1
00:00:00,347 --> 00:00:01,292
Sequoia Capital says

2
00:00:01,667 --> 00:00:03,712
your skills are turning into aluminum

3
00:00:04,267 --> 00:00:05,952
In late April 2026

4
00:00:06,107 --> 00:00:10,631
Sequoia the king of global venture capital held a closed door summit in San Francisco

5
00:00:10,827 --> 00:00:13,791
open only to top AI founders and investors

6
00:00:14,186 --> 00:00:17,032
OpenAI president Greg Brockman was there

7
00:00:17,146 --> 00:00:19,231
AI legend Andrej Karpathy was there too

8
00:00:19,387 --> 00:00:21,711
but the biggest bombshell did not come from them

9
00:00:21,827 --> 00:00:25,152
it came from the opening talk by three Sequoia partners

10
00:00:25,867 --> 00:00:28,371
Here they are first the investor behind Snowflake

11
00:00:28,827 --> 00:00:31,132
Zoom and Notion Pat Grady

12
00:00:31,586 --> 00:00:34,799
then the author of what AI investors worldwide treat as a mental map

13
00:00:34,799 --> 00:00:38,171
the Sequoia generative AI series Sonya Huang

14
00:00:38,707 --> 00:00:40,851
and Konstantine Buhler who backed XBOW Xaira

15
00:00:41,067 --> 00:00:44,631
Waymo and other frontier AI companies

16
00:00:45,551 --> 00:00:45,856
Hello

17
00:00:46,350 --> 00:00:46,856
hi everyone

18
00:00:47,151 --> 00:00:48,155
I am Leo Wang

19
00:00:48,591 --> 00:00:53,356
Today I want to talk about the one story in that Sequoia talk that chilled me the most

20
00:00:53,750 --> 00:00:55,115
not the technical details

21
00:00:55,231 --> 00:00:57,235
but a story about aluminum

22
00:00:57,790 --> 00:01:00,189
This true piece of history from a hundred and forty years ago

23
00:01:00,191 --> 00:01:01,396
is frighteningly precise

24
00:01:01,551 --> 00:01:04,396
it predicts what happens to every one of us in the AI age

25
00:01:04,983 --> 00:01:06,049
Before I tell that story

26
00:01:06,263 --> 00:01:07,334
we need to be clear on one thing

27
00:01:07,344 --> 00:01:10,648
what exactly did Sequoia announce in this talk

28
00:01:10,983 --> 00:01:13,308
In one line AGI is here

29
00:01:13,663 --> 00:01:14,329
You heard that right

30
00:01:14,704 --> 00:01:16,582
this is not clickbait from some influencer

31
00:01:16,584 --> 00:01:18,629
and it is not an AI company hyping itself

32
00:01:18,864 --> 00:01:21,248
this is the firm behind Apple Google Nvidia

33
00:01:21,304 --> 00:01:24,429
and Airbnb the top venture firm in the world Sequoia Capital

34
00:01:24,463 --> 00:01:26,788
putting that statement on the table officially

35
00:01:26,944 --> 00:01:28,269
That is bombshell number one

36
00:01:28,999 --> 00:01:30,924
Pat Grady said on stage

37
00:01:30,999 --> 00:01:34,564
the AI apps of the past few years were like faster horses

38
00:01:34,598 --> 00:01:37,077
ten to forty percent more efficient

39
00:01:37,079 --> 00:01:37,424
nice

40
00:01:37,799 --> 00:01:39,803
but you are still riding a horse

41
00:01:39,999 --> 00:01:40,943
Now it is different

42
00:01:41,199 --> 00:01:42,243
the car has arrived

43
00:01:42,398 --> 00:01:46,083
the new generation of AI apps makes you ten to forty times more productive

44
00:01:46,158 --> 00:01:47,437
what changes is not the speed

45
00:01:47,439 --> 00:01:49,044
it is the nature of the work itself

46
00:01:49,318 --> 00:01:50,883
Then he shifted

47
00:01:50,999 --> 00:01:52,624
and said something that made the whole room hold its breath

