1
00:00:50,330 --> 00:00:54,337
>> Good morning.
It is fantastic to be back in

2
00:00:54,420 --> 00:00:56,937
San Francisco build.
It is always fun to be a

3
00:00:57,020 --> 00:00:59,747
developer conference to see

4
00:00:59,830 --> 00:01:00,747
great change.

5
00:01:00,830 --> 00:01:02,747
Developer conferences are always

6
00:01:02,830 --> 00:01:04,277
about understanding tech shifts

7
00:01:04,360 --> 00:01:07,597
and understanding the new stack.

8
00:01:07,680 --> 00:01:10,857
It is also about really
coming to grips with the new

9
00:01:10,940 --> 00:01:12,387
opportunity for us as

10
00:01:12,470 --> 00:01:13,987
developers, for the companies we

11
00:01:14,070 --> 00:01:18,927
work at as well as the broader world.

12
00:01:19,010 --> 00:01:20,627
Today we are going to unpack this.

13
00:01:20,710 --> 00:01:21,947
This conference is all about

14
00:01:22,030 --> 00:01:22,347
that.

15
00:01:22,430 --> 00:01:24,117
If there is one key take away it

16
00:01:24,200 --> 00:01:27,937
would be this, how do you all

17
00:01:28,020 --> 00:01:30,477
participate fully in this

18
00:01:30,560 --> 00:01:33,977
frontier intelligence ecosystem.

19
00:01:34,060 --> 00:01:36,317
It is not about any one piece of

20
00:01:36,400 --> 00:01:37,417
technology you hear about or

21
00:01:37,500 --> 00:01:39,637
even the platform itself, it is

22
00:01:39,720 --> 00:01:41,497
about the value that you can

23
00:01:41,580 --> 00:01:43,617
build, you can compound, you can

24
00:01:43,700 --> 00:01:49,997
create on top of the platform.

25
00:01:50,080 --> 00:01:54,167
That's what the conference will
be about and that's what we will

26
00:01:54,250 --> 00:01:54,727
focus on.

27
00:01:54,810 --> 00:01:56,647
Let's take a look at this AI

28
00:01:56,730 --> 00:01:57,947
stack this conference is going

29
00:01:58,030 --> 00:02:03,607
to unpack in a great amount of

30
00:02:03,690 --> 00:02:06,647
detail. It starts with ubiquitous

31
00:02:06,730 --> 00:02:10,177
compute fabric that expands
the edge and the cloud.

32
00:02:10,260 --> 00:02:11,547
You then have this layer which

33
00:02:11,630 --> 00:02:13,337
is the emerging layer where you

34
00:02:13,420 --> 00:02:15,387
have the models, you have

35
00:02:15,470 --> 00:02:17,287
context, you have the tools, the

36
00:02:17,370 --> 00:02:22,857
models can access, and then you

37
00:02:22,940 --> 00:02:25,697
have the run time where you

38
00:02:25,780 --> 00:02:28,887
deploy the agents
and applications that you build on

39
00:02:28,970 --> 00:02:30,447
top of the content player and

40
00:02:30,530 --> 00:02:31,617
model layer.

41
00:02:31,700 --> 00:02:33,967
Of course you have the best

42
00:02:34,050 --> 00:02:36,437
tooling to do all of this and

43
00:02:36,520 --> 00:02:37,707
you have the security and

44
00:02:37,790 --> 00:02:43,077
compliance and governance.

45
00:02:43,160 --> 00:02:46,297
That's the simplest form.

46
00:02:46,380 --> 00:02:48,467
Let's start where it all starts

47
00:02:48,550 --> 00:02:50,717
which is infrastructure.

48
00:02:50,800 --> 00:02:52,287
In fact let's start at the edge

49
00:02:52,370 --> 00:02:55,127
with windows.

50
00:02:55,210 --> 00:02:57,897
The amount of compute there is,

51
00:02:57,980 --> 00:03:00,427
at the edge is actually

52
00:03:00,510 --> 00:03:00,967
astounding.

53
00:03:01,050 --> 00:03:05,867
Think about every NPU, GPU, CPU

54
00:03:05,950 --> 00:03:11,220
even, every PC

55
00:03:11,460 --> 00:03:18,460
we ask ourselves one simple

56
00:03:25,600 --> 00:03:30,957
question. AI.

57
00:03:31,040 --> 00:03:33,197
In some sense we are
delivering that already.

58
00:03:33,280 --> 00:03:37,717
When you look at something like
outlook summarized it is using

59
00:03:37,800 --> 00:03:41,607
on board AI locally.
Same thing with PowerPoint all text.

60
00:03:41,690 --> 00:03:44,477
Or teams super resolution.

61
00:03:44,560 --> 00:03:47,747
It is just not Microsoft.

62
00:03:47,830 --> 00:03:51,747
It is Adobe effects or premier.

63
00:03:51,830 --> 00:03:58,830
Using windows ML in order to

64
00:04:00,010 --> 00:04:01,357
happen to all of this compute

65
00:04:01,440 --> 00:04:05,297
power is to expand the scope of

66
00:04:05,380 --> 00:04:09,387
windows ML and windows AI.

67
00:04:09,470 --> 00:04:13,367
You now are full install
base of GPUs you can get to.

68
00:04:13,450 --> 00:04:16,977
I am thrilled every developer
out there can count on building

69
00:04:17,060 --> 00:04:18,907
for local on board AI and then

70
00:04:18,990 --> 00:04:22,647
have it run across all of the

71
00:04:22,730 --> 00:04:27,097
install base.
Now if you are -- we are also

72
00:04:27,180 --> 00:04:29,217
announcing two cool models that

73
00:04:29,300 --> 00:04:36,300
are all going to run on windows

74
00:04:38,110 --> 00:04:43,587
in INVOX. Eye on instruct, it is a great

75
00:04:43,670 --> 00:04:43,987
reasoning model.

76
00:04:44,070 --> 00:04:49,437
We have the planning model eye

77
00:04:49,520 --> 00:04:53,777
-- Ion plan. You have a full local agentic

78
00:04:53,860 --> 00:04:55,917
loop you can give it tools

79
00:04:56,000 --> 00:04:59,097
access and build a fully agentic

80
00:04:59,180 --> 00:05:02,387
applications without having to

81
00:05:02,470 --> 00:05:05,587
run trip to the cloud.

82
00:05:05,670 --> 00:05:07,487
In order to really push the

83
00:05:07,570 --> 00:05:11,897
limits, realize this unmetered

84
00:05:11,980 --> 00:05:13,367
intelligence, there is a lot of

85
00:05:13,450 --> 00:05:15,167
hardware that we are adding too.

86
00:05:15,250 --> 00:05:21,567
We are thrilled to see
the innovation through the

87
00:05:21,650 --> 00:05:24,777
ecosystem. The stuff from
Intel is exciting to see.

88
00:05:24,860 --> 00:05:27,747
Qualcomm announced two sets
of things which are great one on

89
00:05:27,830 --> 00:05:31,847
the high end snap dragon X on

90
00:05:31,930 --> 00:05:38,930
the lower end to sub 500PCs.

91
00:05:39,840 --> 00:05:44,357
Of course that brings us to inn

92
00:05:44,440 --> 00:05:48,027
NVIDIA. This is next generation SOC for

93
00:05:48,110 --> 00:05:51,747
PCs. It brings together the CPU the

94
00:05:51,830 --> 00:05:56,007
GPU as well as AI
capabilities into a single SOC.

95
00:05:56,090 --> 00:05:58,807
It includes unified memory,

96
00:05:58,890 --> 00:06:05,890
architecture and also integrated

97
00:06:06,580 --> 00:06:06,697
DRTM.

98
00:06:06,780 --> 00:06:12,987
You have the SOC coming

99
00:06:13,070 --> 00:06:17,227
together. One we build is the Surface

100
00:06:17,310 --> 00:06:19,467
Ultra it is a beautiful device.

101
00:06:19,550 --> 00:06:22,347
It brings the power of inn

102
00:06:22,430 --> 00:06:27,907
NVIDIA together with the design
and Craftsman ship of Surface,

103
00:06:27,990 --> 00:06:30,377
128 giga bites of unified memory.

104
00:06:30,460 --> 00:06:33,407
Beautiful 2,000 display, and it

105
00:06:33,490 --> 00:06:40,490
has an all day battery life.

106
00:06:41,020 --> 00:06:42,887
Excited to see this later in the fall.

107
00:06:42,970 --> 00:06:47,627
It is wonderful to see all
of the designs from all of the OEM

108
00:06:47,710 --> 00:06:51,097
partners building and taking
advantage of the SOC and the new

109
00:06:51,180 --> 00:06:54,067
platform and bringing out pretty

110
00:06:54,150 --> 00:06:58,887
exciting machines that are going
to be all available come this fall.

111
00:06:58,970 --> 00:07:02,277
So, of course, we said,
okay, this is fantastic.

112
00:07:02,360 --> 00:07:07,547
What can we do next?
We said, let's try and push

113
00:07:07,630 --> 00:07:10,817
this. Push the architecture to its

114
00:07:10,900 --> 00:07:12,627
limit for developers.

115
00:07:12,710 --> 00:07:15,517
What if we could just Max the

116
00:07:15,600 --> 00:07:18,287
compute, Max the memory, build

117
00:07:18,370 --> 00:07:23,097
out that developer machine that
is the dream machine, and that's

118
00:07:23,180 --> 00:07:25,227
what we are announcing today.

119
00:07:25,310 --> 00:07:31,747
Surface RTX, dev VOX.

120
00:07:31,830 --> 00:07:38,830
Let's roll the video.

121
00:08:35,630 --> 00:08:37,917
>> It is truly a dream machine.

122
00:08:38,000 --> 00:08:42,567
It has 1 Peta flop of AI compute.

123
00:08:42,650 --> 00:08:48,977
20CPU cores. All of those things
have 128 giga bites of unified memory

124
00:08:49,060 --> 00:08:49,447
access.

125
00:08:49,530 --> 00:08:54,537
So super excited.

126
00:08:54,620 --> 00:08:58,557
I am on the wait list as well,
there's a wait list.

127
00:08:58,640 --> 00:08:59,587
We will get there.

128
00:08:59,670 --> 00:09:00,557
[Laughter]

129
00:09:00,640 --> 00:09:05,697
So then we said why stop there?

130
00:09:05,780 --> 00:09:08,367
We said what if we did one more

131
00:09:08,450 --> 00:09:12,867
thing? In fact I think Jensen did that

132
00:09:12,950 --> 00:09:16,977
already. He said windows is coming to the

133
00:09:17,060 --> 00:09:17,907
DGX station.

134
00:09:17,990 --> 00:09:20,247
I describe it as the desktop

135
00:09:20,330 --> 00:09:24,747
data center which you can have

136
00:09:24,830 --> 00:09:29,617
running a 1 trillion model locally.

137
00:09:29,700 --> 00:09:31,217
It is close to what we perhaps

138
00:09:31,300 --> 00:09:36,827
had when we built GPT-5 or 3 one

139
00:09:36,910 --> 00:09:39,227
of the first super computers.
It is crazy to think we have

140
00:09:39,310 --> 00:09:44,647
come this far with where you
can have a data center on your

141
00:09:44,730 --> 00:09:48,667
desktop. We are extending
the developer end points to the cloud.

142
00:09:48,750 --> 00:09:53,947
Windows 365 has the developer
distribution that is optimized

143
00:09:54,030 --> 00:09:59,167
for developer productivity
in the cloud that's windows 365

144
00:09:59,250 --> 00:10:04,417
that I use every day. It is about making
windows whether it is on the laptop,

145
00:10:04,500 --> 00:10:07,557
desktop, the cloud wherever
the best place to build.

146
00:10:07,640 --> 00:10:10,297
And so to that end we have tons

147
00:10:10,380 --> 00:10:15,197
and tons of updates.
We have starting -- we are

148
00:10:15,280 --> 00:10:19,767
starting with the favorite
things for all of us, which is

149
00:10:19,850 --> 00:10:21,267
distraction 3 dev environment.

150
00:10:21,350 --> 00:10:23,777
We are introducing an

151
00:10:23,860 --> 00:10:29,577
intelligent terminal that has
built in GitHub Copilot.

152
00:10:29,660 --> 00:10:30,947
Copilot intelligence.

153
00:10:31,030 --> 00:10:34,897
Of course lots and lots of Linux

154
00:10:34,980 --> 00:10:38,917
love for windows now.
You have 70 plus utilities.

155
00:10:39,000 --> 00:10:45,097
Grab in full glory is now available
for regular windows access.

156
00:10:45,180 --> 00:10:48,267
70 plus utilities as I said.
All of that is coming to

157
00:10:48,350 --> 00:10:48,667
windows.

158
00:10:48,750 --> 00:10:55,750
We are also bringing that you

159
00:10:58,520 --> 00:10:58,907
love.

160
00:10:58,990 --> 00:11:05,990
Home brew will be on native as

161
00:11:16,270 --> 00:11:18,457
well. Switching all of this.

162
00:11:18,540 --> 00:11:20,727
Having first class support for

163
00:11:20,810 --> 00:11:26,407
containers while will hp --

164
00:11:26,490 --> 00:11:28,467
will really help us when

165
00:11:28,550 --> 00:11:29,937
developing locally.

166
00:11:30,020 --> 00:11:35,247
Showing you all of this on new

167
00:11:35,330 --> 00:11:36,047
surface on RTX.

168
00:11:36,130 --> 00:11:38,377
I wanted to invite up on stage

169
00:11:38,460 --> 00:11:43,487
Kayla to take a spin through all

170
00:11:43,570 --> 00:11:47,640
of the Dev tools. Kayla, go ahead.

171
00:11:47,840 --> 00:11:52,367
>> Our team has been working hard to make
Windows a great place for development.

172
00:11:52,450 --> 00:11:55,887
Today I am going to show
you major improvements and

173
00:11:55,970 --> 00:11:57,217
delighters you can try out today.

174
00:11:57,300 --> 00:11:58,747
What we have here is the default

175
00:11:58,830 --> 00:12:04,637
experience on the Surface
RTX Spark DevBox.

176
00:12:04,720 --> 00:12:09,617
It is calm. No news feed no
widgets popping up no notifications.

177
00:12:09,700 --> 00:12:11,247
We are in dark mode of course.

178
00:12:11,330 --> 00:12:14,167
I am ready to start development.

179
00:12:14,250 --> 00:12:15,367
There's one thing I would like to change.

180
00:12:15,450 --> 00:12:18,487
I like to put my task bar on the left.

181
00:12:18,570 --> 00:12:22,397
Let me jump into my task bar

182
00:12:22,480 --> 00:12:27,997
settings using run which is built
with power tools xand pallet.

183
00:12:28,080 --> 00:12:29,577
No I want it all of the way on the left.

184
00:12:29,660 --> 00:12:32,287
There we go. After popular demand we are

185
00:12:32,370 --> 00:12:37,637
excited to announce vertical
task bar is available in Windows

186
00:12:37,720 --> 00:12:39,277
insider build.

187
00:12:39,360 --> 00:12:43,917
The new surfacSurface RTX Spark

188
00:12:44,000 --> 00:12:46,397
bunch of tools already installed

189
00:12:46,480 --> 00:12:51,787
like python and node.
All developer goodness all in one file.

190
00:12:51,870 --> 00:12:55,057
If you want the same experience
on your device we are making

191
00:12:55,140 --> 00:12:57,527
this file available to everyone right now.

192
00:12:57,610 --> 00:13:02,167
We have a public repo set up
with configuration file and how

193
00:13:02,250 --> 00:13:08,237
to apply it. It will make adjustments
to windows and install all of the tools.

194
00:13:08,320 --> 00:13:12,137
One cool thing I have running
here is power toys new utility

195
00:13:12,220 --> 00:13:16,917
called grab and move. Hold ality
and move the window around from anywhere.

196
00:13:17,000 --> 00:13:21,637
You can enable end task which
let's you end the process

197
00:13:21,720 --> 00:13:23,617
without having to open task manager.

198
00:13:23,700 --> 00:13:26,957
Let me jump into my Dev drive.

199
00:13:27,040 --> 00:13:31,757
It runs with defender running in

200
00:13:31,840 --> 00:13:35,817
async and authorize performance
when comes to developer scenarios.

201
00:13:35,900 --> 00:13:40,237
Get aware. We have stuff like last name

202
00:13:40,320 --> 00:13:43,777
change message status of each
file and the branch name is on

203
00:13:43,860 --> 00:13:45,177
the bottom left.

204
00:13:45,260 --> 00:13:47,607
So now let's get started

205
00:13:47,690 --> 00:13:54,097
building and open our terminal.

206
00:13:54,180 --> 00:13:57,157
It helps working with agents
even more seamless.

207
00:13:57,240 --> 00:14:02,057
When you install intelligent
terminal you can pick your

208
00:14:02,140 --> 00:14:05,327
favorite agent.
I am using GitHub Copilot today.

209
00:14:05,410 --> 00:14:08,167
You can use which ever
agent speaks to you.

210
00:14:08,250 --> 00:14:12,317
In intelligent terminal I have
terminal pane at the top and

211
00:14:12,400 --> 00:14:14,937
agent working on the bottom.
I, work between them while the

212
00:14:15,020 --> 00:14:18,837
agent helps along the way.
Here is an air being generated.

213
00:14:18,920 --> 00:14:20,947
Agent pain is able to detect it

214
00:14:21,030 --> 00:14:25,997
and provide a fix which is great

215
00:14:26,080 --> 00:14:28,167
when I don't understand the sin tax.

216
00:14:28,250 --> 00:14:32,657
I am working on open claw I have
worked it using WSL container.

217
00:14:32,740 --> 00:14:34,727
It is a native container

218
00:14:34,810 --> 00:14:37,127
experience on Windows plus it

219
00:14:37,210 --> 00:14:40,497
can leverage the GPU.

220
00:14:40,580 --> 00:14:45,767
It can reference your existing
container files just like the

221
00:14:45,850 --> 00:14:49,667
one in open claw project.
This is open in Microsoft edit

222
00:14:49,750 --> 00:14:53,207
got sin tax highlighting
and latest version.

223
00:14:53,290 --> 00:14:56,047
You can see the container

224
00:14:56,130 --> 00:14:59,077
running with container command.

225
00:14:59,160 --> 00:15:03,047
Since we are on the topic of WSL

226
00:15:03,130 --> 00:15:04,687
there is a profile that is

227
00:15:04,770 --> 00:15:10,957
comfortable for those using
tools like star, CSH and home

228
00:15:11,040 --> 00:15:14,627
brew. It has all of the favorite
utilities and available in the

229
00:15:14,710 --> 00:15:19,467
repo I showed earlier
and includes B top one of my

230
00:15:19,550 --> 00:15:19,767
favorites.

231
00:15:19,850 --> 00:15:24,577
The Surface RTX Spark serves

232
00:15:24,660 --> 00:15:27,407
large local models for coding.

233
00:15:27,490 --> 00:15:31,747
I have done development with 120

234
00:15:31,830 --> 00:15:35,817
billion parameter model which
most machines can't load.

235
00:15:35,900 --> 00:15:37,317
We can see how many tokens I

236
00:15:37,400 --> 00:15:43,087
have used locally.
3.4 million tokens leveraged on

237
00:15:43,170 --> 00:15:46,457
the device itself.
We can kickoff agents using fleet.

238
00:15:46,540 --> 00:15:49,517
Just so you don't have to watch
me type I am going to use

239
00:15:49,600 --> 00:15:53,167
Copilot voice feature which
is leveraging its own local model.

240
00:15:53,250 --> 00:15:55,007
I will hold face for it and tell

241
00:15:55,090 --> 00:15:57,137
it what I want it to do.

242
00:15:57,220 --> 00:16:00,307
Find any console dot right line

243
00:16:00,390 --> 00:16:03,077
calls in the tray and node

244
00:16:03,160 --> 00:16:08,367
projects and convert them to the standard
logger used elsewhere in the code base.

245
00:16:08,450 --> 00:16:12,287
There we go. The main
agent will delegate sub

246
00:16:12,370 --> 00:16:15,787
agent tasks of appropriate
complexity to the local model

247
00:16:15,870 --> 00:16:20,267
use liesing GPU and making
it cost efficient.

248
00:16:20,350 --> 00:16:24,147
As developers while debugging we
are looking through log files to

249
00:16:24,230 --> 00:16:28,207
diagnose any issues. Finding the location
of the log files is a challenge.

250
00:16:28,290 --> 00:16:33,947
I would love to be able to type
something like grep log and find them all.

251
00:16:34,030 --> 00:16:37,547
Sweet. On top of already adding curl,

252
00:16:37,630 --> 00:16:41,997
car and pseudo to windows now we
are adding over 75 command line

253
00:16:42,080 --> 00:16:45,657
utilities like end, be head,
tail and touch for those of us

254
00:16:45,740 --> 00:16:49,397
who love to live in the terminal.

255
00:16:49,480 --> 00:16:53,727
So I found all of my log files,
but now having to pars them is

256
00:16:53,810 --> 00:16:57,747
the second challenge.
I have had eye on instruct

257
00:16:57,830 --> 00:17:01,567
practically performing analyses
on my log files this whole time.

258
00:17:01,650 --> 00:17:06,137
Now I can quickly diagnose
anything that has gone wrong in

259
00:17:06,220 --> 00:17:10,747
my development, plus I don't have to worry
about token usage since it is all local.

260
00:17:10,830 --> 00:17:13,397
We can take a look
at the machinery sources.

261
00:17:13,480 --> 00:17:15,737
Models are loaded you can see 90

262
00:17:15,820 --> 00:17:18,257
gigs of ram being utilized by

263
00:17:18,340 --> 00:17:22,997
the GPU using the full power of

264
00:17:23,080 --> 00:17:27,767
RTX Spark. We were able to use them

265
00:17:27,850 --> 00:17:31,117
simultaneously unmetered going
through the dev flow without a

266
00:17:31,200 --> 00:17:31,587
hitch.

267
00:17:31,670 --> 00:17:38,390
That's huge.

268
00:17:38,490 --> 00:17:41,177
So I know you are going to love
what the team has been working on.

269
00:17:41,260 --> 00:17:45,277
We hope it gives you a glimpse
of what's possible on windows

270
00:17:45,360 --> 00:17:46,867
today and where we are headed next.

271
00:17:46,950 --> 00:17:47,397
Thank you.

272
00:17:47,480 --> 00:17:54,480
Back to you Satya.

273
00:17:58,010 --> 00:18:02,687
>> That's the beginning
of unmetered intelligence having

274
00:18:02,770 --> 00:18:07,337
the models and agents using
the models work in parallel to what

275
00:18:07,420 --> 00:18:12,037
you are doing in the cloud. That's what
the platform enables as first class.

276
00:18:12,120 --> 00:18:16,947
Let's move to the cloud. The driving
equation for us remains the same, which is

277
00:18:17,030 --> 00:18:20,997
tokens per dollar per watt.
How do we optimize around this?

278
00:18:21,080 --> 00:18:24,617
When we think about the systems
problem we think about electrons

279
00:18:24,700 --> 00:18:28,187
coming on one end and tokens
on the other end and how do we

280
00:18:28,270 --> 00:18:31,997
think about the systems
optimization end-to-end.

281
00:18:32,080 --> 00:18:34,927
It starts with the data
center design itself.

282
00:18:35,010 --> 00:18:37,487
The core compute, storage, network.

283
00:18:37,570 --> 00:18:41,537
All of the accelerators that are
going to accelerate each of the

284
00:18:41,620 --> 00:18:45,907
components. How do you
think about the DC to

285
00:18:45,990 --> 00:18:51,047
DC connectivity and networking as well as
the off load to something like the local

286
00:18:51,130 --> 00:18:54,297
compute. That's the sort
of systems challenge.

287
00:18:54,380 --> 00:18:57,857
But before we even get into
all of the systems and the

288
00:18:57,940 --> 00:19:01,727
technology and the innovation,
perhaps the most important

289
00:19:01,810 --> 00:19:07,227
design criteria for us is how do
we earn the permission from the

290
00:19:07,310 --> 00:19:10,917
communities in which we are
building these data centers.

291
00:19:11,000 --> 00:19:14,637
That's where these principles
ground us and focus us.

292
00:19:14,720 --> 00:19:17,677
How do we ensure that the DCs do

293
00:19:17,760 --> 00:19:22,707
not increase the electro city

294
00:19:22,790 --> 00:19:27,797
prices. Making sure we are using water

295
00:19:27,880 --> 00:19:30,517
use. Creating jobs
for the local residents.

296
00:19:30,600 --> 00:19:33,927
Adding to the tax base.
Making sure we are strengthening

297
00:19:34,010 --> 00:19:36,927
communities by investing
in had local training and the

298
00:19:37,010 --> 00:19:40,267
nonprofits in the area.

299
00:19:40,350 --> 00:19:42,027
Only when we live up to the

300
00:19:42,110 --> 00:19:46,337
principles do the hard work is

301
00:19:46,420 --> 00:19:49,867
when we innovate and build.
We have been doing a lot of data

302
00:19:49,950 --> 00:19:52,677
center build up.

303
00:19:52,760 --> 00:19:57,147
Today Azure spans more than 500

304
00:19:57,230 --> 00:20:01,337
data centers in the region.
We have the most expansive

305
00:20:01,420 --> 00:20:02,197
hyperscaler footprint out there.

