1
00:00:00,720 --> 00:00:04,517
Hello again. If I keep
doing this, I understand that

2
00:00:04,600 --> 00:00:09,837
there's vacancy at 60 Minutes. Maybe I
can If this investing thing doesn't work

3
00:00:09,920 --> 00:00:14,317
out, maybe I can do that.
Another full disclosure,

4
00:00:14,400 --> 00:00:16,677
Sam is a friend. I'm also on the board

5
00:00:16,760 --> 00:00:19,437
of OpenAI, so

6
00:00:19,520 --> 00:00:23,797
I promise you I will not only ask him
softball questions. I will channel my

7
00:00:23,880 --> 00:00:27,157
inner Brett Bear and and uh

8
00:00:27,240 --> 00:00:28,957
and seek Rupert.

9
00:00:29,040 --> 00:00:31,517
So, Sam, as I said,

10
00:00:31,600 --> 00:00:35,397
CEO of OpenAI, one of the founders.

11
00:00:35,480 --> 00:00:39,157
He previously served as the
president of Y Combinator,

12
00:00:39,240 --> 00:00:43,997
where he helped scale some of the world's
most influential startup accelerators,

13
00:00:44,080 --> 00:00:48,037
and advised companies that will
shape entire industries.

14
00:00:48,120 --> 00:00:49,877
At OpenAI, Sam has obviously been at the

15
00:00:49,960 --> 00:00:51,797
forefront of advancing artificial

16
00:00:51,880 --> 00:00:55,157
intelligence to the benefit of humanity,

17
00:00:55,240 --> 00:01:00,077
helping to bring cutting-edge AI tools
to the hands of millions of people. How

18
00:01:00,160 --> 00:01:01,917
many people? Eight, 900 million people?

19
00:01:02,000 --> 00:01:05,757
>> More, yeah.
>> And businesses around the world.

20
00:01:05,840 --> 00:01:10,437
So, Sam, let me start with what I think
is the question that is on everybody's

21
00:01:10,520 --> 00:01:14,517
mind, okay? Which is

22
00:01:14,600 --> 00:01:16,917
where are we today

23
00:01:17,000 --> 00:01:19,317
in the world of AI?

24
00:01:19,400 --> 00:01:21,997
>> I I think at some point in the last few

25
00:01:22,080 --> 00:01:24,437
months,

26
00:01:24,520 --> 00:01:27,757
we really have crossed a threshold into

27
00:01:27,840 --> 00:01:30,477
major economic utility of these models.

28
00:01:30,560 --> 00:01:33,077
It may have happened a little bit earlier,
but there was such an overhang

29
00:01:33,160 --> 00:01:37,117
before we figured out how to use these,
and we had to not only continue to make

30
00:01:37,200 --> 00:01:40,997
the models get smarter, but figure out
sort of the plumbing to make them easy

31
00:01:41,080 --> 00:01:43,557
to use. Where we're we're now in a world

32
00:01:43,640 --> 00:01:48,877
where the models are
astounding people with the

33
00:01:48,960 --> 00:01:52,677
work they can do.
Uh and I think this has been most

34
00:01:52,760 --> 00:01:54,797
noticeable in coding, yep, but it's also

35
00:01:54,880 --> 00:01:56,877
happening in science. It's happening in

36
00:01:56,960 --> 00:02:00,717
many fields of knowledge work. Uh

37
00:02:00,800 --> 00:02:04,037
at sort of with this orienting speed
where people are saying like, "Man,

38
00:02:04,120 --> 00:02:09,237
these things that I thought were still
years away are happening now and I have

39
00:02:09,320 --> 00:02:12,557
my job shifted from doing

40
00:02:12,640 --> 00:02:16,797
uh you know, direct technical work
or uh you know, legal work to managing a

41
00:02:16,880 --> 00:02:19,077
team of agents doing this work."

42
00:02:19,160 --> 00:02:22,757
This is going to go much further. Uh I
think we're at a very steep part of the

43
00:02:22,840 --> 00:02:23,797
curve

44
00:02:23,880 --> 00:02:27,557
and right now maybe you can trust say a

45
00:02:27,640 --> 00:02:32,077
AI software engineer to do a multi-hour
task. Very soon it'll be a multi-day

46
00:02:32,160 --> 00:02:34,397
task and then a multi-week task.

