1
00:00:00,080 --> 00:00:03,757
Sata, you're the CEO of one of the most
valuable companies in the world. What's

2
00:00:03,840 --> 00:00:06,157
on your mind these days? What's top of

3
00:00:06,240 --> 00:00:09,757
mind for me is twofold, right? One is

4
00:00:09,840 --> 00:00:12,157
I'm grounded in the fact that in our

5
00:00:12,240 --> 00:00:13,997
business, in tech business, now this is

6
00:00:14,080 --> 00:00:17,757
my 32nd year at Microsoft. Um I know

7
00:00:17,840 --> 00:00:21,917
that for a fact that there's no such
thing as a franchise value. And so that

8
00:00:22,000 --> 00:00:25,757
means every day you have to get up
and hopefully you're doing things that are

9
00:00:25,840 --> 00:00:30,157
going to be relevant tomorrow. Um,
and so to me that's sort of perhaps the

10
00:00:30,240 --> 00:00:31,997
biggest lesson learned over all these

11
00:00:32,080 --> 00:00:35,757
decades and years. And so here we are on

12
00:00:35,840 --> 00:00:37,597
what is essentially a complete new

13
00:00:37,680 --> 00:00:39,917
platform shift uh that we are in the

14
00:00:40,000 --> 00:00:42,957
midstop right I kind of say it's my 32nd

15
00:00:43,040 --> 00:00:45,597
year but it's year two of my fourth

16
00:00:45,680 --> 00:00:48,637
platform shift and so what's on mind is

17
00:00:48,720 --> 00:00:51,037
like okay what is this platform shift

18
00:00:51,120 --> 00:00:54,717
really all about and as a company can we

19
00:00:54,800 --> 00:00:59,277
be allin and innovate right I mean at
the end of the day when I say there's no

20
00:00:59,360 --> 00:01:01,197
franchise value it also means that you

21
00:01:01,280 --> 00:01:03,597
get to play for it all up again even the

22
00:01:03,680 --> 00:01:06,477
battles we won and the battles we lost

23
00:01:06,560 --> 00:01:08,477
are all up for grabbed again. And so

24
00:01:08,560 --> 00:01:11,597
therefore, um there's a freshness to it.

25
00:01:11,680 --> 00:01:15,517
So what's on my mind is, you know, that

26
00:01:15,600 --> 00:01:20,717
ability to ground myself back yet
on another platform shift and it's

27
00:01:20,800 --> 00:01:24,237
exciting. The fact that there is no
franchise value, does that make you

28
00:01:24,320 --> 00:01:28,557
nervous? Absolutely. I mean, it should
make every one of us nervous, right? At

29
00:01:28,640 --> 00:01:33,117
least that's why I think tech is so
exciting, right? It's sort of it's kind

30
00:01:33,200 --> 00:01:36,877
of like the two sides to the same coin,
right? One is you're nervous and it's

31
00:01:36,960 --> 00:01:39,677
exciting. So therefore, you can't rest.

32
00:01:39,760 --> 00:01:44,877
Uh but at the same time, hey, who wants
to be in in a business that you know

33
00:01:44,960 --> 00:01:48,957
where you don't get to reinvent yourself
again and again? Absolutely. Do we think

34
00:01:49,040 --> 00:01:53,197
that we look at technology in very
narrow of a sense? I mean, how

35
00:01:53,280 --> 00:01:57,837
technology can be the driver for the
next economical growth? That's a great

36
00:01:57,920 --> 00:02:00,077
great question. I think a lot about that

37
00:02:00,160 --> 00:02:05,037
right which is uh in some sense if you

38
00:02:05,120 --> 00:02:09,037
think at a GDP level uh tech spend

39
00:02:09,120 --> 00:02:11,637
narrowly defined is probably four

40
00:02:11,720 --> 00:02:14,877
5%. Uh so the question is what is

41
00:02:14,960 --> 00:02:17,917
happening with the other 95%. So one of

42
00:02:18,000 --> 00:02:20,397
the ways I've always thought about the

43
00:02:20,480 --> 00:02:23,037
prospects of any new technological

44
00:02:23,120 --> 00:02:26,637
paradigm or platform shift like take AI

45
00:02:26,720 --> 00:02:29,837
let's I if AI is going to be the next

46
00:02:29,920 --> 00:02:31,557
big general purpose

47
00:02:31,640 --> 00:02:34,957
technology for me the real opportunity

48
00:02:35,040 --> 00:02:37,597
is let's say tech spend goes from 5 to

49
00:02:37,680 --> 00:02:40,477
10% over the next you know five years or

50
00:02:40,560 --> 00:02:43,357
what have you or 10 years then what

51
00:02:43,440 --> 00:02:48,317
happens to the other 90% and the pie
does it become bigger right? Does the

52
00:02:48,400 --> 00:02:50,157
you know do we have a breakthrough in

53
00:02:50,240 --> 00:02:53,437
healthcare driven by AI? Do we have a

54
00:02:53,520 --> 00:02:55,917
breakthrough in material science and

55
00:02:56,000 --> 00:03:00,077
energy transition because of uh AI? Uh

56
00:03:00,160 --> 00:03:02,157
and the list goes on, right? So to me

57
00:03:02,240 --> 00:03:05,197
that is fundamentally the way I think

58
00:03:05,280 --> 00:03:07,357
about it, right? is I I think that one

59
00:03:07,440 --> 00:03:09,437
of the things that might be most

60
00:03:09,520 --> 00:03:13,597
important for us is to consider how a

61
00:03:13,680 --> 00:03:15,357
general purpose technology somebody was

62
00:03:15,440 --> 00:03:17,757
telling me this right which is in the

63
00:03:17,840 --> 00:03:20,637
height of the industrial revolution

64
00:03:20,720 --> 00:03:24,317
uh in the United Kingdom they spent 10%

65
00:03:24,400 --> 00:03:28,637
of their GDP building the railroads and
obviously the railroads not just didn't

66
00:03:28,720 --> 00:03:32,957
is not about the railroads it was about
the entire economy of the United Kingdom

67
00:03:33,040 --> 00:03:34,717
and so something like that I think is

68
00:03:34,800 --> 00:03:37,677
what uh that's the unit of analysis at

69
00:03:37,760 --> 00:03:40,317
least for me as to how tech and its

70
00:03:40,400 --> 00:03:42,477
future will impact the you know broader

71
00:03:42,560 --> 00:03:45,117
society and economy. I mean Google had

72
00:03:45,200 --> 00:03:48,077
all the top AI people but suddenly you

73
00:03:48,160 --> 00:03:54,077
got ahead with your partnership
with open AI. How did that came about?