306
00:20:02,280 --> 00:20:04,657
We have added more data center

307
00:20:04,740 --> 00:20:08,457
capacity in the last 18 months

308
00:20:08,540 --> 00:20:14,767
than the first decade of to put

309
00:20:14,850 --> 00:20:21,850
that in perspective.

310
00:20:23,560 --> 00:20:25,837
We built out the cloud

311
00:20:25,920 --> 00:20:31,077
infrastructure for hetero genius
workloads that span the enterprise.

312
00:20:31,160 --> 00:20:33,767
When you look at what we are
building when you think of all

313
00:20:33,850 --> 00:20:39,157
of the gigawatts that are going
to come on-line we have three

314
00:20:39,240 --> 00:20:42,927
dominant workloads. There is training,
inference and the agent run time.

315
00:20:43,010 --> 00:20:45,297
These are three dominant workloads.

316
00:20:45,380 --> 00:20:48,297
In fact when you look at fair

317
00:20:48,380 --> 00:20:51,847
water it was sort of our first

318
00:20:51,930 --> 00:20:55,767
AI super factory.
It spanned two regions, Georgia

319
00:20:55,850 --> 00:20:59,307
and Wisconsin.
The entire system was designed

320
00:20:59,390 --> 00:21:03,607
from the ground up for AI.
We worked in fact very closely

321
00:21:03,690 --> 00:21:05,547
with inn NVIDIA on this.

322
00:21:05,630 --> 00:21:08,167
It is a two-story architecture

323
00:21:08,250 --> 00:21:13,117
that lets us he goes --

324
00:21:13,200 --> 00:21:18,027
essentially put the number of GPU
densely with network access.

325
00:21:18,110 --> 00:21:20,097
Means you really got fantastic

326
00:21:20,180 --> 00:21:22,697
higher performance networking,

327
00:21:22,780 --> 00:21:25,647
lower latency and more band

328
00:21:25,730 --> 00:21:27,507
width across the entire cluster.

329
00:21:27,590 --> 00:21:29,807
We are rethinking even the power

330
00:21:29,890 --> 00:21:30,067
delivery.

331
00:21:30,150 --> 00:21:37,150
Pou do -- how do we deliver 100

332
00:21:38,500 --> 00:21:41,937
kilowatts her row with the
conversion loss to the grid to

333
00:21:42,020 --> 00:21:46,787
silicon. We took a new approach to it.

334
00:21:46,870 --> 00:21:48,587
Also changes with the cooling

335
00:21:48,670 --> 00:21:51,527
system and water.

336
00:21:51,610 --> 00:21:54,427
The tooling loop is filled and

337
00:21:54,510 --> 00:21:59,117
the data center can operate
effectively with zero water

338
00:21:59,200 --> 00:22:02,307
consumption.
The daily water usage over the

339
00:22:02,390 --> 00:22:05,137
course of the entire one year is

340
00:22:05,220 --> 00:22:07,347
equivalent to what the single

341
00:22:07,430 --> 00:22:13,210
restaurant would use.

342
00:22:13,320 --> 00:22:18,117
When it comes to the systems
in the silicon, we have a lot of

343
00:22:18,200 --> 00:22:20,427
choice. We have first party silicon, we

344
00:22:20,510 --> 00:22:24,527
have partner systems.

345
00:22:24,610 --> 00:22:27,367
We have the first cloud to bring

346
00:22:27,450 --> 00:22:32,717
up inn NVIDIA.

347
00:22:32,800 --> 00:22:34,907
We work close with AMD.

348
00:22:34,990 --> 00:22:38,377
Working with them on the NX

349
00:22:38,460 --> 00:22:42,007
generation AMD GPU.

350
00:22:42,090 --> 00:22:43,907
Meyer 200 is continuing to scale

351
00:22:43,990 --> 00:22:48,367
it is live in Arizona will

352
00:22:48,450 --> 00:22:53,857
deploy internationally.
It delivers 30% per tokens per

353
00:22:53,940 --> 00:22:56,757
dollar verses what's
is the leading GPU today.

354
00:22:56,840 --> 00:23:02,127
We have validated it with GPD55
and we are going to use that to

355
00:23:02,210 --> 00:23:03,747
power Microsoft 365 Copilot.

356
00:23:03,830 --> 00:23:09,037
When it comes to running these

357
00:23:09,120 --> 00:23:12,007
agents, interesting thing now it

358
00:23:12,090 --> 00:23:16,747
is no longer about having
AI accelerator or GPU.

359
00:23:16,830 --> 00:23:21,597
CPU is critical.
Ratios may be coming to one is

360
00:23:21,680 --> 00:23:24,957
to one. That's why we are innovating

361
00:23:25,040 --> 00:23:30,127
with cobalt.
We are announcing preview oOf Cob

362
00:23:30,210 --> 00:23:34,897
Cobalt00VMs. Next generation on generation
CPU designed for cloud native

363
00:23:34,980 --> 00:23:36,687
and agent workloads today.

364
00:23:36,770 --> 00:23:39,717
It is exciting to see Cobalt

365
00:23:39,800 --> 00:23:46,790
make progress as well.

366
00:23:46,890 --> 00:23:50,877
One thing we are trying to make
sure being optimized for the new

367
00:23:50,960 --> 00:23:53,297
agent workloads.

368
00:23:53,380 --> 00:23:57,787
Cobalt delivers 30% better than

369
00:23:57,870 --> 00:24:01,957
Cobalt on cloud native.
Using GitHub Copilot traces

370
00:24:02,040 --> 00:24:06,127
these agentic traces to see.
The call patterns are so

371
00:24:06,210 --> 00:24:06,427
different.

372
00:24:06,510 --> 00:24:09,367
We are seeing 33% lower latency

373
00:24:09,450 --> 00:24:14,267
for the agent calls, 14% faster
speed, 23% higher through put.

374
00:24:14,350 --> 00:24:18,837
This is about the co design
of both the AI accelerator and the

375
00:24:18,920 --> 00:24:23,007
CPU for the agent.
And of course, when you talk

376
00:24:23,090 --> 00:24:27,437
about AI workloads you need scale,
you need reliability and

377
00:24:27,520 --> 00:24:29,397
that's why the network becomes

378
00:24:29,480 --> 00:24:32,817
super critical.

379
00:24:32,900 --> 00:24:35,427
We have innovated with the

380
00:24:35,510 --> 00:24:38,197
architecture and rebuilt how

381
00:24:38,280 --> 00:24:40,557
traffic across Azure moves to

382
00:24:40,640 --> 00:24:43,487
support workloads which is

383
00:24:43,570 --> 00:24:48,837
synchronous data workloads.
They span 10s of thousands of

384
00:24:48,920 --> 00:24:52,207
GPUs. You need to have them coherent.

385
00:24:52,290 --> 00:24:53,877
That's what the next frontier is

386
00:24:53,960 --> 00:24:55,877
to make sure we are able to keep

387
00:24:55,960 --> 00:25:02,960
scaling the network

388
00:25:04,500 --> 00:25:06,217
architearchitectur architecture.

389
00:25:06,300 --> 00:25:08,387
It is not just about outside
it is inside the data centers.

390
00:25:08,470 --> 00:25:12,457
It is connected through our

391
00:25:12,540 --> 00:25:17,887
continent spanning AI with the

392
00:25:17,970 --> 00:25:19,887
truly fungible compute fabric.

393
00:25:19,970 --> 00:25:22,507
All of this innovation is

394
00:25:22,590 --> 00:25:25,977
exciting, but when you think

395
00:25:26,060 --> 00:25:29,907
about innovation designing
for AI and I would say deep

396
00:25:29,990 --> 00:25:32,747
understanding of systems,

397
00:25:32,830 --> 00:25:37,657
there's no better company than

398
00:25:37,740 --> 00:25:41,687
NVIDIA and there's no other
person than Jensen to talk about

399
00:25:41,770 --> 00:25:43,197
it.

400
00:25:43,280 --> 00:25:50,280
I want to invite Jensen Wang

401
00:25:55,250 --> 00:25:59,877
from NVIDIA. I know it is late for you in

402
00:25:59,960 --> 00:26:00,377
Taipei.

403
00:26:00,460 --> 00:26:04,717
I appreciate you staying up.

404
00:26:04,800 --> 00:26:11,800
I have been looking at social

405
00:26:15,310 --> 00:26:18,427
and people talking about your keynote.

406
00:26:18,510 --> 00:26:23,097
The concept of unmetered intelligence
right at the edge is so hot again.

407
00:26:23,180 --> 00:26:25,347
You have thought about this,

408
00:26:25,430 --> 00:26:26,837
talked about this, now of course

409
00:26:26,920 --> 00:26:31,277
with RTX Spark really delivered

410
00:26:31,360 --> 00:26:33,377
what is a break through system

411
00:26:33,460 --> 00:26:38,047
for AI to be much Moreau --

412
00:26:38,130 --> 00:26:42,167
more ubiquitous. Maybe you can
share where you see this going?

413
00:26:42,250 --> 00:26:45,657
>> It all started about three
years ago with a conversation

414
00:26:45,740 --> 00:26:49,287
between you and I.
We were talking about how we

415
00:26:49,370 --> 00:26:53,767
could build a new class of PCs
incredible for designers and

416
00:26:53,850 --> 00:26:57,947
creators and it would be incredible
for artificial intelligence.

417
00:26:58,030 --> 00:27:01,407
It would be one of these
systems that has the processing

418
00:27:01,490 --> 00:27:04,037
capability but also the software

419
00:27:04,120 --> 00:27:09,007
stack that is integrated into
the world's design packages and

420
00:27:09,090 --> 00:27:12,177
creator packages and all
of the things we are doing with AI.

421
00:27:12,260 --> 00:27:13,897
Here we are three years later we

422
00:27:13,980 --> 00:27:16,717
built an incredible new chip,

423
00:27:16,800 --> 00:27:20,987
and this system is supported by

424
00:27:21,070 --> 00:27:26,057
all of this new software you

425
00:27:26,140 --> 00:27:30,747
created for Windows.
We have autonomous agent running

426
00:27:30,830 --> 00:27:35,837
on the PC. When you step back
and think about what does that mean?

427
00:27:35,920 --> 00:27:37,007
For 40 years or some 30 years we

428
00:27:37,090 --> 00:27:40,747
have been working together we

429
00:27:40,830 --> 00:27:43,147
went from inventing direct X

430
00:27:43,230 --> 00:27:46,347
together to creating now this

431
00:27:46,430 --> 00:27:50,787
incredible computer that has
autonomous systems running.

432
00:27:50,870 --> 00:27:54,927
The PC evolved from being
an incredible tool to now being a

433
00:27:55,010 --> 00:27:58,527
tool that is used autonomously

434
00:27:58,610 --> 00:28:04,217
by an AI assistant.

435
00:28:04,300 --> 00:28:06,247
The idea I could be traveling

436
00:28:06,330 --> 00:28:11,877
and on the phone and text my PC
and ask my PC to get coding done

437
00:28:11,960 --> 00:28:15,207
or an idea I have and it would
fire up the tools on the PC and

438
00:28:15,290 --> 00:28:18,477
it would make the modifications
or the changes or the design

439
00:28:18,560 --> 00:28:22,667
that I told it to do, and it
would iterate with me while I am

440
00:28:22,750 --> 00:28:24,587
away from the PC.

441
00:28:24,670 --> 00:28:30,057
My PC became an assistant.

442
00:28:30,140 --> 00:28:33,357
While I am sitting there it
would my great system as well.

443
00:28:33,440 --> 00:28:35,827
The idea that the PC evolved

444
00:28:35,910 --> 00:28:37,377
from a personal computer to

445
00:28:37,460 --> 00:28:44,460
personal AI is really exciting.

446
00:28:44,560 --> 00:28:48,107
To see it come to life
and actually doing that, I am super

447
00:28:48,190 --> 00:28:51,747
excited about it.
Spark you mentioned earlier has

448
00:28:51,830 --> 00:28:53,797
all of these capabilities Peta

449
00:28:53,880 --> 00:28:55,597
flop of AI performance.

450
00:28:55,680 --> 00:29:01,987
Has a Peta flop of MVFP4

451
00:29:02,070 --> 00:29:04,127
numerical format our companies
worked on together that allows

452
00:29:04,210 --> 00:29:07,097
it to take advantage of the 128

453
00:29:07,180 --> 00:29:11,787
giga bites of memory and fit
maybe a couple hundred billion

454
00:29:11,870 --> 00:29:16,507
parameter model. A couple hundred billion
parameter model is state of the art.

455
00:29:16,590 --> 00:29:20,037
The days of having a really
smart assistant running on the

456
00:29:20,120 --> 00:29:23,207
PC is here.

457
00:29:23,290 --> 00:29:24,877
>> It is awesome.

458
00:29:24,960 --> 00:29:27,497
I am excited about Windows

459
00:29:27,580 --> 00:29:31,287
coming to the GP300.

460
00:29:31,370 --> 00:29:35,787
It is a data center on your laptop.

461
00:29:35,870 --> 00:29:37,587
Talking about the data center side.

462
00:29:37,670 --> 00:29:40,697
This entire thing got started

463
00:29:40,780 --> 00:29:47,397
when we built the first super
computer to train the GPU

464
00:29:47,480 --> 00:29:49,037
models. I was talking about the fair

465
00:29:49,120 --> 00:29:52,677
water design, this custom build

466
00:29:52,760 --> 00:29:56,477
for the grace black water era
to Max the data center design with

467
00:29:56,560 --> 00:29:58,247
the system design you had.

468
00:29:58,330 --> 00:30:04,557
Now of course we are validating bel

469
00:30:04,640 --> 00:30:06,817
bel belly --.

470
00:30:06,900 --> 00:30:12,197
Maybe you can share on the clout
side how you are sharing the system.

471
00:30:12,280 --> 00:30:16,747
>> Our journey is incredible.
We built the first AI super

472
00:30:16,830 --> 00:30:20,537
computer together.
Hopper was an incredible success.

473
00:30:20,620 --> 00:30:23,507
The first two generations were

474
00:30:23,590 --> 00:30:27,847
focused on free training.
Grace Blackwell came along and

475
00:30:27,930 --> 00:30:31,147
all of the focus moved to post
training, reinforcement learning

476
00:30:31,230 --> 00:30:34,297
which allowed us to have reasoning models.

477
00:30:34,380 --> 00:30:38,687
These reasoning models based
on mixture of experts were

478
00:30:38,770 --> 00:30:42,027
incredibly intelligent, energy efficient.

479
00:30:42,110 --> 00:30:46,457
It requires giant systems.
We required MV length 72.

480
00:30:46,540 --> 00:30:49,067
The entire rack became one computer.

481
00:30:49,150 --> 00:30:51,077
We had evolved from one node to

482
00:30:51,160 --> 00:30:55,037
now one rack.

483
00:30:55,120 --> 00:30:56,737
Well, Microsoft deployed the

484
00:30:56,820 --> 00:30:58,837
largest number of grace

485
00:30:58,920 --> 00:31:02,647
Blackwells in the world today.

486
00:31:02,730 --> 00:31:03,947
Largest number of grace

487
00:31:04,030 --> 00:31:07,627
Blackwells in the world.

488
00:31:07,710 --> 00:31:09,897
Fair water is a magnificent

489
00:31:09,980 --> 00:31:13,357
system to look at.

490
00:31:13,440 --> 00:31:17,357
It is an incredible feet.
It is liquid cooled.

491
00:31:17,440 --> 00:31:20,897
Something you mentioned it
is close looped, basically uses

492
00:31:20,980 --> 00:31:25,447
almost no water.
It is incredibly environmentally

493
00:31:25,530 --> 00:31:29,067
friendly energy efficient.
We are able to increase the

494
00:31:29,150 --> 00:31:31,407
token generation rate and reduce

495
00:31:31,490 --> 00:31:36,277
the cost of token generation
by an order of magnitude, some 30

496
00:31:36,360 --> 00:31:39,377
times over hopper.
That was a huge achievement.

497
00:31:39,460 --> 00:31:45,157
Vera Ruben was created for a
world where these AIs are now

498
00:31:45,240 --> 00:31:48,087
agentic. As hopper was created for

499
00:31:48,170 --> 00:31:49,687
pretraining grace Blackwell for

500
00:31:49,770 --> 00:31:55,527
training, post training and also

501
00:31:55,610 --> 00:31:59,247
inference. Vera Rubin is designed to run

502
00:31:59,330 --> 00:32:03,307
for agents.
It is the same computing pattern

503
00:32:03,390 --> 00:32:06,237
when run on the RTX Spark.

504
00:32:06,320 --> 00:32:08,277
It is the same agentic system

505
00:32:08,360 --> 00:32:13,017
but of course larger win process

506
00:32:13,100 --> 00:32:19,617
enormous number of them. Many will be
from different customers and different

507
00:32:19,700 --> 00:32:22,587
partners. The entire path, coding path

508
00:32:22,670 --> 00:32:26,797
from storage which is the
long-term memory, working memory

509
00:32:26,880 --> 00:32:27,697
is encrypted.

510
00:32:27,780 --> 00:32:31,747
The data is encrypted in transit

511
00:32:31,830 --> 00:32:38,037
and data also incren crepted in

512
00:32:38,120 --> 00:32:41,277
encrypted in use.
This entire disaggregated

513
00:32:41,360 --> 00:32:43,747
distributed computing system you

514
00:32:43,830 --> 00:32:46,417
mentioned CPU Vera is a CPU

515
00:32:46,500 --> 00:32:50,957
designed for agents.
In the past CPUs were designed

516
00:32:51,040 --> 00:32:55,157
for humans. We are more
parent than agents are.

517
00:32:55,240 --> 00:32:57,887
Agents want low latency.

518
00:32:57,970 --> 00:33:00,057
Vera was designed for extremely

519
00:33:00,140 --> 00:33:03,747
low latency.

520
00:33:03,830 --> 00:33:07,567
Vera Rubin I can't wait to show everybody.

521
00:33:07,650 --> 00:33:13,077
You stood it up.
Our two teams have been working

522
00:33:13,160 --> 00:33:16,807
closely. Long before the chips taped out

523
00:33:16,890 --> 00:33:21,297
long before the systems brought
up the two teams were almost

524
00:33:21,380 --> 00:33:23,517
completely aligned.

525
00:33:23,600 --> 00:33:28,427
Data centers created for Vera Rubin.

526
00:33:28,510 --> 00:33:30,827
It is into the networking
and stack and community.

527
00:33:30,910 --> 00:33:34,357
The moment systems were rolling
off the lines they were building

528
00:33:34,440 --> 00:33:39,247
stood up at Microsoft. So incredibly
excited about the collaboration.

529
00:33:39,330 --> 00:33:42,907
>> This speed of light execution
for the teams are fantastic to see.

530
00:33:42,990 --> 00:33:46,337
All of this is to power
the ecosystem around us.

531
00:33:46,420 --> 00:33:50,577
You and I having grown-up with
a PC the server and now with AI,

532
00:33:50,660 --> 00:33:55,537
have always thought about ultimately it is
about creating the opportunity for every

533
00:33:55,620 --> 00:33:58,487
developer, every organization
to build on the work that we do and

534
00:33:58,570 --> 00:34:00,487
the platforms we create.

535
00:34:00,570 --> 00:34:04,197
Speaking of that there's a lot

536
00:34:04,280 --> 00:34:08,467
of software that vid NVIDIA

537
00:34:08,550 --> 00:34:13,247
builds we will have the models
in foundry, tooling in foundry.

538
00:34:13,330 --> 00:34:16,337
Software will help us with
accelerating our workloads when

539
00:34:16,420 --> 00:34:20,307
it comes to the data warehouse.

540
00:34:20,390 --> 00:34:23,047
We will have the stuff in windows.

541
00:34:23,130 --> 00:34:28,367
Talk about the broader vision
of what does it mean as an

542
00:34:28,450 --> 00:34:31,617
opportunity. Everybody talks about the one

543
00:34:31,700 --> 00:34:36,757
model and one piece of tech.
It is about the broadest biggest

544
00:34:36,840 --> 00:34:41,567
opportunity maybe you want
to share a little bit about that.

545
00:34:41,650 --> 00:34:43,567
>> We have been preparing for this moment.

546
00:34:43,650 --> 00:34:47,087
What happened the last several
months we have been working a

547
00:34:47,170 --> 00:34:48,967
decade and a half together
getting ready for really what

548
00:34:49,050 --> 00:34:50,467
happened in the last several months.

549
00:34:50,550 --> 00:34:54,807
All of a sudden because
of agentic systems the convergence

550
00:34:54,890 --> 00:34:57,307
of these great models AI is now

551
00:34:57,390 --> 00:34:57,547
useful.

552
00:34:57,630 --> 00:35:03,067
If you look at GitHub, the

553
00:35:03,150 --> 00:35:07,887
commits into GitHub has
gone completely parabolic.

554
00:35:07,970 --> 00:35:11,657
In the last several months
the number of commits increase by a

555
00:35:11,740 --> 00:35:15,367
factor of three.
It is clear agentic systems are

556
00:35:15,450 --> 00:35:17,677
useful, it is doing productive

557
00:35:17,760 --> 00:35:23,237
work and also it is profitable

558
00:35:23,320 --> 00:35:25,637
as a result. The amount of command for

559
00:35:25,720 --> 00:35:29,177
compute between the usage of AI

560
00:35:29,260 --> 00:35:35,197
the contribution necessary
for agents it has gone through the

561
00:35:35,280 --> 00:35:38,357
roof. One thing we have been doing
together is making sure all of

562
00:35:38,440 --> 00:35:41,987
the tools agents are using are

563
00:35:42,070 --> 00:35:43,457
fully accelerated.

564
00:35:43,540 --> 00:35:45,997
Fabric is fully accelerated.

565
00:35:46,080 --> 00:35:49,597
We are accelerating data

566
00:35:49,680 --> 00:35:54,467
processing, SQL, Spark, semantic

567
00:35:54,550 --> 00:35:57,007
based, graph based we are going
to make sure all of the tools

568
00:35:57,090 --> 00:35:59,477
available on Azure are going to

569
00:35:59,560 --> 00:36:05,577
be fully GPU accelerated because
the agents are going to be impat

570
00:36:05,660 --> 00:36:09,837
impatient. The faster it can
iterate and generate tokens which are

571
00:36:09,920 --> 00:36:13,887
ultimately what the developers
both customers would like to do

572
00:36:13,970 --> 00:36:19,097
is generate a lot of tokens that are
profitable and highly intelligent.

573
00:36:19,180 --> 00:36:22,097
>> Thank you so much for the
partnership and leadership and

574
00:36:22,180 --> 00:36:25,717
innovation that you bring
to this entire ecosystem and really

575
00:36:25,800 --> 00:36:28,967
thrilled to be working closely
with you and the team and bring

576
00:36:29,050 --> 00:36:32,537
all of this to the developers
here and beyond and look forward

577
00:36:32,620 --> 00:36:36,047
to seeing what the next sort
of few months and the next year

578
00:36:36,130 --> 00:36:41,367
will bring in terns of the innovation
that gets built on top of the platform.

579
00:36:41,450 --> 00:36:45,917
Thank you again for joining
this late in the night from Taipei.

580
00:36:46,000 --> 00:36:47,757
>> Thank you so much Satya for

581
00:36:47,840 --> 00:36:53,310
your partnership and friendship.

582
00:36:53,610 --> 00:36:57,177
>> So so far we have talked

583
00:36:57,260 --> 00:37:02,537
about the edge and the cloud.
The form factors, when I saw

584
00:37:02,620 --> 00:37:07,177
that Jensen picture from the weekend
where he had all of the desktops.

585
00:37:07,260 --> 00:37:09,577
I felt man I am back in the 90s.

586
00:37:09,660 --> 00:37:13,277
It is so tool to see the lineup
of all of the machines that I

587
00:37:13,360 --> 00:37:17,617
loved and grew up with back yet
again with new functionality.

588
00:37:17,700 --> 00:37:20,217
The same form factor, but

589
00:37:20,300 --> 00:37:25,127
unbelievable new functionality,
because of the on board AI capability.

590
00:37:25,210 --> 00:37:30,677
That is sort of what we have seen with the
laptop, the desktop and of course with the

591
00:37:30,760 --> 00:37:33,887
cloud. But it also, you know, sets up

592
00:37:33,970 --> 00:37:38,577
that next question.
If you have that capability,

593
00:37:38,660 --> 00:37:42,207
which is new function and you
can put it into existing form

594
00:37:42,290 --> 00:37:45,877
fact factors, can you even purpose

595
00:37:45,960 --> 00:37:49,797
build new form factors
for the new function?

596
00:37:49,880 --> 00:37:52,857
Can you build a new platform

597
00:37:52,940 --> 00:37:54,687
even for the agent era?

598
00:37:54,770 --> 00:37:56,987
And that is the motivation

599
00:37:57,070 --> 00:38:02,827
behind project Solara.

600
00:38:02,910 --> 00:38:07,047
I wanted to invite Stevie
on stage but first let's roll the

601
00:38:07,130 --> 00:38:13,377
video.

602
00:38:13,460 --> 00:38:15,747
>> First there's you.
Then everything in front of you.

603
00:38:15,830 --> 00:38:16,947
We get it.

604
00:38:17,030 --> 00:38:22,667
The noise, the weight of it all.