48
00:01:53,079 --> 00:01:55,663
let us be bold here we believe

49
00:01:55,958 --> 00:01:56,943
this is AGI

50
00:01:58,119 --> 00:01:58,464
Note this

51
00:01:58,759 --> 00:02:00,843
Sequoia is not using the academic definition

52
00:02:01,119 --> 00:02:03,964
Pat put it plainly I studied economics

53
00:02:04,039 --> 00:02:05,518
we are investors

54
00:02:05,518 --> 00:02:08,083
we are not here to write a technical definition of AGI

55
00:02:08,279 --> 00:02:09,318
what we look at is this

56
00:02:09,318 --> 00:02:12,383
can you send an AI off to finish a complex task

57
00:02:12,878 --> 00:02:15,024
when it hits a wall can it figure something out

58
00:02:15,478 --> 00:02:16,904
when it fails can it recover on its own

59
00:02:17,359 --> 00:02:18,904
can it keep going until the job is done

60
00:02:19,359 --> 00:02:19,864
If it can

61
00:02:20,198 --> 00:02:22,063
then to me that is AGI

62
00:02:23,155 --> 00:02:25,075
The scary part of that definition is

63
00:02:25,076 --> 00:02:26,481
it already happened

64
00:02:26,636 --> 00:02:28,421
Claude Code today

65
00:02:28,636 --> 00:02:30,041
OpenAI Codex

66
00:02:30,116 --> 00:02:31,840
and all the agent frameworks out there

67
00:02:31,995 --> 00:02:34,921
you really can send them off to do something complex

68
00:02:34,956 --> 00:02:35,981
then go grab a coffee

69
00:02:36,315 --> 00:02:38,041
and come back to finished work

70
00:02:38,196 --> 00:02:39,101
This is not science fiction

71
00:02:39,435 --> 00:02:41,120
this is what I use every day

72
00:02:41,476 --> 00:02:42,994
On my own YouTube channel

73
00:02:42,995 --> 00:02:46,641
from the script and the images to the voice and the edit

74
00:02:46,755 --> 00:02:49,201
almost all of it is done by AI

75
00:02:49,435 --> 00:02:52,880
Six months ago I still had to watch every step by hand

76
00:02:53,035 --> 00:02:54,714
now I just point a direction

77
00:02:54,715 --> 00:02:56,921
and the agents run the whole thing themselves

78
00:02:57,676 --> 00:02:59,541
This change is not gradual

79
00:02:59,875 --> 00:03:01,201
it is a cliff edge

80
00:03:01,781 --> 00:03:02,446
AGI is here

81
00:03:02,621 --> 00:03:03,207
So what now

82
00:03:03,542 --> 00:03:05,704
Sequoia founder Don Valentine had a famous line

83
00:03:05,714 --> 00:03:06,726
So what

84
00:03:07,302 --> 00:03:08,446
what does all of this actually mean

85
00:03:08,821 --> 00:03:12,707
Pat's answer is driving a car is nothing like riding a horse

86
00:03:12,781 --> 00:03:14,786
you cannot drive a car the way you ride a horse

87
00:03:14,862 --> 00:03:17,107
and you certainly cannot build cars the way you raise horses

88
00:03:17,302 --> 00:03:20,147
if you are still working with horse logic today

89
00:03:20,182 --> 00:03:22,387
you are already on the wrong track

90
00:03:22,873 --> 00:03:23,938
So where is the right one

91
00:03:24,434 --> 00:03:27,119
that brings us to bombshell number two

92
00:03:27,313 --> 00:03:31,438
Pat Grady then dropped a set of numbers that made people gasp

93
00:03:31,714 --> 00:03:32,919
for the past thirty years

94
00:03:32,994 --> 00:03:35,878
all of Silicon Valley the whole golden age of SaaS

95
00:03:35,954 --> 00:03:40,158
every founder was fighting over a six hundred fifty billion dollar software market

96
00:03:40,353 --> 00:03:41,299
that is already huge

97
00:03:41,514 --> 00:03:41,818
right

98
00:03:42,394 --> 00:03:43,239
But he said

99
00:03:43,313 --> 00:03:47,038
the real gold mine in the AI age is the ten trillion dollar services market

100
00:03:47,274 --> 00:03:47,818
Ten trillion

101
00:03:48,114 --> 00:03:49,959
fifteen times the software market

102
00:03:50,761 --> 00:03:53,226
Why could earlier tech revolutions never touch it

103
00:03:53,761 --> 00:03:57,446
because services need judgment reasoning and long term follow through