74
00:03:54,160 --> 00:03:54,917
I

75
00:03:55,000 --> 00:03:59,757
mean to me the way I came at this very

76
00:03:59,840 --> 00:04:02,317
simple which is obviously we mean like I

77
00:04:02,400 --> 00:04:04,557
think the very first thing Microsoft

78
00:04:04,640 --> 00:04:07,917
research did in 1995 when it was formed

79
00:04:08,000 --> 00:04:10,957
was some stuff around speech right in

80
00:04:11,040 --> 00:04:12,797
fact I think we hired a bunch of folks

81
00:04:12,880 --> 00:04:15,597
from CMU and so we've been at this AI

82
00:04:15,680 --> 00:04:18,237
thing in its variety of different forms

83
00:04:18,320 --> 00:04:21,997
forever um and so one of the things that

84
00:04:22,080 --> 00:04:24,717
when I met with the OpenAI folks and Sam

85
00:04:24,800 --> 00:04:28,477
and Greg and crew uh back in you know I

86
00:04:28,560 --> 00:04:29,997
would say when they were working even on

87
00:04:30,080 --> 00:04:33,917
the Dota 2 contest and what have you uh

88
00:04:34,000 --> 00:04:38,397
was to sort of say wow they have a new
different approach to things and we

89
00:04:38,480 --> 00:04:42,157
wanted to partner I mean you know one
of the things that I've always looked at

90
00:04:42,240 --> 00:04:45,277
over my years at Microsoft is look for

91
00:04:45,360 --> 00:04:48,797
high ambition um technology innovation

92
00:04:48,880 --> 00:04:52,797
companies right whether it you know and
partnership like whether it's Intel and

93
00:04:52,880 --> 00:04:55,317
Windows came together and and that was

94
00:04:55,400 --> 00:04:58,797
successful. SAP and SQL server it came

95
00:04:58,880 --> 00:05:03,197
together and it was successful. So I'm
always looking for partners that we can

96
00:05:03,280 --> 00:05:05,517
innovate with and that's what I found in

97
00:05:05,600 --> 00:05:08,637
uh Sam and team and uh and we you know

98
00:05:08,720 --> 00:05:10,797
at that time it was not like it was a it

99
00:05:10,880 --> 00:05:14,477
was a real uh shot in the dark right it

100
00:05:14,560 --> 00:05:17,277
is not like oh wow this is a sure sure

101
00:05:17,360 --> 00:05:22,317
thing everybody now talks about it as if
uh you know this is the issue with tech

102
00:05:22,400 --> 00:05:24,157
right which is uh long before it's

103
00:05:24,240 --> 00:05:25,837
conventional wisdom you have to be all

104
00:05:25,920 --> 00:05:29,437
in and hope it works um and this is one

105
00:05:29,520 --> 00:05:31,517
of those things where we backed it long

106
00:05:31,600 --> 00:05:36,637
before it was conventional wisdom. Um,
and here we are. But I don't take, you

107
00:05:36,720 --> 00:05:39,677
know, like there's going to be severe
amount of competition. You know,

108
00:05:39,760 --> 00:05:44,397
Google's a very competent company
and obviously they have uh both the talent

109
00:05:44,480 --> 00:05:48,797
and the compute and they have, you know,
they're the vertically integrated player

110
00:05:48,880 --> 00:05:50,317
in this, right? They have everything

111
00:05:50,400 --> 00:05:53,757
from data to silicon to models to have

112
00:05:53,840 --> 00:05:58,157
products and distribution and there's
others as well. And so yeah, we will

113
00:05:58,240 --> 00:06:00,397
have significant amount of competition

114
00:06:00,480 --> 00:06:02,877
and I think if anything Microsoft's

115
00:06:02,960 --> 00:06:07,677
partnership with open AI is bringing
more competition to otherwise what would

116
00:06:07,760 --> 00:06:09,437
have been a default Google should have

117
00:06:09,520 --> 00:06:13,437
be the default winner. Uh and if we

118
00:06:13,520 --> 00:06:17,517
partner well and we innovate well we
can bring some competition to that. So if

119
00:06:17,600 --> 00:06:19,357
you look three to five years where do

120
00:06:19,440 --> 00:06:21,117
you think Microsoft's going to be in the

121
00:06:21,200 --> 00:06:24,717
whole AI ecosystem? So to me I think

122
00:06:24,800 --> 00:06:27,357
about this in the fullness of the stack

123
00:06:27,440 --> 00:06:30,637
right so I want us to have first and

124
00:06:30,720 --> 00:06:34,717
foremost the best AI infrastructure

125
00:06:34,800 --> 00:06:39,357
uh so that means when it comes to Azure
whether it's for training whether it's

126
00:06:39,440 --> 00:06:43,357
for inference to have fantastic
infrastructure we'll partner with Nvidia

127
00:06:43,440 --> 00:06:47,757
we'll partner with AMD we'll have
our own silicon uh we will have our own

128
00:06:47,840 --> 00:06:49,997
system architecture we will take the

129
00:06:50,080 --> 00:06:52,317
best system architecture innovation from

130
00:06:52,400 --> 00:06:54,877
Jensen and Lisa and others who may come

131
00:06:54,960 --> 00:06:57,837
along and make sure Azure is serving the

132
00:06:57,920 --> 00:07:01,037
needs of open AI serving the needs of

133
00:07:01,120 --> 00:07:04,957
Mistral uh serving the needs of FI that

134
00:07:05,040 --> 00:07:08,637
we are building right which is the small
language model so that's kind of the

135
00:07:08,720 --> 00:07:11,117
first uh thing that we want to do which

136
00:07:11,200 --> 00:07:14,557
is the best work um in being able to

137
00:07:14,640 --> 00:07:16,477
build the infrastructure out for both

138
00:07:16,560 --> 00:07:18,797
training and inference and then the next

139
00:07:18,880 --> 00:07:21,597
layer up we want to have like the ENT

140
00:07:21,680 --> 00:07:23,997
entire data tier right so you can

141
00:07:24,080 --> 00:07:25,837
imagine as these models and model

142
00:07:25,920 --> 00:07:29,437
capabilities become uh you know more

143
00:07:29,520 --> 00:07:32,237
capable I think the data tier will be

144
00:07:32,320 --> 00:07:35,277
completely uh redone uh right we've

145
00:07:35,360 --> 00:07:39,357
talked with the retrieval augmented
generation already you have all these

146
00:07:39,440 --> 00:07:43,917
things where whether it's embeddings
vector search how do you chunk data such

147
00:07:44,000 --> 00:07:45,917
that retrieval augmented generation can

148
00:07:46,000 --> 00:07:51,037
work well so that's an entire layer or
when context lens become bigger that's a

149
00:07:51,120 --> 00:07:52,877
different sort of data layer like what's

150
00:07:52,960 --> 00:07:55,677
the throughput between data and your

151
00:07:55,760 --> 00:07:58,237
inference fleet uh how do you sort of

152
00:07:58,320 --> 00:08:00,317
think about that so therefore we will

153
00:08:00,400 --> 00:08:02,717
innovate on the data layer and then of

154
00:08:02,800 --> 00:08:05,757
course on top of it is where um we will

155
00:08:05,840 --> 00:08:08,717
innovate on our uh co-pilots one of the

156
00:08:08,800 --> 00:08:10,957
first products we built was GitHub

157
00:08:11,040 --> 00:08:13,437
copilot in fact my entire confidence in

158
00:08:13,520 --> 00:08:15,997
this generation of AI started when I

159
00:08:16,080 --> 00:08:20,957
started seeing from GPD3 to 35
and that implementation in GitHub

