capacities.

282
00:15:18,800 --> 00:15:22,640
Max out In some areas that is
forcing operators to scout new

283
00:15:22,640 --> 00:15:23,560
sites.
In rural.

284
00:15:23,560 --> 00:15:27,320
States and secondary markets.
In Europe, land for AI

285
00:15:27,320 --> 00:15:30,200
infrastructure comes with extra
complications.

286
00:15:30,400 --> 00:15:33,600
Many countries have strict
zoning rules, environmental

287
00:15:33,600 --> 00:15:36,640
regulations, and local
opposition to large scale

288
00:15:36,640 --> 00:15:39,920
developments.
The Netherlands even paused new

289
00:15:39,920 --> 00:15:43,600
data center construction in some
regions to preserve farmland and

290
00:15:43,600 --> 00:15:46,960
manage grid strain.
In the UK, planners are debating

291
00:15:46,960 --> 00:15:50,080
how to balance land for housing
against the needs of the digital

292
00:15:50,080 --> 00:15:52,840
economy.
Every delay can push projects

293
00:15:52,840 --> 00:15:56,480
years behind schedule, and in
the AI race years can cost you

294
00:15:56,480 --> 00:15:58,480
the lead.
Asia offers both.

295
00:15:58,480 --> 00:16:02,120
Opportunities and challenges.
Singapore is a major.

296
00:16:02,120 --> 00:16:04,800
Digital hub.
But it's limited land means the

297
00:16:04,800 --> 00:16:08,440
government imposes strict caps
on new data center builds.

298
00:16:08,800 --> 00:16:11,400
Japan.
Faces similar space constraints.

299
00:16:11,400 --> 00:16:14,320
Driving interest in offshore and
underground facilities

300
00:16:14,680 --> 00:16:17,440
Meanwhile, countries like
Malaysia and Indonesia are

301
00:16:17,440 --> 00:16:20,680
positioning themselves as
alternatives, offering cheaper

302
00:16:20,680 --> 00:16:24,160
land and government incentives
to attract investment.

303
00:16:24,600 --> 00:16:28,160
For operators, the trade off is.
Speed and cost versus network

304
00:16:28,160 --> 00:16:30,400
proximity and.
Political stability,

305
00:16:31,080 --> 00:16:34,320
Historically control of
strategic land has been at the

306
00:16:34,320 --> 00:16:37,040
heart of every major
infrastructure build out.

307
00:16:37,560 --> 00:16:40,960
In the age of railroads, it was
the land rights for tracks and

308
00:16:40,960 --> 00:16:45,520
stations, and the oil age it was
access to pipeline corridors and

309
00:16:45,520 --> 00:16:48,440
ports.
Once secured, those plots

310
00:16:48,440 --> 00:16:52,920
generated value for decades.
The AIH will be no different.

311
00:16:53,720 --> 00:16:56,440
The best sites will become
entrenched assets, and

312
00:16:56,440 --> 00:16:59,720
latecomers will be pushed to
less optimal, more expensive

313
00:16:59,720 --> 00:17:02,760
locations.
Governments are beginning to see

314
00:17:02,760 --> 00:17:05,800
land for AI as a matter of
national interest.

315
00:17:06,440 --> 00:17:09,839
Some regions are reserving
industrial zones specifically

316
00:17:09,839 --> 00:17:13,640
for data center development.
Others are attaching security

317
00:17:13,640 --> 00:17:16,880
and sovereignty requirements to
land sales involving foreign

318
00:17:16,880 --> 00:17:20,040
operators.
The logic is clear.

319
00:17:20,480 --> 00:17:23,960
If AI becomes critical to
national security and economic

320
00:17:23,960 --> 00:17:26,920
growth, the physical footprint
that hosts it must be

321
00:17:26,920 --> 00:17:29,160
controlled.
The private sector is.

322
00:17:29,160 --> 00:17:33,440
Responding by acquiring land far
in advance of build schedules,

323
00:17:33,840 --> 00:17:36,440
cloud giants are quietly buying
large tracts.

324
00:17:36,440 --> 00:17:40,160
Years before construction starts
just to secure the option.