104
00:03:57,522 --> 00:03:59,087
only humans could do that before

105
00:03:59,282 --> 00:04:01,286
but now agents can

106
00:04:01,522 --> 00:04:03,946
Lawyers reviewing contracts accountants checking books

107
00:04:04,242 --> 00:04:06,807
doctors reading scans analysts writing reports

108
00:04:06,921 --> 00:04:09,560
this is no longer AI standing beside you making you faster

109
00:04:09,562 --> 00:04:11,566
this is AI doing the job itself

110
00:04:11,761 --> 00:04:13,552
Then he made a comparison that took the air out of the room

111
00:04:13,562 --> 00:04:17,046
legal services in the United States alone

112
00:04:17,082 --> 00:04:18,927
is a four hundred billion dollar industry

113
00:04:19,162 --> 00:04:21,727
one country one vertical

114
00:04:21,761 --> 00:04:24,607
is worth more than half the entire old software market

115
00:04:25,082 --> 00:04:28,928
So Sequoia's core call is that services are the new software

116
00:04:29,082 --> 00:04:30,027
the past sold tools

117
00:04:30,322 --> 00:04:31,608
the future sells outcomes

118
00:04:31,762 --> 00:04:35,488
if you are still building a product that makes people thirty percent faster

119
00:04:35,563 --> 00:04:37,728
you are competing on the wrong dimension

120
00:04:37,923 --> 00:04:38,707
the real race

121
00:04:38,923 --> 00:04:41,248
is AI replacing the service itself

122
00:04:41,853 --> 00:04:45,173
Follow the logic here you used to hire a lawyer

123
00:04:45,173 --> 00:04:47,059
at five hundred dollars an hour

124
00:04:47,093 --> 00:04:50,899
four hundred of that pays for the law stored in his head

125
00:04:51,334 --> 00:04:54,572
but if AI can pull up the same knowledge in three seconds

126
00:04:54,574 --> 00:04:56,059
and more of it and more accurately

127
00:04:56,093 --> 00:04:57,798
why would you still pay by the hour

128
00:04:58,293 --> 00:05:00,293
You are no longer paying for knowledge

129
00:05:00,293 --> 00:05:01,819
you are paying for trust

130
00:05:01,973 --> 00:05:03,218
and trust

131
00:05:03,254 --> 00:05:05,213
once AI is more accurate than humans

132
00:05:05,214 --> 00:05:06,778
gets repriced too

133
00:05:07,160 --> 00:05:08,765
Bombshell number three hits even harder

134
00:05:08,999 --> 00:05:12,705
Sonya Huang put it beautifully everything you can imagine

135
00:05:12,960 --> 00:05:15,484
that would take a hundred years to build

136
00:05:15,599 --> 00:05:18,244
can now be done in a hundred days

137
00:05:18,840 --> 00:05:21,364
A century of progress compressed into a hundred days

138
00:05:21,520 --> 00:05:22,624
it sounds like a motivational poster

139
00:05:23,080 --> 00:05:25,724
but then she listed three real cases

140
00:05:26,461 --> 00:05:30,586
The first is Nathan Sobo founder of the code editor Zed

141
00:05:31,021 --> 00:05:31,925
over a single vacation

142
00:05:32,220 --> 00:05:32,985
by himself

143
00:05:33,061 --> 00:05:37,386
he used AI coding tools to finish a project planned for three years

144
00:05:37,780 --> 00:05:38,566
Three years of work

145
00:05:38,821 --> 00:05:39,246
one person

146
00:05:39,461 --> 00:05:40,826
one vacation

147
00:05:41,181 --> 00:05:43,066
this is not thirty percent faster

148
00:05:43,181 --> 00:05:45,425
this is time itself compressed a hundred times

149
00:05:45,780 --> 00:05:48,785
The second is Bret Taylor founder of Sierra

150
00:05:49,061 --> 00:05:51,626
the former co CEO of Salesforce

151
00:05:51,660 --> 00:05:55,025
he rebuilt his company's entire core system over one weekend

152
00:05:55,420 --> 00:05:56,465
just one weekend

153
00:05:57,216 --> 00:05:58,502
The third is Notion

154
00:05:58,697 --> 00:06:02,341
they used AI to rewrite their whole product codebase in six weeks