160
00:08:21,040 --> 00:08:25,357
uh and so we now have you know not only
GitHub co-pilot we have co-pilot for m

161
00:08:25,440 --> 00:08:27,837
all knowledge work in Microsoft 365 we

162
00:08:27,920 --> 00:08:32,557
have co-pilots for these functions
whether it's service or sales or finance

163
00:08:32,640 --> 00:08:37,037
uh so we're going to innovate in our app
layer uh on our own and so that's I

164
00:08:37,120 --> 00:08:41,517
think fundamentally how I look at it
it's a full stack approach uh and each

165
00:08:41,600 --> 00:08:45,437
layer by the way we're we will innovate
we will have partners we will have

166
00:08:45,520 --> 00:08:49,437
others innovating there will be
competition even it's not like you know

167
00:08:49,520 --> 00:08:51,437
one of the of being a platform company

168
00:08:51,520 --> 00:08:53,837
is you got to be comfortable uh with

169
00:08:53,920 --> 00:08:56,557
many third parties competing with you on

170
00:08:56,640 --> 00:08:58,957
different layers because that to me is

171
00:08:59,040 --> 00:09:03,917
core otherwise you kind of try to you
know do everything in a monolithic way

172
00:09:04,000 --> 00:09:08,157
and at least what we've learned over the
years is the best thing to do is to keep

173
00:09:08,240 --> 00:09:11,597
each layer competitive on its own.
You said in the beginning that we are year

174
00:09:11,680 --> 00:09:15,197
two in this paradigm shift. How do you
see it compared to other technological

175
00:09:15,280 --> 00:09:17,757
breakthrough that you have witnessed? So

176
00:09:17,840 --> 00:09:20,637
at least the four I've seen is um I

177
00:09:20,720 --> 00:09:23,917
obviously PC client server both what

178
00:09:24,000 --> 00:09:27,677
happened on the you know PC and the
server side that was my first that's

179
00:09:27,760 --> 00:09:29,757
kind of when I joined 92 we're at the

180
00:09:29,840 --> 00:09:32,477
beginning of that uh then there was the

181
00:09:32,560 --> 00:09:35,677
web internet um and then there was

182
00:09:35,760 --> 00:09:40,317
mobile cloud and so AI is the fourth um

183
00:09:40,400 --> 00:09:45,277
I think one of the interesting things
is each one of these built on the previous

184
00:09:45,360 --> 00:09:48,957
right so I don't think the web would
have happened if there was not a

185
00:09:49,040 --> 00:09:50,957
ubiquitous PC. After all, the first time

186
00:09:51,040 --> 00:09:55,357
I saw Mosaic uh you know was as a

187
00:09:55,440 --> 00:09:57,437
browser on top of Windows, right? So,

188
00:09:57,520 --> 00:10:00,557
and then Netscape came about and then uh

189
00:10:00,640 --> 00:10:02,477
IE and what have you. And so therefore,

190
00:10:02,560 --> 00:10:05,957
I think um you sort of see each one of

191
00:10:06,040 --> 00:10:09,437
these births the next uh and then it

192
00:10:09,520 --> 00:10:12,877
goes beyond uh what birthed you, right?

193
00:10:12,960 --> 00:10:16,237
That's I think the the real thing, right?
Right now we're seeing that right

194
00:10:16,320 --> 00:10:19,197
which is the cloud um as we know of it

195
00:10:19,280 --> 00:10:22,717
and mobile uh and PCs on the edge have

196
00:10:22,800 --> 00:10:25,597
really birthed the AI age and the

197
00:10:25,680 --> 00:10:30,477
question is what happens next uh right
which is does it go beyond that and that

198
00:10:30,560 --> 00:10:31,997
I think is going I mean there's going to

199
00:10:32,080 --> 00:10:35,117
be AI that is not just about cognitive

200
00:10:35,200 --> 00:10:37,997
work AI that is also going to accelerate

201
00:10:38,080 --> 00:10:40,237
science so I think that that's exciting

202
00:10:40,320 --> 00:10:43,197
space uh AI that is going to be embodied

203
00:10:43,280 --> 00:10:47,677
in the real world so what may happen
in robotics is an exciting space and so

204
00:10:47,760 --> 00:10:51,757
there is a bunch of things that are
going to happen over time. Uh but

205
00:10:51,840 --> 00:10:54,637
clearly yes this is year two of a

206
00:10:54,720 --> 00:10:59,437
complete new paradigm that obviously
reinforces what happened pre or builds

207
00:10:59,520 --> 00:11:03,757
on what happened previously but also
also is showing early signs of what

208
00:11:03,840 --> 00:11:07,757
happens next. If there is no shortage
in chips would development have gone

209
00:11:07,840 --> 00:11:11,117
faster? That's a good
question in the sense of

210
00:11:11,200 --> 00:11:14,477
scaling laws, right? So there are two
sides to it, right? There's the the

211
00:11:14,560 --> 00:11:16,557
training side and the inference side. On

212
00:11:16,640 --> 00:11:19,597
the training side, clearly uh compute

213
00:11:19,680 --> 00:11:22,637
and compute scale and the scaling laws

214
00:11:22,720 --> 00:11:25,837
have proven to be very successful. U and

215
00:11:25,920 --> 00:11:28,317
the question is how long does that go?

216
00:11:28,400 --> 00:11:30,157
Is there going to be another model

217
00:11:30,240 --> 00:11:32,397
architecture breakthrough or what have

218
00:11:32,480 --> 00:11:35,757
you. Um I I think that one has to see.

219
00:11:35,840 --> 00:11:38,717
It's it's unclear. uh but we definitely

220
00:11:38,800 --> 00:11:42,957
are not going to bet against scaling
laws and nor are we going to bet against

221
00:11:43,040 --> 00:11:47,037
uh or nor are we going to say that
this is the last model architecture

222
00:11:47,120 --> 00:11:51,677
breakthrough. In fact, a great example
of it is even uh what we we've seen even

223
00:11:51,760 --> 00:11:53,677
with uh our small language model

224
00:11:53,760 --> 00:11:55,757
innovation like FI, right? Which is we

225
00:11:55,840 --> 00:11:59,277
are able to create what is significant

226
00:11:59,360 --> 00:12:02,637
um capabilities uh in a small language

227
00:12:02,720 --> 00:12:07,517
model uh which uh doesn't require
obviously the same amount of compute

228
00:12:07,600 --> 00:12:09,437
right which is uh like just like

229
00:12:09,520 --> 00:12:12,077
attention is everything attention is all

230
00:12:12,160 --> 00:12:14,797
you need textbooks is all you need. I

231
00:12:14,880 --> 00:12:17,517
mean that sort of intuitively speaks to

232
00:12:17,600 --> 00:12:19,677
uh how I think co you know learning can

233
00:12:19,760 --> 00:12:22,717
happen and so therefore I would say uh

234
00:12:22,800 --> 00:12:25,917
yes if more more more capacity there is

235
00:12:26,000 --> 00:12:28,637
in the world the chances are that we

236
00:12:28,720 --> 00:12:31,197
will be able to make progress but at the

237
00:12:31,280 --> 00:12:34,237
same time I wouldn't discount a real

238
00:12:34,320 --> 00:12:35,757
breakthrough

239
00:12:35,840 --> 00:12:38,237
uh in model architecture that doesn't

240
00:12:38,320 --> 00:12:41,357
perhaps require uh the same type of

241
00:12:41,440 --> 00:12:45,517
compute. So that's why I think um
at least I want to keep myself open-minded

242
00:12:45,600 --> 00:12:48,717
about it. Talking of small language model,
do you think a small country like

243
00:12:48,800 --> 00:12:52,317
the UEE should develop
their own language model?