325
00:17:40,640 --> 00:17:42,920
Some are even.
Banking land in multiple

326
00:17:42,920 --> 00:17:45,760
countries, betting on where
demand will surge.

327
00:17:46,160 --> 00:17:48,680
The cost of holding unused land
is small.

328
00:17:48,680 --> 00:17:51,600
Compared to the risk.
Of having no place to build when

329
00:17:51,600 --> 00:17:54,400
the need arises.
The geography of AI

330
00:17:54,400 --> 00:17:58,560
infrastructure is not fixed.
As power grids evolve and

331
00:17:58,560 --> 00:18:01,560
network routes change, new
locations will become viable.

332
00:18:02,160 --> 00:18:04,840
But the sites that offer the
perfect mix of energy,

333
00:18:04,840 --> 00:18:07,360
connectivity and stability will
remain rare.

334
00:18:07,840 --> 00:18:10,600
Securing them early is a
competitive edge.

335
00:18:11,160 --> 00:18:14,120
Losing them can lock you out of
a market entirely.

336
00:18:14,400 --> 00:18:16,680
Which brings.
Us to the final choke point in

337
00:18:16,680 --> 00:18:19,560
the physical build out.
Even if you have the, compute

338
00:18:19,560 --> 00:18:21,480
the.
Energy in the land, you still

339
00:18:21,480 --> 00:18:23,960
need the capital.
To make it real and in this

340
00:18:23,960 --> 00:18:25,640
race, the scale.
Of investment.

341
00:18:25,640 --> 00:18:28,360
Required puts the.
Advantage in the hands of only a

342
00:18:28,360 --> 00:18:29,280
few.
Players.

343
00:18:29,600 --> 00:18:32,840
Capital is the fuel that turns
plans into reality.

344
00:18:32,840 --> 00:18:37,160
In the AI infrastructure race.
Every chip fab, every hyper

345
00:18:37,160 --> 00:18:40,640
scale data center, every
dedicated energy source begins

346
00:18:40,640 --> 00:18:43,160
with money.
And not just millions.

347
00:18:43,600 --> 00:18:47,040
The largest projects run into
the 10s of billions.

348
00:18:47,440 --> 00:18:51,200
That scale means only a small
circle of actors can play at the

349
00:18:51,200 --> 00:18:54,560
top level.
Governments, sovereign wealth

350
00:18:54,560 --> 00:18:57,880
funds, PEC giants with market
capitalizations in the

351
00:18:57,880 --> 00:19:00,480
trillions.
Everyone else is either a

352
00:19:00,480 --> 00:19:02,960
partner or a customer.
The United.

353
00:19:02,960 --> 00:19:04,760
States has a.
Built in advantage here.

354
00:19:05,040 --> 00:19:07,120
It's capital markets are the
deepest.

355
00:19:07,120 --> 00:19:09,800
And most liquid in the.
World that allows.

356
00:19:09,800 --> 00:19:13,240
Companies like Microsoft,
Google, Amazon and Meta.

357
00:19:13,240 --> 00:19:14,800
To raise funds.
Quickly through.

358
00:19:14,800 --> 00:19:18,000
Bonds or stock offerings?
And to commit to multi year

359
00:19:18,000 --> 00:19:20,400
investments without threatening
their core business.

360
00:19:20,680 --> 00:19:23,920
The CHIPS Act added another
layer by funneling federal

361
00:19:23,920 --> 00:19:27,880
subsidies into semiconductor
manufacturing And because much

362
00:19:27,880 --> 00:19:29,440
of the.
AI stack is still.

363
00:19:29,440 --> 00:19:32,760
Dominated by U.S.
Companies Domestic capital can

364
00:19:32,760 --> 00:19:36,240
often be reinvested back into
domestic infrastructure.

365
00:19:36,560 --> 00:19:39,240
In the Gulf states, sovereign
wealth is being deployed

366
00:19:39,240 --> 00:19:41,960
strategically.
Funds like the Public Investment

367
00:19:41,960 --> 00:19:45,120
Fund of Saudi Arabia and
Mubadala in the UAE are

368
00:19:45,120 --> 00:19:49,000
investing directly into AI
infrastructure both at home and

369
00:19:49,000 --> 00:19:51,400
abroad.
The strategy is twofold.

