1
00:00:00,160 --> 00:00:03,960
Speaker 1: China will be at the forefront of artificial intelligence. That's

2
00:00:03,960 --> 00:00:06,560
Speaker 1: what the CEO of Alphabet, sun Dar Pashai is saying.

3
00:00:06,600 --> 00:00:09,320
Speaker 1: He says it is important for the US to collaborate

4
00:00:09,440 --> 00:00:13,360
Speaker 1: with China on both regulation and innovation when it comes

5
00:00:13,440 --> 00:00:17,599
Speaker 1: to AI. Pishai, along with executives from Microsoft, City Group

6
00:00:17,640 --> 00:00:20,680
Speaker 1: and Tesla, have been meeting with Chinese President Shi Jinping

7
00:00:20,840 --> 00:00:23,840
Speaker 1: and US President Joe Biden at the APEX summit in

8
00:00:23,880 --> 00:00:28,960
Speaker 1: San Francisco. After those meetings, the Alphabet CEO joined Bloomberg's

9
00:00:28,960 --> 00:00:31,960
Speaker 1: Emily Chang and told her the world's two biggest economies

10
00:00:32,000 --> 00:00:35,920
Speaker 1: do need to work together on AI regulation and development.

11
00:00:36,320 --> 00:00:39,760
Speaker 2: So you've got the presidents of the two biggest economies

12
00:00:39,800 --> 00:00:42,400
Speaker 2: in the world right here in big tech's backyard, and

13
00:00:42,600 --> 00:00:46,400
Speaker 2: everybody has questions about AI. How much do you think

14
00:00:46,440 --> 00:00:49,560
Speaker 2: AI is going to drive economic growth in the Asia

15
00:00:49,600 --> 00:00:52,479
Speaker 2: Pacific region and really the world, let's say, over the

16
00:00:52,520 --> 00:00:53,280
Speaker 2: next decade.

17
00:00:53,479 --> 00:00:56,000
Speaker 3: That's a big question. It's a good, good warmer, I

18
00:00:56,040 --> 00:01:02,960
Speaker 3: think obviously rightfully, I think the excitement around AI is

19
00:01:03,960 --> 00:01:07,800
Speaker 3: well founded. I'm glad it's a big topic. Let me

20
00:01:07,840 --> 00:01:10,760
Speaker 3: start with the Asia Pacific question. One of the things

21
00:01:10,800 --> 00:01:14,880
Speaker 3: that excites me is if you're in a region unlike

22
00:01:14,920 --> 00:01:19,560
Speaker 3: the past technologies take the PC revolution, yet in some

23
00:01:19,600 --> 00:01:23,640
Speaker 3: ways play catch up. Mobile was the first transition. I

24
00:01:23,680 --> 00:01:26,520
Speaker 3: think in Asia, you know, you almost code developed it.

25
00:01:27,200 --> 00:01:29,880
Speaker 3: I look at an Android, It wouldn't have happened without

26
00:01:29,920 --> 00:01:32,120
Speaker 3: all the work we did in Asia, our partners in

27
00:01:32,160 --> 00:01:36,319
Speaker 3: Asia wed Samsung or HDC in Taiwan and so on.

28
00:01:37,120 --> 00:01:40,040
Speaker 3: But I think you started leapfrogging and you don't need

29
00:01:40,080 --> 00:01:43,720
Speaker 3: to play catch up with legacy. I think AI takes

30
00:01:43,760 --> 00:01:46,320
Speaker 3: it the next step further. I think it's the first technology.

31
00:01:47,319 --> 00:01:50,400
Speaker 3: I think you're all in a real position to be

32
00:01:51,040 --> 00:01:53,880
Speaker 3: right there from the start, and I have a seat

33
00:01:53,880 --> 00:01:57,320
Speaker 3: as this technology develops. So I think that has profound implications.

34
00:01:58,880 --> 00:02:02,600
Speaker 3: You know, you all can be as AI natives as

35
00:02:02,640 --> 00:02:04,920
Speaker 3: what's happening in the US and so on. So I

36
00:02:04,960 --> 00:02:07,280
Speaker 3: think in that way, you know it's going to have

37
00:02:07,320 --> 00:02:10,440
Speaker 3: a big impact. I think the way you do that though,

38
00:02:10,560 --> 00:02:15,200
Speaker 3: is by as you know, APEC countries. I would say,

39
00:02:15,240 --> 00:02:18,320
Speaker 3: you have to have a pro innovation mindset, you have

40
00:02:18,360 --> 00:02:21,080
Speaker 3: to embrace it at the start, make sure you have

41
00:02:21,120 --> 00:02:24,560
Speaker 3: the right infrastructure in the country to facilitate AI innovation.