244
00:12:52,400 --> 00:12:56,877
You know, one of the things that I've
studied, there's an economist out of

245
00:12:56,960 --> 00:12:58,397
Dartmouth,

246
00:12:58,480 --> 00:13:01,277
uh, his name is Diego Comman, and he did

247
00:13:01,360 --> 00:13:04,237
one of the best longitudinal studies,

248
00:13:04,320 --> 00:13:07,997
uh, of technology diffusion. Uh and the

249
00:13:08,080 --> 00:13:11,117
fundamental conclusion um paraphrasing

250
00:13:11,200 --> 00:13:13,597
was that any country that wants to get

251
00:13:13,680 --> 00:13:16,557
ahead should one first make sure that

252
00:13:16,640 --> 00:13:19,117
they don't reinvent the wheel which is

253
00:13:19,200 --> 00:13:21,917
they import the best general purpose

254
00:13:22,000 --> 00:13:26,077
technology that is available in and then

255
00:13:26,160 --> 00:13:28,717
on top of it build value add. So there

256
00:13:28,800 --> 00:13:32,397
are ways you can add unique knowledge,

257
00:13:32,480 --> 00:13:36,477
unique value on top of even what exists

258
00:13:36,560 --> 00:13:38,717
uh before you go off and say let's build

259
00:13:38,800 --> 00:13:43,837
all the compute all of it and do the same
pre-training run u you know there's

260
00:13:43,920 --> 00:13:48,237
no nothing stopping any country from doing
any of this uh but the question is

261
00:13:48,320 --> 00:13:52,317
what is the value ad and so therefore
I think that you start from the value ad

262
00:13:52,400 --> 00:13:55,437
and then back into whatever is needed
for it when you think about how

263
00:13:55,520 --> 00:13:59,757
important the big technology companies
have become in geopolitics. What kind of

264
00:13:59,840 --> 00:14:02,237
reflections do you make? Well, I mean at

265
00:14:02,320 --> 00:14:06,957
the end of the day, um two things,
right? One is I'm very very grounded on

266
00:14:07,040 --> 00:14:10,557
the fact that we are a multinational
company that is a definitely in this

267
00:14:10,640 --> 00:14:13,517
case a US-based multinational company

268
00:14:13,600 --> 00:14:16,237
that has to earn permission and license

269
00:14:16,320 --> 00:14:18,637
to operate one country at a time. Right?

270
00:14:18,720 --> 00:14:21,917
So therefore, I I think of the uh we're

271
00:14:22,000 --> 00:14:24,717
not nation states are the ones that have

272
00:14:24,800 --> 00:14:28,557
power. we get to operate in any nation

273
00:14:28,640 --> 00:14:33,357
based on our ability to contribute
to that country's progress right so

274
00:14:33,440 --> 00:14:35,357
whenever I am right whether I'm in

275
00:14:35,440 --> 00:14:39,597
Norway or I'm in Jakarta or or New Delhi

276
00:14:39,680 --> 00:14:43,837
or wherever I I am always grounded on uh

277
00:14:43,920 --> 00:14:46,157
fundamentally the fact is are we able to

278
00:14:46,240 --> 00:14:48,877
make you know look the local politicians

279
00:14:48,960 --> 00:14:51,597
and political leadership and society at

280
00:14:51,680 --> 00:14:56,237
large in the eyes and say that we
contributed to their public sector

281
00:14:56,320 --> 00:15:00,397
becoming more efficient or large
multinational companies in the region

282
00:15:00,480 --> 00:15:04,637
becoming more globally competitive
because of some tech input education

283
00:15:04,720 --> 00:15:07,357
outcome health outcomes small businesses

284
00:15:07,440 --> 00:15:09,677
and their productivity at the end of the

285
00:15:09,760 --> 00:15:11,997
day you have your your social contract

286
00:15:12,080 --> 00:15:15,197
with the country uh comes from your

287
00:15:15,280 --> 00:15:19,517
ability to contribute to their local
progress and that so I don't think we we

288
00:15:19,600 --> 00:15:21,757
can ever be beyond geopolitics will

289
00:15:21,840 --> 00:15:25,037
exist with or without us and our goal

290
00:15:25,120 --> 00:15:29,837
has to be how do you participate and have
permission to operate in your book

291
00:15:29,920 --> 00:15:31,037
hit your fresh you talk about three

292
00:15:31,120 --> 00:15:35,197
areas um AI quantum computing and mixed

293
00:15:35,280 --> 00:15:39,517
reality what kind of opportunities do
you see in the quantum computing it's

294
00:15:39,600 --> 00:15:41,597
fascinating in fact I I think of all

295
00:15:41,680 --> 00:15:43,757
those three still right for example when

296
00:15:43,840 --> 00:15:46,557
I think about um mixed reality I think

297
00:15:46,640 --> 00:15:49,397
of it as that's an embodied AI right

298
00:15:49,480 --> 00:15:52,477
effectively whether it's autonomous uh

299
00:15:52,560 --> 00:15:56,157
vehicles robotics or people with glasses

300
00:15:56,240 --> 00:16:01,037
they're all seeing the real world is
the prompt effectively so your real world

301
00:16:01,120 --> 00:16:05,837
understanding so therefore I think
of in the generation of AI some of these

302
00:16:05,920 --> 00:16:09,677
things become even more interesting and
more important and we do need to broaden

303
00:16:09,760 --> 00:16:12,157
the aperture versus thinking narrowly of

304
00:16:12,240 --> 00:16:15,757
just one device or one form factor

305
00:16:15,840 --> 00:16:20,317
similarly quantum is a fascinating thing
when I I me mention science right one of

306
00:16:20,400 --> 00:16:21,837
the things I look at it and say is in

307
00:16:21,920 --> 00:16:24,077
order to make progress on science You

308
00:16:24,160 --> 00:16:27,757
need great incilico simulation. Quantum

309
00:16:27,840 --> 00:16:31,197
is the ultimate breakthrough. Right?
So when we have a complete new system

310
00:16:31,280 --> 00:16:33,437
architecture that go breaks free of the

311
00:16:33,520 --> 00:16:37,117
one nimon limitation. Uh you are then

312
00:16:37,200 --> 00:16:40,957
finally going to be able to simulate
something like the dynamics of a cell or