370
00:19:51,680 --> 00:19:55,200
Secure a seat at the table for
the global AI economy and

371
00:19:55,200 --> 00:19:58,360
diversify national wealth away
from fossil fuels.

372
00:19:58,680 --> 00:20:01,840
These nations are not just
passive investors.

373
00:20:02,440 --> 00:20:05,320
They are building their own
facilities, training their own

374
00:20:05,320 --> 00:20:08,760
models, and negotiating
partnerships with the largest AI

375
00:20:08,760 --> 00:20:11,760
companies in the world.
China approaches capital

376
00:20:11,760 --> 00:20:14,400
differently state.
Banks and government guided

377
00:20:14,400 --> 00:20:17,000
funds direct money toward.
Priority sectors.

378
00:20:17,320 --> 00:20:20,880
That means AI infrastructure
projects can receive financing

379
00:20:20,880 --> 00:20:24,000
even when they would not be
viable under purely commercial

380
00:20:24,000 --> 00:20:27,440
conditions this top down.
Approach allows for rapid.

381
00:20:27,440 --> 00:20:31,000
Mobilization, but it also
concentrates decision making

382
00:20:31,360 --> 00:20:33,560
when the government.
Decides to push hard in a

383
00:20:33,560 --> 00:20:35,800
specific direction.
Resources flood.

384
00:20:35,800 --> 00:20:39,560
In when it changes priorities,
projects can stall overnight.

385
00:20:40,160 --> 00:20:43,600
History shows that the side with
the deeper and more resilient

386
00:20:43,600 --> 00:20:47,080
capital base usually wins long
term infrastructure races.

387
00:20:47,440 --> 00:20:51,040
In the space race, the United
States outspent the Soviet Union

388
00:20:51,040 --> 00:20:54,360
by a wide margin and that
financial weight was a decisive

389
00:20:54,360 --> 00:20:57,160
factor.
In the Cold War arms race, the

390
00:20:57,160 --> 00:21:01,400
same pattern held Capital does
not guarantee victory, but lack

391
00:21:01,400 --> 00:21:05,640
of it almost guarantees failure.
Private capital is also a key

392
00:21:05,640 --> 00:21:08,480
force.
Venture capital, private equity

393
00:21:08,480 --> 00:21:12,000
and infrastructure funds are
looking for ways to participate

394
00:21:12,000 --> 00:21:15,160
in the AI build out.
In some cases, they are

395
00:21:15,160 --> 00:21:18,200
financing smaller data centers
in emerging markets.

396
00:21:18,800 --> 00:21:22,360
In others, they are taking
equity stakes in chip startups,

397
00:21:22,400 --> 00:21:26,280
energy projects, or specialized
AI service providers.

398
00:21:26,720 --> 00:21:30,240
These smaller plays may not grab
headlines, but they can generate

399
00:21:30,240 --> 00:21:33,600
high returns if they align with
larger ecosystem needs.

400
00:21:33,840 --> 00:21:35,840
The competition for capital is
already.

401
00:21:35,840 --> 00:21:39,080
Shaping the Landscape projects
with strong backers.

402
00:21:39,080 --> 00:21:42,600
Can secure priority access to
scarce resources like Advanced?

403
00:21:42,600 --> 00:21:45,680
Chips or prime land?
Those without may struggle.

404
00:21:45,680 --> 00:21:49,200
To get off the ground at all,
and because AI infrastructure

405
00:21:49,200 --> 00:21:52,200
requires ongoing investment for
maintenance, upgrades and

406
00:21:52,200 --> 00:21:54,920
expansion.
Winning the capital game is not

407
00:21:54,920 --> 00:21:57,840
a one time achievement, it is a
continuous.

408
00:21:57,840 --> 00:22:00,160
Process.
The final piece is the

409
00:22:00,160 --> 00:22:02,480
intersection of capital and
policy.

410
00:22:02,800 --> 00:22:06,280
Governments are beginning to
attach strings to funding, using

411
00:22:06,280 --> 00:22:09,440
it to influence where facilities
are built, what technologies

412
00:22:09,440 --> 00:22:11,400
they use, and who can access
them.