42
00:02:25,560 --> 00:02:28,600
Speaker 3: You're thinking about the right balance regulation both from an

43
00:02:28,600 --> 00:02:32,799
Speaker 3: innovation and safety standpoint, and then as government, maybe lead

44
00:02:32,880 --> 00:02:35,839
Speaker 3: the way. What are the projects you can visibly do.

45
00:02:36,520 --> 00:02:40,640
Speaker 3: How do you use AI to digitize services and serve

46
00:02:40,720 --> 00:02:43,360
Speaker 3: your citizens better? I think if you lead the way

47
00:02:43,880 --> 00:02:46,000
Speaker 3: and then invest in skilling ets, I think there's a

48
00:02:46,040 --> 00:02:48,400
Speaker 3: real chance. So I think it could be more exciting.

49
00:02:48,760 --> 00:02:52,080
Speaker 2: India, your native country, is now the most populous country

50
00:02:52,120 --> 00:02:55,240
Speaker 2: in the world. Do you see AI as a massive

51
00:02:55,360 --> 00:02:58,400
Speaker 2: opportunity for India or could it lead to a massive

52
00:02:58,440 --> 00:03:03,320
Speaker 2: dislocation given the number of lower level coding customer services

53
00:03:03,480 --> 00:03:04,520
Speaker 2: jobs space there.

54
00:03:05,680 --> 00:03:07,800
Speaker 3: I do think it's going to be a massive opportunity,

55
00:03:08,560 --> 00:03:13,560
Speaker 3: you know, across India has India has to make progress

56
00:03:13,560 --> 00:03:17,720
Speaker 3: on many many fundamental areas right, be it healthcare, agriculture,

57
00:03:17,880 --> 00:03:21,960
Speaker 3: food security, climate change, and so on. So an AI

58
00:03:22,120 --> 00:03:24,640
Speaker 3: is going to help us make progress. So it is

59
00:03:24,639 --> 00:03:27,960
Speaker 3: a big opportunity. I think even on the labor market,

60
00:03:28,320 --> 00:03:31,519
Speaker 3: you know a few things. I would say there's a

61
00:03:31,520 --> 00:03:34,480
Speaker 3: study from a mighty economist, David Order, that if you

62
00:03:34,520 --> 00:03:37,800
Speaker 3: look at today, a vast majority of jobs that exist

63
00:03:37,880 --> 00:03:41,800
Speaker 3: today are in new specialties or categories that have come

64
00:03:41,840 --> 00:03:45,600
Speaker 3: about since nineteen forty. So in some ways this can

65
00:03:45,640 --> 00:03:47,560
Speaker 3: be counterint you do. Over the last twenty years, we've

66
00:03:47,560 --> 00:03:51,120
Speaker 3: all worried about automation, and you know, it hasn't quite

67
00:03:51,240 --> 00:03:54,280
Speaker 3: played out the way we have predicted it. Having said that,

68
00:03:54,320 --> 00:03:57,240
Speaker 3: I think your question is important. I think for many

69
00:03:57,360 --> 00:03:59,600
Speaker 3: jobs it also has an opportunity to play it out

70
00:03:59,640 --> 00:04:02,200
Speaker 3: in a month. Whuch better way it make the jobs

71
00:04:02,280 --> 00:04:05,560
Speaker 3: easier to do. Like you know, if you can imagine,

72
00:04:05,560 --> 00:04:10,280
Speaker 3: how do you have radiologists to cover India's breadth of population,

73
00:04:11,000 --> 00:04:15,040
Speaker 3: AI can actually help expand access. But finally, there are

74
00:04:15,080 --> 00:04:17,960
Speaker 3: a category of jobs where there will be a shift,

75
00:04:18,320 --> 00:04:21,440
Speaker 3: and I think that's where the only way to make progress.

76
00:04:21,480 --> 00:04:23,920
Speaker 3: No one can do it alone. You have to invest

77
00:04:23,960 --> 00:04:28,080
Speaker 3: in the hard work of skilling and workforce transition. It's

78
00:04:28,080 --> 00:04:30,960
Speaker 3: true for India, that's true for every country in the world.

79
00:04:30,960 --> 00:04:32,159
Speaker 3: I think you've.