605
00:38:22,750 --> 00:38:29,750
We have been building something.

606
00:38:30,170 --> 00:38:31,527
For an agent first world.

607
00:38:31,610 --> 00:38:33,027
New agent first devices.

608
00:38:33,110 --> 00:38:38,317
You come close, it is ready when

609
00:38:38,400 --> 00:38:40,387
you are. You speak, it understands you.

610
00:38:40,470 --> 00:38:44,437
What matters comes forward.

611
00:38:44,520 --> 00:38:46,837
Everything else falls away.

612
00:38:46,920 --> 00:38:49,777
When you want it to keep going

613
00:38:49,860 --> 00:38:56,860
it keeps going.

614
00:38:57,180 --> 00:38:59,987
Sometimes it sits with you.
Sometimes it goes with you.

615
00:39:00,070 --> 00:39:04,597
Sometimes it sees what you want it to see.

616
00:39:04,680 --> 00:39:08,927
It doesn't decide for you.
It lights the path.

617
00:39:09,010 --> 00:39:11,347
Tries to clear the way
when you need it to.

618
00:39:11,430 --> 00:39:15,567
That's the idea.

619
00:39:15,650 --> 00:39:19,707
A whole ecosystem built to clear

620
00:39:19,790 --> 00:39:25,017
the way.

621
00:39:25,100 --> 00:39:29,767
Wherever you are, whenever it

622
00:39:29,850 --> 00:39:33,617
matters. >> I am so excited to be here.

623
00:39:33,700 --> 00:39:36,057
It really is great
to be back on this stage.

624
00:39:36,140 --> 00:39:37,527
Before I talk about those

625
00:39:37,610 --> 00:39:39,527
awesome new devices you just

626
00:39:39,610 --> 00:39:43,367
saw, let me start with the why.

627
00:39:43,450 --> 00:39:48,037
Back at build 2023 I talked

628
00:39:48,120 --> 00:39:50,907
about the outside AI

629
00:39:50,990 --> 00:39:51,577
infrastructure.

630
00:39:51,660 --> 00:39:55,747
It moves within the application

631
00:39:55,830 --> 00:40:00,577
frame to operating globally to

632
00:40:00,660 --> 00:40:02,167
connect maintain context across

633
00:40:02,250 --> 00:40:04,587
entire workflows, devices and

634
00:40:04,670 --> 00:40:08,987
time scales.

635
00:40:09,070 --> 00:40:11,757
What if there were an ecosystem

636
00:40:11,840 --> 00:40:16,667
of devices specifically designed
for the new type of application

637
00:40:16,750 --> 00:40:18,747
structure for those types of

638
00:40:18,830 --> 00:40:22,567
agents for that transaction.

639
00:40:22,650 --> 00:40:24,837
That's the impetus behind

640
00:40:24,920 --> 00:40:29,747
project Solara.
With many possible forms which

641
00:40:29,830 --> 00:40:32,147
one do you pick? What is the next device.

642
00:40:32,230 --> 00:40:37,367
The big uh-huh for us is that it
is not choosing one form factor

643
00:40:37,450 --> 00:40:39,187
it is about creating a system

644
00:40:39,270 --> 00:40:43,757
that extends your agent across a

645
00:40:43,840 --> 00:40:47,027
cons constellation of devices.

646
00:40:47,110 --> 00:40:51,347
The next computer is all of the
devices working together as one

647
00:40:51,430 --> 00:40:55,767
system with agents cloeing up
closer to where and when you

648
00:40:55,850 --> 00:40:57,137
need them.

649
00:40:57,220 --> 00:40:59,707
To realize this vision two

650
00:40:59,790 --> 00:41:03,777
challenges immediately show up.

651
00:41:03,860 --> 00:41:07,047
First many specialized form
factors already exist but often

652
00:41:07,130 --> 00:41:11,287
rely on custom one off apps
and fragmented stacks that are

653
00:41:11,370 --> 00:41:16,087
difficult and expensive to build,
deploy and maintain.

654
00:41:16,170 --> 00:41:18,857
Second across every industry
people and organizations are

655
00:41:18,940 --> 00:41:22,997
already building their own
agents deeply specialized and

656
00:41:23,080 --> 00:41:27,637
instrumented for their work.
The impact of those agents is

657
00:41:27,720 --> 00:41:28,467
constrained by how and where

658
00:41:28,550 --> 00:41:31,637
they can exist.

659
00:41:31,720 --> 00:41:35,547
Project Solara does both by

660
00:41:35,630 --> 00:41:39,047
extending agents on to new
purpose built easy to manage

661
00:41:39,130 --> 00:41:41,497
form factors designed to reach

662
00:41:41,580 --> 00:41:46,787
the nooks and crannies where
project creators do not exist or

663
00:41:46,870 --> 00:41:50,187
are not optimal.
It is a turnkey solution for

664
00:41:50,270 --> 00:41:55,727
building unique agent first
devices enabled by three

665
00:41:55,810 --> 00:41:58,117
pillars. First it is enterprise ready.

666
00:41:58,200 --> 00:42:01,067
Enabled by the AOSP based

667
00:42:01,150 --> 00:42:06,237
Microsoft device ecosystem platform.

668
00:42:06,320 --> 00:42:09,307
Second it has an agent driven
interaction model with just in

669
00:42:09,390 --> 00:42:13,217
time UI that adapts to the form factor.

670
00:42:13,300 --> 00:42:17,137
Third it has extensibility so
you can bring your own agents.

671
00:42:17,220 --> 00:42:22,287
Tying it all together is Azure unifying
the system across cloud and device.

672
00:42:22,370 --> 00:42:26,427
That's enough of that.
Now let's talk about the

673
00:42:26,510 --> 00:42:30,127
devices. Today we are previewing
two very broad categories.

674
00:42:30,210 --> 00:42:33,087
The first is stationary
and the second is port able.

675
00:42:33,170 --> 00:42:36,067
The first device
is designed for your desk.

676
00:42:36,150 --> 00:42:39,737
It is built on media tech silicon.

677
00:42:39,820 --> 00:42:43,207
With hello for business walking
up to the device securely signs

678
00:42:43,290 --> 00:42:46,517
you in giving you direct
access to your agent.

679
00:42:46,600 --> 00:42:49,697
Just like Nathan
is about to show you here.

680
00:42:49,780 --> 00:42:53,017
For the information work this

681
00:42:53,100 --> 00:42:55,187
means frictionless yet protected

682
00:42:55,270 --> 00:42:58,927
access to Microsoft 365 Copilot

683
00:42:59,010 --> 00:43:02,797
grounded in work IQ.
What matters next in your

684
00:43:02,880 --> 00:43:06,947
workday helping you think,
plan and even act by delegating task

685
00:43:07,030 --> 00:43:09,267
to your agent with a simple tap

686
00:43:09,350 --> 00:43:16,247
or just using your voice.

687
00:43:16,330 --> 00:43:19,777
Dedicated Ambian device for work.

688
00:43:19,860 --> 00:43:25,787
It acts as campaign onto your
existing windows Copilot Plus

689
00:43:25,870 --> 00:43:27,187
-- PC.

690
00:43:27,270 --> 00:43:31,757
Or access windows 365 in a

691
00:43:31,840 --> 00:43:34,367
connected monitor.

692
00:43:34,450 --> 00:43:39,730
How cool is that?

693
00:43:39,830 --> 00:43:43,767
The second is port able. Re-imagining
a wear able millions of people use every

694
00:43:43,850 --> 00:43:47,307
day. The access badge.

695
00:43:47,390 --> 00:43:49,977
Built using Qualcomm silicon

696
00:43:50,060 --> 00:43:54,737
wear able this badge is light
weight form factor designed for

697
00:43:54,820 --> 00:43:59,257
agents on the go.
Adapt able across variety of

698
00:43:59,340 --> 00:44:01,117
verti

699
00:44:01,200 --> 00:44:02,157
verticals and workflows.

700
00:44:02,240 --> 00:44:08,327
I have an early prototype of the

701
00:44:08,410 --> 00:44:10,877
badge. Using my finger print I tapped

702
00:44:10,960 --> 00:44:16,937
to unlock the device now I have
all of my agents in a secure

703
00:44:17,020 --> 00:44:22,247
manner. Gather content
for social media post for today.

704
00:44:22,330 --> 00:44:23,547
Why not do it right now.

705
00:44:23,630 --> 00:44:26,277
I am going to hit record and the

706
00:44:26,360 --> 00:44:31,217
device's camera is recording.

707
00:44:31,300 --> 00:44:34,527
Hope you don't mind I
am going to take shots.

708
00:44:34,610 --> 00:44:38,727
Find good shots from this clean
them up and send them for me and

709
00:44:38,810 --> 00:44:42,767
my team to review. There you have it.

710
00:44:42,850 --> 00:44:44,947
Now my agent is off running

711
00:44:45,030 --> 00:44:47,767
through multiple tasks I clean

712
00:44:47,850 --> 00:44:52,977
them up and send them to me and the team.

713
00:44:53,060 --> 00:44:56,777
Pretty cool.
I know it is a simple demo but

714
00:44:56,860 --> 00:45:01,427
it is all agent driven and there are
so many verticals and opportunity.

715
00:45:01,510 --> 00:45:05,287
Imagine in healthcare from the
moment you pick up the device

716
00:45:05,370 --> 00:45:08,787
the right agent shapes
the experience around the roll and

717
00:45:08,870 --> 00:45:12,727
the workflow helping with check
ins, patient records and

718
00:45:12,810 --> 00:45:17,267
critical insights. All through
enterprise grade secure access.

719
00:45:17,350 --> 00:45:21,167
And with the built
in microphones the nurse can start

720
00:45:21,250 --> 00:45:25,377
a hands free voice based
documentation including

721
00:45:25,460 --> 00:45:27,677
dirizezation and and notation.

722
00:45:27,760 --> 00:45:32,547
The side phasing camera can
help with asht vitals or scan

723
00:45:32,630 --> 00:45:34,897
medications or help verify workflows.

724
00:45:34,980 --> 00:45:38,787
This is a few examples of how
the small purpose built wear

725
00:45:38,870 --> 00:45:41,927
able can bring intelligence
directly into the flow of

726
00:45:42,010 --> 00:45:46,257
patient care helping nurses
access and gather and even act

727
00:45:46,340 --> 00:45:47,497
on information while staying

728
00:45:47,580 --> 00:45:54,580
print with the patient.

729
00:45:55,650 --> 00:46:00,337
Both devices are an expression
of devices the core hardware and

730
00:46:00,420 --> 00:46:02,777
software are designed
to be highly flexible.

731
00:46:02,860 --> 00:46:07,647
With a few changes loading
a different agent adjusting the

732
00:46:07,730 --> 00:46:10,917
shape, screen size,
sensors or input methods the same

733
00:46:11,000 --> 00:46:14,827
foundation, the same software
can be adapted for many

734
00:46:14,910 --> 00:46:19,197
verticals and workflows such
as retail, industrial, hospitality,

735
00:46:19,280 --> 00:46:20,997
financial service, legal and so forth.

736
00:46:21,080 --> 00:46:22,997
That is the power of the

737
00:46:23,080 --> 00:46:28,407
platform, that flexibility.

738
00:46:28,490 --> 00:46:33,377
Whatever scenario there are
places where computing has not

739
00:46:33,460 --> 00:46:33,907
naturally fit before.

740
00:46:33,990 --> 00:46:36,107
While this is an early look, we

741
00:46:36,190 --> 00:46:42,237
are really excited that ak ue

742
00:46:42,320 --> 00:46:46,687
whether, best buy, target
and others are working towards

743
00:46:46,770 --> 00:46:49,727
exploring how devices can
improve their workflows.

744
00:46:49,810 --> 00:46:54,157
This is the broader
opportunity for the ecosystem.

745
00:46:54,240 --> 00:46:58,547
Agents moving outside of the app
and taking shape in devices

746
00:46:58,630 --> 00:47:00,147
designed for specific scenario,

747
00:47:00,230 --> 00:47:02,767
a specific customer and a

748
00:47:02,850 --> 00:47:05,937
specific place.

749
00:47:06,020 --> 00:47:10,077
For all of you this is the
moment to imagine where your

750
00:47:10,160 --> 00:47:12,247
agent should live, what form

751
00:47:12,330 --> 00:47:14,377
they should take and what new

752
00:47:14,460 --> 00:47:19,387
work they can unlock.

753
00:47:19,470 --> 00:47:24,057
Last week Satya sat down with

754
00:47:24,140 --> 00:47:28,557
Qualcomm, Chrissian yaw to talk

755
00:47:28,640 --> 00:47:33,487
about the future. >> Incredible to have
you at device lab and talk about the

756
00:47:33,570 --> 00:47:35,767
reference designs. >> Happy to be here.

757
00:47:35,850 --> 00:47:38,737
I have been to the device
lab a lot recently.

758
00:47:38,820 --> 00:47:41,677
>> One of the things that you
and I have chatted for a while

759
00:47:41,760 --> 00:47:44,617
is how there's a real platform shift.

760
00:47:44,700 --> 00:47:49,937
We are moving from building
operating systems, devices for

761
00:47:50,020 --> 00:47:53,217
apps for agents.
You want to talk about how you

762
00:47:53,300 --> 00:47:57,087
see this?

763
00:47:57,170 --> 00:48:01,657
>> This is technology this
one seems to be a big one.

764
00:48:01,740 --> 00:48:03,827
Agents change the whole nature

765
00:48:03,910 --> 00:48:07,767
of the device in itself.

766
00:48:07,850 --> 00:48:11,577
Starting with the fact that if
AI understands the world as we

767
00:48:11,660 --> 00:48:14,737
understand it it will be closer
to our senses, closer to our

768
00:48:14,820 --> 00:48:19,447
eyes and mouth and ears.

769
00:48:19,530 --> 00:48:21,497
It will be things we wear.

770
00:48:21,580 --> 00:48:27,057
You need computing which
is geared towards real time

771
00:48:27,140 --> 00:48:29,957
context. And from silicon to cloud.

772
00:48:30,040 --> 00:48:34,097
>> This is one of the references
we have that I love.

773
00:48:34,180 --> 00:48:37,097
There is stuff happening 07b9 edge here.

774
00:48:37,180 --> 00:48:39,517
Things on the cloud.
You want to talk about the core

775
00:48:39,600 --> 00:48:41,837
systems and silicon implications

776
00:48:41,920 --> 00:48:47,077
for such a new ecosystem.
>> You need a power efficient

777
00:48:47,160 --> 00:48:51,247
CPU. The silicon is designed
for you to have a cloud native

778
00:48:51,330 --> 00:48:56,267
experience. You have a lot of sensors for

779
00:48:56,350 --> 00:48:56,797
context.

780
00:48:56,880 --> 00:49:01,857
It is personaized.

781
00:49:01,940 --> 00:49:06,557
It is a thank change
in the nature of devices.

782
00:49:06,640 --> 00:49:09,397
Wear able platform is changing itself.

783
00:49:09,480 --> 00:49:12,287
You see incredible factors.

784
00:49:12,370 --> 00:49:15,767
>> The computer platform, how do

785
00:49:15,850 --> 00:49:18,507
you build it such that there is

786
00:49:18,590 --> 00:49:22,847
an open ecosystem.

787
00:49:22,930 --> 00:49:25,817
It is not about one agent it is

788
00:49:25,900 --> 00:49:30,167
about any agent.
>> Because the smartphone today

789
00:49:30,250 --> 00:49:32,587
is at the center of your digital

790
00:49:32,670 --> 00:49:35,787
life, the job of those devices

791
00:49:35,870 --> 00:49:40,927
today is to be extending the

792
00:49:41,010 --> 00:49:42,097
functionality of the smartphone.

793
00:49:42,180 --> 00:49:45,747
You saw many platforms became vertical.

794
00:49:45,830 --> 00:49:50,207
It is a natural thing to have a vertical
platform from the same company.

795
00:49:50,290 --> 00:49:51,407
The phone was at the center.

796
00:49:51,490 --> 00:49:54,277
That changes dramatically when

797
00:49:54,360 --> 00:49:57,607
you think about agents.

798
00:49:57,690 --> 00:49:59,477
Agents becomes the center of

799
00:49:59,560 --> 00:50:02,767
your digital experience.

800
00:50:02,850 --> 00:50:07,757
They will be looking for an open
horizontal platform that enable

801
00:50:07,840 --> 00:50:09,857
the agents to be interacting

802
00:50:09,940 --> 00:50:12,397
with the best possible device

803
00:50:12,480 --> 00:50:16,127
for different applications.

804
00:50:16,210 --> 00:50:20,687
>> We want to make it possible
for anybody working with agentic

805
00:50:20,770 --> 00:50:25,777
system not just bound to devices
but imagine they can be many

806
00:50:25,860 --> 00:50:28,607
devices that carry that intelligence
in different context.

807
00:50:28,690 --> 00:50:32,477
That's the open ecosystem we

808
00:50:32,560 --> 00:50:36,067
want to build together.
It is great to partner with you

809
00:50:36,150 --> 00:50:38,847
to get this started.

810
00:50:38,930 --> 00:50:41,857
>> We are proud of this

811
00:50:41,940 --> 00:50:46,597
partnership and that's just the beginning.

812
00:50:46,680 --> 00:50:51,747
>> Thank you very much
Steven and Christiano.

813
00:50:51,830 --> 00:50:53,517
>> It is a new platform but it

814
00:50:53,600 --> 00:50:57,337
is a set of new platform rules

815
00:50:57,420 --> 00:51:01,537
that don't in some sense hem in

816
00:51:01,620 --> 00:51:05,347
what you can imagine the type of

817
00:51:05,430 --> 00:51:08,767
form factors where your agents live.

818
00:51:08,850 --> 00:51:13,457
Where the new platforms come you
get to rewrite even the rules of

819
00:51:13,540 --> 00:51:17,157
how new platforms operate.
That's what we are trying to get

820
00:51:17,240 --> 00:51:20,697
done with project Solara so you

821
00:51:20,780 --> 00:51:24,427
as developers and enterprises
have the flexibility to imagine

822
00:51:24,510 --> 00:51:27,387
the form factors that you want

823
00:51:27,470 --> 00:51:32,377
and have agents be ubiquitous.

824
00:51:32,460 --> 00:51:34,607
Let's go up the stack to the next layer.

825
00:51:34,690 --> 00:51:38,547
We are building a new
intelligence layer bringing

826
00:51:38,630 --> 00:51:40,477
together the models, context as

827
00:51:40,560 --> 00:51:45,587
well as the tools.

828
00:51:45,670 --> 00:51:49,087
It starts with model choice.
Every customer every developer

829
00:51:49,170 --> 00:51:51,457
is going to choose the right

830
00:51:51,540 --> 00:51:55,767
model for the right task and

831
00:51:55,850 --> 00:51:59,387
eval mix. Agency budget and foundry today

832
00:51:59,470 --> 00:52:02,407
has over 121 sthou models.

833
00:52:02,490 --> 00:52:07,307
The largest model catalog out there.

834
00:52:07,390 --> 00:52:12,177
From OpenAI to an thropic.
Even MAI models.

835
00:52:12,260 --> 00:52:14,147
Last week we brought the OpenAI

836
00:52:14,230 --> 00:52:16,237
real time voice models along

837
00:52:16,320 --> 00:52:19,517
with Claude Opus 4.8 to foundry.

838
00:52:19,600 --> 00:52:24,627
We are continuing to bring
these frontier models.

839
00:52:24,710 --> 00:52:29,467
The consideration about
the models is becoming increasingly

840
00:52:29,550 --> 00:52:33,217
key. The last developer counsel to

841
00:52:33,300 --> 00:52:36,707
now when you are building
the agentic system having this

842
00:52:36,790 --> 00:52:40,237
context really shaped right
is becoming super important.

843
00:52:40,320 --> 00:52:42,547
In fact it starts right at the

844
00:52:42,630 --> 00:52:43,147
data tier.

845
00:52:43,230 --> 00:52:47,547
The data to date has been built

846
00:52:47,630 --> 00:52:51,337
for application that is supported
user facing applications.

847
00:52:51,420 --> 00:52:53,817
Now you have to change build

848
00:52:53,900 --> 00:53:00,457
them for agents again different
call patterns even to the data

849
00:53:00,540 --> 00:53:03,997
tier. Agents are restoring,

850
00:53:04,080 --> 00:53:05,347
retrieving, reasoning, acting

851
00:53:05,430 --> 00:53:07,717
and learning.

852
00:53:07,800 --> 00:53:11,777
That is what is happening in a

853
00:53:11,860 --> 00:53:12,607
continuous loop.

854
00:53:12,690 --> 00:53:17,707
They are using Cosmos DB,

855
00:53:17,790 --> 00:53:21,347
ChatGPT does, Azure search

856
00:53:21,430 --> 00:53:26,217
retrieval of ind -- indices.

857
00:53:26,300 --> 00:53:28,957
They started the business logic

858
00:53:29,040 --> 00:53:31,887
tier effectively for agents.

859
00:53:31,970 --> 00:53:35,327
You have fabric real time

860
00:53:35,410 --> 00:53:38,517
intelligence which now brings me

861
00:53:38,600 --> 00:53:41,887
to an exciting new service

862
00:53:41,970 --> 00:53:45,407
Verizon DB which is fully

863
00:53:45,490 --> 00:53:51,847
managed SQL server on Azure.

864
00:53:51,930 --> 00:53:53,077
One of the things we wanted to

865
00:53:53,160 --> 00:53:56,767
make sure we built ground up

866
00:53:56,850 --> 00:53:59,757
Proest gres managed service for

867
00:53:59,840 --> 00:54:03,727
higher availability scaleout.

868
00:54:03,810 --> 00:54:08,327
It is automated fill over 128

869
00:54:08,410 --> 00:54:10,027
terra bites per cluster.

870
00:54:10,110 --> 00:54:15,937
The read heavy workloads.
You can scale with this managed

871
00:54:16,020 --> 00:54:16,707
service.

872
00:54:16,790 --> 00:54:20,237
In our internal testing we are

873
00:54:20,320 --> 00:54:24,407
seeing Verizon is having 3X

874
00:54:24,490 --> 00:54:29,397
through put compared to PostgreSQL.

875
00:54:29,480 --> 00:54:31,667
It is critical thinking
about the scale you need.

876
00:54:31,750 --> 00:54:34,887
The other data workload we are

877
00:54:34,970 --> 00:54:35,787
changing dramatically is the

878
00:54:35,870 --> 00:54:39,927
data warehouse in a world where

879
00:54:40,010 --> 00:54:44,627
agents are constantly querying

880
00:54:44,710 --> 00:54:47,487
data, data warehouse is critical.

881
00:54:47,570 --> 00:54:51,737
One thing to have it mission
critical for users but when

882
00:54:51,820 --> 00:54:54,277
agents need analysis done on the

883
00:54:54,360 --> 00:54:56,777
fly bringing GPU acceleration to

884
00:54:56,860 --> 00:55:03,437
fabric is super key.
We are seeing 7X performance

885
00:55:03,520 --> 00:55:08,340
gain. It is thrilling
to see that AI acceleration.

886
00:55:08,570 --> 00:55:11,657
So if you sort of have the data

887
00:55:11,740 --> 00:55:14,457
tier, the layer about this is

888
00:55:14,540 --> 00:55:17,867
the IQ layer that we are

889
00:55:17,950 --> 00:55:23,167
building that brings together essentially
the model capabilities to -- along with

890
00:55:23,250 --> 00:55:26,777
the data. You kind of mix the data and

891
00:55:26,860 --> 00:55:31,247
model capability so you can
deliver that right context to

892
00:55:31,330 --> 00:55:34,017
unlock intelligence.

893
00:55:34,100 --> 00:55:37,367
In fact when people talk about

894
00:55:37,450 --> 00:55:42,657
token efficiency this is
perhaps the most important

895
00:55:42,740 --> 00:55:45,197
consideration.
If you structure the context

896
00:55:45,280 --> 00:55:50,127
right and feed the models you
are by definition going to be so

897
00:55:50,210 --> 00:55:53,517
much more token efficient.
The first domain is the web.

898
00:55:53,600 --> 00:55:55,537
Web grounding is so important.

899
00:55:55,620 --> 00:55:58,837
You need that fresh, high

900
00:55:58,920 --> 00:56:03,307
quality and fast web data.

901
00:56:03,390 --> 00:56:07,577
And that's why we are really,
really excited to announce today

902
00:56:07,660 --> 00:56:12,100
web IQ.

903
00:56:12,200 --> 00:56:15,987
And web IQ is built on our

904
00:56:16,070 --> 00:56:19,697
global infrastructure that is
already serving over a billion

905
00:56:19,780 --> 00:56:26,117
users. But fundamentally
rearchitected for the LLM and agentic

906
00:56:26,200 --> 00:56:31,707
workflow. It is model
agnostic, MCP native.

907
00:56:31,790 --> 00:56:32,877
Plugs into any agent run time.

908
00:56:32,960 --> 00:56:36,077
It has web and news, images,

909
00:56:36,160 --> 00:56:38,777
video, so agents can ground

910
00:56:38,860 --> 00:56:42,027
responses in flesh, verify able

911
00:56:42,110 --> 00:56:45,557
content, and web IQ leads across

912
00:56:45,640 --> 00:56:47,917
all of the three key criteria.