313
00:16:41,040 --> 00:16:44,877
a molecule. Right? So that I think when
you can do that then everything else

314
00:16:44,960 --> 00:16:47,117
even in terms of biology becomes more

315
00:16:47,200 --> 00:16:50,077
feasible. The interesting thing is AI is

316
00:16:50,160 --> 00:16:51,997
kind of like an emulator of that

317
00:16:52,080 --> 00:16:54,317
simulator. In other words, you can kind

318
00:16:54,400 --> 00:16:57,677
of simplify the search space. One of the

319
00:16:57,760 --> 00:17:01,517
things we did recently uh was um we have

320
00:17:01,600 --> 00:17:05,357
a of a model uh for material science

321
00:17:05,440 --> 00:17:08,557
which we used to generate a new uh novel

322
00:17:08,640 --> 00:17:10,637
compound which we then went and

323
00:17:10,720 --> 00:17:12,477
manufactured. We worked with the Pacific

324
00:17:12,560 --> 00:17:15,677
National Lab locally here to effectively

325
00:17:15,760 --> 00:17:20,397
go reduce the uh the the lithium content

326
00:17:20,480 --> 00:17:23,917
uh by 70% in a new battery material uh

327
00:17:24,000 --> 00:17:26,157
and produced it right not just

328
00:17:26,240 --> 00:17:28,397
conceptualized it but simulated it

329
00:17:28,480 --> 00:17:31,837
produced it and so to me uh something

330
00:17:31,920 --> 00:17:35,437
where like quantum plus AI I think can

331
00:17:35,520 --> 00:17:38,557
be the ultimate accelerator of science

332
00:17:38,640 --> 00:17:42,477
uh and we are making progress we will
even On quantum we're taking a very full

333
00:17:42,560 --> 00:17:45,117
stack approach. Uh we have our software

334
00:17:45,200 --> 00:17:49,277
stack with uh you know our Q sharp which

335
00:17:49,360 --> 00:17:53,117
is our quantum programming stack. So we
are excited about sort of the progress

336
00:17:53,200 --> 00:17:55,277
we're making on quantum and how it

337
00:17:55,360 --> 00:17:57,677
complements AI and where does gaming fit

338
00:17:57,760 --> 00:18:02,637
into this? Yeah. So to me one of the
things in fact Microsoft was in gaming

339
00:18:02,720 --> 00:18:06,957
long before we you know we were into
Windows. In fact simulator I think was

340
00:18:07,040 --> 00:18:11,117
launched long before Windows was launched.
So uh we are very very excited

341
00:18:11,200 --> 00:18:14,957
obviously now with Activision as part
of Microsoft we have mobile gaming we have

342
00:18:15,040 --> 00:18:17,597
cloud gaming we have console gaming PC

343
00:18:17,680 --> 00:18:19,757
gaming so we're a full stack game pub

344
00:18:19,840 --> 00:18:22,677
publisher uh as well as a game systems

345
00:18:22,760 --> 00:18:26,637
provider um and so our goal there is one

346
00:18:26,720 --> 00:18:30,877
we're in gaming for our love of gaming
right so I always sort of say we should

347
00:18:30,960 --> 00:18:34,477
never be in businesses as a means
to some other end it has to be an end

348
00:18:34,560 --> 00:18:38,637
otherwise it's not a business so to me
gaming uh is something where we want to

349
00:18:38,720 --> 00:18:41,277
ing joy of gaming. That's the one pure

350
00:18:41,360 --> 00:18:44,477
consumer entertainment category. I love

351
00:18:44,560 --> 00:18:48,957
the fact that gaming on a secular basis
is probably going to be if not already

352
00:18:49,040 --> 00:18:53,357
the biggest entertainment category out
there. So that's one. And of course it

353
00:18:53,440 --> 00:18:55,837
has real implications on the rest of it.

354
00:18:55,920 --> 00:18:59,437
Right? If you think about even remember

355
00:18:59,520 --> 00:19:02,877
it's interesting I've not talked
to Jensen about it but one of the greatest

356
00:19:02,960 --> 00:19:05,877
sort of successes of uh

357
00:19:05,960 --> 00:19:09,517
GPUs was fostered by innovation in

358
00:19:09,600 --> 00:19:12,957
gaming right DX which was the Microsoft

359
00:19:13,040 --> 00:19:17,037
graphics stack is what made GPUs uh an

360
00:19:17,120 --> 00:19:19,677
accelerator uh right after all the GPUs

361
00:19:19,760 --> 00:19:22,397
were created for PC gaming you know

362
00:19:22,480 --> 00:19:25,677
gaming I think as both a an application

363
00:19:25,760 --> 00:19:28,957
of AI in terms of like game testing. One

364
00:19:29,040 --> 00:19:33,037
of the first areas I'm very excited
about is games are so complicated right

365
00:19:33,120 --> 00:19:35,037
now. One of the things that we want to

366
00:19:35,120 --> 00:19:37,437
use AI is to be able to find these bugs

367
00:19:37,520 --> 00:19:40,477
in these even closed worlds uh before

368
00:19:40,560 --> 00:19:44,717
they're out there. And so therefore we
have some very great use cases there. Uh

369
00:19:44,800 --> 00:19:47,117
but beyond that I think gaming as uh

370
00:19:47,200 --> 00:19:49,837
data in the context of some of the

371
00:19:49,920 --> 00:19:54,397
innovation uh in our models I think
is going to be important. Do you game

372
00:19:54,480 --> 00:19:56,237
yourself?

373
00:19:56,320 --> 00:19:59,517
I'm a, you know, a light gamer. I used

374
00:19:59,600 --> 00:20:02,557
to, uh, do a lot more of Civ was my

375
00:20:02,640 --> 00:20:04,557
favorite game. Age of Empires is a

376
00:20:04,640 --> 00:20:07,277
another great game that I enjoyed. I

377
00:20:07,360 --> 00:20:12,397
wish I could play more. Uh, but, you know,
from time to time I slip into it.

378
00:20:12,480 --> 00:20:15,517
You have really changed the culture
in Microsoft. When you look back, what are

379
00:20:15,600 --> 00:20:19,357
the most important changes you have made?

380
00:20:19,440 --> 00:20:21,677
Look, I mean, first of all, I've, as I

381
00:20:21,760 --> 00:20:23,997
said, I've grown up. All my professional

382
00:20:24,080 --> 00:20:26,397
career for most part is all Microsoft.

383
00:20:26,480 --> 00:20:29,677
So, uh, you know, when you say I've been

384
00:20:29,760 --> 00:20:31,997
at Microsoft for 32 years, all the good,

385
00:20:32,080 --> 00:20:34,877
the bad, uh, I was part of it, right?