80
00:04:32,000 --> 00:04:35,480
Speaker 2: Met with President Biden on Ai in Washington, Prime Minister

81
00:04:35,600 --> 00:04:38,560
Speaker 2: Sunak on Ai in London. How do we get to

82
00:04:38,680 --> 00:04:42,000
Speaker 2: global consensus on smart AI regulation.

83
00:04:43,960 --> 00:04:46,760
Speaker 3: It's not going to be easy, But I would start

84
00:04:46,800 --> 00:04:52,359
Speaker 3: from this premise that AI will proliferate. So this is

85
00:04:52,400 --> 00:04:56,560
Speaker 3: not the inherent nature of software. AI advances will get

86
00:04:56,560 --> 00:05:01,279
Speaker 3: out to in all countries, so it is naturally the

87
00:05:01,360 --> 00:05:04,960
Speaker 3: kind of technology. I don't think there's any unilateral safety

88
00:05:05,839 --> 00:05:09,880
Speaker 3: to be had. We all have a shared incentive to

89
00:05:09,960 --> 00:05:13,320
Speaker 3: solve for safety. You know, you could have AI go

90
00:05:13,440 --> 00:05:16,839
Speaker 3: wrong in one country that will impact every other country.

91
00:05:17,080 --> 00:05:19,440
Speaker 3: So in some ways, it's like climate change in the planet.

92
00:05:19,440 --> 00:05:22,360
Speaker 3: We all share a planet. I think that's true for AI.

93
00:05:23,160 --> 00:05:26,000
Speaker 3: So now that you know that will be true, I

94
00:05:26,000 --> 00:05:30,400
Speaker 3: think you have to start building the frameworks globally to

95
00:05:30,440 --> 00:05:35,239
Speaker 3: make progress. I've seen encouraging progress when the G seven

96
00:05:35,320 --> 00:05:38,080
Speaker 3: happened in Hiroshima. I think it was a good start.

97
00:05:38,920 --> 00:05:42,320
Speaker 3: You've seen more progress the Uki summit last week. The

98
00:05:42,400 --> 00:05:46,120
Speaker 3: administration here, the White House has been leading the way

99
00:05:46,160 --> 00:05:51,159
Speaker 3: as well, and I saw good, encouraging announcements even yesterday

100
00:05:51,400 --> 00:05:54,919
Speaker 3: for US in China to start having a dialogue on AI.

101
00:05:55,200 --> 00:05:58,080
Speaker 2: Well, that was my next question. Should Chinese regulators be

102
00:05:58,120 --> 00:05:59,960
Speaker 2: part of this conversation on AI regulator?

103
00:06:02,279 --> 00:06:05,480
Speaker 3: You know my senses, there is no way you make

104
00:06:05,560 --> 00:06:09,880
Speaker 3: progress over the long term without China and the US

105
00:06:10,480 --> 00:06:13,359
Speaker 3: deeply talking to each other on something like AI. So

106
00:06:13,400 --> 00:06:15,400
Speaker 3: I think that has got to be an integral part

107
00:06:15,400 --> 00:06:19,240
Speaker 3: of how you make progress, So I think I'm glad

108
00:06:19,279 --> 00:06:21,920
Speaker 3: to see it, and you know, we have to lay

109
00:06:21,960 --> 00:06:24,080
Speaker 3: the foundations. The good thing is we are still in

110
00:06:24,160 --> 00:06:28,839
Speaker 3: early days of the technology, so laying the foundations now

111
00:06:28,920 --> 00:06:31,080
Speaker 3: will allow us to work through the tough issues and

112
00:06:31,080 --> 00:06:32,560
Speaker 3: build a common framework over time.

113
00:06:33,040 --> 00:06:35,839
Speaker 2: Google and other big tech companies have been criticized for

114
00:06:35,880 --> 00:06:39,479
Speaker 2: pushing self serving regulation and pulling up the ladder behind

115
00:06:39,520 --> 00:06:43,760
Speaker 2: them in a way that will stifle AI innovation among startups.

116
00:06:43,760 --> 00:06:44,840
Speaker 2: How do you respond to that?

117
00:06:45,880 --> 00:06:49,400
Speaker 3: Quite the opposite. I mean, pretty much on most areas

118
00:06:49,440 --> 00:06:53,039
Speaker 3: we have worked on, we have deeply supported open innovation.

119
00:06:53,320 --> 00:06:56,040
Speaker 3: You know, if you look at things like Android or Chrome,

120
00:06:56,760 --> 00:06:59,720
Speaker 3: you know, these are all big open source projects. Most

121
00:06:59,760 --> 00:07:03,040
Speaker 3: of the current AI revolution is based on work which

122
00:07:03,040 --> 00:07:05,520
Speaker 3: we've published as a company and shared with the world.