47
00:02:34,480 --> 00:02:37,317
And not long after that, I think the

48
00:02:37,400 --> 00:02:39,637
paradigm will shift again and it'll feel

49
00:02:39,720 --> 00:02:44,517
like these AI systems are just connected
to your life, to your company, whatever

50
00:02:44,600 --> 00:02:48,237
proactively thinking,
working all the time. Uh

51
00:02:48,320 --> 00:02:52,157
and having full context on whatever
they need to know and just sort of doing

52
00:02:52,240 --> 00:02:55,237
stuff like you would trust
a senior employee to do.

53
00:02:55,320 --> 00:02:59,197
>> And and you think
that um companies have

54
00:02:59,280 --> 00:03:01,357
real have a real understanding of how

55
00:03:01,440 --> 00:03:03,677
these systems can help them and

56
00:03:03,760 --> 00:03:07,157
reimagine how they do their businesses?

57
00:03:07,240 --> 00:03:11,157
>> Some do, some don't. Uh certainly
the new generation of startups thinks

58
00:03:11,240 --> 00:03:14,277
differently than any
generation of startups before.

59
00:03:14,360 --> 00:03:17,877
Uh it used to be that
when we would talk to

60
00:03:17,960 --> 00:03:21,917
startups, they would talk about
how many employees they needed.

61
00:03:22,000 --> 00:03:25,997
Uh now they generally don't want to hire
a lot. They think that'll slow them down

62
00:03:26,080 --> 00:03:28,677
and they're all focused on how
much compute they can get.

63
00:03:28,760 --> 00:03:31,117
You know, can I reserve this much
capacity? Can I do a cloud deal for

64
00:03:31,200 --> 00:03:33,877
that? Uh can I get this many tokens?

65
00:03:33,960 --> 00:03:38,317
And I think that is that that
is a mental shift that bigger

66
00:03:38,400 --> 00:03:40,717
companies are going through more slowly

67
00:03:40,800 --> 00:03:44,277
but some are starting to do that. Uh one

68
00:03:44,360 --> 00:03:49,597
place I think you can see this happening
is engineering orgs and product orgs

69
00:03:49,680 --> 00:03:54,357
talking about they're doubling, tripling
what they're planning to ship this year.

70
00:03:54,440 --> 00:03:56,077
And that has not happened before.

71
00:03:56,160 --> 00:03:57,117
>> Yeah.

72
00:03:57,200 --> 00:04:00,597
And you've been quite vocal in saying

73
00:04:00,680 --> 00:04:05,037
that um artificial
general intelligence will

74
00:04:05,120 --> 00:04:06,997
come sooner rather than later.

75
00:04:07,080 --> 00:04:10,677
Want to share your views on how
close we are and how soon it will

76
00:04:10,760 --> 00:04:12,317
come?

77
00:04:12,400 --> 00:04:15,237
>> At at this point,
I think the definition

78
00:04:15,320 --> 00:04:19,477
of AGI really matters. Some people would
say we already got there. Some people

79
00:04:19,560 --> 00:04:22,517
say it's very close. Some people say
we're kind of you know, it's maybe still

80
00:04:22,600 --> 00:04:28,517
a year away, but in any case,
that word has ceased to have much meaning.

81
00:04:28,600 --> 00:04:31,917
>> Mhm. >> Uh there are
maybe two thresholds that we

82
00:04:32,000 --> 00:04:33,597
could talk about that are interesting.

83
00:04:33,680 --> 00:04:35,677
>> Okay. >> Number one,
when is there going to be

84
00:04:35,760 --> 00:04:40,077
more of the world's cognitive capacity
inside of data centers than outside of

85
00:04:40,160 --> 00:04:45,677
them? And that to me feels
like maybe it could happen.

86
00:04:45,760 --> 00:04:48,317
Huge error bars. I could be totally
wrong, but maybe that could happen by

87
00:04:48,400 --> 00:04:51,237
like late 2028. >> Mhm.

88
00:04:51,320 --> 00:04:55,437
>> And that's an extraordinary
shift in the world.