155
00:06:02,457 --> 00:06:04,062
eight million lines of it

156
00:06:04,377 --> 00:06:05,502
Before AI

157
00:06:05,577 --> 00:06:08,942
a strong engineer wrote fifty to a hundred thousand lines a year

158
00:06:09,216 --> 00:06:10,776
what Notion did in six weeks

159
00:06:10,776 --> 00:06:14,062
equals three to five years of work for a large team

160
00:06:14,377 --> 00:06:17,341
Now stack all these compressed timelines together

161
00:06:17,416 --> 00:06:19,502
and it gets genuinely frightening

162
00:06:19,776 --> 00:06:21,981
if one project is compressed a hundred times

163
00:06:22,057 --> 00:06:24,541
and a team runs ten of them at once

164
00:06:24,617 --> 00:06:27,262
and each one calls a hundred agents

165
00:06:27,377 --> 00:06:29,161
that is not a hundred times faster

166
00:06:29,497 --> 00:06:30,622
it is a hundred thousand times

167
00:06:31,017 --> 00:06:33,015
Now you can feel it

168
00:06:33,017 --> 00:06:34,882
why Sequoia says the car has arrived

169
00:06:35,774 --> 00:06:38,613
But the part of the talk that really sank in me

170
00:06:38,613 --> 00:06:41,058
was the story Konstantine Buhler told

171
00:06:41,334 --> 00:06:42,159
The year is 1884

172
00:06:42,493 --> 00:06:44,698
America had just finished the Washington Monument

173
00:06:44,813 --> 00:06:47,178
the tallest building in the world at the time

174
00:06:47,334 --> 00:06:48,852
to show the highest respect

175
00:06:48,854 --> 00:06:53,258
they decided to cap it with the most precious metal on earth

176
00:06:53,334 --> 00:06:54,659
a hundred ounces of it

177
00:06:54,813 --> 00:06:56,493
The piece was so precious

178
00:06:56,493 --> 00:07:00,292
that before installation it went to Tiffany in Manhattan

179
00:07:00,294 --> 00:07:02,818
and sat in the window on public display for a while

180
00:07:03,013 --> 00:07:04,578
people lined up to see it

181
00:07:04,613 --> 00:07:06,573
some even stepped over it

182
00:07:06,573 --> 00:07:10,118
so they could brag I stepped over the tip of the Washington Monument

183
00:07:11,186 --> 00:07:13,691
By now you are probably guessing gold

184
00:07:14,068 --> 00:07:14,530
platinum

185
00:07:14,866 --> 00:07:15,371
diamond

186
00:07:16,025 --> 00:07:17,431
the answer is aluminum

187
00:07:17,547 --> 00:07:18,347
Yes

188
00:07:18,386 --> 00:07:20,251
the same aluminum you wrap leftovers in

189
00:07:20,546 --> 00:07:23,191
wrap sandwiches with and drink soda from

190
00:07:23,705 --> 00:07:24,811
Back in 1884

191
00:07:25,145 --> 00:07:27,551
aluminum cost about as much as silver

192
00:07:27,946 --> 00:07:31,150
it is one of the most abundant metals in the crust of the earth

193
00:07:31,265 --> 00:07:34,671
but nobody could extract it from ore efficiently

194
00:07:34,866 --> 00:07:37,710
so it stood shoulder to shoulder with silver

195
00:07:37,825 --> 00:07:40,590
precious enough to display in a Tiffany window

196
00:07:41,034 --> 00:07:41,459
And then

197
00:07:41,834 --> 00:07:43,032
in 1886

198
00:07:43,034 --> 00:07:45,238
two years after the monument was finished

199
00:07:45,274 --> 00:07:49,318
a twenty two year old named Charles Martin Hall invented electrolysis

200
00:07:49,514 --> 00:07:53,678
in his lab he dissolved alumina in molten cryolite and ran current through it

201
00:07:53,834 --> 00:07:54,918
two hours later

202
00:07:54,993 --> 00:07:57,358
aluminum separated out of the ore

203
00:07:57,993 --> 00:08:00,839
That invention rewrote the fate of aluminum

204
00:08:01,233 --> 00:08:02,199
within a few decades

205
00:08:02,274 --> 00:08:06,512
aluminum went from a treasure in Tiffany's window to candy wrapper material

