1
00:00:00,040 --> 00:00:03,160
Elon Musk told the World 
Economic Forum that chip 

2
00:00:03,160 --> 00:00:06,600
manufacturers will soon produce 
more AI chips than the 

3
00:00:06,600 --> 00:00:10,560
electrical grid can handle. 
He dropped that morning during a

4
00:00:10,560 --> 00:00:14,840
conversation with BlackRock CEO 
Larry Fink in Switzerland, 

5
00:00:14,840 --> 00:00:17,840
stating that electrical power 
has become the fundamental 

6
00:00:17,840 --> 00:00:21,240
bottleneck strangling AI 
deployment worldwide. 

7
00:00:22,000 --> 00:00:24,640
Now, Musk predicted the 
companies will be producing more

8
00:00:24,640 --> 00:00:29,280
chips than we can turn on, 
possibly by the end of 2026. 

9
00:00:29,720 --> 00:00:32,479
He then rattled off a series of 
additional predictions covering 

10
00:00:32,479 --> 00:00:36,680
robo taxis, humanoid robots, 
space travel, artificial 

11
00:00:36,680 --> 00:00:39,560
superintelligence, and even 
human aging. 

12
00:00:39,840 --> 00:00:42,480
His track record on forecasts 
like these includes a Mars 

13
00:00:42,480 --> 00:00:45,880
mission that was supposed to 
launch in 2024 and a Starship 

14
00:00:45,880 --> 00:00:49,040
that was supposed to reach orbit
in 2022, both of which missed 

15
00:00:49,040 --> 00:00:51,880
their deadlines. 
So how seriously should anyone 

16
00:00:51,880 --> 00:00:55,400
take his latest round of 
timeline promises when his 

17
00:00:55,400 --> 00:00:59,520
previous forecasts have 
consistently slipped by years 

18
00:00:59,920 --> 00:01:01,600
now? 
Musk also criticized the Trump 

19
00:01:01,600 --> 00:01:04,239
administration's tariffs on 
solar equipment from Asia, 

20
00:01:04,519 --> 00:01:08,200
calling the tariff barriers 
extremely high and claiming they

21
00:01:08,200 --> 00:01:11,440
make solar deployment 
artificially expensive in the 

22
00:01:11,440 --> 00:01:13,880
United States. 
Now, that statement puts him in 

23
00:01:13,880 --> 00:01:16,400
direct conflict with the 
administration he currently 

24
00:01:16,400 --> 00:01:21,200
advises and will break down 
Musk's power grid warning and 

25
00:01:21,200 --> 00:01:25,160
what it means for AI companies 
racing to deploy new data 

26
00:01:25,160 --> 00:01:28,040
centers across America. 
We'll also examine his 

27
00:01:28,040 --> 00:01:32,720
prediction that AI will surpass 
human intelligence this year and

28
00:01:32,720 --> 00:01:35,720
exceed the collective 
intelligence of humanity by 

29
00:01:35,720 --> 00:01:40,160
2035. 
We will look at his robotaxi and

30
00:01:40,160 --> 00:01:42,400
Optimist robot timelines as 
well. 

31
00:01:42,760 --> 00:01:46,960
And these are against historical
accuracy of these similar 

32
00:01:46,960 --> 00:01:49,000
promises. 
And we'll get right into that 

33
00:01:49,200 --> 00:01:54,120
after this very short break. 
Now, Musk's electrical power 

34
00:01:54,120 --> 00:01:57,160
warning centers on a fundamental
mismatch between silicon 

35
00:01:57,160 --> 00:02:00,320
production capacity and energy 
infrastructure investment. the 

36
00:02:00,320 --> 00:02:03,480
United States has been grappling
with an outdated grid system 

37
00:02:03,480 --> 00:02:06,320
built on decades of 
underinvestment and aging 

38
00:02:06,320 --> 00:02:09,160
infrastructure that was never 
designed for this level of 

39
00:02:09,160 --> 00:02:12,240
concentrated demand. 
Tech companies and utilities 

40
00:02:12,240 --> 00:02:15,600
expected gradual growth in 
electricity consumption, not the

41
00:02:15,600 --> 00:02:19,720
sudden vertical spike that AI 
training clusters require in two

42
00:02:19,720 --> 00:02:23,720
massive data centers in Nvidia's
hometown of Santa Clara, CA Mesa

43
00:02:23,720 --> 00:02:27,280
empty for years waiting for the 
electricity needed to power 

44
00:02:27,280 --> 00:02:30,400
them, according to energy 
experts who have studied the 

45
00:02:30,400 --> 00:02:34,480
regional grid capacity. 
Now, the demand surge from AI 

46
00:02:34,480 --> 00:02:37,920
training and inference workloads
have exposed just how fragile 

47
00:02:37,920 --> 00:02:39,640
the American power supply really
is. 