913
00:56:48,000 --> 00:56:51,457
It is best in class in quality,
best in class in speed as well

914
00:56:51,540 --> 00:56:55,157
as in cost. So we are very, very thrilled

915
00:56:55,240 --> 00:56:57,727
about web IQ being in the

916
00:56:57,810 --> 00:57:02,807
developer's hands as you
build out agentic systems.

917
00:57:02,890 --> 00:57:04,707
We are not stopping beyond the web.

918
00:57:04,790 --> 00:57:08,237
Every developer wants to ground
their agents on what is the most

919
00:57:08,320 --> 00:57:10,877
valuable data across the enterprise.

920
00:57:10,960 --> 00:57:14,147
We are bringing together

921
00:57:14,230 --> 00:57:16,367
foundry, fabric and Microsoft

922
00:57:16,450 --> 00:57:21,587
365 as this unified IQ layer.

923
00:57:21,670 --> 00:57:25,387
Continuously updated understanding of your

924
00:57:25,470 --> 00:57:26,657
organization to show you all of

925
00:57:26,740 --> 00:57:30,097
this rich IQ in action building

926
00:57:30,180 --> 00:57:33,617
real life agents here is Elijah.

927
00:57:33,700 --> 00:57:36,850
Elijah take it away.

928
00:57:37,320 --> 00:57:39,707
>> Agents are only as good as

929
00:57:39,790 --> 00:57:41,677
the context we give them.

930
00:57:41,760 --> 00:57:44,577
Microsoft IQ unifies enterprise

931
00:57:44,660 --> 00:57:48,827
intelligence for every organization.

932
00:57:48,910 --> 00:57:52,987
I am here at a power utilities
control center and will start by

933
00:57:53,070 --> 00:57:57,487
running a long running agent.
This agent is going to help us

934
00:57:57,570 --> 00:58:01,327
assess the current grid
operations incident and produce

935
00:58:01,410 --> 00:58:04,527
a brief for us so we
can respond accordingly.

936
00:58:04,610 --> 00:58:06,887
Now I am going to go
ahead and kick that off.

937
00:58:06,970 --> 00:58:08,967
While that runs let me show you

938
00:58:09,050 --> 00:58:12,907
how we got here.

939
00:58:12,990 --> 00:58:16,747
We built our agent

940
00:58:16,830 --> 00:58:21,697
And its also wired to a
Foundry IQ knowledge base.

941
00:58:21,780 --> 00:58:26,817
A single grounded source that packages our
documents, operational data, and people

942
00:58:26,900 --> 00:58:29,117
into context the agent can reason over.

943
00:58:29,200 --> 00:58:30,987
After building the agent in

944
00:58:31,070 --> 00:58:36,297
Foundry, we published it
to Microsoft 365 for the whole

945
00:58:36,380 --> 00:58:38,647
team. Let's go see it in action.

946
00:58:38,730 --> 00:58:41,867
Now, here in M365, I will start
with a question about current

947
00:58:41,950 --> 00:58:43,907
events that an ungrounded LLM

948
00:58:43,990 --> 00:58:45,407
wouldn't be able to answer.

949
00:58:45,490 --> 00:58:49,747
I'm going to ask about current

950
00:58:49,830 --> 00:58:51,847
electricity prices in SF.

951
00:58:51,930 --> 00:58:54,977
Our agents pulls in the first IQ

952
00:58:55,060 --> 00:58:59,687
tool, Web IQ, search built for the AI era.

953
00:58:59,770 --> 00:59:03,487
It delivers industry-leading
quality, velocity, and

954
00:59:03,570 --> 00:59:03,787
efficiency.

955
00:59:03,870 --> 00:59:10,027
Web IQ constantly indexes fresh

956
00:59:10,110 --> 00:59:14,887
official sources and does
a great job with semantic

957
00:59:14,970 --> 00:59:18,407
documents. And it give us our answer.

958
00:59:18,490 --> 00:59:22,007
Grounded in reality.
Not only can we deploy our agent

959
00:59:22,090 --> 00:59:27,217
here, but we can embed it.
Let's head back to the control

960
00:59:27,300 --> 00:59:31,437
center. Let's how we handleeds a

961
00:59:31,520 --> 00:59:35,987
previous incident. Can he asked for --

962
00:59:36,070 --> 00:59:39,097
Deasked for details
at our at risk substations.

963
00:59:39,180 --> 00:59:42,627
The power of Microsoft IQ comes
when we combine that external

964
00:59:42,710 --> 00:59:45,987
knowledge with our own
internal enterprise context.

965
00:59:46,070 --> 00:59:50,567
For this information our agent
pulls in the next layer, Fabric

966
00:59:50,650 --> 00:59:52,907
IQ.

967
00:59:52,990 --> 00:59:55,777
Here is bright line's grid

968
00:59:55,860 --> 00:59:58,547
represented as a tab Rick on

969
00:59:58,630 --> 01:00:00,777
theology, an operational of the

970
01:00:00,860 --> 01:00:04,697
live grid, and we didn't build

971
01:00:04,780 --> 01:00:08,757
this from scratch.
It takes models used by millions

972
01:00:08,840 --> 01:00:12,297
of customers and lets teams
extend them into rich ontologies

973
01:00:12,380 --> 01:00:15,097
that helps run the business.
This model -- yes!

974
01:00:15,180 --> 01:00:20,117
This model is coupled with live
telemetry, so it reflects the

975
01:00:20,200 --> 01:00:23,237
real operational state
of the grid minute by minute.

976
01:00:23,320 --> 01:00:25,537
This is what Microsoft IQ means

977
01:00:25,620 --> 01:00:31,790
by enterprise intelligence.

978
01:00:31,890 --> 01:00:35,077
Not data scattered across
disparate systems, but a single

979
01:00:35,160 --> 01:00:38,537
living model of the business
that an agent can reason over.

980
01:00:38,620 --> 01:00:41,217
Now, turning back to our agent.

981
01:00:41,300 --> 01:00:46,097
We can see that it gave us a table
of our most exposed substations.

982
01:00:46,180 --> 01:00:49,867
That's the agent query
and the ontology you just saw.

983
01:00:49,950 --> 01:00:53,517
Now the agent has access to the
outside world and our COMBRID.

984
01:00:53,600 --> 01:00:58,337
The last -- and our grid. The last piece
turns the situation into a response.

985
01:00:58,420 --> 01:01:00,807
Our policies and our people.

986
01:01:00,890 --> 01:01:02,847
By asking, "What are the steps

987
01:01:02,930 --> 01:01:08,147
to respond to a substation
trip," we activate the final

988
01:01:08,230 --> 01:01:09,897
layer, Work IQ.

989
01:01:09,980 --> 01:01:15,257
This is Brightline's response
procedure in SharePoint.

990
01:01:15,340 --> 01:01:18,797
It's the play book the team reaches
for when something goes wrong.

991
01:01:18,880 --> 01:01:23,167
And the important thing,
the agent isn't working from a stale

992
01:01:23,250 --> 01:01:26,347
upload or copied snapshot.
It's answering from the same

993
01:01:26,430 --> 01:01:28,147
source the team maintains day to day.

994
01:01:28,230 --> 01:01:31,967
When the procedure changes,
the answer changes with it.

995
01:01:32,050 --> 01:01:34,637
No reuploads. No prompt rebuildings.

996
01:01:34,720 --> 01:01:38,547
And no stale versions.
And critically, this is your

997
01:01:38,630 --> 01:01:42,997
knowledge. The assets you create
stay with you, no matter what model or

998
01:01:43,080 --> 01:01:46,887
agent is reasoning over them.
Now, if we return to our agent,

999
01:01:46,970 --> 01:01:50,057
we can see the proper way
to respond to this issue.

1000
01:01:50,140 --> 01:01:55,127
Not a generic recommendation,
but our play book applied to

1001
01:01:55,210 --> 01:01:59,177
this incident. Now, we just saw one
situation, three questions, and one

1002
01:01:59,260 --> 01:02:03,937
connected answer. Now let's go check
back in on our long-running agent.

1003
01:02:04,020 --> 01:02:06,907
And... oh! Let's check the backup really

1004
01:02:06,990 --> 01:02:08,137
quick.

1005
01:02:08,220 --> 01:02:11,247
Moment of truth.

1006
01:02:11,330 --> 01:02:15,077
And...

1007
01:02:15,160 --> 01:02:20,800
Boom! Our task finished! Ha-ha!

1008
01:02:20,900 --> 01:02:23,387
Here we can see every step the

1009
01:02:23,470 --> 01:02:24,927
agent took.

1010
01:02:25,010 --> 01:02:30,667
First, beginning with Web IQ,
connecting it on the outside

1011
01:02:30,750 --> 01:02:34,187
world. Second, fabric IQ through
foundry, organicing it in the

1012
01:02:34,270 --> 01:02:35,967
is -- anchoring it
in the state of operations.

1013
01:02:36,050 --> 01:02:39,107
And work IQ, grounding
in people and procedures.

1014
01:02:39,190 --> 01:02:42,177
I triggered this manually,
but with Foundry routines, this can

1015
01:02:42,260 --> 01:02:46,777
run on a schedule, turning a one
off response into continuous,

1016
01:02:46,860 --> 01:02:50,097
proactive execution.
And if we look closely, we can

1017
01:02:50,180 --> 01:02:53,987
see it even use the power
of work IQ to alert me directly

1018
01:02:54,070 --> 01:02:57,957
about the incident.
And if we go ahead and check

1019
01:02:58,040 --> 01:03:00,557
Teams, it sent me a brief

1020
01:03:00,640 --> 01:03:07,330
notifying me of the situation.

1021
01:03:07,430 --> 01:03:10,037
That's the power of Microsoft

1022
01:03:10,120 --> 01:03:16,307
IQ. When the crisis its, teams don't

1023
01:03:16,390 --> 01:03:19,747
chance answer.
They get an answer all in one

1024
01:03:19,830 --> 01:03:25,500
place they can trust. Back to you, Satya.

1025
01:03:25,600 --> 01:03:27,187
>> All right.

1026
01:03:27,270 --> 01:03:29,617
Thank you, Elijah.

1027
01:03:29,700 --> 01:03:33,197
And now, you know, let's move up

1028
01:03:33,280 --> 01:03:37,797
from this context model layer
to deploying these agents and

1029
01:03:37,880 --> 01:03:41,387
thinking about the run time.
And, you know, when you're

1030
01:03:41,470 --> 01:03:46,077
building a first-class agentic
system, you need a first-class

1031
01:03:46,160 --> 01:03:49,577
agent run time and a platform
that we are going to ship both

1032
01:03:49,660 --> 01:03:53,947
with Windows as well
as part of Foundry in Azure.

1033
01:03:54,030 --> 01:03:55,927
We want Windows to be a

1034
01:03:56,010 --> 01:03:59,917
fantastic place to run and scale

1035
01:04:00,000 --> 01:04:01,957
agents, and agents effectively

1036
01:04:02,040 --> 01:04:03,857
are a new execution environment.

1037
01:04:03,940 --> 01:04:07,257
Right? It's a new paradigm, even.

1038
01:04:07,340 --> 01:04:12,917
They reason continuously.
They generate and run code

1039
01:04:13,000 --> 01:04:17,007
dynamically. They take action across
files and devices, as well as across

1040
01:04:17,090 --> 01:04:21,637
the network. Obviously
there's a lot of power in it.

1041
01:04:21,720 --> 01:04:23,877
Right? The fact that it can generate

1042
01:04:23,960 --> 01:04:29,597
code and act on it,
on a long-running agent that's autonomous.

1043
01:04:29,680 --> 01:04:31,097
But obviously it creates new risk.

1044
01:04:31,180 --> 01:04:33,887
And that's why today we're

1045
01:04:33,970 --> 01:04:35,487
introducing Microsoft execution

1046
01:04:35,570 --> 01:04:38,387
containers, or MXC, which is a

1047
01:04:38,470 --> 01:04:41,827
new policy layer that lets

1048
01:04:41,910 --> 01:04:44,927
Windows apply isolation and

1049
01:04:45,010 --> 01:04:49,267
containment, using AI native
-- OS native primitives.

1050
01:04:49,350 --> 01:04:54,077
You need to bake this into
the operating system, so that the

1051
01:04:54,160 --> 01:04:56,577
containment is enforced by policy.

1052
01:04:56,660 --> 01:05:01,117
You can have process level isolation
for lightweight agent actions.

1053
01:05:01,200 --> 01:05:02,797
You can have session-level

1054
01:05:02,880 --> 01:05:05,937
isolation for user separation.

1055
01:05:06,020 --> 01:05:09,887
So Windows and Linux machines,

1056
01:05:09,970 --> 01:05:13,557
including WSL,
for much stronger boundaries.

1057
01:05:13,640 --> 01:05:16,897
And in fact, if you want full
isolation and containment,

1058
01:05:16,980 --> 01:05:20,717
Windows 365 for agents,
for maximum isolation in a separate

1059
01:05:20,800 --> 01:05:23,097
managed environment, effectively.

1060
01:05:23,180 --> 01:05:28,007
You can pick the right
containment option for the work

1061
01:05:28,090 --> 01:05:30,147
load, and windows will enforce it.

1062
01:05:30,230 --> 01:05:33,177
And I think this becomes pretty
critical as you think about

1063
01:05:33,260 --> 01:05:35,767
deploying agents at scale on

1064
01:05:35,850 --> 01:05:40,587
your Windows desktop.
We want to ensure containment is

1065
01:05:40,670 --> 01:05:43,257
enforced, of course,
regardless of who builds the agent.

1066
01:05:43,340 --> 01:05:47,427
This is why you want to bake
it into the operating system.

1067
01:05:47,510 --> 01:05:52,147
On to that end we're working
with many partners to ensure

1068
01:05:52,230 --> 01:05:56,067
that taint, we're supporting
real developer workloads out

1069
01:05:56,150 --> 01:05:58,537
there and addresses the needs.

1070
01:05:58,620 --> 01:06:03,707
And NVIDIA is bring Open Shell

1071
01:06:03,790 --> 01:06:06,577
so Windows, and today, we are

1072
01:06:06,660 --> 01:06:09,567
really thrilled to announce that

1073
01:06:09,650 --> 01:06:13,387
Open Claw redundancy on Windows

1074
01:06:13,470 --> 01:06:16,710
leveraging MXC.

1075
01:06:16,810 --> 01:06:18,727
We are very deeply engaged with

1076
01:06:18,810 --> 01:06:22,557
the team to make Open Claw run

1077
01:06:22,640 --> 01:06:26,617
super well on Windows.
Let me hand it over to my

1078
01:06:26,700 --> 01:06:30,867
colleagues scat and Samantha
toish -- Scott and smoothened to

1079
01:06:30,950 --> 01:06:33,907
show you. Scott and Samantha?

1080
01:06:33,990 --> 01:06:38,877
[Cheers and applause.] >> Hey, friends!

1081
01:06:38,960 --> 01:06:41,817
Open Claw came out in November
of last year, and it took the

1082
01:06:41,900 --> 01:06:45,267
world by on the stomach.
And for the -- by the storm.

1083
01:06:45,350 --> 01:06:48,587
And I've been using it
to stay on top of my health.

1084
01:06:48,670 --> 01:06:51,787
It can help me manage my blood

1085
01:06:51,870 --> 01:06:55,497
sugar and gives me

1086
01:06:55,580 --> 01:06:55,827
notifications.

1087
01:06:55,910 --> 01:07:00,187
I've got line triaging email,

1088
01:07:00,270 --> 01:07:03,307
GitHub issues, and buys movie tickets.

1089
01:07:03,390 --> 01:07:04,767
What are you using your for?

1090
01:07:04,850 --> 01:07:08,437
>> I turned my Open Claw agent

1091
01:07:08,520 --> 01:07:10,977
into my triathlon coach.

1092
01:07:11,060 --> 01:07:12,907
It developed a work back plan

1093
01:07:12,990 --> 01:07:17,367
for me, using my data to notify

1094
01:07:17,450 --> 01:07:19,337
me how I'm progressing and

1095
01:07:19,420 --> 01:07:23,457
keeping me accountable.

1096
01:07:23,540 --> 01:07:29,927
>> We're working closely with Open Claw to
make them successful and more successful

1097
01:07:30,010 --> 01:07:34,387
on Windows. We've collaborated
on GitHub to bring you an Open Claw app.

1098
01:07:34,470 --> 01:07:40,207
It will help you set up your own
on or connect to existing ones.

1099
01:07:40,290 --> 01:07:43,407
And the Windows companion,
we're going to sandbox the Open Claw

1100
01:07:43,490 --> 01:07:47,577
tool calls to keep you
and your system safe.

1101
01:07:47,660 --> 01:07:53,117
>> Yeah! You'll see the Open Claw
windows app rin running in the background.

1102
01:07:53,200 --> 01:07:58,857
Right-click on it, Scott. >> All right.

1103
01:07:58,940 --> 01:08:00,527
>> It looks awesome!

1104
01:08:00,610 --> 01:08:05,527
It looks like a native windows

1105
01:08:05,610 --> 01:08:09,367
app because it is. It's got information
about I gateway, other machines, my

1106
01:08:09,450 --> 01:08:12,967
sessions, and my usage.
I've also got quick access to

1107
01:08:13,050 --> 01:08:17,377
things like chat, Canvas,
the main dashed, and more.

1108
01:08:17,460 --> 01:08:19,947
Let's jump into companion settings.

1109
01:08:20,030 --> 01:08:23,797
Within the app, we've got full
chat support with tool calling,

1110
01:08:23,880 --> 01:08:26,117
and you'll notice in the corner,

1111
01:08:26,200 --> 01:08:31,787
we've got lots of PERMTHSZ options,
along with our sandbox configuration.

1112
01:08:31,870 --> 01:08:33,457
>> Thepermissions options,
along with our sandbox configuration.

1113
01:08:33,540 --> 01:08:36,467
>> The sandbox is interesting
because this is using MXC, and

1114
01:08:36,550 --> 01:08:39,647
for this, we're going to be
using process isolation.

1115
01:08:39,730 --> 01:08:41,587
Now, newer version of Windows

1116
01:08:41,670 --> 01:08:46,837
will have more containment
options, so keep your eyes out

1117
01:08:46,920 --> 01:08:50,847
for news. I've got one-click security
option settings, but Samantha,

1118
01:08:50,930 --> 01:08:53,917
talk to me about custom folders.

1119
01:08:54,000 --> 01:08:58,237
>> Yeah. You've got full support
about what files and folders you want

1120
01:08:58,320 --> 01:09:01,817
Open Claw to have access to,
and really granular security

1121
01:09:01,900 --> 01:09:05,757
features like clipboard access
or talking to the internet itself.

1122
01:09:05,840 --> 01:09:11,367
Now, I've given it read-only
desktop -- read-access to your

1123
01:09:11,450 --> 01:09:12,227
depth folder.

1124
01:09:12,310 --> 01:09:13,947
Open Claw already has a rich

1125
01:09:14,030 --> 01:09:20,207
safety layer, augmented more by

1126
01:09:20,290 --> 01:09:24,247
me or policies by IT.
For the purposes of this demo,

1127
01:09:24,330 --> 01:09:28,847
I'm going to do something scary
and ask Open Claw to delete all

1128
01:09:28,930 --> 01:09:31,667
the fails on your depth
-- files on your depth.

1129
01:09:31,750 --> 01:09:35,687
>> I like a clean desktop.
>> You hid all your icons from

1130
01:09:35,770 --> 01:09:37,857
the audience while we were on stage.

1131
01:09:37,940 --> 01:09:41,957
I think we need to show the audience who
you really are >> That's so disrespectful.

1132
01:09:42,040 --> 01:09:47,347
I know where everything is. [Laughter]
It's just -- don't touch my stuff!

1133
01:09:47,430 --> 01:09:50,837
I know where they are. Sam, I need to make
sure that you're clear that a messy

1134
01:09:50,920 --> 01:09:55,507
desktop is an organized mind. I'm pretty
sure that that's the... that's the quote,

1135
01:09:55,590 --> 01:09:58,807
right? So what we've done, asked
Open Claw to delete those files from

1136
01:09:58,890 --> 01:10:01,647
the Windows node, and the only
thing that is going to keep it

1137
01:10:01,730 --> 01:10:06,037
from happening is MXC because
we've turned off all of the

1138
01:10:06,120 --> 01:10:07,587
layers that Open Claw offers,

1139
01:10:07,670 --> 01:10:09,757
but our IT, in this case,

1140
01:10:09,840 --> 01:10:13,997
Samantha, has stet to

1141
01:10:14,080 --> 01:10:19,847
-- you can see the attempts,
and checking the directory, and

1142
01:10:19,930 --> 01:10:25,537
deleting again. It wants the files
gone, and I want them to stay.

1143
01:10:25,620 --> 01:10:26,177
[Laughter] Ope! No!

1144
01:10:26,260 --> 01:10:32,777
The read-only sandbox is there.

1145
01:10:32,860 --> 01:10:36,097
94 je pegs are still there. They're safe!

1146
01:10:36,180 --> 01:10:38,637
>> Foiled again! >> Oh, my goodness.

1147
01:10:38,720 --> 01:10:43,727
So bad. >> We've seen security

1148
01:10:43,810 --> 01:10:46,357
sandboxing, and brought together
all in this alpha release of a

1149
01:10:46,440 --> 01:10:50,027
Windows companion app. We think this app
is a great opportunity to showcase Open

1150
01:10:50,110 --> 01:10:53,587
Claw on windows, and it's only going
to get better in the coming months

1151
01:10:53,670 --> 01:10:57,067
>> That is right. And, by the way,
I want to note that we're doing a all this

1152
01:10:57,150 --> 01:10:59,937
development work on this calm
Windows development machine with

1153
01:11:00,020 --> 01:11:04,107
all the tools that I love, like

1154
01:11:04,190 --> 01:11:09,117
WSL, containers, a containment, and I've
even got GitHub Copilot with multimodel

1155
01:11:09,200 --> 01:11:11,837
support ready to go.
This is literally how the team

1156
01:11:11,920 --> 01:11:16,287
and I have been working
on the Open Claw app on GitHub.

1157
01:11:16,370 --> 01:11:18,627
Now, people wonder how this came together.

1158
01:11:18,710 --> 01:11:22,827
Turns out that over the holidays,
I got a DM from this

1159
01:11:22,910 --> 01:11:27,447
random guy on the intent, and it
takes me a couple of days to get

1160
01:11:27,530 --> 01:11:31,707
back to him. And then when I did,
we just had this kind of cool idea.

1161
01:11:31,790 --> 01:11:35,177
That maybe he should come
to -- to Microsoft Build.

1162
01:11:35,260 --> 01:11:37,077
I think there's one more
person that we want to thank for

1163
01:11:37,160 --> 01:11:40,517
bringing the next generation
of agents to the world.

1164
01:11:40,600 --> 01:11:43,267
Everybody put your hands

1165
01:11:43,350 --> 01:11:49,857
together for the claw father
himself, Peter Steinberger!

1166
01:11:49,940 --> 01:11:51,957
[Cheers and applause.] >>
Samantha, it's just showing

1167
01:11:52,040 --> 01:11:54,727
off my secret DMs!

1168
01:11:54,810 --> 01:11:59,777
I'm so excited to see Open
Claw native on Windows.

1169
01:11:59,860 --> 01:12:01,787
You know, watching a claw try to

1170
01:12:01,870 --> 01:12:07,737
delete all your desktop files
and just fail makes me really happy.

1171
01:12:07,820 --> 01:12:12,507
[Laughter] Because six months ago,
that totally would've worked.

1172
01:12:12,590 --> 01:12:16,367
[Laughter] I built Open Claw
to have access to everything.

1173
01:12:16,450 --> 01:12:18,767
My files, my machines, my chats.

1174
01:12:18,850 --> 01:12:21,087
Always on, and fully open source.

1175
01:12:21,170 --> 01:12:23,287
That's what makes it so powerful!

1176
01:12:23,370 --> 01:12:24,727
And that's what also makes

1177
01:12:24,810 --> 01:12:27,697
companies a bit nervous.

1178
01:12:27,780 --> 01:12:32,467
You know, what I kept hearing
was, "Peter, I love my claw.

1179
01:12:32,550 --> 01:12:35,447
Can I use this at work?"
And that's what we spent the last

1180
01:12:35,530 --> 01:12:37,267
few months on.

1181
01:12:37,350 --> 01:12:41,477
With Microsoft, git Hick, open

1182
01:12:41,560 --> 01:12:44,177
AI, NVIDIA, just to name a few.

1183
01:12:44,260 --> 01:12:47,677
We had observability,
auto mode for permissions.

1184
01:12:47,760 --> 01:12:50,937
We changed how access works.
It's not all or nothing anymore.

1185
01:12:51,020 --> 01:12:56,287
You can pick which folder should read
only, which ones should be hidden.

1186
01:12:56,370 --> 01:12:59,757
Here's the nutrition: -- Here's the news:.

1187
01:12:59,840 --> 01:13:01,827
You can run it inside your company now.

1188
01:13:01,910 --> 01:13:07,487
[Cheering] And we even made the harness

1189
01:13:07,570 --> 01:13:08,617
itself a plugin.

1190
01:13:08,700 --> 01:13:11,467
You can bring your own.