89
00:04:55,520 --> 00:04:56,677
Um

90
00:04:56,760 --> 00:05:00,477
the other one is when can

91
00:05:00,560 --> 00:05:03,157
a CEO of major company, a president of a

92
00:05:03,240 --> 00:05:05,437
major country, a Nobel Prize winning

93
00:05:05,520 --> 00:05:07,637
scientist, when can they not do their

94
00:05:07,720 --> 00:05:10,957
job without making heavy use of AI? This

95
00:05:11,040 --> 00:05:14,677
doesn't mean that there will be
an AI CEO or an AI president, uh

96
00:05:14,760 --> 00:05:18,317
but it does mean that
the the role of let's

97
00:05:18,400 --> 00:05:23,277
say a human CEO, when I think about my
job, it's really quite different. You you

98
00:05:23,360 --> 00:05:25,277
still do need a person to

99
00:05:25,360 --> 00:05:27,277
stand behind decisions and kind of

100
00:05:27,360 --> 00:05:30,677
exercise human judgment and

101
00:05:30,760 --> 00:05:33,037
all of the understanding

102
00:05:33,120 --> 00:05:35,597
that we expect out of someone running an

103
00:05:35,680 --> 00:05:39,437
an important organization to do, but

104
00:05:39,520 --> 00:05:43,037
the actual parts of my role that I will

105
00:05:43,120 --> 00:05:47,477
increasingly have to rely on an AI to do
because no human can. You know, no human

106
00:05:47,560 --> 00:05:51,477
can talk to no human CEO can talk
to every employee at a company, every

107
00:05:51,560 --> 00:05:53,557
customer, be in every meeting, be an

108
00:05:53,640 --> 00:05:56,357
expert in every field. And so,

109
00:05:56,440 --> 00:05:59,317
more and more I think of these jobs will

110
00:05:59,400 --> 00:06:03,637
be supervising a bunch of AI, um providing

111
00:06:03,720 --> 00:06:06,197
oversight, deciding how to like trust

112
00:06:06,280 --> 00:06:08,037
the the outputs, how to provide

113
00:06:08,120 --> 00:06:12,317
guidance. And that threshold of when you

114
00:06:12,400 --> 00:06:16,637
really wouldn't want to be doing your job,

115
00:06:16,720 --> 00:06:20,877
running a large organization
without heavily heavy reliance on AI,

116
00:06:20,960 --> 00:06:23,357
I think that's another sort
of interesting threshold. That may take a

117
00:06:23,440 --> 00:06:26,277
little bit longer,
but probably not a lot longer.

118
00:06:26,360 --> 00:06:29,077
>> And and as you do
your job, how much are

119
00:06:29,160 --> 00:06:32,357
you finding yourself relying on

120
00:06:32,440 --> 00:06:36,117
some of the agents and some of the
artificial intelligence that we're

121
00:06:36,200 --> 00:06:39,317
developing at OpenAI? >> Uh

122
00:06:39,400 --> 00:06:42,117
it's ramping incredibly

123
00:06:42,200 --> 00:06:45,757
quickly. Um if I have like a new idea

124
00:06:45,840 --> 00:06:51,520
for a business model,
a strategy shift, a uh

125
00:06:51,680 --> 00:06:55,317
a product offering we should do,
the very first thing I do before I even

126
00:06:55,400 --> 00:06:57,517
bounce it off somebody else is to ask

127
00:06:57,600 --> 00:07:02,517
our tools. And as they get more context,
and I think this really is like

128
00:07:02,600 --> 00:07:06,117
the next big thing to happen is they
can get close to full context of our

129
00:07:06,200 --> 00:07:10,717
company, access to all of our
internal docs, communication, code,

130
00:07:10,800 --> 00:07:16,117
customer data, everything, the quality of
the answers, thought, whatever you want to

131
00:07:16,200 --> 00:07:17,237
call it, gets better and better.

132
00:07:17,320 --> 00:07:21,437
>> Right. Okay. So, let's
let's shift um a little bit. 2

133
00:07:21,520 --> 00:07:26,477
weeks ago, um you announced
a $110 billion funding

134
00:07:26,560 --> 00:07:31,877
round. I asked ChatGPT,
how does that compare compare to any

135
00:07:31,960 --> 00:07:34,437
other fundraising that's been
done in the public markets?

136
00:07:34,520 --> 00:07:37,917
>> I actually don't know.
>> Uh four times as large, okay? The

137
00:07:38,000 --> 00:07:42,357
largest public offering ever done
was roughly 25 billion that Aramco did

138
00:07:42,440 --> 00:07:45,477
several years ago.
And the public markets are supposed to

139
00:07:45,560 --> 00:07:50,317
be the broadest and deepest
sources of capital, okay?

140
00:07:50,400 --> 00:07:53,677
Three strategic partners, Amazon,

141
00:07:53,760 --> 00:07:56,997
Nvidia, and oh and SoftBank.

142
00:07:57,080 --> 00:07:59,477
Tell us a little bit about this.

143
00:07:59,560 --> 00:08:02,397
How is this an inflection
point of the company?

