1
00:00:02,400 --> 00:00:06,800
Speaker 1: Bloomberg Audio Studios, podcasts, radio news.

2
00:00:07,240 --> 00:00:10,120
Speaker 2: Welcome to our Bloomberg television and radio audiences. I'm here

3
00:00:10,119 --> 00:00:13,600
Speaker 2: in Redmend, Washington with Microsoft CEO Sacha Nadala with some

4
00:00:13,680 --> 00:00:15,720
Speaker 2: big unveilings at Build today.

5
00:00:15,760 --> 00:00:17,440
Speaker 1: Sacha, thank you so much for joining us.

6
00:00:17,920 --> 00:00:21,080
Speaker 2: You've got a new version of Windows powered by AI

7
00:00:21,239 --> 00:00:24,400
Speaker 2: that basically means you have your own AI agent on

8
00:00:24,440 --> 00:00:28,520
Speaker 2: your own device. How will this AI powered hardware experience

9
00:00:28,600 --> 00:00:30,840
Speaker 2: be totally different from what we've seen before?

10
00:00:31,320 --> 00:00:33,080
Speaker 3: First, well, thank you so much for having me Emily

11
00:00:33,159 --> 00:00:33,519
Speaker 3: end up.

12
00:00:33,680 --> 00:00:38,159
Speaker 4: Yeah, I mean it's really a ground up, you know,

13
00:00:38,200 --> 00:00:44,559
Speaker 4: reimagination from Silicon So we have a new SoC that

14
00:00:44,720 --> 00:00:47,680
Speaker 4: has the NPU. So think about it as we've always

15
00:00:47,680 --> 00:00:51,200
Speaker 4: had CPUs and GPUs and now you have newler processing

16
00:00:51,280 --> 00:00:54,920
Speaker 4: units to in fact locally process these AI models. We

17
00:00:55,080 --> 00:00:58,200
Speaker 4: have forty plus AI models that are local.

18
00:00:59,480 --> 00:01:00,000
Speaker 3: On Window.

19
00:01:00,040 --> 00:01:02,960
Speaker 4: Those machines that then are being used by a variety

20
00:01:03,040 --> 00:01:06,000
Speaker 4: of experiences that we are built into Windows, starting with

21
00:01:06,120 --> 00:01:09,840
Speaker 4: this photographic memory feature called Recall, which is just phenomenal.

22
00:01:10,560 --> 00:01:13,760
Speaker 4: And of course Copilot is built in to Windows and

23
00:01:13,800 --> 00:01:15,800
Speaker 4: Copilot is just not an app, it's sort of a

24
00:01:15,840 --> 00:01:18,880
Speaker 4: shell affordance that's going to be there assisting you as

25
00:01:18,880 --> 00:01:22,200
Speaker 4: you be. Showed a demo of somebody playing Minecraft, and

26
00:01:22,319 --> 00:01:25,160
Speaker 4: you know me just sharing my Minecraft screen with Copilot,

27
00:01:25,160 --> 00:01:27,600
Speaker 4: and Copilot helps me finally me as competitive with my

28
00:01:27,680 --> 00:01:30,880
Speaker 4: daughter on Minecraft has always dreamed of, right and so,

29
00:01:31,440 --> 00:01:33,440
Speaker 4: and then of course the third parties, I mean, Adobe

30
00:01:33,480 --> 00:01:35,840
Speaker 4: showed how they're take bringing all of what they do

31
00:01:36,240 --> 00:01:40,760
Speaker 4: to Copilot plus PC's and many many more other developers.

32
00:01:40,760 --> 00:01:42,480
Speaker 4: So it's an exciting day for us to have a

33
00:01:42,520 --> 00:01:45,880
Speaker 4: complete rethink on the full stack of windows for the AI.

34
00:01:46,000 --> 00:01:48,600
Speaker 2: Ah, I bet you're pretty good on Minecraft just a hunch.

35
00:01:49,120 --> 00:01:52,040
Speaker 2: This new feature recall means your computer basically has a

36
00:01:52,080 --> 00:01:53,240
Speaker 2: photographic memory of.