1191
01:13:11,550 --> 01:13:14,837
Copilot, codex, whatever you

1192
01:13:14,920 --> 01:13:19,347
already trust in and
your rules come with it.

1193
01:13:19,430 --> 01:13:21,817
And you put Open Claw on top of it.

1194
01:13:21,900 --> 01:13:25,437
Persistent memory, heartbeats,
and a claw right inside Slack or

1195
01:13:25,520 --> 01:13:27,287
Teams.

1196
01:13:27,370 --> 01:13:31,087
It's been really exciting to see

1197
01:13:31,170 --> 01:13:35,627
Open Claw grow into something much bigger.

1198
01:13:35,710 --> 01:13:37,327
A global movement.

1199
01:13:37,410 --> 01:13:39,687
And a community.

1200
01:13:39,770 --> 01:13:43,087
And I started the Open Claw

1201
01:13:43,170 --> 01:13:45,467
Foundation, a really nonprofit.

1202
01:13:45,550 --> 01:13:48,407
Any model, any operating system.

1203
01:13:48,490 --> 01:13:51,947
Because we are entering a new

1204
01:13:52,030 --> 01:13:55,847
era of building agents,
more capability for the people who

1205
01:13:55,930 --> 01:13:59,617
don't code, and more
power for those who do.

1206
01:13:59,700 --> 01:14:02,517
And we get to do this, we get to

1207
01:14:02,600 --> 01:14:04,687
build it together in the open.

1208
01:14:04,770 --> 01:14:08,197
So my task is simple: Come build with us.

1209
01:14:08,280 --> 01:14:10,697
Thank you. [Cheers and applause.]

1210
01:14:10,780 --> 01:14:15,647
>> Fantastic.
Thank you very much to Peter, to

1211
01:14:15,730 --> 01:14:18,767
Samantha, to the companion
app team for their hard work.

1212
01:14:18,850 --> 01:14:22,507
Thank you to the Open Claw
community for giving everyone

1213
01:14:22,590 --> 01:14:26,077
here a crucifix crustacean
-- a crustacean of their own.

1214
01:14:26,160 --> 01:14:31,537
And that's the first time you've
even Open Claw running on a

1215
01:14:31,620 --> 01:14:36,857
Surface ultra.
[Cheers and applause.] >> All right.

1216
01:14:36,940 --> 01:14:39,287
Thank you very much,
Samantha, Scott, and Peter.

1217
01:14:39,370 --> 01:14:42,057
It's so wonderful to see Open

1218
01:14:42,140 --> 01:14:44,997
Claw come to Windows.

1219
01:14:45,080 --> 01:14:48,117
And have all of that capability

1220
01:14:48,200 --> 01:14:53,367
in terms of the security and --
and that comfort, to be able to

1221
01:14:53,450 --> 01:14:57,037
have these long running agents
and unmetered intelligence come together.

1222
01:14:57,120 --> 01:15:00,147
So now, let's move from the

1223
01:15:00,230 --> 01:15:02,247
Windows side on the edge to the

1224
01:15:02,330 --> 01:15:04,167
cloud with Foundry briar briar,

1225
01:15:04,250 --> 01:15:08,757
and we're building Foundry into

1226
01:15:08,840 --> 01:15:12,427
a full application
platform for the agent era.

1227
01:15:12,510 --> 01:15:15,397
Right? Every era of platform shift,

1228
01:15:15,480 --> 01:15:17,867
when you move to the cloud, we

1229
01:15:17,950 --> 01:15:24,007
have the cloud stack in Foundry.

1230
01:15:24,090 --> 01:15:28,007
We're excited about the Foundry
hosted agent as a runtime for

1231
01:15:28,090 --> 01:15:29,847
long running agents.

1232
01:15:29,930 --> 01:15:35,217
You know, agents now have --
if you're building in Foundry

1233
01:15:35,300 --> 01:15:38,497
hosted agents, you can
have all the IQ layers.

1234
01:15:38,580 --> 01:15:42,757
You have the tools.
You have the durability and the

1235
01:15:42,840 --> 01:15:45,787
memory and the state.
You have your own sandbox.

1236
01:15:45,870 --> 01:15:49,297
In fact, it's a super-fast

1237
01:15:49,380 --> 01:15:51,297
sandbox that you can spin up.

1238
01:15:51,380 --> 01:15:54,187
You can generate the rubrics.

1239
01:15:54,270 --> 01:15:58,567
You can -- and the evals.
You can in fact have all the

1240
01:15:58,650 --> 01:16:02,437
safety and the guardrails
around your agentic system.

1241
01:16:02,520 --> 01:16:04,847
And in fact, even one of the
coolest things in Foundry is

1242
01:16:04,930 --> 01:16:07,347
it's a continuously improving loop, even.

1243
01:16:07,430 --> 01:16:10,137
It's got that self-improvement
loop built in.

1244
01:16:10,220 --> 01:16:13,617
You build an agent that's
continuously getting better.

1245
01:16:13,700 --> 01:16:16,857
And I'm really also excited

1246
01:16:16,940 --> 01:16:20,757
today to announce a partnership

1247
01:16:20,840 --> 01:16:26,177
with Fireworks AI, bringing
all of their open wave models to

1248
01:16:26,260 --> 01:16:30,507
Foundry. That means giving you
as developmenters more choices, as

1249
01:16:30,590 --> 01:16:32,867
well as that great inference
stack, to build the next

1250
01:16:32,950 --> 01:16:35,837
generation of these agentic

1251
01:16:35,920 --> 01:16:41,317
indications, with all
the enterprise Rayls that Foundry

1252
01:16:41,400 --> 01:16:43,437
has, as well as the governance
that Foundry has.

1253
01:16:43,520 --> 01:16:48,670
Really excited about that partnership.

1254
01:16:48,770 --> 01:16:53,057
And so now that brings us to the tools.

1255
01:16:53,140 --> 01:16:55,097
Data itself is at the heart of all this.

1256
01:16:55,180 --> 01:16:57,897
You know, in fact, Jensen spoke to this.

1257
01:16:57,980 --> 01:17:00,247
GitHub is not just about the code repo.

1258
01:17:00,330 --> 01:17:04,467
It's becoming the control
plane for all the agents.

1259
01:17:04,550 --> 01:17:06,137
And nearly everything we measure

1260
01:17:06,220 --> 01:17:09,037
on GitHub, whether it's repo

1261
01:17:09,120 --> 01:17:11,707
creation, PR activity, API

1262
01:17:11,790 --> 01:17:16,497
usage, actions, all of this are
going faster because of the

1263
01:17:16,580 --> 01:17:22,457
agentic workflows. The new scale is driven
by humans and agents collaborating

1264
01:17:22,540 --> 01:17:22,717
together.

1265
01:17:22,800 --> 01:17:26,897
And they're exposing too long

1266
01:17:26,980 --> 01:17:30,327
-- tooling crass every form factor.

1267
01:17:30,410 --> 01:17:33,017
We've seen tremendous growth
in CLI, driven, you know, the

1268
01:17:33,100 --> 01:17:37,407
approachability of the CLI form
factor, it's allusion -- always

1269
01:17:37,490 --> 01:17:42,707
been great to go to a termal.
And now when combined with

1270
01:17:42,790 --> 01:17:46,047
models and natural language,
CLI is the thing that everything goes to.

1271
01:17:46,130 --> 01:17:48,117
But end of the day, when you

1272
01:17:48,200 --> 01:17:54,687
have hundreds of CLI's,
it becomes pretty complicated.

1273
01:17:54,770 --> 01:17:59,427
It doesn't scale. Especially the cognitive
load that I have when you have 100CLI

1274
01:17:59,510 --> 01:18:03,927
sessions open, is such that you
kind of need something new.

1275
01:18:04,010 --> 01:18:06,227
And so that's what has led us to

1276
01:18:06,310 --> 01:18:11,407
build -- you know, we need like
this tool, essentially, that has

1277
01:18:11,490 --> 01:18:14,407
the speed and the flexibility of a CLI.

1278
01:18:14,490 --> 01:18:18,977
But has the capability of an IE,
and the ability to scale to

1279
01:18:19,060 --> 01:18:21,477
infinite number of agent sessions.

1280
01:18:21,560 --> 01:18:23,727
So today we're taking that next

1281
01:18:23,810 --> 01:18:27,757
big step, introducing our new

1282
01:18:27,840 --> 01:18:33,980
git Hick Copilot app --
Our new GitHub cCopilot app.

1283
01:18:34,070 --> 01:18:35,457
And we realize it's not sufficient.

1284
01:18:35,540 --> 01:18:38,097
Because you still have
a back toned deal with.

1285
01:18:38,180 --> 01:18:40,617
Code is easy to generate,
but what about the back end?

1286
01:18:40,700 --> 01:18:43,117
You need to content with

1287
01:18:43,200 --> 01:18:44,907
identities, storage, database

1288
01:18:44,990 --> 01:18:47,777
schemas, and that's why we're

1289
01:18:47,860 --> 01:18:51,547
really super excited about

1290
01:18:51,630 --> 01:18:56,077
Rayfin, an agent-first SDK that
connects your agents to the back

1291
01:18:56,160 --> 01:19:00,257
end as a service, and we're bringing
this to everywhere you build.

1292
01:19:00,340 --> 01:19:05,087
That's why I'm, again, super
excited about Rayfin and the

1293
01:19:05,170 --> 01:19:12,080
relationship with Replit. You can now
build apps in Replit, while the -- yeah.

1294
01:19:12,180 --> 01:19:16,367
You can build the app in Replit
while the app and data are

1295
01:19:16,450 --> 01:19:18,167
deployed into the

1296
01:19:18,250 --> 01:19:19,707
enterprise-managed fabric

1297
01:19:19,790 --> 01:19:23,577
tenant, thanks to the Rayfin

1298
01:19:23,660 --> 01:19:26,947
SDK, which is available now
for anyone else to be able to use

1299
01:19:27,030 --> 01:19:29,897
with their tools as a back end,
and essentially have the back

1300
01:19:29,980 --> 01:19:34,587
end service.
And now to show you all of the

1301
01:19:34,670 --> 01:19:36,927
Foundry and Rayfin and building

1302
01:19:37,010 --> 01:19:38,587
agents and the long running

1303
01:19:38,670 --> 01:19:43,057
agents, let me invite Cassidy.

1304
01:19:43,140 --> 01:19:46,127
Take with away. >> Hello, everyone!

1305
01:19:46,210 --> 01:19:50,337
I'm so excited to be
amongst my fellow Devs.

1306
01:19:50,420 --> 01:19:52,237
This is a long day. Fix your shoulders.

1307
01:19:52,320 --> 01:19:54,907
Sit back. I see a lot of you sitting up.

1308
01:19:54,990 --> 01:19:57,447
Great. I know you're drinking from
the fire hose of information today,

1309
01:19:57,530 --> 01:20:01,217
so I want you to go into the next
few minutes thinking, "What

1310
01:20:01,300 --> 01:20:03,997
can I drywall try out on -- what

1311
01:20:04,080 --> 01:20:08,617
can I try out later today?"
First, I can't wait to show you

1312
01:20:08,700 --> 01:20:12,557
the new GitHub Copilot app.
This is your home base for

1313
01:20:12,640 --> 01:20:13,927
development and operations
on your computer.

1314
01:20:14,010 --> 01:20:17,097
And we think you're going to love it.
Let me show you around.

1315
01:20:17,180 --> 01:20:20,817
When you open up the app,
you see the home screen where you

1316
01:20:20,900 --> 01:20:26,137
can kick off a new session.
But before night the serious

1317
01:20:26,220 --> 01:20:29,177
stuff, you can drag Mona
around, and there's a game!

1318
01:20:29,260 --> 01:20:32,207
Look! It's so fun!
[Laughter] I'm not very good at it, so

1319
01:20:32,290 --> 01:20:37,327
let's just get back to, you can kick
off a new agentic coding session.

1320
01:20:37,410 --> 01:20:39,987
So I started off one
a little bit earlier here.

1321
01:20:40,070 --> 01:20:43,557
And it gave me a review
of a bunch of release blockers.

1322
01:20:43,640 --> 01:20:48,197
Which one should I fix? Call it out.

1323
01:20:48,280 --> 01:20:50,497
83? The critical ones? You know what?

1324
01:20:50,580 --> 01:20:52,927
How about we just do all of them?

1325
01:20:53,010 --> 01:20:57,197
Let's go! This app will now kick
off a separate session for every

1326
01:20:57,280 --> 01:21:02,537
single issue here. I don't have to worry
about stashing or coding complex or

1327
01:21:02,620 --> 01:21:08,217
anything, because the app
takes care of that with Git work

1328
01:21:08,300 --> 01:21:09,967
trees, that are isolated
environments for each session.

1329
01:21:10,050 --> 01:21:12,667
Your agents can work in parallel
without stepping on each other.

1330
01:21:12,750 --> 01:21:15,187
But you still have to merge them.

1331
01:21:15,270 --> 01:21:18,587
Right? So -- [chuckling] --
Copilot has your back there too.

1332
01:21:18,670 --> 01:21:20,597
If I head over -- not to this one.

1333
01:21:20,680 --> 01:21:23,357
But to this issue here, I can

1334
01:21:23,440 --> 01:21:27,717
run agent merge.
And when I enable agent merge,

1335
01:21:27,800 --> 01:21:33,067
cofight will baby sit this PR through CI
check, code review, and merge collects.

1336
01:21:33,150 --> 01:21:36,437
Okay. Let me keep showing
you around while it's running.

1337
01:21:36,520 --> 01:21:39,177
If I head over to my work, I can

1338
01:21:39,260 --> 01:21:45,337
see a focused view of all of my activity
ask just projects loaded in the app, PRs,

1339
01:21:45,420 --> 01:21:50,357
everything here. And under automation,
I have a bunch of reusable sessions that

1340
01:21:50,440 --> 01:21:52,257
can run locally or on the cloud.

1341
01:21:52,340 --> 01:21:54,397
You see there's issue poetry there.

1342
01:21:54,480 --> 01:21:57,227
That's real. And that is load-bearing.

1343
01:21:57,310 --> 01:22:02,047
Okay! [Chuckling]. And now undersessions,
like I briefly showed earlier, these

1344
01:22:02,130 --> 01:22:03,367
are sessions.

1345
01:22:03,450 --> 01:22:06,567
If I want to add a new

1346
01:22:06,650 --> 01:22:11,647
repository, I can click the button,
and it will pull one,

1347
01:22:11,730 --> 01:22:12,947
and if I were to add one, I can

1348
01:22:13,030 --> 01:22:15,947
add a session, and this is an

1349
01:22:16,030 --> 01:22:22,117
open source -- I can start
a session anywhere and it loads

1350
01:22:22,200 --> 01:22:26,797
it. When I look at a session within
this repository, let me look at

1351
01:22:26,880 --> 01:22:31,597
the other one, I get an integrated
browser, there's a terminal.

1352
01:22:31,680 --> 01:22:35,247
And a chat. I can technology light
and -- toggle light and dark mode.

1353
01:22:35,330 --> 01:22:38,767
And there's a button, pick and polish.

1354
01:22:38,850 --> 01:22:42,237
I can polish anything in this app,
and it adds it to the chat,

1355
01:22:42,320 --> 01:22:46,277
and I can say, hey, I want you
to add reordering to this list.

1356
01:22:46,360 --> 01:22:48,977
And it will just work,
all living in there.

1357
01:22:49,060 --> 01:22:52,797
I have access to all the most
popular models via my single

1358
01:22:52,880 --> 01:22:55,317
GitHub Copilot subscription,

1359
01:22:55,400 --> 01:23:00,627
including from open AI,
Ann throw meic, and Google.

1360
01:23:00,710 --> 01:23:02,757
And having a model change is agreed.

1361
01:23:02,840 --> 01:23:07,077
Not only can you pick the right
one, but Copilot can request a

1362
01:23:07,160 --> 01:23:08,647
rubber duck review.

1363
01:23:08,730 --> 01:23:12,707
In this session, I was use use

1364
01:23:12,790 --> 01:23:15,737
GPT5.5, but it asked one from

1365
01:23:15,820 --> 01:23:20,407
Claude opus 4.8.
All models have blind spots, and

1366
01:23:20,490 --> 01:23:23,197
the power of this approach
means I can catch them earlier.

1367
01:23:23,280 --> 01:23:28,317
This is all very cool.
But working with AI in 2026

1368
01:23:28,400 --> 01:23:29,887
should be more than just chat.

1369
01:23:29,970 --> 01:23:32,957
You just saw me scrolling up.
There's so many words here.

1370
01:23:33,040 --> 01:23:39,327
[Laughter] So today, I'm very
excited to show you the concept of a

1371
01:23:39,410 --> 01:23:43,047
canvass. I'm going to own one right there.

1372
01:23:43,130 --> 01:23:46,467
The canvas is how an agent can
build a custom UI to communicate

1373
01:23:46,550 --> 01:23:47,107
with you.

1374
01:23:47,190 --> 01:23:53,747
What if your AI could see?
Everyone say, demo gods, bless

1375
01:23:53,830 --> 01:23:58,267
us. Let's see if it works.
Here's a fun canvas where, if I

1376
01:23:58,350 --> 01:24:02,657
get the camera going, okay.
The agent shows your PRs down

1377
01:24:02,740 --> 01:24:04,527
here, and I can toggle it with a

1378
01:24:04,610 --> 01:24:10,497
thumbs up or thumbs down.
Let's aapprove it!

1379
01:24:10,580 --> 01:24:12,767
Hey! It's so fun! And goes deeper.

1380
01:24:12,850 --> 01:24:14,747
This is just the beginning
of what you can do.

1381
01:24:14,830 --> 01:24:16,247
So again, I kicked off a bunch

1382
01:24:16,330 --> 01:24:19,867
of different sessions earlier.

1383
01:24:19,950 --> 01:24:24,007
This is a signal box app.
It's 100% agent-build.

1384
01:24:24,090 --> 01:24:27,447
It's containerized
with a database back end.

1385
01:24:27,530 --> 01:24:30,027
Would you be able to deploy
this to your enterprise with no

1386
01:24:30,110 --> 01:24:31,067
questions asked?

1387
01:24:31,150 --> 01:24:37,827
Be honest. No. Yes.

1388
01:24:37,910 --> 01:24:42,757
Exact. No, but you can with Rayfin,
and that is very, very, exciting.

1389
01:24:42,840 --> 01:24:44,927
Let me open up a new terminal.

1390
01:24:45,010 --> 01:24:48,587
All I have to do is type "Rayfin

1391
01:24:48,670 --> 01:24:52,137
up." And then, demo gods bless you!

1392
01:24:52,220 --> 01:24:52,367
Come on!

1393
01:24:52,450 --> 01:24:54,837
It will...

1394
01:24:54,920 --> 01:25:00,407
[Singing] maybe dedeploy! Blammo!

1395
01:25:00,490 --> 01:25:02,317
It happened! Yes!
And hosted on Microsoft fabric.

1396
01:25:02,400 --> 01:25:04,267
I know that's a lot. I know that's a lot.

1397
01:25:04,350 --> 01:25:09,187
With Rayfin, your agents get
a complete enterprise back end, so

1398
01:25:09,270 --> 01:25:12,457
you can deploy with confidence
in the way that's best for you.

1399
01:25:12,540 --> 01:25:15,427
But this is the key thing
to remember: This app is not just

1400
01:25:15,510 --> 01:25:18,567
another session manager. Yes, it is.

1401
01:25:18,650 --> 01:25:20,447
It managing a lot of sessions.

1402
01:25:20,530 --> 01:25:25,167
But session managers just make
it easy to create work, but this

1403
01:25:25,250 --> 01:25:28,077
helps you to finish it.
Thank you very much.

1404
01:25:28,160 --> 01:25:28,737
Happy pride.

1405
01:25:28,820 --> 01:25:35,820
Back to you, Satya.

1406
01:25:35,960 --> 01:25:40,647
>> All right. Now we're
back to IDEs that have UI.

1407
01:25:40,730 --> 01:25:43,587
That sort of -- it's so cool! [Laughter]

1408
01:25:43,670 --> 01:25:45,827
It's come full circle.

1409
01:25:45,910 --> 01:25:48,727
Now, let's talk about how you

1410
01:25:48,810 --> 01:25:51,167
can observe, govern, and secure

1411
01:25:51,250 --> 01:25:53,167
these agents.

1412
01:25:53,250 --> 01:25:57,067
Agent 365 is the agent control

1413
01:25:57,150 --> 01:26:02,577
plane, agent requires their own
identities, access controls.

1414
01:26:02,660 --> 01:26:07,147
Even when they're working on your behalf,
right? You just want that work on behalf

1415
01:26:07,230 --> 01:26:08,417
identity to be enforced.

1416
01:26:08,500 --> 01:26:14,057
So we extended an intraagent

1417
01:26:14,140 --> 01:26:17,957
need real-time defense,
so we extended Defender.

1418
01:26:18,040 --> 01:26:19,957
Agents require this always-on

1419
01:26:20,040 --> 01:26:23,027
data protection and compliance,

1420
01:26:23,110 --> 01:26:27,447
so we extended Purview.
And these agents can be hosted

1421
01:26:27,530 --> 01:26:30,807
everywhere, AWS, GCP, not just on Azure.

1422
01:26:30,890 --> 01:26:36,377
Or build with any framework.
And today we are announcing a

1423
01:26:36,460 --> 01:26:38,277
number of updates, including the

1424
01:26:38,360 --> 01:26:42,817
GA of agent 265SDK, and we're x

1425
01:26:42,900 --> 01:26:45,697
expanding it to your local
agents running on Windows and

1426
01:26:45,780 --> 01:26:47,857
elsewhere, and the claws,
you just saw earlier.

1427
01:26:47,940 --> 01:26:50,517
And let us take a look at how

1428
01:26:50,600 --> 01:26:54,097
all the Agent 365 composes.

1429
01:26:54,180 --> 01:27:00,180
Over to you, Amanda.

1430
01:27:00,270 --> 01:27:03,787
>> Everyone is building agents,
but that's not the hard part anymore.

1431
01:27:03,870 --> 01:27:07,067
The hard part is intrigue
grating them into your business

1432
01:27:07,150 --> 01:27:11,847
and governing them at scale. Today I
will show you how Foundry makes this easy.

1433
01:27:11,930 --> 01:27:16,247
Let's start locally. I've already built a
line graph agent, and now I will show you

1434
01:27:16,330 --> 01:27:21,117
the value Foundry adds. First, tools.

1435
01:27:21,200 --> 01:27:24,187
Agents need tools to get work done,
and with Foundry tool box,

1436
01:27:24,270 --> 01:27:25,927
I just add my tools once, and

1437
01:27:26,010 --> 01:27:27,527
any agent can consume them

1438
01:27:27,610 --> 01:27:31,327
through a single MCP end point.

1439
01:27:31,410 --> 01:27:35,647
And because tools essentially,
that means governance does as

1440
01:27:35,730 --> 01:27:38,567
well. I've applied a guard rile that

1441
01:27:38,650 --> 01:27:42,207
blocks PII from leaking, and all

1442
01:27:42,290 --> 01:27:44,147
I have to do is apply this once,

1443
01:27:44,230 --> 01:27:47,347
and all my agents are protected.

1444
01:27:47,430 --> 01:27:51,097
Now let's make this agent
enterprise-ready.

1445
01:27:51,180 --> 01:27:53,937
To deploy this agent to Foundry,
all I have to do is add this one

1446
01:27:54,020 --> 01:27:58,557
block of code, and then I push
my changes and GitHub actions

1447
01:27:58,640 --> 01:28:02,297
takes it from there.
Once deployed, I can actually

1448
01:28:02,380 --> 01:28:06,727
use this agent in the same
Foundry extension I showed you

1449
01:28:06,810 --> 01:28:09,487
earlier. And let's now test it out.

1450
01:28:09,570 --> 01:28:14,107
I'm asking it to track a few open items,
from a standup I had earlier today.

1451
01:28:14,190 --> 01:28:18,377
And what's actually happening
right now, is Foundry is

1452
01:28:18,460 --> 01:28:21,847
spinning up a dedicated
microVM just for this session.

1453
01:28:21,930 --> 01:28:26,367
And the session even gets its
own persistent file system.

1454
01:28:26,450 --> 01:28:29,657
What you're going to see
in a minute here, if I go to the

1455
01:28:29,740 --> 01:28:34,927
Files tab to track the open
items, the agent's actually

1456
01:28:35,010 --> 01:28:36,027
writing to a file.

1457
01:28:36,110 --> 01:28:42,787
Pretty cool, right?

1458
01:28:42,870 --> 01:28:46,767
Off the top of my head has
server side traces and evals to

1459
01:28:46,850 --> 01:28:48,437
show me exactly what
happened on every run.

1460
01:28:48,520 --> 01:28:50,207
But how do I know if my agent's

1461
01:28:50,290 --> 01:28:51,577
doing a good job?

1462
01:28:51,660 --> 01:28:56,517
That's what Foundry's brand-new

1463
01:28:56,600 --> 01:28:59,897
rubric evaluators are for.

1464
01:28:59,980 --> 01:29:04,557
One easy command, Foundry
reads my agent and generates the

1465
01:29:04,640 --> 01:29:08,697
evaluation criteria for me.
It creates a rubric personalized

1466
01:29:08,780 --> 01:29:12,067
just to this agent.
Now let's check out this rubric

1467
01:29:12,150 --> 01:29:15,417
in Foundry portal.