386
00:20:34,960 --> 00:20:39,917
So, I don't sort of somehow think
that I'm I don't represent all errors of

387
00:20:40,000 --> 00:20:43,517
Microsoft. Yeah. Which even makes it
more incredible that you have made all

388
00:20:43,600 --> 00:20:47,277
these changes. Yeah. And and the way
though I came about it is quite frankly

389
00:20:47,360 --> 00:20:49,437
as a consumate insider I basically

390
00:20:49,520 --> 00:20:51,997
pattern matched as to hey what were the

391
00:20:52,080 --> 00:20:55,037
thing when were we at our best uh what

392
00:20:55,120 --> 00:20:57,677
was the cultural set of attributes that

393
00:20:57,760 --> 00:21:00,077
helped us succeed and then when we

394
00:21:00,160 --> 00:21:02,637
failed what were the cultural attributes

395
00:21:02,720 --> 00:21:05,117
that caused that failure and then you

396
00:21:05,200 --> 00:21:08,397
know dampen this ladder and amplify the

397
00:21:08,480 --> 00:21:10,717
former that was as simple as that. So

398
00:21:10,800 --> 00:21:13,997
one of the things I look back even in my

399
00:21:14,080 --> 00:21:17,997
you know career at Microsoft when we
first became the largest market cap

400
00:21:18,080 --> 00:21:22,717
company I forget I think in the early
2000s I think we crossed GE you know

401
00:21:22,800 --> 00:21:26,637
people on our campus probably were
walking around including me thinking oh

402
00:21:26,720 --> 00:21:31,277
we must be God's gifts uh because we're
so brilliant and what have you except

403
00:21:31,360 --> 00:21:33,277
what we needed to be grounded on that

404
00:21:33,360 --> 00:21:37,077
day was to say wow we now have a real

405
00:21:37,160 --> 00:21:40,397
responsibility to regground ourselves to

406
00:21:40,480 --> 00:21:44,717
innovate again uh so that we are
relevant in the future right and so

407
00:21:44,800 --> 00:21:47,357
that's why I I was lucky enough to have

408
00:21:47,440 --> 00:21:50,717
read Carol Dwek's book on mindset which

409
00:21:50,800 --> 00:21:55,597
is around child psychology called growth
mindset and I love that book I read it

410
00:21:55,680 --> 00:21:57,357
more on the context of sort of our

411
00:21:57,440 --> 00:22:00,077
children's education but I must say I

412
00:22:00,160 --> 00:22:03,117
got educated because I felt like this is

413
00:22:03,200 --> 00:22:06,397
uh what makes uh individuals children in

414
00:22:06,480 --> 00:22:10,957
school it's very clear right it's
better to be a learn it all versus a

415
00:22:11,040 --> 00:22:12,957
know-it-all because even if the knowit

416
00:22:13,040 --> 00:22:16,477
all has great innate capability

417
00:22:16,560 --> 00:22:21,037
uh the learn it all you know even if
they start from behind they will surpass

418
00:22:21,120 --> 00:22:23,437
the knowit-all right that's sort of uh

419
00:22:23,520 --> 00:22:25,517
you know true for children in school

420
00:22:25,600 --> 00:22:28,797
it's true for CEOs in my seat uh it's

421
00:22:28,880 --> 00:22:31,757
true for companies and so we took that

422
00:22:31,840 --> 00:22:35,197
approach we said let's be a learn it all

423
00:22:35,280 --> 00:22:39,437
versus a know-it-all uh and the A you

424
00:22:39,520 --> 00:22:41,557
say you've achieved that cultural

425
00:22:41,640 --> 00:22:43,757
transformation means you're a

426
00:22:43,840 --> 00:22:47,277
know-it-all. So therefore it is a good

427
00:22:47,360 --> 00:22:51,437
way to sort of say every day you make
a bunch of mistakes you at least have the

428
00:22:51,520 --> 00:22:56,317
courage to acknowledge them and continue
on it. And so it's a it's not a

429
00:22:56,400 --> 00:23:00,077
destination you ever reach. But how do
you get the organization to buy into it?

430
00:23:00,160 --> 00:23:02,877
It's a beautiful it's a great point. So

431
00:23:02,960 --> 00:23:07,437
I think the way I I I think see
the problem of corporations especially for

432
00:23:07,520 --> 00:23:11,997
non-founder companies right founders
have great power and great followership

433
00:23:12,080 --> 00:23:16,317
and that's why I think they're so
successful or at least you know at least

434
00:23:16,400 --> 00:23:18,077
we only talk about the successful

435
00:23:18,160 --> 00:23:20,237
founders but if you set that that class

436
00:23:20,320 --> 00:23:24,077
aside for me mortals like me uh it can't

437
00:23:24,160 --> 00:23:27,117
be like okay new dogma from a new CEO

438
00:23:27,200 --> 00:23:29,117
and more corporate speak it has to

439
00:23:29,200 --> 00:23:31,597
appeal to us as human beings that's why

440
00:23:31,680 --> 00:23:36,157
I you I credit more of this work uh by

441
00:23:36,240 --> 00:23:39,997
Carol and team and so on because it's
not like it was not like I don't think

442
00:23:40,080 --> 00:23:42,077
anybody at Microsoft views growth

443
00:23:42,160 --> 00:23:44,797
mindset as some Microsoft dogma or

444
00:23:44,880 --> 00:23:47,757
definitely not Satia Nadella dogma it is

445
00:23:47,840 --> 00:23:49,917
something that speaks to them as humans

446
00:23:50,000 --> 00:23:51,917
right which is it's good for them as

447
00:23:52,000 --> 00:23:55,157
friends colleagues partners parents

448
00:23:55,240 --> 00:23:58,077
neighbors it integrates work and life

449
00:23:58,160 --> 00:24:00,797
they can bring their own personality and

450
00:24:00,880 --> 00:24:04,237
passion to it uh and benefit from it,

451
00:24:04,320 --> 00:24:07,837
right? I I always say like you're not
do like this Microsoft culture of growth

452
00:24:07,920 --> 00:24:11,677
mindset is not for Microsoft, it's for
you. And if you you should only practice

453
00:24:11,760 --> 00:24:13,997
it if you feel like it speaks to your

454
00:24:14,080 --> 00:24:17,837
own thriving at Microsoft and in life.

455
00:24:17,920 --> 00:24:22,477
Uh that's I think what I attribute it to,
right? It wouldn't have taken off if

456
00:24:22,560 --> 00:24:24,597
it was just another thing that is a

457
00:24:24,680 --> 00:24:27,197
top-down slogan.