37
00:01:53,240 --> 00:01:54,440
Speaker 1: Everything you do on it.

38
00:01:54,840 --> 00:01:57,320
Speaker 2: And the speed at which this is all happening is

39
00:01:57,440 --> 00:02:00,840
Speaker 2: just so fast. It seems so unpressed. How will this

40
00:02:00,920 --> 00:02:04,160
Speaker 2: really make things different for consumers and change our daily routines?

41
00:02:04,280 --> 00:02:06,760
Speaker 4: Yeah, I think if you think about the PC, it

42
00:02:06,840 --> 00:02:11,400
Speaker 4: has always been about you know, creating things right, which

43
00:02:11,440 --> 00:02:14,639
Speaker 4: is it's about being able to synthesize lots of information,

44
00:02:15,040 --> 00:02:20,520
Speaker 4: create new information, pass better judgment, gain insights, and act. Right,

45
00:02:20,560 --> 00:02:23,760
Speaker 4: that's what you sort of do on a PC. AI

46
00:02:24,080 --> 00:02:28,440
Speaker 4: is an assistant AI helps you, as a as an

47
00:02:28,520 --> 00:02:31,320
Speaker 4: end user, do anything right. One of the most beautiful

48
00:02:31,320 --> 00:02:33,720
Speaker 4: demos today was even this thing called co Create. Right.

49
00:02:33,760 --> 00:02:36,960
Speaker 4: I am in paint and I'm painting a nice spring

50
00:02:37,120 --> 00:02:41,080
Speaker 4: Seattle day and I want mountains with wildflowers, and I

51
00:02:41,120 --> 00:02:43,640
Speaker 4: can now have the AI assistant that painter and me

52
00:02:43,760 --> 00:02:46,280
Speaker 4: now can be inspired by AI. But ultimately I am

53
00:02:46,280 --> 00:02:49,160
Speaker 4: the one painting and I get that satisfactor like I feel.

54
00:02:49,440 --> 00:02:51,239
Speaker 4: One of the things that's been one of my coworkers

55
00:02:51,400 --> 00:02:53,520
Speaker 4: was telling me when her five year old is saying,

56
00:02:53,560 --> 00:02:56,840
Speaker 4: for the first time, give me designer time. That's what felt,

57
00:02:56,919 --> 00:02:58,880
Speaker 4: you know, it feels like in the late eighties and

58
00:02:58,919 --> 00:03:01,200
Speaker 4: the early nineties when we all got fell in love

59
00:03:01,240 --> 00:03:04,720
Speaker 4: with PCs, that was what really gods going, which is

60
00:03:04,760 --> 00:03:08,240
Speaker 4: the ability to use computers to create. And I think

61
00:03:08,280 --> 00:03:10,160
Speaker 4: this AI is just going to take it to the

62
00:03:10,200 --> 00:03:10,760
Speaker 4: next level.

63
00:03:11,000 --> 00:03:13,480
Speaker 2: Microsoft has had trouble keeping up with some of the

64
00:03:13,480 --> 00:03:15,359
Speaker 2: new cutting edge Max and you did do a lot

65
00:03:15,360 --> 00:03:19,080
Speaker 2: of side by side comparisons today. Have you narrowed the

66
00:03:19,080 --> 00:03:21,960
Speaker 2: gap today truly? And how much share do you think

67
00:03:22,000 --> 00:03:22,919
Speaker 2: you can take from Apple?

68
00:03:23,200 --> 00:03:25,320
Speaker 4: Well, I mean, I think today was a day where

69
00:03:25,320 --> 00:03:28,200
Speaker 4: we brought some of the efforts that we've had for

70
00:03:28,240 --> 00:03:32,840
Speaker 4: a long time all into focus and really brought cohesiveness

71
00:03:32,840 --> 00:03:35,080
Speaker 4: to them. For example, we've been on this journey to

72
00:03:35,120 --> 00:03:38,680
Speaker 4: bring I call it arm based PCs for a while.