123
00:07:06,920 --> 00:07:10,000
Speaker 3: And I think as we move ahead, open source is

124
00:07:10,040 --> 00:07:13,280
Speaker 3: going to play critical role in driving AI innovation forward.

125
00:07:14,120 --> 00:07:16,400
Speaker 3: And I think we have to be careful not to

126
00:07:16,880 --> 00:07:20,840
Speaker 3: do regulation in a way that you know, either harms

127
00:07:20,880 --> 00:07:24,160
Speaker 3: open source technologies or smaller companies. So I actually think

128
00:07:24,200 --> 00:07:28,600
Speaker 3: we care about that, you know. Having said that, I

129
00:07:28,640 --> 00:07:33,560
Speaker 3: think we're also being asked to contribute for how would

130
00:07:33,600 --> 00:07:37,240
Speaker 3: you tackle AI safety right and so we have to

131
00:07:37,280 --> 00:07:39,200
Speaker 3: think through issues like that. I don't see this as

132
00:07:39,200 --> 00:07:43,320
Speaker 3: being at odds with each other. For example, next year,

133
00:07:43,360 --> 00:07:45,440
Speaker 3: about two and a half billion people around the world

134
00:07:45,480 --> 00:07:47,200
Speaker 3: would participate in an election.

135
00:07:47,320 --> 00:07:50,440
Speaker 2: Right, the biggest election in history is going to be

136
00:07:50,480 --> 00:07:51,000
Speaker 2: happening in.

137
00:07:50,920 --> 00:07:54,720
Speaker 3: India, that's right, and you know, and many more countries

138
00:07:54,720 --> 00:07:58,080
Speaker 3: around the world, including the US, So maybe almost one

139
00:07:58,120 --> 00:08:00,240
Speaker 3: in three people in the world maybe part of paid

140
00:08:00,280 --> 00:08:03,640
Speaker 3: in an electoral process next year. So how do you

141
00:08:03,680 --> 00:08:08,320
Speaker 3: think about you know, generative AI, you know, causing misinformation

142
00:08:08,440 --> 00:08:12,520
Speaker 3: there These are real problems. So we're thinking through areas

143
00:08:12,560 --> 00:08:14,560
Speaker 3: like that. But I think that that is not the

144
00:08:15,720 --> 00:08:19,720
Speaker 3: that shouldn't hinder innovation because of the opportunity, you know,

145
00:08:19,760 --> 00:08:23,000
Speaker 3: opportunity we talked about earlier. So we you know, we

146
00:08:23,000 --> 00:08:25,720
Speaker 3: we think about this framework as being bold and responsible

147
00:08:25,760 --> 00:08:28,440
Speaker 3: at the same time. And I think it's important to

148
00:08:28,480 --> 00:08:30,800
Speaker 3: do both. And I think that's true for governments as well.

149
00:08:31,280 --> 00:08:34,960
Speaker 2: We've already seen, you know, as it pertains to India

150
00:08:35,920 --> 00:08:41,600
Speaker 2: three D generated holograms of Prime Minister Modi AI generated

151
00:08:41,679 --> 00:08:45,720
Speaker 2: voice and voice and songs. How do you think AI,

152
00:08:45,880 --> 00:08:48,400
Speaker 2: and obviously the US presidential election coming up as well,

153
00:08:48,440 --> 00:08:50,240
Speaker 2: how do you think AI is going to further test

154
00:08:50,400 --> 00:08:51,600
Speaker 2: election integrity?

155
00:08:53,679 --> 00:08:56,920
Speaker 3: I think, you know, over time, it's going to lower

156
00:08:57,000 --> 00:09:01,760
Speaker 3: the barrier for creating you know, artificial information which may

157
00:09:01,840 --> 00:09:04,360
Speaker 3: or may not mirror what's happening in the real world, right,

158
00:09:05,040 --> 00:09:08,560
Speaker 3: and that barrier will come down. So in this cat

159
00:09:08,640 --> 00:09:12,160
Speaker 3: and mouse game, how do we amp up our defenses

160
00:09:12,920 --> 00:09:16,800
Speaker 3: against that. We are in early stages, right. You know,

161
00:09:16,800 --> 00:09:19,320
Speaker 3: we were one of the first companies to announce a

162
00:09:19,400 --> 00:09:23,719
Speaker 3: watermarking technology for image generation. It's called Synthide, done by

163
00:09:23,760 --> 00:09:26,520
Speaker 3: deep Mind, and we are providing API access to it.