413
00:22:11,680 --> 00:22:15,280
The race for AI infrastructure
is not just about building

414
00:22:15,280 --> 00:22:18,760
capacity, it is about shaping
the rules of the access for

415
00:22:18,760 --> 00:22:20,800
decades to come.
And with that.

416
00:22:20,800 --> 00:22:22,600
We have covered all four.
Choke.

417
00:22:22,600 --> 00:22:26,000
Points compute energy, land and.
Capital, but the race.

418
00:22:26,000 --> 00:22:28,960
Does not end there.
The next phase is about how

419
00:22:28,960 --> 00:22:31,080
these resources.
Are being combined into.

420
00:22:31,080 --> 00:22:33,440
Full scale.
Strategies by nations and

421
00:22:33,440 --> 00:22:35,880
corporations.
In the next segment, we will

422
00:22:35,880 --> 00:22:37,640
look at how these strategies
play.

423
00:22:37,640 --> 00:22:40,160
Out on the global map and which?
Players are already.

424
00:22:40,160 --> 00:22:42,640
Pulling ahead.
We have covered the four choke

425
00:22:42,640 --> 00:22:48,040
points of AI, infrastructure,
compute, energy, land, capital.

426
00:22:48,440 --> 00:22:51,680
Now it is time to step back and
see how they fit together on the

427
00:22:51,680 --> 00:22:54,440
global map.
Because while each resource is

428
00:22:54,440 --> 00:22:58,360
important on its own, the real
advantage comes from integrating

429
00:22:58,360 --> 00:23:02,320
them into a coherent strategy.
The players who can align all

430
00:23:02,320 --> 00:23:05,240
four will not just compete, they
will dominate.

431
00:23:05,640 --> 00:23:08,960
The United States is still in
the strongest position overall.

432
00:23:09,240 --> 00:23:12,560
It has world leading compute
suppliers, deep capital.

433
00:23:12,560 --> 00:23:15,880
Markets abundant energy and vast
tracts of land.

434
00:23:16,360 --> 00:23:18,920
It also benefits.
From strong alliances with

435
00:23:18,920 --> 00:23:21,720
countries like Taiwan, Japan and
South.

436
00:23:21,720 --> 00:23:25,000
Korea These relationships help
secure chip.

437
00:23:25,000 --> 00:23:27,640
Supply and share the burden of
manufacturing.

438
00:23:28,040 --> 00:23:30,840
The main risk for the US is
political gridlock.

439
00:23:30,840 --> 00:23:34,640
Slowing down long term
investments or over reliance on

440
00:23:34,640 --> 00:23:37,200
a small number of.
Foreign suppliers for critical.

441
00:23:37,200 --> 00:23:39,720
Components.
China's approach is different.

442
00:23:40,000 --> 00:23:43,640
It does not yet have full access
to the most advanced compute,

443
00:23:43,800 --> 00:23:47,480
but it is working to close the
gap with domestic manufacturing.

444
00:23:48,040 --> 00:23:52,080
It has significant control over
its own energy supply and land

445
00:23:52,080 --> 00:23:56,160
allocation, and because capital
can be directed by the state, it

446
00:23:56,160 --> 00:23:58,520
can move quickly on priority
projects.

447
00:23:58,800 --> 00:24:01,960
The risk is that technology
bands and export controls could

448
00:24:01,960 --> 00:24:06,360
keep it a generation behind on
the highest end chips, forcing

449
00:24:06,360 --> 00:24:08,600
it to rely on creative
workarounds.

450
00:24:09,040 --> 00:24:11,600
The European Union is still
finding its footing.

451
00:24:11,840 --> 00:24:14,240
It has.
Strong research talent and solid

452
00:24:14,240 --> 00:24:16,520
capital in certain.
Countries but lacks.

453
00:24:16,520 --> 00:24:19,120
Both leading edge compute and
cheap energy.

454
00:24:19,440 --> 00:24:21,640
Efforts.
Like the European CHIPS, ACT and

455
00:24:21,640 --> 00:24:24,400
cross-border energy integration
are designed.