206
00:08:06,514 --> 00:08:09,318
crumpled up and thrown in the trash

207
00:08:10,179 --> 00:08:12,105
After telling that story Konstantine

208
00:08:12,139 --> 00:08:14,264
said something that ran a chill down my spine

209
00:08:14,740 --> 00:08:14,925
aluminum

210
00:08:15,340 --> 00:08:16,304
is intelligence

211
00:08:16,699 --> 00:08:17,285
electrolysis

212
00:08:17,660 --> 00:08:18,624
is AI

213
00:08:19,019 --> 00:08:22,324
We are entering a world where skills that took ten

214
00:08:22,540 --> 00:08:25,064
or twenty years to reach a PhD level

215
00:08:25,139 --> 00:08:27,585
can be summoned in an instant

216
00:08:27,660 --> 00:08:30,225
and crumpled up and tossed when you are done

217
00:08:30,816 --> 00:08:32,340
Think about what that means

218
00:08:32,576 --> 00:08:33,520
if you are a lawyer

219
00:08:33,775 --> 00:08:35,040
seven years of studying law

220
00:08:35,535 --> 00:08:36,441
if you are a doctor

221
00:08:36,735 --> 00:08:37,880
ten years of studying medicine

222
00:08:38,375 --> 00:08:40,100
if you are a financial analyst

223
00:08:40,135 --> 00:08:43,815
fifteen years in the industry you think that is your moat

224
00:08:43,816 --> 00:08:45,220
you think nobody can take it

225
00:08:45,455 --> 00:08:47,380
but aluminum thought so too

226
00:08:47,576 --> 00:08:49,541
Aluminum was once as expensive as silver

227
00:08:49,576 --> 00:08:51,740
it stood on top of the Washington Monument

228
00:08:51,896 --> 00:08:54,894
then a twenty two year old spent two hours

229
00:08:54,896 --> 00:08:56,860
and rewrote its fate completely

230
00:08:57,772 --> 00:08:59,096
This is not fear mongering

231
00:08:59,132 --> 00:09:00,976
this is a pattern history keeps confirming

232
00:09:01,132 --> 00:09:03,490
when technology fully solves a scarcity

233
00:09:03,491 --> 00:09:04,777
that scarcity stops being worth anything

234
00:09:05,052 --> 00:09:06,356
Oil lamps gave way to electric light

235
00:09:06,652 --> 00:09:08,257
carriages gave way to cars

236
00:09:08,292 --> 00:09:10,336
long distance calls gave way to the internet

237
00:09:10,611 --> 00:09:13,051
every time people said this time is different

238
00:09:13,052 --> 00:09:14,356
my skill cannot be replaced

239
00:09:14,772 --> 00:09:15,356
and every time

240
00:09:15,652 --> 00:09:17,537
technology did not care what people thought

241
00:09:17,971 --> 00:09:18,970
Only this time

242
00:09:18,971 --> 00:09:22,297
the scarcity being solved is not a material or an energy source

243
00:09:22,331 --> 00:09:25,177
it is the intelligence we are so proud of

244
00:09:25,731 --> 00:09:28,811
Stop for a second and ask in the work you do now

245
00:09:28,811 --> 00:09:31,116
how much of it is selling an information gap in your head

246
00:09:32,091 --> 00:09:34,716
how much is charging because you read a few more books

247
00:09:35,132 --> 00:09:36,956
or handled a few more cases than the next person

248
00:09:37,571 --> 00:09:40,537
if AI closes those gaps within three years

249
00:09:40,811 --> 00:09:41,757
what is left of you

250
00:09:43,106 --> 00:09:44,790
Konstantine shared another detail

251
00:09:45,106 --> 00:09:48,190
in 2006 NASA needed an antenna for a satellite

252
00:09:48,506 --> 00:09:50,584
traditional antennas are drawn by engineers

253
00:09:50,586 --> 00:09:52,751
geometrically symmetric with elegant lines

254
00:09:52,785 --> 00:09:55,430
a perfect marriage of human taste and physics

255
00:09:55,706 --> 00:09:56,890
but that time NASA said

256
00:09:57,185 --> 00:09:58,670
let the computer design it

257
00:09:58,905 --> 00:10:00,711
They used an evolutionary algorithm

258
00:10:00,746 --> 00:10:04,190
letting the program try fail optimize and evolve