458
00:24:27,280 --> 00:24:29,997
uh it is ultim I always have believed

459
00:24:30,080 --> 00:24:33,517
that which is ultimately people work um

460
00:24:33,600 --> 00:24:38,077
and find meaning in work uh only if they

461
00:24:38,160 --> 00:24:42,717
can find some true deeper meaning
for themselves and so that's I think what

462
00:24:42,800 --> 00:24:46,237
I've been always trying to invoke
and well they clearly also found a true

463
00:24:46,320 --> 00:24:50,637
deeper meaning in the concept of empathy
because you talk a lot about it and you

464
00:24:50,720 --> 00:24:54,477
say it's the key for innovation
and leadership and why do you think it's

465
00:24:54,560 --> 00:24:59,037
very important to you I think Most
people think of empathy as some kind of

466
00:24:59,120 --> 00:25:01,117
a soft skill that's interesting in the

467
00:25:01,200 --> 00:25:03,677
context of your family or personal life

468
00:25:03,760 --> 00:25:06,957
and work's all about hardcore you know

469
00:25:07,040 --> 00:25:09,277
whatever right but that but I I look at

470
00:25:09,360 --> 00:25:12,557
that and say again what is where does

471
00:25:12,640 --> 00:25:14,717
innovation come from right innovation

472
00:25:14,800 --> 00:25:18,557
comes from us being able to drive the

473
00:25:18,640 --> 00:25:22,637
solutions to unmet unarticulated needs

474
00:25:22,720 --> 00:25:25,197
of customers out there right so the key

475
00:25:25,280 --> 00:25:29,757
being unmet meta and unarticulated.
So that means you have to have a better

476
00:25:29,840 --> 00:25:31,597
sense when you're even looking at some

477
00:25:31,680 --> 00:25:35,517
log data or you know some customer

478
00:25:35,600 --> 00:25:40,397
interview data or whatever you it's not
just the words that they're saying but

479
00:25:40,480 --> 00:25:42,117
you got to be able to walk in their

480
00:25:42,200 --> 00:25:45,197
shoes. Uh and the good news there is

481
00:25:45,280 --> 00:25:47,517
this is innate in us all human beings.

482
00:25:47,600 --> 00:25:50,637
We have uh the ability to empathize with

483
00:25:50,720 --> 00:25:54,077
the other person. Um in fact design

484
00:25:54,160 --> 00:25:58,397
thinking is that right? So when people
go and say let's do learn about design

485
00:25:58,480 --> 00:26:00,357
thinking design thinking is applied

486
00:26:00,440 --> 00:26:04,397
empathy. Uh and so to me uh that's what

487
00:26:04,480 --> 00:26:08,157
I pulled the thread on which is let's
not think of empathy as something that

488
00:26:08,240 --> 00:26:11,677
you know is just a soft skill that you
reserve for your friends and family but

489
00:26:11,760 --> 00:26:13,677
it's I think it's at the root of all

490
00:26:13,760 --> 00:26:16,637
innovation. Uh it's about being able to

491
00:26:16,720 --> 00:26:19,437
meet the unmet unarticulated needs that

492
00:26:19,520 --> 00:26:21,837
comes from your unique I mean your

493
00:26:21,920 --> 00:26:24,477
innate ability to have curiosity to

494
00:26:24,560 --> 00:26:28,397
learn about others walk in their shoes

495
00:26:28,480 --> 00:26:31,117
innovate on their behalf uh and that I

496
00:26:31,200 --> 00:26:35,757
think is I think what we have to do
and that's why I I think empathy is an

497
00:26:35,840 --> 00:26:37,597
important important skill for all of us.

498
00:26:37,680 --> 00:26:39,277
When did you first discover the power of

499
00:26:39,360 --> 00:26:42,157
empathy? Yeah. And you know, I think

500
00:26:42,240 --> 00:26:45,277
that one of the things that I I feel

501
00:26:45,360 --> 00:26:49,037
like all of us learn how to turn on this

502
00:26:49,120 --> 00:26:50,957
bit of empathy through life's

503
00:26:51,040 --> 00:26:53,357
experience, right? So, in some sense,

504
00:26:53,440 --> 00:26:58,157
every day you get confronted with

505
00:26:58,240 --> 00:27:00,157
different circumstances, not just

506
00:27:00,240 --> 00:27:02,397
yourself personally, but people around

507
00:27:02,480 --> 00:27:06,077
you. Uh for me obviously the birth of my

508
00:27:06,160 --> 00:27:09,757
son uh for both my wife and me was uh

509
00:27:09,840 --> 00:27:11,997
was a life-changing event and it was

510
00:27:12,080 --> 00:27:14,877
something that uh you know over the

511
00:27:14,960 --> 00:27:17,997
years I at least learned a lot because I

512
00:27:18,080 --> 00:27:22,397
I I remember in the early days it
was all about sort of my son was born with

513
00:27:22,480 --> 00:27:24,157
cerebal palsy. He passed away a few

514
00:27:24,240 --> 00:27:27,917
years ago. Uh and um uh but you know in

515
00:27:28,000 --> 00:27:32,157
the when the when he was born it
was a lot about what happened to me. I was

516
00:27:32,240 --> 00:27:33,997
sort of you know essentially quote

517
00:27:34,080 --> 00:27:38,077
unquote um you know in you know all

518
00:27:38,160 --> 00:27:42,637
about why did this happen to us,
why did it happen to me or how and then I

519
00:27:42,720 --> 00:27:44,957
realized after watching in some sense my

520
00:27:45,040 --> 00:27:48,157
wife uh who was there as a caregiver, as

521
00:27:48,240 --> 00:27:51,517
a parent uh you know taking him up and

522
00:27:51,600 --> 00:27:54,237
down Seattle to every therapy possible.

523
00:27:54,320 --> 00:27:55,997
Uh quite frankly you know it took me

524
00:27:56,080 --> 00:27:58,637
years not you know days or months or

525
00:27:58,720 --> 00:28:02,637
weeks. Um and um and then I realized

526
00:28:02,720 --> 00:28:05,837
that it nothing had happened to me
but something had happened to my son and I

527
00:28:05,920 --> 00:28:09,197
needed to be there for my son and that

528
00:28:09,280 --> 00:28:13,837
is the experience I talk a lot about but
there's experiences like that every day

529
00:28:13,920 --> 00:28:15,917
right some colleague of mine comes with

530
00:28:16,000 --> 00:28:19,277
some you know parent of theirs who needs

531
00:28:19,360 --> 00:28:22,877
care right that sort of you know I learn

532
00:28:22,960 --> 00:28:25,037
from it I or at least let me put it this

533
00:28:25,120 --> 00:28:28,717
way I am more attuned to learning from

534
00:28:28,800 --> 00:28:33,597
other people's experiences today than I
was in the beginning of my career and I

535
00:28:33,680 --> 00:28:35,117
think that happens to all of us right

536
00:28:35,200 --> 00:28:38,637
which is life's experiences if they they

537
00:28:38,720 --> 00:28:42,877
acrewue uh that ability to build a

538
00:28:42,960 --> 00:28:45,037
deeper empathy for other people and that

539
00:28:45,120 --> 00:28:47,277
also helps you be a better manager a

540
00:28:47,360 --> 00:28:50,317
better co-worker a better uh innovator

541
00:28:50,400 --> 00:28:52,317
thank you for sharing but do you think

542
00:28:52,400 --> 00:28:54,397
there's a contradiction between empathy

543
00:28:54,480 --> 00:28:58,157
and execution I don't right I think that

544
00:28:58,240 --> 00:29:01,677
at the end end of the day to me you have

545
00:29:01,760 --> 00:29:04,157
to take accountability right so this is

546
00:29:04,240 --> 00:29:07,997
one of the things that uh in business um

547
00:29:08,080 --> 00:29:11,277
like in life too right uh you have to be

548
00:29:11,360 --> 00:29:14,157
accountable for making real progress uh

549
00:29:14,240 --> 00:29:16,317
one of the things I think about why do

550
00:29:16,400 --> 00:29:21,277
businesses exist businesses exist at least
I like this colin mayor definition