144
00:08:02,480 --> 00:08:05,237
And one of the questions somebody asked
me is, what are we spending all that

145
00:08:05,320 --> 00:08:08,797
money on? The

146
00:08:08,880 --> 00:08:12,357
>> There is many hard parts of this
business, but one of the hardest ones is

147
00:08:12,440 --> 00:08:15,157
the infrastructure is so expensive. You

148
00:08:15,240 --> 00:08:20,360
need so much of it, and you
have to commit so far in advance.

149
00:08:20,960 --> 00:08:25,757
I have never seen any other
industry quite like this. I

150
00:08:25,840 --> 00:08:29,637
mean, there have been clearly many
capital intensive industries throughout

151
00:08:29,720 --> 00:08:31,757
history, but

152
00:08:31,840 --> 00:08:34,317
as I look at the what's to come in front

153
00:08:34,400 --> 00:08:36,717
of us in the future in future years, if

154
00:08:36,800 --> 00:08:39,597
the ramp stays as steep as it looks like

155
00:08:39,680 --> 00:08:44,197
it is right now, the demand is growing
as fast as it's growing, you have to do

156
00:08:44,280 --> 00:08:47,277
some pretty unusual things. Uh

157
00:08:47,360 --> 00:08:51,237
OpenAI does a lot
of things that look weird.

158
00:08:51,320 --> 00:08:53,757
We spend a ton of money on

159
00:08:53,840 --> 00:08:56,837
infrastructure in advance of revenue. We

160
00:08:56,920 --> 00:08:59,517
do new business models like ads that

161
00:08:59,600 --> 00:09:04,477
seem like, you know, maybe not the most
profitable thing we could do. Um

162
00:09:04,560 --> 00:09:08,277
a long list of other things, but we have
this fundamental belief in abundance of

163
00:09:08,360 --> 00:09:12,357
intelligence. And that one of the most
important things in the future is that

164
00:09:12,440 --> 00:09:16,997
we make intelligence, you know,
to borrow an old phrase from

165
00:09:17,080 --> 00:09:19,757
the energy industry that didn't quite
work, too cheap to meter. We want to

166
00:09:19,840 --> 00:09:21,997
flood the world with intelligence.
We want people to just use it for

167
00:09:22,080 --> 00:09:24,997
everything. We want this to just be
something that the future generation

168
00:09:25,080 --> 00:09:29,637
doesn't think about. They expect
everywhere, and everybody has access to

169
00:09:29,720 --> 00:09:33,157
like geniuses, as many as they need in any

170
00:09:33,240 --> 00:09:37,240
area that they need. And

171
00:09:37,640 --> 00:09:41,917
this principle, which is one of our kind
of like top guiding principles, does

172
00:09:42,000 --> 00:09:45,437
lead to a lot of behavior that would look

173
00:09:45,520 --> 00:09:49,197
less natural for other companies.
And one of those is

174
00:09:49,280 --> 00:09:52,477
we really want to get out of this world
that we have been in that we still think

175
00:09:52,560 --> 00:09:55,637
we're on a trajectory to stay on without
changing what we do of always being

176
00:09:55,720 --> 00:09:58,237
capacity constrained. >> Right.

177
00:09:58,320 --> 00:10:01,477
And capacity constrained
you mean compute? Yeah.

178
00:10:01,560 --> 00:10:06,917
And I've often heard you say
a lot that compute is revenues.

179
00:10:07,000 --> 00:10:11,317
You want to talk a little bit about
how you think how you think about that?

180
00:10:11,400 --> 00:10:14,797
>> Fundamentally, our business and I
think the business of every other model

181
00:10:14,880 --> 00:10:19,237
provider is going to look
like selling tokens.

182
00:10:19,320 --> 00:10:23,437
You know, they may come from bigger
or smaller models which makes them more or

183
00:10:23,520 --> 00:10:26,917
less expensive. They may use more
or less reasoning which also makes them

184
00:10:27,000 --> 00:10:30,037
more or less expensive. They may be
running all the time in the background

185
00:10:30,120 --> 00:10:33,437
trying to help you out. Uh they may run
only when you need them if you want to

186
00:10:33,520 --> 00:10:38,637
pay less. They may work super hard.
You know, spend tens of

187
00:10:38,720 --> 00:10:41,877
millions, hundreds of millions, someday
billions of dollars on a single problem.