259
00:10:04,466 --> 00:10:07,145
the antenna it produced was bent and crooked

260
00:10:07,145 --> 00:10:08,270
with no symmetry at all

261
00:10:08,425 --> 00:10:10,751
it looked like a piece of wire someone had crumpled

262
00:10:11,065 --> 00:10:13,471
no engineer would ever draw that

263
00:10:13,706 --> 00:10:16,951
but it performed far better than the human designed ones

264
00:10:17,706 --> 00:10:19,111
In March 2006

265
00:10:19,145 --> 00:10:22,751
that evolved antenna flew to space on the satellite

266
00:10:22,905 --> 00:10:25,810
becoming the first hardware evolved by a computer

267
00:10:26,025 --> 00:10:27,790
ever to actually fly in space

268
00:10:28,002 --> 00:10:31,200
Konstantine said when we hand cognition to machines

269
00:10:31,201 --> 00:10:34,206
what comes back may defy human intuition entirely

270
00:10:34,482 --> 00:10:36,480
AI does not think the way we do

271
00:10:36,482 --> 00:10:38,367
it has an alien design language

272
00:10:38,522 --> 00:10:40,247
That line stayed with me for a long time

273
00:10:40,737 --> 00:10:44,801
we instinctively judge AI output by human taste is it pretty

274
00:10:45,577 --> 00:10:46,521
does it read like a human wrote it

275
00:10:47,056 --> 00:10:48,041
does it match our experience

276
00:10:48,896 --> 00:10:51,501
but that evolved antenna tells us

277
00:10:51,656 --> 00:10:54,862
the optimal answer may live outside the human map entirely

278
00:10:55,016 --> 00:10:56,901
that bent crooked wire

279
00:10:56,936 --> 00:11:00,622
simply outperformed the antenna the best human engineers could design

280
00:11:01,709 --> 00:11:03,674
There is something deeper here

281
00:11:03,949 --> 00:11:05,814
why does human design chase symmetry

282
00:11:06,309 --> 00:11:08,034
because the brain likes symmetry

283
00:11:08,309 --> 00:11:11,554
symmetry feels safe stable and predictable

284
00:11:11,829 --> 00:11:14,515
but the universe does not care what your brain likes

285
00:11:14,669 --> 00:11:16,909
the optimum is computed by the laws of physics

286
00:11:16,909 --> 00:11:19,314
not voted on by human taste

287
00:11:19,470 --> 00:11:22,714
When you offload cognition from the brain to algorithms

288
00:11:22,789 --> 00:11:25,434
what comes back may make no sense to you

289
00:11:25,510 --> 00:11:27,154
and it is still better

290
00:11:28,018 --> 00:11:29,563
If you are starting to feel anxious

291
00:11:29,778 --> 00:11:30,423
hold on

292
00:11:30,818 --> 00:11:34,663
Konstantine closed with something that quieted the whole room

293
00:11:34,939 --> 00:11:35,740
he said

294
00:11:35,898 --> 00:11:40,222
tens of thousands of years of art history were a chase after realism

295
00:11:40,538 --> 00:11:43,023
from cave paintings to Egyptian murals

296
00:11:43,138 --> 00:11:45,862
from Greek pottery to Renaissance oil painting

297
00:11:46,058 --> 00:11:49,023
we spent tens of thousands of years making pictures look more real

298
00:11:49,298 --> 00:11:52,183
Then in 1839 photography arrived

299
00:11:52,298 --> 00:11:54,482
a machine did in one second

300
00:11:54,538 --> 00:11:57,023
what generations of painters spent lifetimes learning

301
00:11:57,337 --> 00:12:00,697
everyone said painting was finished art was dead

302
00:12:00,697 --> 00:12:02,423
the machine renders better than any human

303
00:12:02,658 --> 00:12:03,923
Sound familiar

304
00:12:04,898 --> 00:12:07,582
it is exactly the AI conversation we are having today

305
00:12:08,016 --> 00:12:09,200
But what actually happened

306
00:12:09,656 --> 00:12:11,460
humans did not abandon art

307
00:12:11,495 --> 00:12:13,220
they redefined it

308
00:12:13,335 --> 00:12:17,421
painters asked themselves is art about capturing what the eye sees