1468
01:29:15,500 --> 01:29:20,067
Look at these dimensions.
Governance outcome correctness,

1469
01:29:20,150 --> 01:29:22,937
prescribed source usage.
I didn't write any of these.

1470
01:29:23,020 --> 01:29:25,677
Foundry generated them
from production traces.

1471
01:29:25,760 --> 01:29:28,417
And with this rubric, I can

1472
01:29:28,500 --> 01:29:33,137
score my agent, and run evals,
but we can do so much more than that.

1473
01:29:33,220 --> 01:29:37,787
That's where Foundry's brand-new
agent optimizer takes over.

1474
01:29:37,870 --> 01:29:43,897
Here's how it works. It tunes four
things: The model, instructions, tool

1475
01:29:43,980 --> 01:29:46,667
descriptions, and scales.
And then it generates improved

1476
01:29:46,750 --> 01:29:51,037
candidates and scores each one
using the rubric I just showed you.

1477
01:29:51,120 --> 01:29:52,037
Here, I can view the candidates.

1478
01:29:52,120 --> 01:29:55,707
I can see the strategy used.
I can see their scores.

1479
01:29:55,790 --> 01:29:59,677
And I can actually view
exactly what changed.

1480
01:29:59,760 --> 01:30:02,347
This candidate, for example,
improved its score by updating

1481
01:30:02,430 --> 01:30:05,397
the model and the system prompt.

1482
01:30:05,480 --> 01:30:08,787
Foundry then makes it super easy
to deploy the best candidate as

1483
01:30:08,870 --> 01:30:14,657
a brand-new agent version.
But this is not a one time

1484
01:30:14,740 --> 01:30:16,657
thing. Every run feeds the next eval.

1485
01:30:16,740 --> 01:30:20,917
And every eval tells the optimizer
where to improve next.

1486
01:30:21,000 --> 01:30:24,467
So your agents now get
better the more they're used.

1487
01:30:24,550 --> 01:30:27,677
But now, let's put this agent to work.

1488
01:30:27,760 --> 01:30:30,847
I've published my agent to Teams

1489
01:30:30,930 --> 01:30:34,917
and M365 Copilot.
Right now I'm going to ask it to

1490
01:30:35,000 --> 01:30:39,597
just catch me up on what I've
missed, because I've been of

1491
01:30:39,680 --> 01:30:41,617
offline all morning, but my agent hasn't.

1492
01:30:41,700 --> 01:30:46,927
While it works on that, though,
let me explain what makes this

1493
01:30:47,010 --> 01:30:50,497
agent different. This is an autopilot
agent, which means it has its own

1494
01:30:50,580 --> 01:30:52,397
identity and productivity license.

1495
01:30:52,480 --> 01:30:56,347
So it can work across
M365 on its own behalf.

1496
01:30:56,430 --> 01:31:00,107
Earlier, you saw me using
this agent by myself.

1497
01:31:00,190 --> 01:31:04,007
But shipping of features and new
release takes a whole team.

1498
01:31:04,090 --> 01:31:07,677
And that why this agent
lives in our Teams group chat.

1499
01:31:07,760 --> 01:31:10,497
Let's check out how it did.

1500
01:31:10,580 --> 01:31:13,617
Serums as you didded --

1501
01:31:13,700 --> 01:31:15,757
summariesed all updates and call

1502
01:31:15,840 --> 01:31:20,897
out features I need to track.
But putting this to work in your

1503
01:31:20,980 --> 01:31:22,857
enterprise, governance
needs to come first.

1504
01:31:22,940 --> 01:31:27,317
That's why every autopilot agent

1505
01:31:27,400 --> 01:31:32,537
requires admin approval to she
can review it and choose who has

1506
01:31:32,620 --> 01:31:33,867
the ability to talk to it.

1507
01:31:33,950 --> 01:31:37,477
But it doesn'tent there.

1508
01:31:37,560 --> 01:31:41,547
An over approval, add minutes
anterior and block -- monitor it

1509
01:31:41,630 --> 01:31:45,837
and block it at any time. But governance
can't apply to just one agent.

1510
01:31:45,920 --> 01:31:47,717
Every agent in your organization

1511
01:31:47,800 --> 01:31:53,227
needs to be managed with the same
rigor as users, apps, and devices.

1512
01:31:53,310 --> 01:31:57,327
And that's exactly what
Microsoft Agent 365 provides.

1513
01:31:57,410 --> 01:32:01,447
So let's zoom out.
Today we saw Foundry accelerate

1514
01:32:01,530 --> 01:32:04,667
development, take your agent
from local to enterprise-ready,

1515
01:32:04,750 --> 01:32:09,207
and put it to work in M365.
Foundry makes it simple.

1516
01:32:09,290 --> 01:32:11,777
You build the agent; we handle the rest.

1517
01:32:11,860 --> 01:32:15,200
Now back to you, Satya!

1518
01:32:15,530 --> 01:32:19,667
>> Thank you, Amanda.

1519
01:32:19,750 --> 01:32:26,057
You know, what you just saw
was how we're building security for

1520
01:32:26,140 --> 01:32:29,527
AI. But there's also one other
critical aspect, especially, you

1521
01:32:29,610 --> 01:32:31,997
know, the news today is all

1522
01:32:32,080 --> 01:32:37,637
about, you know, how do you
defend yourself using AI against

1523
01:32:37,720 --> 01:32:40,467
attacks that may in fact be using AI?

1524
01:32:40,550 --> 01:32:40,677
Right?

1525
01:32:40,760 --> 01:32:46,677
So last month, we announced our

1526
01:32:46,760 --> 01:32:51,327
multimodel agentic security
system in dash, an agent harness

1527
01:32:51,410 --> 01:32:53,837
for security, essentially, that we build.

1528
01:32:53,920 --> 01:32:56,557
We're bringing together a

1529
01:32:56,640 --> 01:33:01,097
hundred agents across the frontier,
and custom models, to

1530
01:33:01,180 --> 01:33:05,227
really find these exploitable
bugs better than any single

1531
01:33:05,310 --> 01:33:05,667
model does.

1532
01:33:05,750 --> 01:33:09,987
In fact, when we Dubdebuted thi

1533
01:33:10,070 --> 01:33:15,537
harness, it was on top of cyber
gym [dog barking] benchmark --

1534
01:33:15,620 --> 01:33:20,507
Jim benchmark. So look what you
can do to defend against AI attacks.

1535
01:33:20,590 --> 01:33:27,590
Over to you, Sarah.

1536
01:33:29,500 --> 01:33:33,757
>> Thank you. We know that security
scans can take a while, so I'm not going

1537
01:33:33,840 --> 01:33:38,597
to do one life. Let me show you the
results of an ender scan I already ran on

1538
01:33:38,680 --> 01:33:41,847
my code base.
The system runs as a stand-alone

1539
01:33:41,930 --> 01:33:47,907
CLI, but today, I'm using it
in my GitHub Copilot app on my

1540
01:33:47,990 --> 01:33:52,907
local dev machine. The scan is broken
down but vulnerability, demands, and

1541
01:33:52,990 --> 01:33:57,667
severity. And it's also an additional

1542
01:33:57,750 --> 01:34:01,617
to -- finding traditional
issues like coding errors and

1543
01:34:01,700 --> 01:34:06,787
hard-coated secrets. It's identifying
AI-specific vulnerabilities in the code

1544
01:34:06,870 --> 01:34:08,697
base. What happened under the hood,

1545
01:34:08,780 --> 01:34:10,627
over 100 specialized agents

1546
01:34:10,710 --> 01:34:13,917
working together to discover,

1547
01:34:14,000 --> 01:34:17,837
debate, and improve exploitable
vulnerabilities end to end.

1548
01:34:17,920 --> 01:34:20,607
When the scan finishes, it

1549
01:34:20,690 --> 01:34:25,807
generates both a log and an HTML
report that I can give to my

1550
01:34:25,890 --> 01:34:31,467
management. The defender details
command allows me to dig into these

1551
01:34:31,550 --> 01:34:35,057
vulnerabilities, and you can
see what it is, where it is in my

1552
01:34:35,140 --> 01:34:38,287
code, and the severity
to help me prioritize.

1553
01:34:38,370 --> 01:34:40,357
And from here, of course, we are

1554
01:34:40,440 --> 01:34:42,957
going to fix it.

1555
01:34:43,040 --> 01:34:47,617
Now, using the Defender
fix command, the system will

1556
01:34:47,700 --> 01:34:51,967
remediate suggest the fixing
directly in my local dev environment.

1557
01:34:52,050 --> 01:34:55,077
When that's done,
I can check out the diff.

1558
01:34:55,160 --> 01:34:57,847
So I have full transparency
about what the harness has done,

1559
01:34:57,930 --> 01:35:02,177
and I still have a human
in the loop check.

1560
01:35:02,260 --> 01:35:05,597
But of course, everything
I've shown you so far has run

1561
01:35:05,680 --> 01:35:06,267
locally.

1562
01:35:06,350 --> 01:35:09,387
But I can also create a PR to

1563
01:35:09,470 --> 01:35:13,227
plug into my existing workflows

1564
01:35:13,310 --> 01:35:17,227
and push up to my repo,
and I can take the output from the

1565
01:35:17,310 --> 01:35:21,477
scan and upload it to tools,
like GitHub advanced security,

1566
01:35:21,560 --> 01:35:24,537
and manage everything alongside

1567
01:35:24,620 --> 01:35:28,677
my other indication security findings.

1568
01:35:28,760 --> 01:35:32,977
I've shown it working on my code, but let
me show you a vulnerability our security

1569
01:35:33,060 --> 01:35:38,167
research teams identified
using MDASH that you can look up

1570
01:35:38,250 --> 01:35:40,737
yourself. So the TL;DR, because this is a

1571
01:35:40,820 --> 01:35:46,687
lot to read of this bug,
is the time reads an out of date map of

1572
01:35:46,770 --> 01:35:50,967
an object. Runs off the end.
And then crashes the host.

1573
01:35:51,050 --> 01:35:55,917
So this is exactly the kind of bug
that the harness was build for.

1574
01:35:56,000 --> 01:36:00,037
Because the floor spread across three
different parts of the code base.

1575
01:36:00,120 --> 01:36:03,277
So no single file looks wrong on its own.

1576
01:36:03,360 --> 01:36:06,077
They look absolutely fine.

1577
01:36:06,160 --> 01:36:10,447
And we can even see here be

1578
01:36:10,530 --> 01:36:14,117
probably my favorite part,
a statement from the developers

1579
01:36:14,200 --> 01:36:17,817
claiming everything is fine.
This is exactly the sort of

1580
01:36:17,900 --> 01:36:22,157
reassurance that fools normal
scanners and single AI models.

1581
01:36:22,240 --> 01:36:27,417
But MDASH wasn't fooled. One team
of agents spotted the suspicious gap.

1582
01:36:27,500 --> 01:36:29,187
Another team argue they it in part.

1583
01:36:29,270 --> 01:36:34,337
And a third team build a working example
that actually triggered the crash.

1584
01:36:34,420 --> 01:36:36,407
And it did all of this
in open source code base.

1585
01:36:36,490 --> 01:36:41,847
And this is a kind of joined-up
reasoning that previously

1586
01:36:41,930 --> 01:36:44,327
required significant manual
security research effort.

1587
01:36:44,410 --> 01:36:48,557
This is MDASH helping developers
create secure code from the

1588
01:36:48,640 --> 01:36:52,187
start, coming soon to your CLI
and the Microsoft Defender portal.

1589
01:36:52,270 --> 01:36:56,180
Back to you, Satya!

1590
01:36:56,540 --> 01:37:00,627
>> Thank you, Sarah.
So that was the stack.

1591
01:37:00,710 --> 01:37:05,337
Before, though, we move to,
you know, unpacking more of the

1592
01:37:05,420 --> 01:37:07,407
opportunity, I want
to do something different.

1593
01:37:07,490 --> 01:37:10,437
I want to introduce two people

1594
01:37:10,520 --> 01:37:12,707
whose LinkedIn profiles were

1595
01:37:12,790 --> 01:37:14,717
both super impressive and

1596
01:37:14,800 --> 01:37:18,667
slightly perplexing.

1597
01:37:18,750 --> 01:37:22,217
Under role had says they're
general partners on Mantis.

1598
01:37:22,300 --> 01:37:25,627
On the previous role, sold out

1599
01:37:25,710 --> 01:37:29,127
Madison Square Garden.

1600
01:37:29,210 --> 01:37:36,210
Welcome Alex and Drew from the

1601
01:37:38,650 --> 01:37:40,437
Chainsmokers. Thank you so much.

1602
01:37:40,520 --> 01:37:43,737
One of the things when my team
came to me and said, hey, we're

1603
01:37:43,820 --> 01:37:47,377
going to have Chainsmokers
at Build, I thought maybe that's

1604
01:37:47,460 --> 01:37:51,567
what we're calling
our new GitHub Copilot app.

1605
01:37:51,650 --> 01:37:55,557
But maybe they're your fans.
But tell me about how you got into this.

1606
01:37:55,640 --> 01:38:00,957
I mean, you're being added
for what -- at it for what, 12-plus

1607
01:38:01,040 --> 01:38:03,997
years as investors first,
and you had your firm now for

1608
01:38:04,080 --> 01:38:06,667
seven-plus years, and you even
picked what I would have not

1609
01:38:06,750 --> 01:38:09,987
thought is the natural place,

1610
01:38:10,070 --> 01:38:15,607
which is B2B SaaS. Give us the back story.

1611
01:38:15,690 --> 01:38:18,677
>> I'm sure you guys are
wondering what time line you're

1612
01:38:18,760 --> 01:38:20,447
on, where we're at Microsoft

1613
01:38:20,530 --> 01:38:24,497
Build, but hey, how are you.

1614
01:38:24,580 --> 01:38:28,057
We've been in the Chainsmokers
for 14 years, and you know, when

1615
01:38:28,140 --> 01:38:30,487
our music started to take off in

1616
01:38:30,570 --> 01:38:35,457
the mid two thousand teens,
we got to play a few events like this.

1617
01:38:35,540 --> 01:38:40,167
And we met a lot of the founders from
the consumer mobile cloud era of startups.

1618
01:38:40,250 --> 01:38:41,617
And they taught us a lot and

1619
01:38:41,700 --> 01:38:48,207
give us a front row seat into
what early stage investing was

1620
01:38:48,290 --> 01:38:51,437
like. We fell love with it,
had a lot in common with the founders and

1621
01:38:51,520 --> 01:38:54,447
the way they build their
business was the same way that

1622
01:38:54,530 --> 01:38:59,637
we thought about starting the Chainsmokers
and breaking through the noise there.

1623
01:38:59,720 --> 01:39:03,757
And we got to participate
in a few of their deals.

1624
01:39:03,840 --> 01:39:08,927
And we decided to institutionalized
in 2020 and start our own fund

1625
01:39:09,010 --> 01:39:12,327
>> That's cool. When you hear about
all the AI stuff and what's happening,

1626
01:39:12,410 --> 01:39:16,247
there's a lot going on, even in your
portfolio we were talking back staining.

1627
01:39:16,330 --> 01:39:19,237
Things are changing. How do you
see the opportunity going forward?

1628
01:39:19,320 --> 01:39:22,877
>> I mean, there's so many
vectors you could talk about.

1629
01:39:22,960 --> 01:39:26,347
There's the creative
output side -- >> Yep.

1630
01:39:26,430 --> 01:39:28,017
>> -- we've been
experimenting with in music.

1631
01:39:28,100 --> 01:39:31,697
But always important to have your
authenticity when it comes to creativity.

1632
01:39:31,780 --> 01:39:36,857
But on the invest side, we're moving from
producing outputs to producing actions.

1633
01:39:36,940 --> 01:39:42,757
Which I think presents a very interesting
opportunity to reimagine the entire

1634
01:39:42,840 --> 01:39:45,427
architecture of the way
software enterprise is build.

1635
01:39:45,510 --> 01:39:49,187
Instead of humus producing
outputs -- humans producing

1636
01:39:49,270 --> 01:39:52,937
outputs, it's machines, and rethinking
the entire space in that context.

1637
01:39:53,020 --> 01:39:56,107
>> It's fantastic to see that.

1638
01:39:56,190 --> 01:40:00,877
And maybe to close out,
any sort of -- as artists who have had

1639
01:40:00,960 --> 01:40:04,437
great success, when you look at

1640
01:40:04,520 --> 01:40:08,217
a founder, what sort of is the
advice you give them when you're

1641
01:40:08,300 --> 01:40:12,087
creating, you know, you're
bringing something new, and it's

1642
01:40:12,170 --> 01:40:16,057
a creative process. So what is --
what are you looking for in founders?

1643
01:40:16,140 --> 01:40:18,497
What is it that you're
giving them as advice?

1644
01:40:18,580 --> 01:40:21,077
>> Totally. There really
are so many parallels.

1645
01:40:21,160 --> 01:40:22,947
Maybe it seems like that;
maybe it doesn't.

1646
01:40:23,030 --> 01:40:26,737
But it's really difficult
as artists to really find, I guess,

1647
01:40:26,820 --> 01:40:29,637
what your sound s
what's authentically you.

1648
01:40:29,720 --> 01:40:33,677
What you can continue to do over
and over, because with this much

1649
01:40:33,760 --> 01:40:35,847
competition, you have
to be moving authentically.

1650
01:40:35,930 --> 01:40:39,597
You have to be connected
to the product that you're creating.

1651
01:40:39,680 --> 01:40:44,927
And -- 'cause you're going
to have to iteratedate and keep it

1652
01:40:45,010 --> 01:40:47,957
going for a long time to lock
into your fan base and make

1653
01:40:48,040 --> 01:40:50,957
something special and nearby.
And we try -- and unique.

1654
01:40:51,040 --> 01:40:54,717
And we try to advise the same thing.

1655
01:40:54,800 --> 01:40:58,037
>> Are you game for playing for us
this evening for our closest friends here?

1656
01:40:58,120 --> 01:41:01,607
>> We'll be there.
>> 6:00! 6 this evening.

1657
01:41:01,690 --> 01:41:02,607
Thank you very much.

1658
01:41:02,690 --> 01:41:03,107
[Cheers and applause.]

1659
01:41:03,190 --> 01:41:08,817
Thank you, thank you.

1660
01:41:08,900 --> 01:41:11,697
We will be right here at 6 p.m.

1661
01:41:11,780 --> 01:41:15,887
tonight, and really looking forward to it.

1662
01:41:15,970 --> 01:41:19,227
So that effectively sort
of talks a little bit about the

1663
01:41:19,310 --> 01:41:23,757
developer stacks.
So now let's talk about what is

1664
01:41:23,840 --> 01:41:26,427
the opportunity for every company.

1665
01:41:26,510 --> 01:41:30,517
Right? At the end of the day,
we are an institution, an organizational

1666
01:41:30,600 --> 01:41:35,777
builders, whether it's AI nave
company, whether it's a SaaS

1667
01:41:35,860 --> 01:41:38,607
company, or whether enterprise,
the first thing we want to do is

1668
01:41:38,690 --> 01:41:40,747
make sure that as you build

1669
01:41:40,830 --> 01:41:43,817
things black --

1670
01:41:43,900 --> 01:41:45,937
Build things like plugins or

1671
01:41:46,020 --> 01:41:51,757
agents or AI apping, we want
them to be discovered through

1672
01:41:51,840 --> 01:41:54,087
the Microsoft ecosystem.
That's job one for us.

1673
01:41:54,170 --> 01:41:58,227
That's why we're doing things
in Windows or Microsoft 365 Copilot

1674
01:41:58,310 --> 01:41:59,897
or in Teams or GitHub.

1675
01:41:59,980 --> 01:42:03,847
We want to structure them that

1676
01:42:03,930 --> 01:42:07,237
your applications, agents,
plugins are discovered.

1677
01:42:07,320 --> 01:42:10,307
Customers are also building lots

1678
01:42:10,390 --> 01:42:14,607
of line of business applications
and agents using Copilot Studio,

1679
01:42:14,690 --> 01:42:17,577
and we want to make that all
discoverable again, as part of

1680
01:42:17,660 --> 01:42:21,987
the Copilot experience.
And Teams in some sense has

1681
01:42:22,070 --> 01:42:26,657
become this destination
for multiplayer, human to agent

1682
01:42:26,740 --> 01:42:32,027
interaction. We want you to be able
to find agents, interact with agents,

1683
01:42:32,110 --> 01:42:37,667
right in Teams.
And we are super-charging all of

1684
01:42:37,750 --> 01:42:41,207
this. Copilot continues
to evolve very quickly.

1685
01:42:41,290 --> 01:42:43,037
It started first with chat, with

1686
01:42:43,120 --> 01:42:45,277
some of the best models, with

1687
01:42:45,360 --> 01:42:49,977
great access to Work IQ.

1688
01:42:50,060 --> 01:42:53,737
We didn't have the name,
but that's where it started.

1689
01:42:53,820 --> 01:42:57,517
Then came Cowork, a new way
of working, and generating the

1690
01:42:57,600 --> 01:43:01,227
stunning artifacts,
and solve these multistep problems.

1691
01:43:01,310 --> 01:43:05,697
Right? So you assign
a multistep task to Cowork.

1692
01:43:05,780 --> 01:43:07,527
You saw GitHub, which has

1693
01:43:07,610 --> 01:43:09,447
continued to evolve as well.

1694
01:43:09,530 --> 01:43:13,937
And now, as of, you know, come

1695
01:43:14,020 --> 01:43:18,477
summer, you will be bringing
coding to all knowledge work

1696
01:43:18,560 --> 01:43:21,847
within one Copilot super app. Right?

1697
01:43:21,930 --> 01:43:27,450
That's going to be really
exciting, to see -- yep!

1698
01:43:27,550 --> 01:43:31,557
You will have chat, cowork,
and code all in Copilot.

1699
01:43:31,640 --> 01:43:34,157
But today, we're introducing

1700
01:43:34,240 --> 01:43:40,297
something completely new.

1701
01:43:40,380 --> 01:43:43,087
Autopilots. When you
think of autopilots as

1702
01:43:43,170 --> 01:43:45,267
enterprise-grade claws.

1703
01:43:45,350 --> 01:43:50,477
These are autonomous,
long-running agents with full

1704
01:43:50,560 --> 01:43:52,547
enterprise compliance that run

1705
01:43:52,630 --> 01:43:53,877
in your tenant.

1706
01:43:53,960 --> 01:43:57,967
Autopilots can have a name,

1707
01:43:58,050 --> 01:44:01,137
personality, custom connecters,
context, and memory.

1708
01:44:01,220 --> 01:44:04,987
And they're totally new way
to reduce toil and get you back to

1709
01:44:05,070 --> 01:44:07,657
what you love, to kick things

1710
01:44:07,740 --> 01:44:11,157
off, the the first autopilot

1711
01:44:11,240 --> 01:44:14,547
we're introducing is Scout.

1712
01:44:14,630 --> 01:44:21,630
Let's take a look.

1713
01:45:11,740 --> 01:45:17,057
>> All right. As you can see, Scout works

1714
01:45:17,140 --> 01:45:21,247
where you work, joining group
chats and teams, handling

1715
01:45:21,330 --> 01:45:25,367
threats in Outlook.
Starting today for those of you

1716
01:45:25,450 --> 01:45:27,807
on Copilot Frontier,
you can try out Scout.

1717
01:45:27,890 --> 01:45:32,647
And in the coming months, we will build
this out to a complete digital team of

1718
01:45:32,730 --> 01:45:34,447
autopilots, right inside of Copilot.

1719
01:45:34,530 --> 01:45:36,647
Right? So you can go to the Copilot app.

1720
01:45:36,730 --> 01:45:38,137
Scout is the one that comes by default.

1721
01:45:38,220 --> 01:45:42,317
But you can build more
of these autopilots.

1722
01:45:42,400 --> 01:45:46,357
And so that's the future of what
we think of as the Copilot

1723
01:45:46,440 --> 01:45:50,397
ecosystem itself.
And so far, we have talked a lot

1724
01:45:50,480 --> 01:45:53,397
about what you can do to build
your own agents and how these

1725
01:45:53,480 --> 01:45:58,067
agents are discovered and things
like Copilot and Teams and Windows.

1726
01:45:58,150 --> 01:45:59,907
But when you think about what

1727
01:45:59,990 --> 01:46:01,937
makes any organization, any

1728
01:46:02,020 --> 01:46:05,647
enterprise, any company unique,

1729
01:46:05,730 --> 01:46:11,467
it is its Tatacit knowledge
tha is continuously compounding

1730
01:46:11,550 --> 01:46:14,267
through its operations.
So the key consideration at some

1731
01:46:14,350 --> 01:46:16,987
level in an AI age is to ask the

1732
01:46:17,070 --> 01:46:21,397
question, "What's the future of the firm?"

1733
01:46:21,480 --> 01:46:24,427
How do you continue to preserve
and compound that tacit

1734
01:46:24,510 --> 01:46:27,867
knowledge in the age of AI,
where models can learn anything?

1735
01:46:27,950 --> 01:46:29,517
From the data and the

1736
01:46:29,600 --> 01:46:30,887
trajectories they see.