73
00:03:39,160 --> 00:03:42,360
Speaker 4: Finally we now have the best ARM processors, which are

74
00:03:42,440 --> 00:03:46,800
Speaker 4: really the performance of the CPU itself is cutting edge.

75
00:03:47,440 --> 00:03:51,920
Speaker 4: Then you combine it with the NPU and these experiences.

76
00:03:52,080 --> 00:03:54,080
Speaker 4: I think we are now going to start just like

77
00:03:54,160 --> 00:03:55,600
Speaker 4: you know, when it comes to the cloud.

78
00:03:55,680 --> 00:03:58,040
Speaker 3: Let me take you another one, right, which is.

79
00:03:58,040 --> 00:04:01,680
Speaker 4: When we were We came from second and what became

80
00:04:01,800 --> 00:04:03,720
Speaker 4: a leader in the AI age on the cloud.

81
00:04:04,080 --> 00:04:06,440
Speaker 3: I hope on the PC we bring good competition. I mean,

82
00:04:06,480 --> 00:04:08,000
Speaker 3: Apple has done a fantastic job.

83
00:04:08,400 --> 00:04:11,920
Speaker 4: We now want to bring real competition back to the

84
00:04:11,960 --> 00:04:13,080
Speaker 4: Windows versus Mac.

85
00:04:13,640 --> 00:04:16,040
Speaker 3: When it comes to the aihe.

86
00:04:15,760 --> 00:04:18,880
Speaker 2: You've been doing a lot of unique deals to move quickly,

87
00:04:18,920 --> 00:04:21,679
Speaker 2: for example, this six hundred and fifty million dollar investment

88
00:04:21,680 --> 00:04:22,200
Speaker 2: in Inflection.

89
00:04:22,320 --> 00:04:24,560
Speaker 1: You hire the team, you hire the co founder, must

90
00:04:24,600 --> 00:04:27,440
Speaker 1: office Suliman. Are you at all.

91
00:04:27,279 --> 00:04:29,680
Speaker 2: Worried that regulators are going to crack down on investments

92
00:04:29,680 --> 00:04:31,599
Speaker 2: and aqua hires the way they have cracked down on

93
00:04:31,720 --> 00:04:32,120
Speaker 2: M and A like.

94
00:04:32,160 --> 00:04:34,600
Speaker 1: Do you wish you could be buying these companies outright.

95
00:04:34,680 --> 00:04:36,600
Speaker 4: I mean, look, I mean the regulators have to do

96
00:04:36,640 --> 00:04:39,320
Speaker 4: what they have to do to make sure that ultimately

97
00:04:39,360 --> 00:04:41,719
Speaker 4: there's competition, right, I mean, if that is the case,

98
00:04:41,880 --> 00:04:44,520
Speaker 4: then I think I think one of the fundamental things

99
00:04:44,520 --> 00:04:46,800
Speaker 4: that I think all regulators need to stay rounded on

100
00:04:47,839 --> 00:04:48,960
Speaker 4: is where is competition?

101
00:04:49,040 --> 00:04:51,400
Speaker 3: What are the sources of competition, and how.

102
00:04:51,240 --> 00:04:54,240
Speaker 4: To foster them? To me, that's what we're doing right today.

103
00:04:54,560 --> 00:04:57,960
Speaker 4: There is formidable amount of competition, right, whether it is

104
00:04:58,000 --> 00:05:00,800
Speaker 4: between the big players or the small pla and there's

105
00:05:00,880 --> 00:05:04,240
Speaker 4: partnerships are also a source of competition. So I don't

106
00:05:04,240 --> 00:05:08,000
Speaker 4: think that this is about any one company acquiring or

107
00:05:08,000 --> 00:05:10,640
Speaker 4: not acquiring, But it's about competition. And as long as

108
00:05:10,680 --> 00:05:13,800
Speaker 4: regulators focused on that and we focus on it, I

109
00:05:13,839 --> 00:05:14,679
Speaker 4: think we'll be fine.

110
00:05:14,800 --> 00:05:17,240
Speaker 2: Microsoft just took a big stake in the Abu Dhabi

111
00:05:17,279 --> 00:05:18,680
Speaker 2: AI company G forty two.