309
00:12:17,575 --> 00:12:19,840
or capturing what the heart and soul see

310
00:12:20,455 --> 00:12:21,681
So impressionism came

311
00:12:21,936 --> 00:12:23,094
expressionism came

312
00:12:23,095 --> 00:12:24,301
cubism came

313
00:12:24,416 --> 00:12:29,061
these new art forms were humanity's greatest answer to that revolution

314
00:12:29,375 --> 00:12:31,141
Photography did not kill art

315
00:12:31,176 --> 00:12:33,100
it only killed realism

316
00:12:33,176 --> 00:12:36,661
the territory of art was blown wide open instead

317
00:12:37,225 --> 00:12:39,609
Konstantine ended by quoting a line from twenty five hundred years ago

318
00:12:39,864 --> 00:12:42,147
from the ancient Greek philosopher Protagoras

319
00:12:42,157 --> 00:12:44,470
man is the measure of all things

320
00:12:44,585 --> 00:12:47,543
Nothing has value to us in a vacuum

321
00:12:47,543 --> 00:12:48,305
not aluminum

322
00:12:48,305 --> 00:12:48,970
not art

323
00:12:49,144 --> 00:12:50,229
and not intelligence

324
00:12:50,345 --> 00:12:51,864
they have value

325
00:12:51,864 --> 00:12:53,989
because of how people experience them

326
00:12:54,225 --> 00:12:58,029
Then he said what I think was the most powerful line of the whole talk

327
00:12:58,264 --> 00:12:59,289
AI can do the work

328
00:12:59,624 --> 00:13:01,023
AI will do the work

329
00:13:01,024 --> 00:13:04,830
but only human connection gives you a reason to care

330
00:13:05,545 --> 00:13:07,550
Work will look completely different in ten years

331
00:13:07,784 --> 00:13:09,389
everything will change violently

332
00:13:09,624 --> 00:13:10,667
but one thing will not

333
00:13:10,677 --> 00:13:15,909
the relationship you build today with the person next to you will last

334
00:13:16,185 --> 00:13:19,509
that is what you will find truly valuable when you look back

335
00:13:20,617 --> 00:13:22,543
I have been an angel investor for over a decade

336
00:13:22,617 --> 00:13:26,303
I have watched wave after wave of technology wash over society

337
00:13:26,617 --> 00:13:27,896
every time a wave arrives

338
00:13:27,898 --> 00:13:31,622
most people react with fear afraid of being left behind

339
00:13:31,697 --> 00:13:32,602
afraid their skills lose value

340
00:13:32,938 --> 00:13:35,023
afraid there is no place for them in the new world

341
00:13:35,377 --> 00:13:37,742
That fear is real and reasonable

342
00:13:37,898 --> 00:13:41,102
but fear has never been a good guide to action

343
00:13:42,128 --> 00:13:44,607
The biggest takeaway for me from this Sequoia talk

344
00:13:44,608 --> 00:13:46,814
is not how big a ten trillion dollar market is

345
00:13:46,848 --> 00:13:49,534
nor how fast a century compresses into a hundred days

346
00:13:49,728 --> 00:13:52,934
it is the deepest lesson hidden in the aluminum story

347
00:13:53,449 --> 00:13:55,654
Aluminum went from treasure to trash

348
00:13:55,728 --> 00:13:57,574
not because aluminum changed

349
00:13:57,929 --> 00:13:58,833
it is the same aluminum

350
00:13:59,208 --> 00:14:00,367
same chemical formula

351
00:14:00,368 --> 00:14:01,314
same physical properties

352
00:14:01,649 --> 00:14:02,694
same uses

353
00:14:03,128 --> 00:14:05,333
what changed was the cost of getting it

354
00:14:05,728 --> 00:14:09,373
When that cost falls from astronomical to almost nothing

355
00:14:09,608 --> 00:14:11,294
the scarcity disappears

356
00:14:11,728 --> 00:14:13,394
value is not in the thing itself

357
00:14:13,649 --> 00:14:15,013
it is in the scarcity

358
00:14:16,050 --> 00:14:16,915
So in the AI age

359
00:14:17,250 --> 00:14:18,594
what is actually scarce

360
00:14:19,010 --> 00:14:20,289
Knowledge is not scarce

361
00:14:20,290 --> 00:14:22,035
AI remembers more and looks it up faster

362
00:14:22,530 --> 00:14:23,809
reasoning is not scarce

363
00:14:23,809 --> 00:14:25,594
AI calculates more precisely and thinks more completely