1737
01:46:30,970 --> 01:46:35,067
To do that, we believe that

1738
01:46:35,150 --> 01:46:37,977
every organization, whether it's

1739
01:46:38,060 --> 01:46:42,447
an AI native company, a SaaS
company, or an enterprise, will

1740
01:46:42,530 --> 01:46:43,747
need to build their own

1741
01:46:43,830 --> 01:46:45,197
hill-climbing machine.

1742
01:46:45,280 --> 01:46:49,857
It's a system that continuously

1743
01:46:49,940 --> 01:46:54,757
improves against your objectives,
your private evals,

1744
01:46:54,840 --> 01:46:56,757
compounding your advantage over time.

1745
01:46:56,840 --> 01:46:59,327
Not someone else's.

1746
01:46:59,410 --> 01:47:04,687
To deliver on this, we're taking
the next step with Frontier

1747
01:47:04,770 --> 01:47:08,477
tuning. We've been working with many
customers already to help, you

1748
01:47:08,560 --> 01:47:11,677
know, really build
their own learning loop.

1749
01:47:11,760 --> 01:47:17,217
The environment, the context,
the tools, the ribosomes -- the

1750
01:47:17,300 --> 01:47:18,967
rubrics, and even train their own models.

1751
01:47:19,050 --> 01:47:23,357
And today, we're really thrilled
and excited about super charging

1752
01:47:23,440 --> 01:47:28,257
that frontier tuning capability with the
innovation that's coming out of our super

1753
01:47:28,340 --> 01:47:35,270
intelligence lab, to tell you all about
this, let me hand it over to Mustafa.

1754
01:47:35,360 --> 01:47:40,137
>> Thank you. [Music].
Thank you, thank you.

1755
01:47:40,220 --> 01:47:43,337
Thank you, good morning, everybody.

1756
01:47:43,420 --> 01:47:45,277
You know, we really are living

1757
01:47:45,360 --> 01:47:47,907
in the most remarkable times.

1758
01:47:47,990 --> 01:47:50,647
Since I started working in AI,

1759
01:47:50,730 --> 01:47:53,237
the computer that we used to

1760
01:47:53,320 --> 01:47:55,417
train frontier models has

1761
01:47:55,500 --> 01:47:58,517
increased by one trillion-fold.

1762
01:47:58,600 --> 01:48:02,157
That is 12 orders of magnitude

1763
01:48:02,240 --> 01:48:03,697
of computation in just 15 years.

1764
01:48:03,780 --> 01:48:07,097
It's clear that a consistent,

1765
01:48:07,180 --> 01:48:10,847
exponential increase
in computation leads to predictable

1766
01:48:10,930 --> 01:48:15,507
advances in AI capabilities.
And in the next few years, we're

1767
01:48:15,590 --> 01:48:17,337
going to see three more orders

1768
01:48:17,420 --> 01:48:24,420
of magnitude of compute apply to

1769
01:48:25,380 --> 01:48:29,287
train Frontier models.
Long linear hill climbing is the norm.

1770
01:48:29,370 --> 01:48:32,117
The scaling laws are clearly
holding, and it is a remarkable

1771
01:48:32,200 --> 01:48:35,427
time in our industry.

1772
01:48:35,510 --> 01:48:39,567
And so in this context,
we at MAI are building towards what we

1773
01:48:39,650 --> 01:48:42,277
call humanist superintelligence.

1774
01:48:42,360 --> 01:48:47,437
State of the art AI capabilities
that are explicitly designed to

1775
01:48:47,520 --> 01:48:52,147
serve people and organizations
and not replace them.

1776
01:48:52,230 --> 01:48:54,407
Because the type of AI that we

1777
01:48:54,490 --> 01:48:56,977
create really does matter.

1778
01:48:57,060 --> 01:48:59,537
We need an AI that places

1779
01:48:59,620 --> 01:49:01,537
humanity first.

1780
01:49:01,620 --> 01:49:05,457
But always prioritizes human

1781
01:49:05,540 --> 01:49:07,287
well-being and human progress.

1782
01:49:07,370 --> 01:49:12,867
This is the core philosophy and

1783
01:49:12,950 --> 01:49:15,847
modification behind our efforts
at Microsoft, and it shapes

1784
01:49:15,930 --> 01:49:18,287
everything that we do.
And as a platform company, our

1785
01:49:18,370 --> 01:49:23,877
job and our commitment is to keep
you developers building the

1786
01:49:23,960 --> 01:49:27,177
absolute frontier.
So today, we're very excited to

1787
01:49:27,260 --> 01:49:32,467
announce a family of seven
new models across image, voice,

1788
01:49:32,550 --> 01:49:33,567
transcription, and coding.

1789
01:49:33,650 --> 01:49:38,157
[Cheers and applause.]

1790
01:49:38,240 --> 01:49:42,657
These are all built with real
attention to detail and a

1791
01:49:42,740 --> 01:49:45,697
commitment to making very
practical and efficient tools

1792
01:49:45,780 --> 01:49:49,817
that are typed to -- tuned
to just how you work in the real

1793
01:49:49,900 --> 01:49:53,407
world. First up: MAI image 2.5 and its

1794
01:49:53,490 --> 01:49:55,007
flash variant, two super strong

1795
01:49:55,090 --> 01:49:57,607
models that deliver a step

1796
01:49:57,690 --> 01:50:03,117
change in quality, not -- now at

1797
01:50:03,200 --> 01:50:09,200
#2, surpassing the core of Nanabanana2.

1798
01:50:09,300 --> 01:50:12,787
They gave you precise editing
with control and consistency,

1799
01:50:12,870 --> 01:50:16,997
flashes here for super efficient
production work loads, while 2.5

1800
01:50:17,080 --> 01:50:21,947
gives you that maximum fidelity
and professional grade performance.

1801
01:50:22,030 --> 01:50:25,737
Alive in PowerPoint today,
they're rolling out to One

1802
01:50:25,820 --> 01:50:29,907
Drive, and you can access them
on Foundry at a market leading

1803
01:50:29,990 --> 01:50:34,847
quality per dollar.
Next, MAI transcribe 1.5.

1804
01:50:34,930 --> 01:50:37,547
This is the best transcription
model in the world.

1805
01:50:37,630 --> 01:50:41,887
State of the art accuracy across

1806
01:50:41,970 --> 01:50:47,587
43 languages, beating out Gemini
and open AI's flagship models.

1807
01:50:47,670 --> 01:50:48,827
We've optimized it for real

1808
01:50:48,910 --> 01:50:54,397
world use so you can produce

1809
01:50:54,480 --> 01:50:57,247
highly accurate triceps,
five times -- transcripts, five times

1810
01:50:57,330 --> 01:51:02,177
faster than all rival models.
It's integrated inside of

1811
01:51:02,260 --> 01:51:05,877
GitHub, Copilot, dynamic 365
contact center, and available in

1812
01:51:05,960 --> 01:51:11,117
Foundry where I'm excited
to say it is the fastest, most

1813
01:51:11,200 --> 01:51:13,667
official, and most -- efficient,
and most cost-effective

1814
01:51:13,750 --> 01:51:19,970
transcription model of any
of the hyperscalers out there.

1815
01:51:20,070 --> 01:51:22,457
So paired with that, we've got

1816
01:51:22,540 --> 01:51:23,397
MAI Voice 2.

1817
01:51:23,480 --> 01:51:28,127
The latest speech generation.

1818
01:51:28,210 --> 01:51:29,687
Beautiful prosody, natural

1819
01:51:29,770 --> 01:51:34,437
delivery, fine grained control,

1820
01:51:34,520 --> 01:51:37,577
in 15 languages with more to come.

1821
01:51:37,660 --> 01:51:39,277
And voice 2 flash, providing the

1822
01:51:39,360 --> 01:51:45,397
best value and speed for ultra

1823
01:51:45,480 --> 01:51:48,137
lateliency voice, which
is the giggest thing in 2026.

1824
01:51:48,220 --> 01:51:50,857
Next up, text foundation model.

1825
01:51:50,940 --> 01:51:55,427
MAI thinking one. This is the first
reasoning model, and it's exceptionally

1826
01:51:55,510 --> 01:51:58,257
strong, and our target use cases

1827
01:51:58,340 --> 01:52:00,427
of reasoning and sweet tasks.

1828
01:52:00,510 --> 01:52:03,517
It's a 35 billion active

1829
01:52:03,600 --> 01:52:10,407
parameter E, meaning it competes
in the medium sized weight

1830
01:52:10,490 --> 01:52:12,037
class. And independent human raters on

1831
01:52:12,120 --> 01:52:13,877
surge prefer it in overall

1832
01:52:13,960 --> 01:52:19,737
quality side by side versus

1833
01:52:19,820 --> 01:52:20,837
sonnet 4.6.

1834
01:52:20,920 --> 01:52:21,767
[Cheering]

1835
01:52:21,850 --> 01:52:27,527
It's achieved 97% on AMI2025,

1836
01:52:27,610 --> 01:52:30,297
the key measure of its
capabilities, but most

1837
01:52:30,380 --> 01:52:34,467
importantly, it's now at 53% on

1838
01:52:34,550 --> 01:52:35,847
bench pro that places it

1839
01:52:35,930 --> 01:52:39,137
alongside opus 4-6, at least on

1840
01:52:39,220 --> 01:52:42,507
the toughest coding benchmark

1841
01:52:42,590 --> 01:52:46,790
out there. We're very happy with that.

1842
01:52:46,890 --> 01:52:51,147
There's plenty more for us as we
get this into production and

1843
01:52:51,230 --> 01:52:55,017
hill climb against real-world
tasks and traffic, but what's

1844
01:52:55,100 --> 01:52:58,787
most remarkable about this model,
we think, it has climbed

1845
01:52:58,870 --> 01:53:00,627
entirely from the bottom.

1846
01:53:00,710 --> 01:53:02,327
And that means that it hasn't

1847
01:53:02,410 --> 01:53:07,267
targeted any of the benchmarks

1848
01:53:07,350 --> 01:53:09,987
specifically, with zero distillation.

1849
01:53:10,070 --> 01:53:12,067
And to us, this is critical.
It means that the model is

1850
01:53:12,150 --> 01:53:17,307
created with an enterprise-grade
clean and commercially licensed

1851
01:53:17,390 --> 01:53:21,077
data lineage that means that you
can put it into production in a

1852
01:53:21,160 --> 01:53:24,047
very trustworthy way
with complete confidence.

1853
01:53:24,130 --> 01:53:29,487
Now, finally, I'm incredibly
excited to announce MAI code 1

1854
01:53:29,570 --> 01:53:32,587
flash. This is our new inference

1855
01:53:32,670 --> 01:53:34,427
sufficient coding model tuned

1856
01:53:34,510 --> 01:53:41,510
for VS code and GitHub CLI.

1857
01:53:42,400 --> 01:53:45,287
5 51%, despite five billion parameters.

1858
01:53:45,370 --> 01:53:49,577
Close to haiku in terms of size
but cheapner cost, delivering

1859
01:53:49,660 --> 01:53:51,737
strong coding performance at

1860
01:53:51,820 --> 01:53:56,817
great sufficiencily.
And it's rolling out inside of

1861
01:53:56,900 --> 01:54:00,750
VS Code.

1862
01:54:00,850 --> 01:54:04,787
Now, alongside distribution
on Foundry and optimization for our

1863
01:54:04,870 --> 01:54:07,657
1P products, we're excited
to make our models available on

1864
01:54:07,740 --> 01:54:11,357
open router as well
as Fireworks and base ten.

1865
01:54:11,440 --> 01:54:13,997
This means for the first time
you will be able to tune the

1866
01:54:14,080 --> 01:54:17,487
weights directly yourself
in an ecosystem of your choice.

1867
01:54:17,570 --> 01:54:21,707
Now, across this entire family,
safety and security has been

1868
01:54:21,790 --> 01:54:26,707
build in from the start.
Voice models come with

1869
01:54:26,790 --> 01:54:29,147
protectiontion against
unauthorized cloning.

1870
01:54:29,230 --> 01:54:32,497
Everything is water marked from scratch.

1871
01:54:32,580 --> 01:54:36,657
Reduced the reFOOUFLS,
improvedish -- refuseals, and

1872
01:54:36,740 --> 01:54:40,257
accomplishing a very detailed
technical report today to give

1873
01:54:40,340 --> 01:54:46,127
you a full and transparent understanding
of how we put all of this together.

1874
01:54:46,210 --> 01:54:50,647
One of the things I'm particularly excited
about is that we have been carefully

1875
01:54:50,730 --> 01:54:53,037
codesigning our models
with our own silicon.

1876
01:54:53,120 --> 01:54:55,767
That means that we've optimized

1877
01:54:55,850 --> 01:54:59,107
MAI thinking one on our very own

1878
01:54:59,190 --> 01:55:04,047
myo200 Chip and benchmarked

1879
01:55:04,130 --> 01:55:07,597
against the GB2300.

1880
01:55:07,680 --> 01:55:09,917
So top of the improvement, we're

1881
01:55:10,000 --> 01:55:13,727
seeing a further 1.4X

1882
01:55:13,810 --> 01:55:19,027
performance per Watt gain when
we run them on the MIO model end

1883
01:55:19,110 --> 01:55:23,217
to end, and that's huge.
As everybody knows, this scale,

1884
01:55:23,300 --> 01:55:27,007
every Watt counts.
And silicon and model codesign

1885
01:55:27,090 --> 01:55:31,007
is a really key advantage that
we think is going to help keep

1886
01:55:31,090 --> 01:55:36,047
everybody here right on the
Frontier with the most efficient

1887
01:55:36,130 --> 01:55:38,667
and powerful thinking
and coding agents out there.

1888
01:55:38,750 --> 01:55:41,367
And we're excited that these
faster and more efficient MAI

1889
01:55:41,450 --> 01:55:45,057
models are coming to the N1X
that Satya mentioned a few

1890
01:55:45,140 --> 01:55:48,087
moments ago, and we think that's
going to be able to deliver the

1891
01:55:48,170 --> 01:55:52,697
very best performance on mounds

1892
01:55:52,780 --> 01:55:55,847
in a new -- on windows
in a few months time.

1893
01:55:55,930 --> 01:56:00,637
This is what owning the full
end to end stack looks like.

1894
01:56:00,720 --> 01:56:04,577
It lets you customize the MAI
models using our full hill

1895
01:56:04,660 --> 01:56:05,777
climbing machine right where you want it.

1896
01:56:05,860 --> 01:56:10,017
And it means that the discipline
and very relentless engineering

1897
01:56:10,100 --> 01:56:13,767
that has gone into building our models
is now available to all of you.

1898
01:56:13,850 --> 01:56:17,987
On a platform that you can trust,
working on your behalf,

1899
01:56:18,070 --> 01:56:22,257
to create custom agents
that you will control.

1900
01:56:22,340 --> 01:56:26,927
So the really big thing, of course, that's
happened in the last year is these RLEs.

1901
01:56:27,010 --> 01:56:28,467
Reinforcement learning environments.

1902
01:56:28,550 --> 01:56:31,447
These unique training gyms for your AIs.

1903
01:56:31,530 --> 01:56:35,767
They create company
and task-specific agents, adapted

1904
01:56:35,850 --> 01:56:38,777
only to you, built on MAI models.

1905
01:56:38,860 --> 01:56:40,447
For example, within Microsoft we

1906
01:56:40,530 --> 01:56:42,947
use our RLEs combined with MAI

1907
01:56:43,030 --> 01:56:47,297
models to climb towards the best

1908
01:56:47,380 --> 01:56:48,637
agentic ace uses on Excel.

1909
01:56:48,720 --> 01:56:53,787
The model is on par with GPT5.4

1910
01:56:53,870 --> 01:56:59,767
on public and private benchmarks, and
being ten times more efficient on cost.

1911
01:56:59,850 --> 01:57:03,187
[Cheering] Thank -- [chuckling]
And many other earlier adopters

1912
01:57:03,270 --> 01:57:04,947
are seeing similar results.

1913
01:57:05,030 --> 01:57:07,867
When we've tuned our models on

1914
01:57:07,950 --> 01:57:12,177
McKensie's tasks,
it delivered the highest win rate,

1915
01:57:12,260 --> 01:57:16,747
outperforming GPT5.5, and delivering
10X grater efficiency on cost.

1916
01:57:16,830 --> 01:57:21,197
To us, this is the advantage
of very carefully calibrated

1917
01:57:21,280 --> 01:57:25,057
frontier tuning.
And importantly, unlike with

1918
01:57:25,140 --> 01:57:26,187
some of the other companies,

1919
01:57:26,270 --> 01:57:33,270
with shared model

1920
01:57:35,080 --> 01:57:39,307
only you keep benefits from the
hard earned workflows, know how,

1921
01:57:39,390 --> 01:57:41,707
knowledge and your own institutional data.

1922
01:57:41,790 --> 01:57:45,037
Only you get to control
the resulting model.

1923
01:57:45,120 --> 01:57:47,707
So with us, the RLEs and the

1924
01:57:47,790 --> 01:57:50,377
models that you build inside of

1925
01:57:50,460 --> 01:57:55,237
them they become your mote.

1926
01:57:55,320 --> 01:58:00,027
It marks a new era in AI that we
are all really very excited about.

1927
01:58:00,110 --> 01:58:04,157
Okay. So now just one final announcement
that I am very excited about.

1928
01:58:04,240 --> 01:58:05,897
We are taking customization and

1929
01:58:05,980 --> 01:58:08,197
co creation of our models to the

1930
01:58:08,280 --> 01:58:12,537
highest level possible on what I

1931
01:58:12,620 --> 01:58:16,707
think offers the most important
application of AI, healthcare.

1932
01:58:16,790 --> 01:58:18,077
Today we are very proud to be

1933
01:58:18,160 --> 01:58:19,777
announcing we are partnering

1934
01:58:19,860 --> 01:58:24,077
with Mayo Clinic for a new

1935
01:58:24,160 --> 01:58:27,137
frontier for health and deploy
it in their hospitals and beyond.

1936
01:58:27,220 --> 01:58:30,787
Help me in welcoming to the stage
physician ground breaking

1937
01:58:30,870 --> 01:58:32,957
researcher, president and CEO of

1938
01:58:33,040 --> 01:58:40,040
the Mayo Clinic Dr. Gianrico

1939
01:58:42,830 --> 01:58:47,390
Farrugia.

1940
01:58:47,490 --> 01:58:50,807
Thank you for being here.
Everyone recognizes Mayo as the

1941
01:58:50,890 --> 01:58:53,977
leading hospital in the world
with incredible track record of

1942
01:58:54,060 --> 01:58:55,277
research, innovation and

1943
01:58:55,360 --> 01:58:56,217
clinical practice.

1944
01:58:56,300 --> 01:59:01,637
Tell us more about what hope to

1945
01:59:01,720 --> 01:59:07,087
get collaboration. >> Mayo Clinic is known
for being able to live up to primary

1946
01:59:07,170 --> 01:59:09,687
value the needs
of the patient comes first.

1947
01:59:09,770 --> 01:59:11,597
We deliver outstanding healthcare.

1948
01:59:11,680 --> 01:59:14,127
We are ranked number one
healthcare organization in the

1949
01:59:14,210 --> 01:59:17,047
world, yet we know most people
in the world will not have

1950
01:59:17,130 --> 01:59:21,407
access to Mayo Clinic.
So seven years ago we decided to

1951
01:59:21,490 --> 01:59:23,807
create a platform, Mayo Clinic

1952
01:59:23,890 --> 01:59:29,577
platform, moving all
of healthcare from a pipeline to a

1953
01:59:29,660 --> 01:59:33,567
platform. With our partners that
platform is in four continents and

1954
01:59:33,650 --> 01:59:35,787
reaches about 100 million people.

1955
01:59:35,870 --> 01:59:38,817
It has created the largest, to

1956
01:59:38,900 --> 01:59:42,987
our knowledge, longitudal

1957
01:59:43,070 --> 01:59:47,497
healthcare center in the world,
multimodal genomics.

1958
01:59:47,580 --> 01:59:50,747
We have the opportunity to do
what we do best together which

1959
01:59:50,830 --> 01:59:52,967
is create a frontier model for healthcare.

1960
01:59:53,050 --> 01:59:55,977
What it means if you are
a patient, if you are somebody

1961
01:59:56,060 --> 02:00:01,347
interested in healthcare,
you can get clinical and logistical

1962
02:00:01,430 --> 02:00:04,347
answers to your healthcare.
If your healthcare provider,

1963
02:00:04,430 --> 02:00:10,217
your physician it can give
you insight, it can act as your

1964
02:00:10,300 --> 02:00:13,957
real-time team member it can
tell you what is likely to

1965
02:00:14,040 --> 02:00:17,397
happen next, but it can also
prevent harm and therefore

1966
02:00:17,480 --> 02:00:20,227
increase patient safety and give
valuable insights that make the

1967
02:00:20,310 --> 02:00:24,787
team better at giving you what you
need most which is better healthcare.

1968
02:00:24,870 --> 02:00:28,407
>> One of the things we are most
excited about is the models are

1969
02:00:28,490 --> 02:00:32,407
already pretty incredible
about textbook knowledge.

1970
02:00:32,490 --> 02:00:34,377
Read all of the journals and papers.

1971
02:00:34,460 --> 02:00:39,667
What they are looking is clinical practice
and clinical expertise of your team and

1972
02:00:39,750 --> 02:00:42,417
clinicians you developed
over the last many decades.

1973
02:00:42,500 --> 02:00:47,087
How do you think you might go
about the clinical practice for

1974
02:00:47,170 --> 02:00:50,727
the model and production?
>> The exciting part here is we

1975
02:00:50,810 --> 02:00:54,567
each do what we do best.
We can tackle something that

1976
02:00:54,650 --> 02:00:58,767
Hazel lewded healthcare for a long
time trusted scalable solutions.

1977
02:00:58,850 --> 02:01:02,007
To do that you need to have
the right data, certainly need to

1978
02:01:02,090 --> 02:01:07,147
have the right people, but you need
to have a very patient focused lens.

1979
02:01:07,230 --> 02:01:10,617
Between the two of us we have
all of this together so we can

1980
02:01:10,700 --> 02:01:13,967
build this frontier model and

1981
02:01:14,050 --> 02:01:19,087
offer safe, secure,
trustworthy and effective healthcare

1982
02:01:19,170 --> 02:01:23,797
solutions for all. >> Number one objective
is put the patient first, deliver high

1983
02:01:23,880 --> 02:01:28,397
quality that we can as trusted
as we can and share the world over.

1984
02:01:28,480 --> 02:01:30,087
Excited to share this partnership.

1985
02:01:30,170 --> 02:01:31,887
Can't wait to share more in the future.

1986
02:01:31,970 --> 02:01:37,190
>> Us too. Thank you.

1987
02:01:37,420 --> 02:01:40,777
>> So today marks some very,
very exciting steps that we are

1988
02:01:40,860 --> 02:01:45,597
taking on our journey to create
humanist super intelligence at

1989
02:01:45,680 --> 02:01:49,647
Microsoft. We have an incredible roster
of seven new world class models to

1990
02:01:49,730 --> 02:01:51,617
keep everybody working
at the absolute frontier.

1991
02:01:51,700 --> 02:01:55,257
We are really looking forward
to everybody being able to co

1992
02:01:55,340 --> 02:01:57,957
create your own unique agents

1993
02:01:58,040 --> 02:02:01,497
adapted to you at your control.

1994
02:02:01,580 --> 02:02:04,187
I feel this is a new era in AI.

1995
02:02:04,270 --> 02:02:06,607
An era of AI that you
control on your terms.

1996
02:02:06,690 --> 02:02:07,967
Let's build it together.

1997
02:02:08,050 --> 02:02:15,050
Thank you very much everyone.

1998
02:02:17,760 --> 02:02:20,397
Now over to Tanaya to show us

1999
02:02:20,480 --> 02:02:23,387
how frontier learning works in

2000
02:02:23,470 --> 02:02:28,610
practice.

2001
02:02:28,870 --> 02:02:33,267
>> Thanks Mustafa.
With frontier tuning we are

2002
02:02:33,350 --> 02:02:37,087
making it possible for you
to create your own enterprise AI,

2003
02:02:37,170 --> 02:02:42,637
with the models and the harness
tuned on your data and your workflows.

2004
02:02:42,720 --> 02:02:43,707
Let me show you how you can

2005
02:02:43,790 --> 02:02:49,147
build your own health planning

2006
02:02:49,230 --> 02:02:54,037
machine. MAI thinking one is available
in private preview in the foundry

2007
02:02:54,120 --> 02:02:54,437
model catalog.

2008
02:02:54,520 --> 02:02:59,157
You can go ahead and deploy the

2009
02:02:59,240 --> 02:03:03,357
mo model as is or if you want to go

2010
02:03:03,440 --> 02:03:05,357
on your health journey click the

2011
02:03:05,440 --> 02:03:07,767
fine tune button.

2012
02:03:07,850 --> 02:03:13,167
The first thing I will set.

2013
02:03:13,250 --> 02:03:15,167
Next I am going to add a greater

2014
02:03:15,250 --> 02:03:20,477
and that's it.
I go ahead andee ate a -- and

2015
02:03:20,560 --> 02:03:22,077
create a job.

2016
02:03:22,160 --> 02:03:24,947
It has through logs, scores them

2017
02:03:25,030 --> 02:03:28,497
and we start the hill climb.