188
00:10:41,960 --> 00:10:44,437
>> Right. >> That's really valuable. But

189
00:10:44,520 --> 00:10:46,997
we

190
00:10:47,080 --> 00:10:49,437
see a future where

191
00:10:49,520 --> 00:10:51,597
intelligence is a utility like

192
00:10:51,680 --> 00:10:56,517
electricity or water
and people buy it from us

193
00:10:56,600 --> 00:11:00,957
um on a meter and use it
for whatever they want to use it for.

194
00:11:01,040 --> 00:11:04,997
The demand that we see for that seems
like it's going to continue to just go

195
00:11:05,080 --> 00:11:08,797
like this. And

196
00:11:08,880 --> 00:11:14,197
if we don't have enough, we either can't
sell it or the price gets really high

197
00:11:14,280 --> 00:11:19,397
and it you know, it kind of goes to rich
people or society makes a bunch of sort

198
00:11:19,480 --> 00:11:22,557
of central planning decisions that I
think almost always go badly about, you

199
00:11:22,640 --> 00:11:25,997
know, we're going to use our limited
compute supply for this and not that.

200
00:11:26,080 --> 00:11:30,397
So, the best thing to me throughout all
the history of capitalism, innovation,

201
00:11:30,480 --> 00:11:32,637
whatever you want
is to just flood the market.

202
00:11:32,720 --> 00:11:35,757
>> Yeah. And obviously a critical part of

203
00:11:35,840 --> 00:11:37,637
addressing that

204
00:11:37,720 --> 00:11:41,477
compute demand is Stargate.

205
00:11:41,560 --> 00:11:45,757
And this is an infrastructure conference
and you know, you announced that several

206
00:11:45,840 --> 00:11:49,837
months ago, or almost a year
ago, I guess it was, yeah.

207
00:11:49,920 --> 00:11:53,997
How's it going in the US?
And then how is it going, because I

208
00:11:54,080 --> 00:11:59,680
also started in in Abu Dhabi.
>> Yeah. Yeah. Um

209
00:12:00,280 --> 00:12:03,037
Honestly, there's many
cool parts of the job. One

210
00:12:03,120 --> 00:12:06,317
of the coolest is getting to go
visit these mega data centers under

211
00:12:06,400 --> 00:12:07,837
construction and operation. >> Yeah.

212
00:12:07,920 --> 00:12:11,317
>> Just the scale is, you know, of like
these gigawatt campuses. It's it's it's

213
00:12:11,400 --> 00:12:15,237
really hard to explain and you see these
photos and it's like, okay, it does look

214
00:12:15,320 --> 00:12:18,157
big. And then you go there and you're
walking through it and building to

215
00:12:18,240 --> 00:12:21,597
building and it's just, you know,
10,000 people there, all these different

216
00:12:21,680 --> 00:12:24,197
skilled trades doing all
these different things and

217
00:12:24,280 --> 00:12:28,797
you know, it looks like a spaceship
inside. It's really quite incredible. Um

218
00:12:28,880 --> 00:12:32,557
we are training right now
on the first site in Abilene,

219
00:12:32,640 --> 00:12:33,917
what I think will be the best model in

220
00:12:34,000 --> 00:12:38,797
the world, hopefully by a lot. And it is

221
00:12:38,880 --> 00:12:42,597
it's so amazing to have gone from,
you know, many visits while it was under

222
00:12:42,680 --> 00:12:46,997
construction, and just really
internalizing the scale and incredible

223
00:12:47,080 --> 00:12:50,477
complexity to the day
that one researcher at OpenAI

224
00:12:50,560 --> 00:12:54,157
types in like one command
and it pushes enter,

225
00:12:54,240 --> 00:12:58,557
unbelievable number of GPUs spin up
and start doing this one huge computation

226
00:12:58,640 --> 00:13:02,517
all together. >> Yeah. It's very And and

227
00:13:02,600 --> 00:13:06,757
what have been the most
pleasant surprises and what have

228
00:13:06,840 --> 00:13:11,237
been the hardest things about getting
it up and running and scaling it going

229
00:13:11,320 --> 00:13:16,157
forward? >> Uh I mean,
there have been all the expected

230
00:13:16,240 --> 00:13:19,517
challenges and then the unknown unknowns.
Like there was a crazy weather

231
00:13:19,600 --> 00:13:22,677
event in Abilene that we hadn't,
you know, was outside

232
00:13:22,760 --> 00:13:25,277
what we had planned for and that brought
things down for a little while.