164
00:09:27,400 --> 00:09:30,880
Speaker 3: But all of us need to tackle it. These are

165
00:09:30,920 --> 00:09:33,760
Speaker 3: areas where regulation will have to play a role, right.

166
00:09:33,800 --> 00:09:37,600
Speaker 3: I think governments will have to over time pass regulations

167
00:09:37,600 --> 00:09:41,199
Speaker 3: about what is okay for you some of this synthetic

168
00:09:41,280 --> 00:09:43,440
Speaker 3: content and so, which is why I think you have

169
00:09:43,440 --> 00:09:45,880
Speaker 3: to think about it, you know, together.

170
00:09:46,320 --> 00:09:48,920
Speaker 2: Open AI CEO Sam Moment has said repeatedly he wants

171
00:09:48,960 --> 00:09:51,240
Speaker 2: to know more about what's happening with AI in China.

172
00:09:51,800 --> 00:09:53,480
Speaker 2: What do you know and what do you not know

173
00:09:53,720 --> 00:09:56,040
Speaker 2: about where China is on AI?

174
00:09:57,880 --> 00:10:01,120
Speaker 3: Okay? I think they are making from what I can tell,

175
00:10:01,160 --> 00:10:05,400
Speaker 3: they're making deep investments in AI. The scale of AI

176
00:10:05,520 --> 00:10:09,000
Speaker 3: research talent in China is just simply astounding to see.

177
00:10:10,000 --> 00:10:13,520
Speaker 3: So I think, you know, in some ways this question,

178
00:10:13,840 --> 00:10:15,880
Speaker 3: China is going to be at the forefront of AI,

179
00:10:16,200 --> 00:10:19,839
Speaker 3: and you know, I think that's a given. And so

180
00:10:19,880 --> 00:10:23,679
Speaker 3: the question is how do we work over time, both

181
00:10:23,800 --> 00:10:26,679
Speaker 3: for you know, other countries to make sure you're making

182
00:10:26,720 --> 00:10:29,480
Speaker 3: progress in AI and over time, how do we develop

183
00:10:29,559 --> 00:10:34,480
Speaker 3: the frameworks where you know, countries can coexist peacefully in

184
00:10:34,480 --> 00:10:37,120
Speaker 3: a world in which AI will be you know, everywhere.

185
00:10:38,120 --> 00:10:41,120
Speaker 2: You know, President Biden actually just said he doesn't see

186
00:10:41,160 --> 00:10:43,760
Speaker 2: the USD coupling with China, but the world does seem

187
00:10:43,760 --> 00:10:47,480
Speaker 2: to be on a path to two separate internets. Do

188
00:10:47,559 --> 00:10:49,880
Speaker 2: we continue in that direction? And what does that mean?

189
00:10:51,160 --> 00:10:54,319
Speaker 3: It's tough to say. You know, things go through in phases.

190
00:10:55,040 --> 00:10:57,079
Speaker 3: I think we are definitely in a phase where there

191
00:10:57,080 --> 00:11:02,240
Speaker 3: are more forces pulling it apart. But you know, inherently

192
00:11:02,320 --> 00:11:08,320
Speaker 3: these technologies also facilitate easy exchange of information, so I

193
00:11:08,320 --> 00:11:10,959
Speaker 3: think there are countervailing forces as well, So I think

194
00:11:11,000 --> 00:11:14,000
Speaker 3: it's tough to predict. I do think information wants to

195
00:11:14,000 --> 00:11:17,880
Speaker 3: flow freely by nature, So you know, my hope is

196
00:11:17,960 --> 00:11:21,320
Speaker 3: over time, you know, things to couple back again.

197
00:11:21,840 --> 00:11:24,800
Speaker 2: Could AI or cloud some of these newer businesses that

198
00:11:24,880 --> 00:11:27,280
Speaker 2: Google has been building, could that be a path back

199
00:11:27,320 --> 00:11:28,720
Speaker 2: into China for Google?

200
00:11:29,600 --> 00:11:32,640
Speaker 3: Yeah? Today we you know, our presence in China is

201
00:11:32,679 --> 00:11:38,560
Speaker 3: limited limited, and we are definitely focused on We deeply

202
00:11:38,600 --> 00:11:41,319
Speaker 3: partner with governments around the world. In fact, one of

203
00:11:41,360 --> 00:11:43,840
Speaker 3: the big opportunities we have with cloud and AI is

204
00:11:44,480 --> 00:11:47,520
Speaker 3: many governments are working and they are thinking about how

205
00:11:47,559 --> 00:11:50,640
Speaker 3: to incorporate AI to transform their services to their citizens,

206
00:11:51,160 --> 00:11:53,480
Speaker 3: improve their infrastructure, et cetera. So it's the AA where

207
00:11:53,480 --> 00:11:57,160
Speaker 3: we really focus on, but not messly visa VI China.