112
00:05:18,800 --> 00:05:20,640
Speaker 1: Brad Smith joined the board.

113
00:05:21,000 --> 00:05:22,680
Speaker 2: This is a company that at least at one point

114
00:05:22,720 --> 00:05:24,479
Speaker 2: had ties to China. So there do seem to be

115
00:05:24,520 --> 00:05:26,559
Speaker 2: some big geopolitical implications here.

116
00:05:26,960 --> 00:05:29,760
Speaker 1: Is this your play in a tech cold war?

117
00:05:30,560 --> 00:05:32,680
Speaker 4: You know, first of all, you know it is important

118
00:05:32,720 --> 00:05:35,279
Speaker 4: for us to be grounded on the fact that we're

119
00:05:35,880 --> 00:05:38,279
Speaker 4: a multinational company, but that also happens to be an

120
00:05:38,320 --> 00:05:41,559
Speaker 4: American company. So we did this partnership with G forty

121
00:05:41,560 --> 00:05:46,839
Speaker 4: two and the UAE, with sort of the real involvement

122
00:05:46,880 --> 00:05:49,120
Speaker 4: of the US government as well, where the US government

123
00:05:49,160 --> 00:05:52,520
Speaker 4: has now an assurance program with the UAE, which both

124
00:05:52,600 --> 00:05:54,719
Speaker 4: the government, you know, it's a state to state agreement,

125
00:05:55,120 --> 00:05:57,680
Speaker 4: and that allowed us in fact created the conditions for

126
00:05:57,760 --> 00:05:58,440
Speaker 4: G forty.

127
00:05:58,240 --> 00:05:59,599
Speaker 3: Two and Microsoft to come together.

128
00:05:59,640 --> 00:06:03,040
Speaker 4: But I think it's important for us as a company

129
00:06:03,120 --> 00:06:05,119
Speaker 4: to be grounded in the fact that we need great

130
00:06:05,160 --> 00:06:08,160
Speaker 4: partners who then allow us to expand what we are

131
00:06:08,160 --> 00:06:10,159
Speaker 4: doing globally, right because at the end of the day,

132
00:06:10,360 --> 00:06:12,719
Speaker 4: one of my dreams has always been Emily that can

133
00:06:13,080 --> 00:06:16,520
Speaker 4: the you know, the information revolution, perhaps after the Industrial

134
00:06:16,560 --> 00:06:20,320
Speaker 4: Revolution was magical in the sense that it diffused fast

135
00:06:20,400 --> 00:06:23,760
Speaker 4: around the world, but this intelligence revolution, to me, can

136
00:06:23,839 --> 00:06:26,200
Speaker 4: even take it even to the next level, right, which

137
00:06:26,240 --> 00:06:28,640
Speaker 4: is when people in the Middle East, or whether it's

138
00:06:28,640 --> 00:06:31,479
Speaker 4: in Asia or Africa or Latin America can enjoy the

139
00:06:31,520 --> 00:06:35,119
Speaker 4: benefits of this technological wave. That's not going to that's

140
00:06:35,200 --> 00:06:38,159
Speaker 4: going to require real partnerships, And I'm very, very excited

141
00:06:38,160 --> 00:06:38,800
Speaker 4: about this one.

142
00:06:38,920 --> 00:06:42,400
Speaker 2: Open Ai is your closest partner in Microsoft hit history.

143
00:06:42,920 --> 00:06:46,040
Speaker 2: They just released a chat GPT app for Mac before Windows,

144
00:06:46,080 --> 00:06:48,080
Speaker 2: and we're reporting the same as going to happen for

145
00:06:48,480 --> 00:06:51,480
Speaker 2: the iPhone. How do you balance partnering with them and

146
00:06:51,560 --> 00:06:53,279
Speaker 2: competing for their attention?

147
00:06:53,920 --> 00:06:54,080
Speaker 3: Yeah?