48
00:02:39,960 --> 00:02:43,200
Tech companies increasingly rely
on grid operators who cannot 

49
00:02:43,200 --> 00:02:46,400
deliver consistent power at the 
scale these massive facilities 

50
00:02:46,400 --> 00:02:48,480
require for continuous 
operation. 

51
00:02:49,000 --> 00:02:52,240
A single modern AI training 
cluster can consume as much 

52
00:02:52,280 --> 00:02:55,480
electricity as a small city, and
dozens of these facilities are 

53
00:02:55,480 --> 00:02:58,320
currently under construction or 
planned across the United 

54
00:02:58,320 --> 00:03:01,400
States. 
In earlier this month, the Trump

55
00:03:01,400 --> 00:03:04,760
administration had 13 bipartisan
governors pressured operators of

56
00:03:04,760 --> 00:03:08,240
PJM Interconnection, the 
country's largest grid serving 

57
00:03:08,240 --> 00:03:11,840
65 million people, to boost 
power supplies and hold an 

58
00:03:11,880 --> 00:03:15,640
auction allowing tech firms to 
bid on 15 year contracts to 

59
00:03:15,640 --> 00:03:18,640
build new power plants. 
Now, that proposal would 

60
00:03:18,640 --> 00:03:21,080
transfer the cost of electricity
infrastructure away from 

61
00:03:21,080 --> 00:03:24,640
residential consumers and under 
the data center operators who 

62
00:03:24,640 --> 00:03:28,520
are driving the demand increase.
Interior Secretary Doug Bergbach

63
00:03:29,000 --> 00:03:32,120
framed the issue as a national 
security threat. 

64
00:03:32,520 --> 00:03:35,480
And they're racing against 
China, seeing the country needs 

65
00:03:35,480 --> 00:03:38,960
to power AI infrastructure to 
compete in a technology contest 

66
00:03:39,160 --> 00:03:42,520
that will transform every job 
and every company in every 

67
00:03:42,520 --> 00:03:44,960
industry. 
Now, China has already solved 

68
00:03:44,960 --> 00:03:48,800
this problem through massive 
solar deployment that dwarfs 

69
00:03:48,840 --> 00:03:50,600
anything the United States has 
attempted. 

70
00:03:50,880 --> 00:03:54,120
According to the Global Energy 
Monitor Solar Tracker, China 

71
00:03:54,120 --> 00:03:56,920
operates nearly four times the 
solar power capacity of the 

72
00:03:56,920 --> 00:04:00,040
United States, with room to grow
even further. 

73
00:04:00,400 --> 00:04:02,600
Including projects under 
development and construction, 

74
00:04:02,880 --> 00:04:09,200
China is expected to reach 1.18 
million megawatts of solar 

75
00:04:09,200 --> 00:04:14,960
output compared to America's 
237. 1947 megawatts pales in 

76
00:04:14,960 --> 00:04:17,320
comparison. 
Musk claimed the entire United 

77
00:04:17,320 --> 00:04:21,399
States could run out of 100 mile
by 100 mile square of solar 

78
00:04:21,399 --> 00:04:24,280
fields, which would represent a 
fraction of the available land 

79
00:04:24,280 --> 00:04:27,880
in states like Nevada, Arizona 
or even New Mexico. 

80
00:04:28,160 --> 00:04:31,000
He called solar by far the 
biggest source of energy on a 

81
00:04:31,000 --> 00:04:34,000
global scale and pointed to 
China's tremendous growth and 

82
00:04:34,000 --> 00:04:37,840
electrical capabilities as 
evidence the technology works 

83
00:04:37,840 --> 00:04:42,320
reliably when deployed at scale.
Now, the contradiction in Musk's

84
00:04:42,320 --> 00:04:45,760
position becomes obvious when 
you examine recent American 

85
00:04:45,760 --> 00:04:48,360
trade policy and his role in the
current administration. 