551
00:29:21,360 --> 00:29:23,157
that you have to create profitable

552
00:29:23,240 --> 00:29:25,677
solution to challenges of people and

553
00:29:25,760 --> 00:29:28,637
planet uh because that's at least a good

554
00:29:28,720 --> 00:29:31,437
way to allocate the global resources

555
00:29:31,520 --> 00:29:36,237
that are available, right? The profit
motive is a good motive because it means

556
00:29:36,320 --> 00:29:38,157
you're competing and allocating

557
00:29:38,240 --> 00:29:41,357
resources in the most efficient ways and

558
00:29:41,440 --> 00:29:45,757
uh and face competition and so therefore
you have to have great execution. have

559
00:29:45,840 --> 00:29:46,837
to have great

560
00:29:46,920 --> 00:29:48,957
accountability. And so I think of

561
00:29:49,040 --> 00:29:53,077
empathy as a necessary condition to

562
00:29:53,160 --> 00:29:57,437
create great solutions that are
profitable and that are competitive

563
00:29:57,520 --> 00:30:01,117
solutions that are winning in the
marketplace as opposed to somehow this

564
00:30:01,200 --> 00:30:04,637
being a trade-off. When you look at your
skills and personality as a leader, what

565
00:30:04,720 --> 00:30:07,757
do you think that makes you most
effective as a leader? I mean, first of

566
00:30:07,840 --> 00:30:10,997
all, right, I I I don't think of this

567
00:30:11,080 --> 00:30:15,477
as I I I don't know. I have causality

568
00:30:15,560 --> 00:30:19,677
here well understood because quite
frankly it's so much easier for others

569
00:30:19,760 --> 00:30:21,597
to opine on this than or rather it's

570
00:30:21,680 --> 00:30:24,077
it's just really for others to judge and

571
00:30:24,160 --> 00:30:27,357
assign uh causality there but the way at

572
00:30:27,440 --> 00:30:31,277
least I come at this is I don't start

573
00:30:31,360 --> 00:30:37,717
with uh what am I good at I am very

574
00:30:37,800 --> 00:30:42,077
keenly shaped by what am I not good at

575
00:30:42,160 --> 00:30:45,277
uh in other words. I'm always looking

576
00:30:45,360 --> 00:30:49,837
what can I learn from someone else.
So if there's one attribute I have I don't

577
00:30:49,920 --> 00:30:51,757
start each day by thinking oh all the

578
00:30:51,840 --> 00:30:53,597
stuff I know and I'm good at. I'm like

579
00:30:53,680 --> 00:30:56,877
wow what am I weak at whom do I talk to

580
00:30:56,960 --> 00:30:59,917
whom do I meet? How do I really shape

581
00:31:00,000 --> 00:31:02,557
the colleagues around me who have better

582
00:31:02,640 --> 00:31:08,157
skills than me on many fronts? uh that's
what I'm wired like maybe that helps but

583
00:31:08,240 --> 00:31:10,957
I don't know whether that's
the causality but I don't start each day

584
00:31:11,040 --> 00:31:14,637
with saying wow I'm so good at this so
therefore I'm going to go do this is

585
00:31:14,720 --> 00:31:17,197
that something that you learned from
Bill Gates because he said something

586
00:31:17,280 --> 00:31:19,277
very similar it's a good point both Bill

587
00:31:19,360 --> 00:31:23,277
and Steve uh there is uh there's a sense

588
00:31:23,360 --> 00:31:26,797
at Microsoft I think that that's an
interesting thing there is you know what

589
00:31:26,880 --> 00:31:28,797
Andy Grove would talk the paranoid

590
00:31:28,880 --> 00:31:32,237
survive or what have you um and I don't

591
00:31:32,320 --> 00:31:37,037
come at it though with paranoia right
I I mean I don't like paranoia. I like

592
00:31:37,120 --> 00:31:41,277
this just that's why I go back to my own
words for this is that's why the the

593
00:31:41,360 --> 00:31:45,757
growth mindset or confronting your own
fixed mindset having confidence that wow

594
00:31:45,840 --> 00:31:50,077
what an unbelievable world we live
in that every customer can teach me every

595
00:31:50,160 --> 00:31:51,997
partner can teach me every colleague can

596
00:31:52,080 --> 00:31:55,037
teach me like what more can I ask for in

597
00:31:55,120 --> 00:31:58,157
life uh so it's not paranoia it's not

598
00:31:58,240 --> 00:32:01,757
like oh wow we have to go in every day

599
00:32:01,840 --> 00:32:05,597
uh that if I don't learn something I'll
you know fail I'm more about like what

600
00:32:05,680 --> 00:32:10,477
do I learn so that I can innovate maybe
uh that's how I come at it and that's

601
00:32:10,560 --> 00:32:15,037
right Bill and Steve in their own unique
way uh had that mindset and so I've

602
00:32:15,120 --> 00:32:19,197
grown up around it how do you solve this
in an organization you know humbleness

603
00:32:19,280 --> 00:32:20,957
is an interesting word right I I always

604
00:32:21,040 --> 00:32:24,317
say that you you need confidence

605
00:32:24,400 --> 00:32:28,477
um and uh humility

606
00:32:28,560 --> 00:32:31,517
uh and not hubris right because there is

607
00:32:31,600 --> 00:32:34,797
sometimes confidence with humility can

608
00:32:34,880 --> 00:32:39,277
allow you to really make great progress
but confidence that translates into

609
00:32:39,360 --> 00:32:41,117
hubris can bring you know it's a

610
00:32:41,200 --> 00:32:43,837
downfall of you know civilizations

611
00:32:43,920 --> 00:32:46,477
empires and individuals right and from

612
00:32:46,560 --> 00:32:48,397
ancient Greece to modern Silicon Valley

613
00:32:48,480 --> 00:32:53,597
and so that's why I think uh you have to
sort of really get that calibration that

614
00:32:53,680 --> 00:32:57,357
you got to have some confidence in your
own capability you said in a podcast

615
00:32:57,440 --> 00:32:59,997
with Adam Grad that your father has a

616
00:33:00,080 --> 00:33:05,277
list of people that he met and a list
of ideas generated do you have a list of

617
00:33:05,360 --> 00:33:08,957
people that you would like to meet.
you know he had this note you know his

618
00:33:09,040 --> 00:33:11,917
diaries were full of that schema uh

619
00:33:12,000 --> 00:33:15,277
which is uh people met uh ideas

620
00:33:15,360 --> 00:33:17,517
generated and tasks completed which I

621
00:33:17,600 --> 00:33:20,077
love which is a beautiful way each day

622
00:33:20,160 --> 00:33:23,277
to keep a count of uh and absolutely uh

623
00:33:23,360 --> 00:33:27,037
so that's sort of like I I took that u

624
00:33:27,120 --> 00:33:33,040
to heart and that's essentially
how that's my framework for life as