148
00:06:54,120 --> 00:06:56,480
Speaker 4: First of all, you know, like open Ai, you know,

149
00:06:56,520 --> 00:07:00,440
Speaker 4: will decide how to prioritize platforms and and how to

150
00:07:00,640 --> 00:07:02,520
Speaker 4: really go about managing their roadmap.

151
00:07:02,520 --> 00:07:03,479
Speaker 3: Would be very excited.

152
00:07:03,480 --> 00:07:05,599
Speaker 4: In fact, you know, I'm thrilled to you know today

153
00:07:05,640 --> 00:07:09,040
Speaker 4: when you saw the launch, we basically took gpto and

154
00:07:09,120 --> 00:07:11,280
Speaker 4: made it part of Core Pilot. So Windows is going

155
00:07:11,320 --> 00:07:14,120
Speaker 4: to have some of the latest innovation from open Ai

156
00:07:14,240 --> 00:07:17,800
Speaker 4: and the latest innovation from Microsoft built into Windows. So therefore,

157
00:07:18,560 --> 00:07:21,360
Speaker 4: I'm thrilled about their prioritization of the Mac.

158
00:07:21,520 --> 00:07:24,520
Speaker 1: Chat GPT four. Oh is you know, it's pretty big?

159
00:07:24,560 --> 00:07:24,800
Speaker 4: Wow.

160
00:07:24,920 --> 00:07:26,679
Speaker 1: Have you tried chat GPT five?

161
00:07:27,560 --> 00:07:28,840
Speaker 3: So that the latest? Oh?

162
00:07:28,880 --> 00:07:29,440
Speaker 4: Yeah, of course?

163
00:07:29,480 --> 00:07:32,400
Speaker 1: What about the next one? We hear reasons you mean,

164
00:07:32,720 --> 00:07:33,160
Speaker 1: so the.

165
00:07:33,120 --> 00:07:34,960
Speaker 4: Next version, which I believe is not been named, so

166
00:07:35,000 --> 00:07:37,960
Speaker 4: I don't want to talk for them, but you know, look,

167
00:07:38,080 --> 00:07:43,040
Speaker 4: I think open Ai as a company has really focused

168
00:07:43,080 --> 00:07:45,640
Speaker 4: a lot on not only moving the frontier forward, but

169
00:07:45,680 --> 00:07:48,720
Speaker 4: also doing it with really safety first, and so they're

170
00:07:49,160 --> 00:07:51,880
Speaker 4: very carefully thinking about what's the next big model. But

171
00:07:52,000 --> 00:07:54,680
Speaker 4: the thing that GPTO like, I don't want to understate

172
00:07:55,120 --> 00:07:57,680
Speaker 4: the innovation in what they just launched last week, right,

173
00:07:57,720 --> 00:07:59,680
Speaker 4: which is just the first time they've been able to

174
00:07:59,680 --> 00:08:03,080
Speaker 4: bring ultimodal capabilities natively into the model. And you're seeing

175
00:08:03,080 --> 00:08:06,280
Speaker 4: it like in the wild as you use copilot or

176
00:08:06,360 --> 00:08:08,440
Speaker 4: chat GPT you can see the power.

177
00:08:08,880 --> 00:08:09,040
Speaker 3: You know.

178
00:08:09,080 --> 00:08:11,600
Speaker 4: It's most people think, oh, GPT four came out what

179
00:08:11,720 --> 00:08:15,520
Speaker 4: fifteen sixteen months ago, but it's not been static. It

180
00:08:15,600 --> 00:08:18,440
Speaker 4: has been continuously getting better and we're excited about it.

181
00:08:18,520 --> 00:08:20,920
Speaker 2: So on that note, you stress trust a lot in

182
00:08:20,960 --> 00:08:23,440
Speaker 2: your keynote today, But there have been renewed concerns over

183
00:08:23,440 --> 00:08:25,840
Speaker 2: the week and about safety at open AI. More key

184
00:08:25,840 --> 00:08:29,600
Speaker 2: employees have left, one saying safety has taken a backseat to.

185
00:08:29,560 --> 00:08:30,640
Speaker 1: Shiny new products.