86
00:04:48,800 --> 00:04:52,200
Tariffs on solar equipment from 
Asia took effect last May, with 

87
00:04:52,200 --> 00:04:57,000
import taxes reaching as high as
3500% on countries like Cambodia

88
00:04:57,000 --> 00:04:59,640
following an International Trade
Commission ruling. 

89
00:05:00,040 --> 00:05:02,800
Now, the Commission determined 
that imports from Malaysia, 

90
00:05:02,800 --> 00:05:06,480
Thailand, Vietnam and Cambodia 
were detrimental to American 

91
00:05:06,480 --> 00:05:12,000
solar manufacturers and they 
could not compete on a price a 

92
00:05:12,000 --> 00:05:13,640
must. 
Called those terrier barriffs 

93
00:05:13,840 --> 00:05:17,000
extremely high and blame them 
for making solar economics 

94
00:05:17,000 --> 00:05:20,080
artificially expensive in the 
United States of America. 

95
00:05:20,680 --> 00:05:24,120
He said this while serving as a 
senior advisor to Trump. 

96
00:05:24,720 --> 00:05:29,280
And that's when they implemented
these exact policies. 

97
00:05:29,680 --> 00:05:32,440
And he will continue to defend 
those people. 

98
00:05:32,800 --> 00:05:36,000
Now, the Trump administration 
has actively opposed pivoting to

99
00:05:36,000 --> 00:05:40,080
solar and strip subsidies for 
renewable sources, claiming they

100
00:05:40,080 --> 00:05:43,320
compromise your electric grid, 
despite Musk's public statements

101
00:05:43,320 --> 00:05:46,640
suggesting the opposite 
conclusion that Trump offered a 

102
00:05:46,680 --> 00:05:50,120
different solution During his 
own remarks at Davos, he 

103
00:05:50,120 --> 00:05:52,840
encouraged tech companies to 
build their own nuclear plants 

104
00:05:53,080 --> 00:05:55,600
and claimed his administration 
would approve them in just three

105
00:05:55,600 --> 00:05:57,640
weeks. 
Nuclear plant approvals 

106
00:05:57,640 --> 00:06:01,440
historically take years, maybe 
decades of environmental review,

107
00:06:01,680 --> 00:06:04,520
safety certification and 
construction permitting before a

108
00:06:04,520 --> 00:06:06,400
single foundation is even 
poured. 

109
00:06:06,640 --> 00:06:10,600
And the Nuclear Regulatory 
Commission process also takes 

110
00:06:11,320 --> 00:06:14,920
probably 42 to 60 months, which 
is required. 

111
00:06:15,120 --> 00:06:17,520
And that timeline assumes no 
challenges or design 

112
00:06:17,520 --> 00:06:19,880
modifications arise during this 
review. 

113
00:06:19,880 --> 00:06:24,760
It could take up to 8090 months 
if it's all said and done now. 

114
00:06:24,760 --> 00:06:27,720
Trump's three-week claim appears
entirely disconnected from the 

115
00:06:27,720 --> 00:06:30,920
regulatory and construction 
realities from nuclear energy 

116
00:06:30,920 --> 00:06:33,640
deployment. 
Now, Musk did not endure it's 

117
00:06:33,640 --> 00:06:36,280
this timeline during his own 
remarks and instead continues 

118
00:06:36,280 --> 00:06:40,440
emphasizing solar as the fastest
path to additional power 

119
00:06:40,440 --> 00:06:42,440
capacity. 
Now, I've been digging through 

120
00:06:42,440 --> 00:06:45,000
our analytics here and 37% of 
you have been following the 

121
00:06:45,000 --> 00:06:46,680
channel for you. 
I love you. 

122
00:06:46,720 --> 00:06:49,560
I'm forever grateful for the 
other 63% we haven't hit the 

123
00:06:49,560 --> 00:06:51,200
follow or subscribe button. 
I get it. 