233
00:13:25,360 --> 00:13:27,837
>> Mhm. >> Um there's all of supply chain

234
00:13:27,920 --> 00:13:32,437
challenges. Anything at this scale is just
like so much stuff goes wrong. Lots

235
00:13:32,520 --> 00:13:39,080
goes right, but the just try trying
to build something around the clock

236
00:13:39,280 --> 00:13:42,397
with such complexity, there's
all the stuff that goes wrong.

237
00:13:42,480 --> 00:13:46,757
One of the biggest surprises
on the upside is how

238
00:13:46,840 --> 00:13:50,877
many different organizations had to come
together to do this in an incredibly

239
00:13:50,960 --> 00:13:53,197
short period of time. Uh, and how much

240
00:13:53,280 --> 00:13:54,997
we all like ended up working together as

241
00:13:55,080 --> 00:13:58,517
one team uh, under a lot of pressure.

242
00:13:58,600 --> 00:14:03,117
>> And of course, the power demand part
of it is the one that a lot of people are

243
00:14:03,200 --> 00:14:06,197
focused on. Um,

244
00:14:06,280 --> 00:14:10,357
are you optimistic that we
will solve that challenge

245
00:14:10,440 --> 00:14:15,077
first in the US and then we can
talk about in in other places?

246
00:14:15,160 --> 00:14:19,077
>> Uh, in the long term, I am.

247
00:14:19,160 --> 00:14:23,517
>> Okay. >> Uh, I have no
doubt we will figure out how

248
00:14:23,600 --> 00:14:27,797
to build huge amounts of power generation.
AI will help, of course.

249
00:14:27,880 --> 00:14:31,437
Uh, but I think the portfolio
the world has in front of it

250
00:14:31,520 --> 00:14:35,197
gas, solar, nuclear fission,
nuclear fusion, more

251
00:14:35,280 --> 00:14:39,237
like we're I feel good
about what what we will be

252
00:14:39,320 --> 00:14:41,397
capable of and what we will eventually do.

253
00:14:41,480 --> 00:14:42,157
>> Mhm.

254
00:14:42,240 --> 00:14:43,957
>> Um,

255
00:14:44,040 --> 00:14:48,637
given the demand growth we are seeing,

256
00:14:48,720 --> 00:14:52,997
I am sort of hoping for a miracle in terms
of figuring out how we can get way

257
00:14:53,080 --> 00:14:54,957
more efficient per watt >> Uh, okay.

258
00:14:55,040 --> 00:14:59,877
>> models to give us time to build
out all of this infrastructure. Now,

259
00:14:59,960 --> 00:15:03,477
the track record there
has been incredible. Um,

260
00:15:03,560 --> 00:15:07,317
people cite whatever amazing statistic
they like about how much more efficient

261
00:15:07,400 --> 00:15:10,877
our models have gotten over our industries
models have gotten over time,

262
00:15:10,960 --> 00:15:14,477
but one that I think is incredible,

263
00:15:14,560 --> 00:15:18,237
our first reasoning model
was called 01, came out

264
00:15:18,320 --> 00:15:22,277
like 16 months ago.

265
00:15:22,360 --> 00:15:28,800
Uh, and latest model where we're
now integrated reasoning is 5.4.

266
00:15:29,040 --> 00:15:31,357
To get the same answer

267
00:15:31,440 --> 00:15:34,437
to a hard problem from that first model

268
00:15:34,520 --> 00:15:40,800
to 5.4 has been a reduction
in cost of about 1,000 x.

269
00:15:41,320 --> 00:15:42,997
Maybe I was a little bit wrong
on the timeline. Maybe it was a little bit

270
00:15:43,080 --> 00:15:45,197
longer. But in any case,
since 01 till now.

271
00:15:45,280 --> 00:15:50,317
>> Yeah. >> 1,000 x. Uh,
that is unbelievable in a relatively

272
00:15:50,400 --> 00:15:54,677
short period of time. And

273
00:15:54,760 --> 00:15:58,397
the thing that that the two
things that I think that points to

274
00:15:58,480 --> 00:16:02,277
one, we are still so early and in this

275
00:16:02,360 --> 00:16:05,597
paradigm and we're still so we
have so much more to gain about our

276
00:16:05,680 --> 00:16:08,957
understanding of how to develop these
models and train them and run them

277
00:16:09,040 --> 00:16:13,597
efficiently >> Mhm. >> that we have

278
00:16:13,680 --> 00:16:17,757
were then and still are doing things
in dumb ways and will get better and