186
00:08:30,920 --> 00:08:33,199
Speaker 2: What kind of conversations have you had with Sam Molman

187
00:08:33,240 --> 00:08:34,320
Speaker 2: about this look.

188
00:08:34,320 --> 00:08:36,960
Speaker 4: I mean, at least one of the fundamental things that

189
00:08:37,040 --> 00:08:40,480
Speaker 4: broad Open Air and Microsoft together way back even in

190
00:08:40,520 --> 00:08:44,240
Speaker 4: twenty nineteen was that focus on how do we make

191
00:08:44,320 --> 00:08:46,560
Speaker 4: sure that we could make progress. At that time, it

192
00:08:46,600 --> 00:08:49,400
Speaker 4: is not even clearest whether things will even work the

193
00:08:49,440 --> 00:08:52,360
Speaker 4: way they work, right, But even there, that company was

194
00:08:52,440 --> 00:08:54,720
Speaker 4: very grounded on their mission around Hey, we want to

195
00:08:54,720 --> 00:08:56,679
Speaker 4: bring the benefits of this to the broadest set of

196
00:08:56,720 --> 00:08:59,720
Speaker 4: audience and do it safely. So and that requires, at

197
00:08:59,800 --> 00:09:01,520
Speaker 4: least in my mind, I fall into the camp of

198
00:09:01,640 --> 00:09:04,200
Speaker 4: iterative development of it. Right, you got you can't just

199
00:09:04,240 --> 00:09:06,160
Speaker 4: be in the lab. You have to get it into

200
00:09:06,200 --> 00:09:09,760
Speaker 4: the real world. See the adversarial attacks, make the safety

201
00:09:09,840 --> 00:09:13,040
Speaker 4: harnesses better. That doesn't mean you don't take safety first

202
00:09:13,080 --> 00:09:15,840
Speaker 4: as the most important feature. In fact, I say, even

203
00:09:15,840 --> 00:09:19,240
Speaker 4: at Microsoft, which is the most important feature of Azure

204
00:09:19,240 --> 00:09:22,240
Speaker 4: AI is our safety service, because that's what even gives

205
00:09:22,320 --> 00:09:24,080
Speaker 4: us permission to deploy anything.

206
00:09:24,320 --> 00:09:25,280
Speaker 3: You can't just deploy.

207
00:09:25,320 --> 00:09:27,640
Speaker 4: I mean, if I take the base model, like even

208
00:09:27,679 --> 00:09:31,400
Speaker 4: take today's announcement, it's just so important, right, Windows now

209
00:09:31,440 --> 00:09:34,680
Speaker 4: will have forty plus models that are going to be

210
00:09:34,760 --> 00:09:37,760
Speaker 4: onboard on Windows. But guess what, in order to keep

211
00:09:37,800 --> 00:09:41,640
Speaker 4: them safe, we will use the Azure AI service because

212
00:09:41,679 --> 00:09:44,880
Speaker 4: we will be learning every hour of every day what

213
00:09:45,040 --> 00:09:48,920
Speaker 4: is the new adversarial attack, and that knowledge, that learning

214
00:09:49,040 --> 00:09:51,559
Speaker 4: will keep all the AI models safe. Because the next

215
00:09:51,679 --> 00:09:53,959
Speaker 4: jail break that happens, we'll learn about it and make

216
00:09:53,960 --> 00:09:55,880
Speaker 4: sure that that doesn't happen on the local model. And

217
00:09:55,920 --> 00:09:57,600
Speaker 4: so that's I think the only way we're going to

218
00:09:57,679 --> 00:09:58,640
Speaker 4: keep AI.

219
00:09:58,440 --> 00:10:01,080
Speaker 2: Safe, all right, Microsoft CEA, oh saucha nidala, thank you

220
00:10:01,120 --> 00:10:02,719
Speaker 2: so much for joining us. I'm going to send it

221
00:10:02,760 --> 00:10:05,239
Speaker 2: back to you guys, and thanks to our Bloomberg television

222
00:10:05,240 --> 00:10:06,360
Speaker 2: and radio audiences,