208
00:11:58,160 --> 00:12:01,960
Speaker 2: Now, the conversation around AI is super polarized. Either we're

209
00:12:01,960 --> 00:12:05,400
Speaker 2: headed toward human extinction and the robocalypse or we're all

210
00:12:05,400 --> 00:12:10,320
Speaker 2: going to have superpowers? Which one is it? You know?

211
00:12:11,120 --> 00:12:13,920
Speaker 3: I am I am optimistic, right. I think as humanity,

212
00:12:13,960 --> 00:12:18,320
Speaker 3: we've harnessed every technology to our benefits, you know, and

213
00:12:19,280 --> 00:12:22,320
Speaker 3: more than any other technology, I've seen us be worried

214
00:12:22,360 --> 00:12:26,120
Speaker 3: about it at its earliest stages. So in some ways

215
00:12:26,120 --> 00:12:29,320
Speaker 3: that gives me hope. I do think we have to

216
00:12:29,360 --> 00:12:33,760
Speaker 3: be very optimistic because it can really you know, drive progress,

217
00:12:34,200 --> 00:12:37,880
Speaker 3: you know, at a fundamental level, be its scientific discovery,

218
00:12:37,960 --> 00:12:39,880
Speaker 3: like we are seeing with the alpha fold at Google.

219
00:12:40,679 --> 00:12:43,440
Speaker 3: So I think there's a lot of optimism to be had,

220
00:12:44,600 --> 00:12:46,800
Speaker 3: but I think we have to work hard to harness it.

221
00:12:47,360 --> 00:12:50,160
Speaker 3: But that is true of every other technological advance we've

222
00:12:50,160 --> 00:12:53,480
Speaker 3: had before. It was true of the Industrial Revolution, and

223
00:12:53,520 --> 00:12:56,120
Speaker 3: I think we can learn from those things. Act earlier,

224
00:12:56,760 --> 00:12:59,400
Speaker 3: to work hard making sure there's a better outcome. Some

225
00:12:59,600 --> 00:13:00,800
Speaker 3: optimist stick it.

226
00:13:01,360 --> 00:13:04,760
Speaker 2: You pivoted Google to an AI first company in twenty sixteen,

227
00:13:04,960 --> 00:13:07,640
Speaker 2: and yet there's still this perception that Google is somehow

228
00:13:07,640 --> 00:13:10,920
Speaker 2: missing the boat on AI. Are you behind? Are you

229
00:13:11,000 --> 00:13:12,080
Speaker 2: behind Open Ai?

230
00:13:13,400 --> 00:13:16,280
Speaker 3: You know, look coming credit to them for I think

231
00:13:16,960 --> 00:13:20,400
Speaker 3: Chatchipt was a great product with a great product market fit.

232
00:13:21,400 --> 00:13:25,240
Speaker 3: But when I look at me, we've been incorporating AI

233
00:13:25,280 --> 00:13:29,800
Speaker 3: in our products for a long time. A lot of

234
00:13:29,800 --> 00:13:32,880
Speaker 3: the underlying technology is built from Google. We are building

235
00:13:32,880 --> 00:13:36,839
Speaker 3: our next generation of models with Gemini, and I am

236
00:13:37,000 --> 00:13:41,480
Speaker 3: extraordinarily excited at the innovation coming ahead. If anything, I

237
00:13:41,520 --> 00:13:44,880
Speaker 3: look ahead, I think we build a company for this moment.

238
00:13:45,200 --> 00:13:47,640
Speaker 3: I expected to be a golden age of innovation ahead

239
00:13:48,400 --> 00:13:51,600
Speaker 3: and can wait to bring out the innovations to more people.

240
00:13:51,720 --> 00:13:54,000
Speaker 2: So tell us more about Gemini, give us a glimpse.

241
00:13:54,000 --> 00:13:55,960
Speaker 2: I mean, this is something that's been touted as a

242
00:13:56,000 --> 00:13:59,679
Speaker 2: magic ingredient for search. What have you seen in how

243
00:13:59,720 --> 00:14:01,280
Speaker 2: trans formative ook.

244
00:14:01,360 --> 00:14:03,680
Speaker 3: Going to be? I think all of us are trying

245
00:14:03,720 --> 00:14:05,680
Speaker 3: to push the state of the art of r generatio AI.