124
00:06:51,400 --> 00:06:53,600
You listen to a podcast a bunch 
of times, you forget to hit the 

125
00:06:53,600 --> 00:06:55,040
subscribe button or the follow 
button. 

126
00:06:55,480 --> 00:06:57,600
I've done it too. 
I've been doing this for about 6

127
00:06:57,600 --> 00:06:59,080
years, going to do it for the 
next 10. 

128
00:06:59,360 --> 00:07:02,000
All I ask from you is one second
your time to hit that follow 

129
00:07:02,000 --> 00:07:03,600
button. 
Thank you very much. 

130
00:07:04,160 --> 00:07:06,800
Now Musk's AI timeline 
predictions drew the most 

131
00:07:06,800 --> 00:07:10,000
attention and skepticism from 
observers who have watched his 

132
00:07:10,000 --> 00:07:13,200
forecast slip repeatedly. 
He claimed AI will become 

133
00:07:13,200 --> 00:07:17,040
smarter than any individual 
human this year or no later the 

134
00:07:17,040 --> 00:07:19,600
next year if development hits 
unexpected obstacles. 

135
00:07:20,040 --> 00:07:22,440
He projected that AI will 
surpass the collective 

136
00:07:22,440 --> 00:07:27,040
intelligence of humanity by 
2035, creating what he had 

137
00:07:27,040 --> 00:07:30,040
previously described as a 
potential existential risk to 

138
00:07:30,040 --> 00:07:33,840
human civilization. 
Now the Nvidia's CEO offered a 

139
00:07:33,840 --> 00:07:36,360
starkly different view earlier 
this month, stating that 

140
00:07:36,360 --> 00:07:40,360
researchers are nowhere close to
creating what he called God AI, 

141
00:07:40,600 --> 00:07:43,640
capable of general reasoning 
across all domains. 

142
00:07:44,360 --> 00:07:48,080
He said the technology excels at
discrete tasks, but achieving 

143
00:07:48,080 --> 00:07:52,440
general intelligence operates on
biblical scales, on galactic 

144
00:07:52,440 --> 00:07:55,360
scales that current 
architectures cannot even get 

145
00:07:55,360 --> 00:07:57,560
close to. 
The gap between Musk's 

146
00:07:57,560 --> 00:08:01,240
aggressive timeline and his 
skepticism shows just how 

147
00:08:01,240 --> 00:08:04,280
uncertain these forecasts 
remain, even among the experts, 

148
00:08:04,280 --> 00:08:07,200
the people that are building the
underlying hardware and software

149
00:08:07,200 --> 00:08:10,800
for these things and Musk's robo
taxi predictions and followed a 

150
00:08:10,800 --> 00:08:14,080
familiar path of deadline 
extensions that stretch back 

151
00:08:14,080 --> 00:08:16,840
nearly a decade. 
Tesla finally launched a limited

152
00:08:16,840 --> 00:08:20,640
robo taxi service in Austin, TX 
in 2025, but each vehicle still 

153
00:08:20,640 --> 00:08:24,160
carries a human is sitting in 
the passenger seat, ready to 

154
00:08:24,160 --> 00:08:26,400
take control. 
And that launch came after years

155
00:08:26,400 --> 00:08:29,040
of Musk promising fully 
autonomous vehicles were 

156
00:08:29,040 --> 00:08:33,320
imminent, including 2019 claim 
that Tesla would have 1,000,000 

157
00:08:33,320 --> 00:08:37,520
robo taxis on the road by 2020. 
Now with Davos, Musk moved the 

158
00:08:37,520 --> 00:08:40,039
goal posts again. 
Let me Tesla's Robotech without 

159
00:08:40,039 --> 00:08:43,200
human Monitors company is 
currently working to expand 

160
00:08:43,200 --> 00:08:46,600
service to states with more 
permissive autonomous vehicle 

161
00:08:46,600 --> 00:08:50,520
regulations, including Arizona, 
Florida and Nevada, where 

162
00:08:50,520 --> 00:08:54,440
regulators have created 
friendlier frameworks for 

163
00:08:54,440 --> 00:08:56,600
testing. 
Whether very widespread 

164
00:08:56,600 --> 00:09:00,200
materializes by December remains
an open question, given the 

165
00:09:00,200 --> 00:09:03,520
consistent pattern of delays on 
this specific promise. 