279
00:16:17,840 --> 00:16:21,437
better. And two, is that

280
00:16:21,520 --> 00:16:25,477
human ingenuity and the ability to operate

281
00:16:25,560 --> 00:16:29,877
in constraints and to find ways to solve
problems almost always surprise you on the

282
00:16:29,960 --> 00:16:32,797
upside. So, it's not just that the models
have gotten better, it's that we

283
00:16:32,880 --> 00:16:36,277
have figured out, you know, kernel
engineers came to help figure out how to

284
00:16:36,360 --> 00:16:40,237
write more efficient kernels and like
power engineers, the people that design

285
00:16:40,320 --> 00:16:43,077
data centers found more efficient ways
to do that. So, people are answering the

286
00:16:43,160 --> 00:16:45,237
call well beyond just
the model side >> Right.

287
00:16:45,320 --> 00:16:46,277
>> to make this more efficient.

288
00:16:46,360 --> 00:16:48,837
>> Right. And then of course, um,

289
00:16:48,920 --> 00:16:53,717
obviously OpenAI is a big customer
of the Nvidias and AMDs of the world, but

290
00:16:53,800 --> 00:16:57,117
um, you know, I think we signed
a big contract with Amazon.

291
00:16:57,200 --> 00:16:58,357
But we're building our own chip.

292
00:16:58,440 --> 00:17:01,517
>> Yeah. >> What's what's
the thinking behind that?

293
00:17:01,600 --> 00:17:03,717
>> Uh, so the chip we're
doing is inference only.

294
00:17:03,800 --> 00:17:08,037
>> Okay. >> And the thinking behind
it is that on this, you know, come

295
00:17:08,120 --> 00:17:10,957
rise up to solve problems in front of us,

296
00:17:11,040 --> 00:17:15,957
um, we think that a specialized chip to be

297
00:17:16,040 --> 00:17:19,717
not necessarily the fastest inference
chip, but the cheapest inference chip,

298
00:17:19,800 --> 00:17:23,597
the most efficient per watt given
the constraints we see in front of us, is

299
00:17:23,680 --> 00:17:27,197
going to be important for all the agent
demand we see in the future. So, it's a

300
00:17:27,280 --> 00:17:31,277
it's an opinionated bet, it's a limited
chip, but the thing that it does,

301
00:17:31,360 --> 00:17:34,157
in a world where we're energy constrained,
I think it'll be very important.

302
00:17:34,240 --> 00:17:38,197
>> Maybe you can explain because I am
not sure everybody in the audience is

303
00:17:38,280 --> 00:17:43,197
AI proficient with the difference between
an inference chip and a training chip.

304
00:17:43,280 --> 00:17:44,557
>> Sorry, I should have done that first.

305
00:17:44,640 --> 00:17:45,517
So,

306
00:17:45,600 --> 00:17:49,957
there there are two main phases to uh

307
00:17:50,040 --> 00:17:53,477
AI workloads. Now, they're going
to blend together eventually and it'll all

308
00:17:53,560 --> 00:17:55,837
be one continuous thing. But right now,

309
00:17:55,920 --> 00:17:58,757
first we train a model,

310
00:17:58,840 --> 00:18:00,717
you know, a gigantic number of GPUs

311
00:18:00,800 --> 00:18:03,157
crunch for weeks or months on a bunch of

312
00:18:03,240 --> 00:18:05,837
data and you can think of that like

313
00:18:05,920 --> 00:18:08,597
maybe it takes you 22 years in life to

314
00:18:08,680 --> 00:18:13,237
get your education and you learn a bunch
of stuff starting when you're a baby and

315
00:18:13,320 --> 00:18:16,317
you drop things and see they fall and
then eventually like in college physics

316
00:18:16,400 --> 00:18:20,157
understanding like at a very
detailed level what's going on there.

317
00:18:20,240 --> 00:18:24,157
Um And then after that, if you ask a model

318
00:18:24,240 --> 00:18:26,637
to solve a physics question, it that's

319
00:18:26,720 --> 00:18:31,197
called inference uh and that's quite
efficient. But this is true for people

320
00:18:31,280 --> 00:18:34,397
as well and when people talk
about all these AI models are so

321
00:18:34,480 --> 00:18:37,557
efficient, they're usually comparing the

322
00:18:37,640 --> 00:18:40,597
22 years that a human takes to train to

323
00:18:40,680 --> 00:18:45,517
the 1 second that an adult takes to solve
a physics problem. If you compare the

324
00:18:45,600 --> 00:18:48,717
you know, the model solving the physics
problem to the human solving physics

325
00:18:48,800 --> 00:18:52,917
problem, actually the models already
probably are more energy efficient. Um

326
00:18:53,000 --> 00:18:57,637
but training is this hugely massive
amount of processing power that then

327
00:18:57,720 --> 00:19:01,037
produces really just like
a file of numbers that

328
00:19:01,120 --> 00:19:03,877
you can then pose a question
to and get a response.