246
00:14:05,920 --> 00:14:09,840
Speaker 3: Is our goal with Gemini, you know, is to put

247
00:14:09,840 --> 00:14:12,480
Speaker 3: out a state of the art model, right, That's where

248
00:14:12,480 --> 00:14:15,280
Speaker 3: we would start with Gemini one point zero and then

249
00:14:15,440 --> 00:14:20,080
Speaker 3: add more innovations, you know, truly make the multimodel, bring

250
00:14:20,120 --> 00:14:22,280
Speaker 3: in features like memory and planning in and so on.

251
00:14:22,840 --> 00:14:24,920
Speaker 3: But we have focused on getting it out, you know,

252
00:14:24,960 --> 00:14:27,800
Speaker 3: I viewed Gemini as a series of models. We are

253
00:14:27,800 --> 00:14:30,120
Speaker 3: focused on getting Gemini one point zero out as soon

254
00:14:30,160 --> 00:14:33,200
Speaker 3: as possible, make sure it's competitive, state of the art,

255
00:14:33,760 --> 00:14:34,880
Speaker 3: and build from there on.

256
00:14:35,880 --> 00:14:39,560
Speaker 2: Google just invested two billion dollars in anthropic Microsoft, of course,

257
00:14:39,560 --> 00:14:42,680
Speaker 2: has billions in open AI. I recently sat down with

258
00:14:42,760 --> 00:14:45,440
Speaker 2: Lena Khan, who's the chair of the US Federal Trade Commission,

259
00:14:45,440 --> 00:14:48,680
Speaker 2: and she said they're hearing concerns that big tech companies

260
00:14:48,720 --> 00:14:52,600
Speaker 2: are extending their power by investing not just money, but

261
00:14:52,680 --> 00:14:56,360
Speaker 2: also sharing their clouds with these AI startups. Is that

262
00:14:56,440 --> 00:15:01,760
Speaker 2: fair can innovator succeed and start up succeed without allying

263
00:15:01,800 --> 00:15:02,760
Speaker 2: themselves with big tech.

264
00:15:03,920 --> 00:15:06,880
Speaker 3: Well, you know, I mean it all depens on the details.

265
00:15:06,920 --> 00:15:09,000
Speaker 3: And you know, in these deals when we do with

266
00:15:09,040 --> 00:15:12,200
Speaker 3: these companies, we are also enabling them because of our

267
00:15:12,240 --> 00:15:16,200
Speaker 3: cloud technology. And we you know, we don't have a

268
00:15:16,240 --> 00:15:18,480
Speaker 3: controlling stake in any of these companies, so they are

269
00:15:18,520 --> 00:15:23,000
Speaker 3: independent companies in which we are a technology provider, right,

270
00:15:23,080 --> 00:15:26,760
Speaker 3: And I think I actually would make the argument the

271
00:15:26,800 --> 00:15:28,760
Speaker 3: opposite way if you just look at the last year,

272
00:15:28,800 --> 00:15:30,760
Speaker 3: look at all the new names we are talking about.

273
00:15:31,400 --> 00:15:33,680
Speaker 3: To me, it shows that we are in an incredibly

274
00:15:33,760 --> 00:15:38,040
Speaker 3: dynamic moment again, you know, and so and yes, we

275
00:15:38,520 --> 00:15:40,520
Speaker 3: I think we have to play a role in enabling

276
00:15:40,560 --> 00:15:43,080
Speaker 3: the next generation of companies, which is what Cloud does.

277
00:15:44,000 --> 00:15:46,800
Speaker 3: Cloud allows us to take the same technology that Google

278
00:15:46,880 --> 00:15:49,640
Speaker 3: is built on and actually share it with everyone else.

279
00:15:50,160 --> 00:15:52,640
Speaker 3: So I think, if anything, it's pro innovation, and so

280
00:15:52,720 --> 00:15:54,280
Speaker 3: I think there's a lot of good things about it.

281
00:15:54,800 --> 00:15:57,560
Speaker 2: You've been spending some time in Washington lately. There are

282
00:15:57,600 --> 00:16:00,880
Speaker 2: a couple of big trials underway. The US Justice Department,

283
00:16:00,920 --> 00:16:03,280
Speaker 2: for one, is trying to prove that Google is a monopoly,

284
00:16:03,400 --> 00:16:07,760
Speaker 2: a search monopoly. Google does dominate ninety percent of search,

285
00:16:08,600 --> 00:16:10,920
Speaker 2: which is ninety percent of like how we see and

286
00:16:11,000 --> 00:16:13,400
Speaker 2: experience in the world. It's hard to get information and

287
00:16:13,440 --> 00:16:17,400
Speaker 2: not use Google. Why should any one company have that

288
00:16:17,520 --> 00:16:18,200
Speaker 2: much power?