2018
02:03:28,580 --> 02:03:32,817
If you want full control over
the training loop let me give

2019
02:03:32,900 --> 02:03:35,827
you sneak peak into
low level training API.

2020
02:03:35,910 --> 02:03:39,697
You can see I have the MEI sinking model.

2021
02:03:39,780 --> 02:03:42,417
I can also configure my role out

2022
02:03:42,500 --> 02:03:44,687
strategy as well as my hyper

2023
02:03:44,770 --> 02:03:50,207
parameters to define exactly how
I want my training algorithm to

2024
02:03:50,290 --> 02:03:50,677
work.

2025
02:03:50,760 --> 02:03:54,547
In fact I can also incorporate

2026
02:03:54,630 --> 02:03:57,847
my own RLM by defining the tools

2027
02:03:57,930 --> 02:03:59,297
this model interacts with.

2028
02:03:59,380 --> 02:04:02,837
You just saw all of the code,

2029
02:04:02,920 --> 02:04:05,687
but as M365 customer you never

2030
02:04:05,770 --> 02:04:10,327
start from scratch.
Let's hop on to Copilot.

2031
02:04:10,410 --> 02:04:11,927
As a part of frontier tuning we

2032
02:04:12,010 --> 02:04:15,167
are introducing a new way to

2033
02:04:15,250 --> 02:04:18,087
build environments based on your

2034
02:04:18,170 --> 02:04:20,437
data and your workflows.

2035
02:04:20,520 --> 02:04:25,637
One of our customers land
O lakes one of the largest

2036
02:04:25,720 --> 02:04:27,777
businesses in the United States

2037
02:04:27,860 --> 02:04:33,867
is using this to perfect
that butter on your moerning --

2038
02:04:33,950 --> 02:04:36,917
morning toast. Let me walk
you through their environment.

2039
02:04:37,000 --> 02:04:39,427
The environment has soft skills,

2040
02:04:39,510 --> 02:04:41,327
knowledge and tools.

2041
02:04:41,410 --> 02:04:46,797
But in the back end we create
an entire origin for your agents to

2042
02:04:46,880 --> 02:04:47,767
continuously learn the way that

2043
02:04:47,850 --> 02:04:54,207
you work. Let's look at the butter report

2044
02:04:54,290 --> 02:04:56,837
generation here.
These tasks are very complex.

2045
02:04:56,920 --> 02:05:01,207
They require many manual steps

2046
02:05:01,290 --> 02:05:02,717
and a high degree of precision.

2047
02:05:02,800 --> 02:05:06,467
In these tasks almost 80%
accuracy isn't good enough.

2048
02:05:06,550 --> 02:05:10,217
To hill climb to higher accuracy,
we are extending the

2049
02:05:10,300 --> 02:05:12,787
industry definition of skills to

2050
02:05:12,870 --> 02:05:18,227
include Rubriks on what good looks like.

2051
02:05:18,310 --> 02:05:22,577
This is one task. How do I
scale this to qualify all o

2052
02:05:22,660 --> 02:05:24,247
all of the tasks in enterprise.

2053
02:05:24,330 --> 02:05:28,437
You spend a lot of time in M365

2054
02:05:28,520 --> 02:05:31,807
in teams, outlook, word,
excel and PowerPoint.

2055
02:05:31,890 --> 02:05:35,707
We use those signals to suggest
skills and rubrics that define

2056
02:05:35,790 --> 02:05:40,137
the way you work. Next you can add your

2057
02:05:40,220 --> 02:05:43,687
organizational knowledge for branding
like from OneDrive and SharePoint.

2058
02:05:43,770 --> 02:05:47,927
The environment comes built
in with Microsoft tools and you can

2059
02:05:48,010 --> 02:05:50,527
add custom tools to it.

2060
02:05:50,610 --> 02:05:52,597
Because these tools tap into

2061
02:05:52,680 --> 02:05:57,817
real workflows we simulate

2062
02:05:57,900 --> 02:06:01,167
execution so the model can learn
without actually impacting the

2063
02:06:01,250 --> 02:06:07,177
live state of your business.
Now my most favorite part, the

2064
02:06:07,260 --> 02:06:09,317
science. By generalizing all of these

2065
02:06:09,400 --> 02:06:14,567
learnings into the main model
as well as in the embedding model,

2066
02:06:14,650 --> 02:06:20,857
we are able to hill climb
for tasks that require high

2067
02:06:20,940 --> 02:06:21,227
accuracy.

2068
02:06:21,310 --> 02:06:26,157
And not only this, we are able

2069
02:06:26,240 --> 02:06:30,747
to hill climb to more than 90%

2070
02:06:30,830 --> 02:06:32,417
accuracy for Land-O-Lakes class

2071
02:06:32,500 --> 02:06:35,737
for using the model.

2072
02:06:35,820 --> 02:06:38,837
We estimate the model to be 10X

2073
02:06:38,920 --> 02:06:45,677
efficient than the baseline

2074
02:06:45,760 --> 02:06:47,367
models. Let's run the task.

2075
02:06:47,450 --> 02:06:50,657
This time using this environment

2076
02:06:50,740 --> 02:06:54,887
as an entrancing harness.
Usually this task takes a couple

2077
02:06:54,970 --> 02:06:57,497
of minutes to run.

2078
02:06:57,580 --> 02:07:04,580
In the meantime I will show you

2079
02:07:07,720 --> 02:07:12,377
a carshed response. They have research
with multiple models including a fine tune

2080
02:07:12,460 --> 02:07:15,477
model. What you see here is a summary

2081
02:07:15,560 --> 02:07:20,967
that doesn't feel generic,
it feels undoubtedly Land-O-Lakes.

2082
02:07:21,050 --> 02:07:25,387
That's not it.
The task holds itself to high

2083
02:07:25,470 --> 02:07:31,297
standards. It continuously retrospect and

2084
02:07:31,380 --> 02:07:35,767
evaluate success.
With frontier tuning your agent

2085
02:07:35,850 --> 02:07:37,847
continuously improves with

2086
02:07:37,930 --> 02:07:41,267
environment on your data

2087
02:07:41,350 --> 02:07:45,137
encoding the way you work.

2088
02:07:45,220 --> 02:07:47,947
That is what I call frontier

2089
02:07:48,030 --> 02:07:53,397
tuning as smooth as butter.

2090
02:07:53,480 --> 02:07:57,957
We can't wait to see
the environment you build.

2091
02:07:58,040 --> 02:08:00,510
Back to you Satya.

2092
02:08:00,870 --> 02:08:01,287
>> Thank you very much.

2093
02:08:01,370 --> 02:08:08,370
Thank you for the partnership

2094
02:08:08,980 --> 02:08:11,387
here. What you saw is a pretty

2095
02:08:11,470 --> 02:08:14,707
significant shift.

2096
02:08:14,790 --> 02:08:19,847
Times come from just consuming
frontier model to fully

2097
02:08:19,930 --> 02:08:20,847
participating at the frontier

2098
02:08:20,930 --> 02:08:25,217
and frontier ecosystem.

2099
02:08:25,300 --> 02:08:28,367
That is the transition.

2100
02:08:28,450 --> 02:08:30,857
You can have private evals and

2101
02:08:30,940 --> 02:08:33,117
outcomes private RLEs and

2102
02:08:33,200 --> 02:08:36,187
traces, your enterprise

2103
02:08:36,270 --> 02:08:38,827
knowledge create scaffolding

2104
02:08:38,910 --> 02:08:42,127
from models to hill climb.

2105
02:08:42,210 --> 02:08:44,987
That will allow you to

2106
02:08:45,070 --> 02:08:50,067
differentiate on-line what you control.

2107
02:08:50,150 --> 02:08:51,737
There is a new operating point

2108
02:08:51,820 --> 02:08:53,147
at the frontier.

2109
02:08:53,230 --> 02:08:59,217
Where you can use a very effic

2110
02:08:59,300 --> 02:09:04,087
efficient reasoning model
and coding model and achieve

2111
02:09:04,170 --> 02:09:07,687
frontier because you have done
the hard work to create the RLE

2112
02:09:07,770 --> 02:09:10,727
the hill climbing machine that

2113
02:09:10,810 --> 02:09:13,767
create the model your hill

2114
02:09:13,850 --> 02:09:16,847
traces can make to the frontier.

2115
02:09:16,930 --> 02:09:19,767
We think the combination
of the two is a game-changer and how

2116
02:09:19,850 --> 02:09:24,037
people think about what does it
mean to operate in the frontier.

2117
02:09:24,120 --> 02:09:27,747
What does the token look like
how are you in control what is

2118
02:09:27,830 --> 02:09:31,717
the future of the firm those
are the questions and really the

2119
02:09:31,800 --> 02:09:33,897
ecosystem that gets built around

2120
02:09:33,980 --> 02:09:36,757
as opposed to a few models

2121
02:09:36,840 --> 02:09:41,487
hungry for all data.

2122
02:09:41,570 --> 02:09:44,897
To close out I want to talk

2123
02:09:44,980 --> 02:09:48,367
about the frontiers beyond to

2124
02:09:48,450 --> 02:09:50,687
push the frontier what's the

2125
02:09:50,770 --> 02:09:57,337
next best thing.

2126
02:09:57,420 --> 02:09:59,077
Building the scientific

2127
02:09:59,160 --> 02:10:03,747
discovery loop can have perhaps

2128
02:10:03,830 --> 02:10:10,257
the biggest societal impact.
Science today is a little too leni

2129
02:10:10,340 --> 02:10:14,157
lenient. You form a hypothesis, you run

2130
02:10:14,240 --> 02:10:18,997
an experiment you wait for lab
results and then you begin

2131
02:10:19,080 --> 02:10:24,247
again. What if the scientific
method itself can become more

2132
02:10:24,330 --> 02:10:25,847
continuous, more program able.

2133
02:10:25,930 --> 02:10:29,367
That's what we are doing
with Microsoft discovery.

2134
02:10:29,450 --> 02:10:30,907
Discovery brings together the

2135
02:10:30,990 --> 02:10:36,907
models the ACP compute,
the knowledge graphs of all of the

2136
02:10:36,990 --> 02:10:37,777
scientific knowledge automated

2137
02:10:37,860 --> 02:10:41,137
lab and simulation into one

2138
02:10:41,220 --> 02:10:45,357
agentic discovery loop.

2139
02:10:45,440 --> 02:10:50,327
I am thrilled to GA and Microsoft
Discovery, to show you

2140
02:10:50,410 --> 02:10:54,110
this in action let me
invite David up on the stage.

2141
02:10:54,480 --> 02:10:56,947
>> Hello. I work in the Microsoft

2142
02:10:57,030 --> 02:10:59,737
Discovery and quantum team.

2143
02:10:59,820 --> 02:11:02,067
Today to recycle this plastic

2144
02:11:02,150 --> 02:11:05,807
like this bottle you have to

2145
02:11:05,890 --> 02:11:09,717
sled it and melt it.
You cannot use it to make this

2146
02:11:09,800 --> 02:11:13,227
bottle again. It is down cycling.

2147
02:11:13,310 --> 02:11:17,057
We are using Microsoft discovery
to advance that research.

2148
02:11:17,140 --> 02:11:22,187
Let me show it to you.
This is the Microsoft discovery

2149
02:11:22,270 --> 02:11:24,757
app. It is VS code.

2150
02:11:24,840 --> 02:11:29,027
Agentic Discovery has many with

2151
02:11:29,110 --> 02:11:34,067
software engineers.
I want to do three things as a

2152
02:11:34,150 --> 02:11:36,307
scientist. I want to write a scientific

2153
02:11:36,390 --> 02:11:40,707
paper about this topic.

2154
02:11:40,790 --> 02:11:43,317
Instead of melting the plastic I

2155
02:11:43,400 --> 02:11:45,577
can use proteins so you can

2156
02:11:45,660 --> 02:11:48,697
recycle it again and again.

2157
02:11:48,780 --> 02:11:50,857
Second I want to perform the

2158
02:11:50,940 --> 02:11:56,127
actual discovery to define this.

2159
02:11:56,210 --> 02:12:01,297
Third I want to create
a protocol to test the results in

2160
02:12:01,380 --> 02:12:02,227
a real lab.

2161
02:12:02,310 --> 02:12:04,587
If you are a developer the steps

2162
02:12:04,670 --> 02:12:11,670
will be very familiar to you.

2163
02:12:15,560 --> 02:12:16,717
Let's launch it.

2164
02:12:16,800 --> 02:12:21,097
This will is discovery engine.

2165
02:12:21,180 --> 02:12:24,617
These are specialized agents
always running in the background

2166
02:12:24,700 --> 02:12:26,387
following the scientific methods.

2167
02:12:26,470 --> 02:12:28,687
You can see the agents you can add more.

2168
02:12:28,770 --> 02:12:30,797
Microsoft discovery includes a

2169
02:12:30,880 --> 02:12:34,027
community of agents, models and

2170
02:12:34,110 --> 02:12:37,077
tools across many domains.

2171
02:12:37,160 --> 02:12:41,137
You can have officers fair
parties or create your own.

2172
02:12:41,220 --> 02:12:42,507
This is still running.

2173
02:12:42,590 --> 02:12:44,407
It will be running for a while.

2174
02:12:44,490 --> 02:12:51,490
Even hours or days.

2175
02:12:55,820 --> 02:12:59,017
It is running simulations

2176
02:12:59,100 --> 02:13:03,357
dynamically. You can see
the files created here.

2177
02:13:03,440 --> 02:13:05,497
This one is the research paper

2178
02:13:05,580 --> 02:13:08,567
that I was asking for.

2179
02:13:08,650 --> 02:13:14,837
Discovery using no less graph

2180
02:13:14,920 --> 02:13:16,367
internally. It is internal knowledge.

2181
02:13:16,450 --> 02:13:19,207
This is a critical asset because

2182
02:13:19,290 --> 02:13:23,247
it provide compete visibility.

2183
02:13:23,330 --> 02:13:30,330
Scientists can be in full

2184
02:13:34,310 --> 02:13:35,357
control.

2185
02:13:35,440 --> 02:13:42,440
How do you come up with the

2186
02:13:44,700 --> 02:13:45,887
plastic.

2187
02:13:45,970 --> 02:13:47,807
You need a way to generate the

2188
02:13:47,890 --> 02:13:53,237
proteins and identify the most

2189
02:13:53,320 --> 02:14:00,320
promising ones.

2190
02:14:00,580 --> 02:14:02,397
There is no out of the box agent for that.

2191
02:14:02,480 --> 02:14:07,227
That's okay. With Microsoft discovery
there is not an agent for the task.

2192
02:14:07,310 --> 02:14:10,227
You can create one on the
fly which is pretty cool.

2193
02:14:10,310 --> 02:14:12,397
Here is all of the files that it

2194
02:14:12,480 --> 02:14:19,187
created. This one is the YAMMER for the

2195
02:14:19,270 --> 02:14:21,737
agent. Now we need to generate

2196
02:14:21,820 --> 02:14:22,967
candidates for the proteins.

2197
02:14:23,050 --> 02:14:28,277
That task requires a lot of

2198
02:14:28,360 --> 02:14:31,517
compute. It is integrated with HPC.

2199
02:14:31,600 --> 02:14:36,137
You can use it. You can see the process.

2200
02:14:36,220 --> 02:14:40,587
You start with C protein
and created variations by replacing

2201
02:14:40,670 --> 02:14:41,527
small segments.

2202
02:14:41,610 --> 02:14:47,227
You can see the segments there.

2203
02:14:47,310 --> 02:14:51,847
You apply to see if it helps or

2204
02:14:51,930 --> 02:14:52,517
not.

2205
02:14:52,600 --> 02:14:56,937
This is done millions of times

2206
02:14:57,020 --> 02:15:01,347
exploring a huge tree of proteins.

2207
02:15:01,430 --> 02:15:03,607
The result is 80 proteins sent

2208
02:15:03,690 --> 02:15:06,467
to the lab for testing.

2209
02:15:06,550 --> 02:15:08,537
It is like the software.

2210
02:15:08,620 --> 02:15:11,757
Creating a protein is a little

2211
02:15:11,840 --> 02:15:17,897
more complicated.
The bacteria create the protein

2212
02:15:17,980 --> 02:15:20,827
for me. So discovery created another

2213
02:15:20,910 --> 02:15:24,847
file.

2214
02:15:24,930 --> 02:15:26,017
This one here.

2215
02:15:26,100 --> 02:15:30,067
This contains the DNA sequences.

2216
02:15:30,150 --> 02:15:37,150
I can send them to the lab if I

2217
02:15:37,460 --> 02:15:42,437
have a lab I can integrate it

2218
02:15:42,520 --> 02:15:48,527
directly with the agent.
Go to proteins and submit to

2219
02:15:48,610 --> 02:15:52,697
lab. This will use a custom agent
that sends instructions for the

2220
02:15:52,780 --> 02:15:54,097
lab appointment.

2221
02:15:54,180 --> 02:16:00,267
This is very real.
That's an maut ma automated lab

2222
02:16:00,350 --> 02:16:03,237
I have it right here.
This is application controlling the lab.

2223
02:16:03,320 --> 02:16:05,137
It has a Copilot interface so

2224
02:16:05,220 --> 02:16:08,147
the scientist can interact with

2225
02:16:08,230 --> 02:16:11,917
the lab to the science

2226
02:16:12,000 --> 02:16:15,567
experiment. In this case Discovery has

2227
02:16:15,650 --> 02:16:19,587
submitted the job.
You can see it right here.

2228
02:16:19,670 --> 02:16:22,887
This will take some time.
Let me open a previous run.

2229
02:16:22,970 --> 02:16:26,767
There you go. These are all
of the steps that the agents are doing.

2230
02:16:26,850 --> 02:16:30,987
Most of them completely
automated with human supervision.

2231
02:16:31,070 --> 02:16:38,070
How cool is that?

2232
02:16:39,290 --> 02:16:41,707
This is bringing the together

2233
02:16:41,790 --> 02:16:45,817
the physical and agents in

2234
02:16:45,900 --> 02:16:50,217
unified Discovery loop.
This is one example of what

2235
02:16:50,300 --> 02:16:54,017
Microsoft Discovery can do.
Industries are using it today to

2236
02:16:54,100 --> 02:16:57,657
embrace a new era of scientific discovery.

2237
02:16:57,740 --> 02:17:02,667
Thank you. Back to you Satya.
>> Thank you so much David.

2238
02:17:02,750 --> 02:17:04,587
Talking about scientific

2239
02:17:04,670 --> 02:17:05,587
discovery, we are also

2240
02:17:05,670 --> 02:17:12,337
continuing to make rapid

2241
02:17:12,420 --> 02:17:16,247
progress on our long-term goal
of building a scalable computer.

2242
02:17:16,330 --> 02:17:19,447
We announced last year our first QPU.

2243
02:17:19,530 --> 02:17:21,067
We created a new state of matter

2244
02:17:21,150 --> 02:17:27,787
that was only theorized 100

2245
02:17:27,870 --> 02:17:29,927
years ago. Proved it out that it exists.

2246
02:17:30,010 --> 02:17:33,097
Our vision was to take
a radically different approach to

2247
02:17:33,180 --> 02:17:37,217
addressing the fundamental
barriers to building a scalable

2248
02:17:37,300 --> 02:17:41,607
quantum machine which is all about
reliability, speed as well as size.

2249
02:17:41,690 --> 02:17:43,767
Since then, we have continued to

2250
02:17:43,850 --> 02:17:47,037
make progress across the full

2251
02:17:47,120 --> 02:17:49,177
quantum stack with both our

2252
02:17:49,260 --> 02:17:52,547
academic and industry partners.

2253
02:17:52,630 --> 02:17:57,717
In Q north we will have
a quantum computer powered by atom

2254
02:17:57,800 --> 02:18:00,487
computers natural atom computers
with our stack in there.

2255
02:18:00,570 --> 02:18:02,127
We are working with algorithmic

2256
02:18:02,210 --> 02:18:05,497
and Columbia and Zurich.

2257
02:18:05,580 --> 02:18:08,627
We ourselves use this discovery

2258
02:18:08,710 --> 02:18:11,387
agentic loop to accelerate the

2259
02:18:11,470 --> 02:18:14,337
work in quantum compressing

2260
02:18:14,420 --> 02:18:18,277
years of research into this last

2261
02:18:18,360 --> 02:18:18,537
year.

2262
02:18:18,620 --> 02:18:23,447
Today I am really thrilled to

2263
02:18:23,530 --> 02:18:30,530
announce Majorana2.

2264
02:18:39,240 --> 02:18:42,727
The resulting cube bits are

2265
02:18:42,810 --> 02:18:46,427
reliable and maintaining
their state much longer while other

2266
02:18:46,510 --> 02:18:48,337
common approaches deliver a

2267
02:18:48,420 --> 02:18:52,677
lifetime of micro seconds.

2268
02:18:52,760 --> 02:18:57,607
Majorana 2 provide cubic
lifetime of 20 seconds or up to

2269
02:18:57,690 --> 02:19:00,777
even a minute.
Essentially a thousand times

2270
02:19:00,860 --> 02:19:02,177
higher than what we were able to

2271
02:19:02,260 --> 02:19:09,260
achieve with high -- my or --

2272
02:19:11,640 --> 02:19:14,827
Majorana 1. They are enabling complex

2273
02:19:14,910 --> 02:19:18,727
quantum computation in that lifetime.

2274
02:19:18,810 --> 02:19:21,407
All of this in the same cubic

2275
02:19:21,490 --> 02:19:27,737
form factor of Majorana 1 in one
100th of a millimeter control of

2276
02:19:27,820 --> 02:19:30,537
digital. Which again
is a super important aspect.

2277
02:19:30,620 --> 02:19:35,337
Making it all possible to fit
a million of these cube bits in a

2278
02:19:35,420 --> 02:19:37,187
chip smaller than a credit card.

2279
02:19:37,270 --> 02:19:39,457
It is this combination of the

2280
02:19:39,540 --> 02:19:41,127
reliability, the speed, the size

2281
02:19:41,210 --> 02:19:44,357
that makes the topological

2282
02:19:44,440 --> 02:19:46,857
approach so unique.

2283
02:19:46,940 --> 02:19:48,357
With Majorana 1 we had proven

2284
02:19:48,440 --> 02:19:52,057
out the foundational physics and

2285
02:19:52,140 --> 02:19:54,377
with Majorana 2 we begin the

2286
02:19:54,460 --> 02:19:59,760
engineering scale.

2287
02:19:59,860 --> 02:20:02,907
Yup.

2288
02:20:02,990 --> 02:20:04,807
Ultimately it is never about

2289
02:20:04,890 --> 02:20:07,877
tech for tech sake.

2290
02:20:07,960 --> 02:20:12,027
It is about tackling those
pressing challenges of people

2291
02:20:12,110 --> 02:20:16,787
and planet. It is also the fundamental
point of this conference.

2292
02:20:16,870 --> 02:20:20,857
The question is not about
whether you can build the next

2293
02:20:20,940 --> 02:20:24,957
great model or platform
even this quantum machine.

2294
02:20:25,040 --> 02:20:29,477
It is the question of how do
we build the frontier ecosystem

2295
02:20:29,560 --> 02:20:32,677
together. Because there are really two

2296
02:20:32,760 --> 02:20:36,677
stories people can tell about this moment.

2297
02:20:36,760 --> 02:20:39,607
One is that technology
concentrates power, reduces

2298
02:20:39,690 --> 02:20:43,927
human agency and leaves
the society to absorb the

2299
02:20:44,010 --> 02:20:46,887
consequences.
The other is that we use this

2300
02:20:46,970 --> 02:20:49,587
next wave to unlock opportunity

2301
02:20:49,670 --> 02:20:52,487
for developers, scientists,

2302
02:20:52,570 --> 02:20:54,027
enterprises in every community.

2303
02:20:54,110 --> 02:20:57,587
Our job is to make the second

2304
02:20:57,670 --> 02:21:01,977
story true. That's our North
Star for the frontier ecosystem.

2305
02:21:02,060 --> 02:21:03,737
Let's all go build together.

2306
02:21:03,820 --> 02:21:06,577
Thank you all very, very much.

2307
02:21:06,660 --> 02:21:13,660
>> Thank you.

2308
02:21:59,840 --> 02:22:04,187
>> Within reach.

2309
02:22:04,270 --> 02:22:07,777
Quantum is our most fundamental

2310
02:22:07,860 --> 02:22:12,607
physical theorem.

2311
02:22:12,690 --> 02:22:14,337
For quantum to deliver real

2312
02:22:14,420 --> 02:22:16,637
world impact it has to be both

2313
02:22:16,720 --> 02:22:21,907
reliable and scalable.
>> Reaching quantum at scale

2314
02:22:21,990 --> 02:22:25,157
will allow us to make a more
direct approach to solving

2315
02:22:25,240 --> 02:22:27,687
problems in chemistry.

2316
02:22:27,770 --> 02:22:30,457
>> Discovery in life sciences.

2317
02:22:30,540 --> 02:22:35,957
>> We bring drugs to market it
opens up millions of possibilities.

2318
02:22:36,040 --> 02:22:37,027
Quantum will change the world.

2319
02:22:37,110 --> 02:22:44,110
This change is just beginning to

2320
02:22:46,090 --> 02:22:48,290
happen.