329
00:19:03,960 --> 00:19:07,757
>> And are you optimistic about
the progress we're making on the chip?

330
00:19:07,840 --> 00:19:12,037
>> Yeah, um we should have
the first chips deployed at scale

331
00:19:12,120 --> 00:19:15,077
uh by the end of this year. We should
have the first chips back just in a few

332
00:19:15,160 --> 00:19:18,237
months now. Um and looks
like it'll be really good.

333
00:19:18,320 --> 00:19:21,877
>> Fantastic. So,
you announced a new partnership this

334
00:19:21,960 --> 00:19:25,957
morning, actually, with the North
American Building Trades Unions to

335
00:19:26,040 --> 00:19:30,357
expand training pathways for skilled
construction, which has been a subject

336
00:19:30,440 --> 00:19:34,677
of my many of the discussions today. Can
you tell us a little bit about what you

337
00:19:34,760 --> 00:19:38,517
and uh Sean McGarvey >> Yes.
>> agreed and What's he And he's going to

338
00:19:38,600 --> 00:19:41,997
be on stage later on. So, let's hear
your version of it and then see if his

339
00:19:42,080 --> 00:19:46,640
version is the same. >> So,

340
00:19:48,200 --> 00:19:52,557
we talked earlier and we've talked
The world has talked many times

341
00:19:52,640 --> 00:19:53,997
in the past about the need for AI

342
00:19:54,080 --> 00:19:56,437
infrastructure, the need for the world

343
00:19:56,520 --> 00:20:00,357
to have the physical infrastructure,

344
00:20:00,440 --> 00:20:03,117
power plants, transmission lines, data

345
00:20:03,200 --> 00:20:05,997
center halls, chillers,

346
00:20:06,080 --> 00:20:10,477
obviously the racks and the GPUs and
everything that goes inside of those. Um

347
00:20:10,560 --> 00:20:14,757
and I sort of wish everyone would go visit

348
00:20:14,840 --> 00:20:17,157
at some point one of these

349
00:20:17,240 --> 00:20:21,997
mega scale data centers
because it it's hard,

350
00:20:22,080 --> 00:20:25,837
you know, when you ask ChatGPT a question,
you get your answer back and it's really

351
00:20:25,920 --> 00:20:29,877
hard to visualize the scale of what it
took to make that or what it takes to

352
00:20:29,960 --> 00:20:34,397
make that happen. Uh people
talk about all of these different

353
00:20:34,480 --> 00:20:36,397
places that were limited. >> Mhm.

354
00:20:36,480 --> 00:20:38,477
>> We are limited in the, you know, number

355
00:20:38,560 --> 00:20:43,837
of turbines or now it's the voltage
transformers or it's the memory fabs or

356
00:20:43,920 --> 00:20:46,557
it's building the data centers,
whatever. All of these things have one

357
00:20:46,640 --> 00:20:49,757
thing in common, which is they
are massively complex physical

358
00:20:49,840 --> 00:20:52,077
infrastructure that require

359
00:20:52,160 --> 00:20:53,797
skilled tradespeople

360
00:20:53,880 --> 00:20:57,077
uh and a lot of them to do. And

361
00:20:57,160 --> 00:20:59,357
no matter where the choke point in the

362
00:20:59,440 --> 00:21:02,677
supply chain is at any one time,

363
00:21:02,760 --> 00:21:04,397
when I talk to people about what it

364
00:21:04,480 --> 00:21:07,597
would take to accelerate it, um

365
00:21:07,680 --> 00:21:11,997
it is more skilled trades workers
to build out all of this infrastructure

366
00:21:12,080 --> 00:21:15,557
that we all depend on.
I think this will be incredible jobs.

367
00:21:15,640 --> 00:21:19,997
Uh more than that, I will I think they
will lay the foundation for the next

368
00:21:20,080 --> 00:21:23,837
generation of American infrastructure
and economic prosperity. And we are

369
00:21:23,920 --> 00:21:28,197
thrilled to get to work uh
together to to drive that faster.

370
00:21:28,280 --> 00:21:32,600
>> Uh just so you know,
we at BlackRock share that