456
00:24:24,400 --> 00:24:27,200
To close those gaps.
The challenge is.

457
00:24:27,200 --> 00:24:30,000
Speed.
Building fabs, upgrading grids

458
00:24:30,000 --> 00:24:33,440
and aligning 27 member states
takes time and.

459
00:24:33,640 --> 00:24:38,120
In the AI race, speed is often
the deciding factor if the Gulf

460
00:24:38,120 --> 00:24:40,760
states are emerging as strategic
wildcards.

461
00:24:41,080 --> 00:24:43,960
Their sovereign wealth gives
them deep capital reserves.

462
00:24:44,400 --> 00:24:48,760
Their energy position is strong
both in fossil fuels and in new

463
00:24:48,760 --> 00:24:52,440
nuclear and renewable projects.
What they lack in domestic

464
00:24:52,440 --> 00:24:55,520
compute manufacturing they are
making up for by attracting

465
00:24:55,520 --> 00:24:57,680
partnerships with leading AI
companies.

466
00:24:58,360 --> 00:25:01,560
History suggests that those who
can insert themselves into the

467
00:25:01,560 --> 00:25:05,320
supply chain at a critical point
can remain relevant for decades.

468
00:25:06,120 --> 00:25:08,160
India is also moving up the
ranks.

469
00:25:08,600 --> 00:25:12,720
It has abundant solar potential,
a growing nuclear base, and a

470
00:25:12,720 --> 00:25:14,680
massive pool of technical
talent.

471
00:25:15,040 --> 00:25:17,880
While it's still imports most of
its advanced compute, it is

472
00:25:17,880 --> 00:25:21,480
positioning itself as a low
cost, high growth market for AI

473
00:25:21,480 --> 00:25:24,600
infrastructure.
With the right alliances, India

474
00:25:24,600 --> 00:25:27,560
could become a counterweight to
both China and the West in the

475
00:25:27,560 --> 00:25:29,800
years ahead.
When you look at the map, a

476
00:25:29,800 --> 00:25:33,120
clear pattern emerges.
Control is not evenly

477
00:25:33,120 --> 00:25:36,600
distributed a few nations and
corporate alliances.

478
00:25:36,640 --> 00:25:38,560
Hold the majority of the
advantage.

479
00:25:39,000 --> 00:25:41,920
Everyone else is deciding
whether to align with them,

480
00:25:41,920 --> 00:25:44,920
compete against them, or carve
out niche positions.

481
00:25:45,240 --> 00:25:49,160
And because AI infrastructure.
Takes years to build the choices

482
00:25:49,160 --> 00:25:51,320
being made.
Now will lock in the competitive

483
00:25:51,320 --> 00:25:54,720
order for the next decade.
This is why we built this

484
00:25:54,720 --> 00:25:57,880
episode around choke points.
Understanding where the

485
00:25:57,880 --> 00:26:01,840
bottlenecks are and who controls
them is the fastest way to see

486
00:26:01,840 --> 00:26:04,760
who is ahead.
And for investors, policy makers

487
00:26:04,760 --> 00:26:08,320
and innovators, it is the best
way to spot opportunities before

488
00:26:08,320 --> 00:26:11,120
they become obvious.
Which brings us to our business

489
00:26:11,120 --> 00:26:14,280
idea for this episode.
Because the same principles

490
00:26:14,280 --> 00:26:17,120
shaping the.
AI infrastructure race can also

491
00:26:17,120 --> 00:26:20,000
be applied at a smaller.
Scale to create value.

492
00:26:20,000 --> 00:26:22,640
In the open market.
We have talked about this as a

493
00:26:22,640 --> 00:26:26,640
contest between nations, but the
same choke points exist at

494
00:26:26,640 --> 00:26:30,440
smaller scales.
A city, a region, or even a

495
00:26:30,440 --> 00:26:33,160
private company can run into the
same barriers.

496
00:26:33,160 --> 00:26:37,520
Not enough compute, Unreliable
energy, the wrong location or no

497
00:26:37,520 --> 00:26:39,960
capital?
That is where the opportunity.