166
00:09:04,000 --> 00:09:06,560
Let's talk about Optimus 
humanoid robot next. 

167
00:09:07,280 --> 00:09:09,520
It's a similar treatment by 
Musk. 

168
00:09:09,520 --> 00:09:14,720
He first stated that there were 
promises of Optimus in 2021, 

169
00:09:14,720 --> 00:09:18,120
initially showcasing basically a
person in a robot suit before 

170
00:09:18,120 --> 00:09:20,200
revealing an actual prototype 
hardware. 

171
00:09:20,520 --> 00:09:24,040
He recently said Tesla would 
manufacture thousands of Optimus

172
00:09:24,040 --> 00:09:28,000
units in 2025 for eternal 
factory use before selling them 

173
00:09:28,000 --> 00:09:30,600
externally. 
Reports indicate the company 

174
00:09:30,600 --> 00:09:33,440
still struggles to make 
Optimus's hands work properly 

175
00:09:33,640 --> 00:09:36,720
for the fine manipulation tasks 
required in manufacturing 

176
00:09:36,720 --> 00:09:39,680
environments. 
Musk predicted the billions of 

177
00:09:39,760 --> 00:09:42,640
AI powered robots would 
eventually outnumber humans and 

178
00:09:42,640 --> 00:09:45,760
saturate all human needs to the 
point where a scarcity 

179
00:09:45,760 --> 00:09:48,560
effectively disappears. 
He said consumers would reach a 

180
00:09:48,560 --> 00:09:51,680
point where they could not think
of anything left to ask a robot 

181
00:09:51,680 --> 00:09:54,800
for because goods and services 
would be so abundant. 

182
00:09:55,920 --> 00:09:57,840
Never has to have. 
They never have to go to work. 

183
00:09:57,960 --> 00:10:01,040
Everything works for you. 
He also announced Optimist would

184
00:10:01,040 --> 00:10:04,800
go on sale to external customers
late next year, adding another 

185
00:10:04,800 --> 00:10:07,320
specific deadline to his 
collection of promises. 

186
00:10:07,920 --> 00:10:10,440
Now SpaceX predictions follow 
the same trajectory of 

187
00:10:10,440 --> 00:10:12,720
optimistic timelines, meaning 
engineering reality. 

188
00:10:13,000 --> 00:10:16,840
Musk promised in 2020 in that a 
crude mission to Mars would 

189
00:10:16,840 --> 00:10:20,160
launch by 2024. 
That deadline came and went 

190
00:10:20,320 --> 00:10:22,160
without a mission. 
He said Starship would reach 

191
00:10:22,160 --> 00:10:25,080
orbit by 2022, though the 
company did not achieve that 

192
00:10:25,080 --> 00:10:28,200
milestone until last year after 
multiple test failures and 

193
00:10:28,200 --> 00:10:29,840
redesigns. 
They didn't get a full orbit, 

194
00:10:29,840 --> 00:10:32,800
though, that Davos. 
Musk repeated his promise that 

195
00:10:32,800 --> 00:10:35,680
Starship would become fully 
reusable by the end of this 

196
00:10:35,680 --> 00:10:38,800
year, with both stages returning
intact for reflay. 

197
00:10:39,720 --> 00:10:44,320
He claimed full reusability 
would cut space cost travel by a

198
00:10:44,320 --> 00:10:47,520
factor of 100 and eventually 
make space freight competitive 

199
00:10:47,760 --> 00:10:52,160
with airplane freight prices on 
a per pound basis. 

200
00:10:52,560 --> 00:10:55,920
Now Musk concluded his 
appearance with something 

201
00:10:55,920 --> 00:10:58,960
resembling self-awareness about 
his forecasting habits and the 

202
00:10:58,960 --> 00:11:02,080
optimism bias that drives them. 
He offered what sounded like a 

203
00:11:02,080 --> 00:11:04,880
personal mission statement, 
saying it is better for quality 

204
00:11:04,880 --> 00:11:08,760
of life to err on the side of 
being an optimist and wrong than

205
00:11:08,760 --> 00:11:10,320
a pessimist and right.