364
00:14:26,209 --> 00:14:28,529
even judgment is being eaten away by AI

365
00:14:28,530 --> 00:14:32,415
because AI has seen more cases than any expert sees in a lifetime

366
00:14:32,689 --> 00:14:34,275
So what can AI not do

367
00:14:34,609 --> 00:14:36,474
what will never turn into aluminum

368
00:14:37,170 --> 00:14:38,833
I think Konstantine already gave the answer

369
00:14:38,843 --> 00:14:41,255
it is why you do this at all

370
00:14:41,290 --> 00:14:42,354
it is what you care about

371
00:14:42,609 --> 00:14:44,094
it is who you are with

372
00:14:44,130 --> 00:14:48,135
it is the pain the joy the confusion and the insight you have lived

373
00:14:49,089 --> 00:14:50,834
AI can generate ten thousand love poems

374
00:14:51,050 --> 00:14:51,875
but it has never loved

375
00:14:52,370 --> 00:14:54,655
AI can analyze ten thousand business models

376
00:14:54,729 --> 00:14:56,434
but it never loses sleep over a product

377
00:14:56,969 --> 00:14:59,054
AI can simulate the language of empathy

378
00:14:59,130 --> 00:15:00,689
but it has never lost anyone

379
00:15:00,689 --> 00:15:03,934
Your scarcity is not how much knowledge you carry

380
00:15:04,010 --> 00:15:06,375
it is that you are someone who has lived

381
00:15:07,352 --> 00:15:11,077
I once read an analogy fearing AI will replace human connection

382
00:15:11,112 --> 00:15:13,558
is like fearing calculators will replace mathematicians

383
00:15:13,793 --> 00:15:15,378
calculators replaced arithmetic

384
00:15:15,673 --> 00:15:16,597
not mathematics

385
00:15:16,832 --> 00:15:18,837
That line is exactly right

386
00:15:18,952 --> 00:15:21,758
AI replaces the mechanical part of cognition

387
00:15:21,832 --> 00:15:25,398
the parts that can be encoded compressed and copied at scale

388
00:15:25,712 --> 00:15:27,437
but the understanding between two people

389
00:15:27,592 --> 00:15:29,918
the moment something clicks late at night

390
00:15:30,072 --> 00:15:32,817
the thing that floods your chest the first time you become a parent

391
00:15:33,393 --> 00:15:34,792
these are not data

392
00:15:34,793 --> 00:15:35,957
they are existence itself

393
00:15:36,224 --> 00:15:36,929
I often say

394
00:15:37,385 --> 00:15:38,829
all knowledge is a raft

395
00:15:39,064 --> 00:15:40,109
science is a raft

396
00:15:40,265 --> 00:15:41,010
Buddhism is a raft

397
00:15:41,385 --> 00:15:42,370
philosophy is a raft

398
00:15:42,864 --> 00:15:44,030
AI is a raft too

399
00:15:44,464 --> 00:15:45,650
A raft is for crossing the river

400
00:15:45,944 --> 00:15:46,910
not for collecting

401
00:15:47,224 --> 00:15:49,343
wise people do not cling to the raft

402
00:15:49,344 --> 00:15:50,630
they look at the far shore

403
00:15:51,024 --> 00:15:52,869
Sequoia says the car has arrived

404
00:15:53,624 --> 00:15:53,969
fine

405
00:15:54,305 --> 00:15:55,310
the car has arrived

406
00:15:55,705 --> 00:15:57,369
do you keep clutching your saddle

407
00:15:57,825 --> 00:15:58,690
or do you learn to drive

408
00:15:59,504 --> 00:16:03,769
or do you answer the deeper question where are you actually going

409
00:16:04,464 --> 00:16:06,030
What do you think in the AI age

410
00:16:06,104 --> 00:16:08,210
what skill will never turn into aluminum

411
00:16:08,785 --> 00:16:10,430
I am curious about your answer

412
00:16:10,825 --> 00:16:11,489
I am Leo Wang

413
00:16:11,744 --> 00:16:12,738
see you next time