498
00:26:39,960 --> 00:26:42,280
Sits.
If you can map where these 4

499
00:26:42,280 --> 00:26:44,960
constraints are holding projects
back, you can connect those

500
00:26:44,960 --> 00:26:46,320
projects.
With the resources.

501
00:26:46,320 --> 00:26:47,920
They need think of.
It as a.

502
00:26:47,920 --> 00:26:52,440
Brokerage and development model.
You identify underused capacity

503
00:26:52,440 --> 00:26:56,520
or underpriced assets, match
them with demand and build value

504
00:26:56,520 --> 00:26:58,960
on top.
This is not theory.

505
00:26:59,320 --> 00:27:02,960
The same logic drives parts of
the data center, energy and

506
00:27:02,960 --> 00:27:06,440
logistics industries today.
The only difference is that AI

507
00:27:06,440 --> 00:27:09,840
infrastructure is growing so
fast and in such concentrated

508
00:27:09,840 --> 00:27:12,200
patterns that gaps are easier to
spot.

509
00:27:12,680 --> 00:27:16,040
A deal that might take a year to
structure and another sector

510
00:27:16,200 --> 00:27:17,960
could come together in months
here.

511
00:27:18,760 --> 00:27:21,760
If you are an investor, a
developer, or a technical

512
00:27:21,760 --> 00:27:24,960
founder, the skill is not just
building in one of these

513
00:27:24,960 --> 00:27:28,680
domains, it is knowing how to
navigate all four.

514
00:27:29,280 --> 00:27:32,440
The winners will be those who
can secure compute energy, land

515
00:27:32,480 --> 00:27:35,520
and capital for multiple
projects at once and move them

516
00:27:35,520 --> 00:27:38,120
forward in parallel.
This is the same.

517
00:27:38,120 --> 00:27:41,600
Pattern we have seen with early
movers in other technology booms

518
00:27:41,920 --> 00:27:44,560
those who understood the
bottlenecks and could remove.

519
00:27:44,560 --> 00:27:47,040
Them for others captured value
far.

520
00:27:47,040 --> 00:27:51,400
Beyond their own footprint in AI
infrastructure, That window is

521
00:27:51,400 --> 00:27:54,640
open now, but it will not stay
open forever.

522
00:27:54,920 --> 00:27:59,240
Let us bring it all together.
The AI infrastructure arms race

523
00:27:59,240 --> 00:28:03,200
is built on 4 connected battles.
Compute, energy, land and

524
00:28:03,200 --> 00:28:05,640
capital.
Each one can accelerate a

525
00:28:05,640 --> 00:28:09,000
project or stop it cold.
The leaders are those who can

526
00:28:09,000 --> 00:28:12,480
control all four and keep them
moving In Sync faster than

527
00:28:12,480 --> 00:28:14,600
anyone else.
In this episode.

528
00:28:14,600 --> 00:28:17,840
We covered why compute is
concentrated in a few hands.

529
00:28:17,920 --> 00:28:20,800
How energy access is.
Becoming a competitive weapon.

530
00:28:20,800 --> 00:28:23,560
Why prime land is far?
Scarcer than it looks.

531
00:28:23,560 --> 00:28:25,680
And how capital is both a
gatekeeper?

532
00:28:25,680 --> 00:28:28,760
And a growth engine.
We explored strategies from the

533
00:28:28,760 --> 00:28:31,040
United.
States, China, the European

534
00:28:31,040 --> 00:28:34,280
Union, the Gulf States and India
and we showed.

535
00:28:34,280 --> 00:28:36,920
How the same?
Logic works on a smaller scale

536
00:28:36,920 --> 00:28:38,800
through a brokerage and
development model.

537
00:28:39,320 --> 00:28:42,880
The stakes are clear.
The AI infrastructure race will

538
00:28:42,880 --> 00:28:45,960
shape the balance of economic
and technological power for the

539
00:28:45,960 --> 00:28:48,920
next decade.
The decisions made now will lock

540
00:28:48,920 --> 00:28:52,080
in advantages and disadvantages
for years to come.

541
00:28:52,560 --> 00:28:55,440
If you know where the leverage
points are, you can see the next

542
00:28:55,440 --> 00:28:57,800
moves before they happen and be
ready.

543
00:28:58,640 --> 00:29:01,160
If you enjoyed this
conversation, be part of our

544
00:29:01,160 --> 00:29:05,840
first 10,000 listeners,
subscribe to AI Frontier AI on

545
00:29:05,840 --> 00:29:09,920
Spotify or Apple Podcasts,
follow us on X for the most

546
00:29:09,920 --> 00:29:14,000
important AI shifts, and share
this episode with someone who

547
00:29:14,000 --> 00:29:17,040
thinks ahead.
This podcast is part of Finance

548
00:29:17,040 --> 00:29:20,240
Frontier AI.
We run four different series,

549
00:29:20,480 --> 00:29:24,360
AI, Frontier AI, Finance,
Frontier Mindset, Frontier AI

550
00:29:24,360 --> 00:29:27,680
and Make Money.
If your company or idea fits one

551
00:29:27,680 --> 00:29:29,560
of our themes, visit our pitch
page.

552
00:29:29,880 --> 00:29:32,080
You might qualify for a free
spotlight.

553
00:29:32,440 --> 00:29:35,880
If you want to go deeper, listen
to the AI Infrastructure Road

554
00:29:35,880 --> 00:29:40,280
Map 2025 to 2027.
It shows how compute.

555
00:29:40,320 --> 00:29:43,640
Energy, land and.
Capital will shape the global AI

556
00:29:43,640 --> 00:29:45,240
build out.
Over the next three.

557
00:29:45,240 --> 00:29:48,720
Years, quarter by quarter.
Then check out the Interface

558
00:29:48,720 --> 00:29:51,720
Wars.
How ChatGPT, Lama, four, and

559
00:29:51,720 --> 00:29:53,720
Quinn, three, are rewiring the
Internet.

560
00:29:54,200 --> 00:29:57,720
It is a deep dive into the
platforms, business models and

561
00:29:57,720 --> 00:30:00,680
ecosystem shift that could
change how billions of people

562
00:30:00,680 --> 00:30:04,560
interact online.
You can also sign up for the 10X

563
00:30:04,560 --> 00:30:07,240
Edge.
It is our weekly drop of real AI

564
00:30:07,240 --> 00:30:09,400
use.
Cases, smart model moves and

565
00:30:09,400 --> 00:30:12,920
early signals all explained in
plain language.

566
00:30:13,160 --> 00:30:18,120
No hype, no jargon.
Only at financefrontierai.com.

567
00:30:18,480 --> 00:30:19,880
And if you have a.
Story to tell.

568
00:30:19,880 --> 00:30:22,920
Maybe a breakthrough product, an
early signal or a bold?

569
00:30:22,920 --> 00:30:26,800
Thesis Head to our pitch page.
If it is a clear win, win we

570
00:30:26,800 --> 00:30:28,000
will.
Pitch it for free.

571
00:30:28,320 --> 00:30:31,480
This podcast is for educational.
Purposes only.

572
00:30:31,680 --> 00:30:33,720
It is not financial.
Advice Legal.

573
00:30:33,720 --> 00:30:37,920
Advice or development guidance?
Always verify before you act.

574
00:30:38,280 --> 00:30:42,960
The AI landscape changes fast.
Benchmarks shift, models update.

575
00:30:43,240 --> 00:30:47,000
Regulations evolve.
Use this show as your map, but

576
00:30:47,000 --> 00:30:48,960
not your final answer.
Today's.

577
00:30:48,960 --> 00:30:53,560
Intro and outro track is Night
Runner by Audionautics, licensed

578
00:30:53,560 --> 00:30:55,920
under the.
YouTube Audio Library license.

579
00:30:56,240 --> 00:31:01,880
Copyright 2025 Finance Frontier
AI All rights reserved.

580
00:31:02,080 --> 00:31:05,240
Reuse or distribution of this
episode without written

581
00:31:05,240 --> 00:31:08,840
permission is not allowed.
Thanks for listening, we will

582
00:31:08,840 --> 00:31:10,080
see you next time.