289
00:16:19,160 --> 00:16:22,960
Speaker 3: You know, I genuinely think we don't have that position.

290
00:16:23,720 --> 00:16:26,440
Speaker 3: We are not ninety percent of users information needs. You know,

291
00:16:26,480 --> 00:16:30,280
Speaker 3: I look at my kids generation on their phone. They

292
00:16:30,360 --> 00:16:34,000
Speaker 3: have access to the entire world's information at their fingertips.

293
00:16:34,400 --> 00:16:36,760
Speaker 3: They're getting a lot of their information from social media.

294
00:16:36,880 --> 00:16:39,960
Speaker 3: They get to go to wherever they want, and so

295
00:16:40,600 --> 00:16:43,080
Speaker 3: you know, those are you know, people don't always go

296
00:16:43,160 --> 00:16:46,520
Speaker 3: to a general search engine for information, and that plays

297
00:16:46,520 --> 00:16:49,880
Speaker 3: out in the ads market too. I think we've just

298
00:16:49,920 --> 00:16:52,560
Speaker 3: talked about when I look at how much tech is

299
00:16:52,600 --> 00:16:55,840
Speaker 3: constantly innovating, and we have to work hard every year.

300
00:16:56,920 --> 00:16:59,720
Speaker 3: You know, there's irony in your questions of aren't you

301
00:16:59,720 --> 00:17:02,680
Speaker 3: in this position? And R and behind I mean ai,

302
00:17:02,720 --> 00:17:04,840
Speaker 3: I mean so both can be true at the same time.

303
00:17:05,560 --> 00:17:08,240
Speaker 3: So it's clearly a dynamic moment, and I'm glad we

304
00:17:08,280 --> 00:17:10,920
Speaker 3: get a chance to make that case. And so I

305
00:17:11,000 --> 00:17:12,880
Speaker 3: think the facts will speak for themselves.

306
00:17:14,040 --> 00:17:16,800
Speaker 2: Some of Google's work with governments around the world has

307
00:17:16,840 --> 00:17:19,600
Speaker 2: gotten pushed back from your own employees. Obviously, we've got

308
00:17:19,640 --> 00:17:23,160
Speaker 2: the world watching the Israel Hamas war right now. You've

309
00:17:23,160 --> 00:17:26,840
Speaker 2: got a contract project Nimbus, where by Google provides cloud

310
00:17:26,880 --> 00:17:30,120
Speaker 2: and AI services along with Amazon's Aws to the Israeli

311
00:17:30,160 --> 00:17:32,879
Speaker 2: military and government. Can you give us an update on

312
00:17:32,920 --> 00:17:33,560
Speaker 2: that project.

313
00:17:34,520 --> 00:17:37,440
Speaker 3: It's very much in line with what I said earlier.

314
00:17:37,520 --> 00:17:40,320
Speaker 3: We work with governments around the world. You know, we

315
00:17:40,359 --> 00:17:44,200
Speaker 3: think it's a responsibility. AI is an important technology. How

316
00:17:44,200 --> 00:17:48,000
Speaker 3: do you use it to modernize your country's infrastructure and

317
00:17:48,040 --> 00:17:53,440
Speaker 3: your services? Project Nimbers is this was an RFP from

318
00:17:53,560 --> 00:17:58,840
Speaker 3: Israel's Ministry of Finance to modernize their digital infrastructure. And

319
00:17:58,920 --> 00:18:02,280
Speaker 3: that's the project. And proud to be doing project numbers

320
00:18:02,320 --> 00:18:04,160
Speaker 3: like we do with many governments around the world.

321
00:18:05,520 --> 00:18:09,000
Speaker 2: There was you know, Russia, Ukraine, Now we've got Israel

322
00:18:09,040 --> 00:18:12,920
Speaker 2: and Hamas. How do you think about wielding Google's geopolitical

323
00:18:12,960 --> 00:18:14,400
Speaker 2: power in a time of conflict?

324
00:18:14,960 --> 00:18:18,159
Speaker 3: Okay? I mean I view us as a partner to

325
00:18:19,240 --> 00:18:22,840
Speaker 3: you know, like minded governments which share democratic values around

326
00:18:22,840 --> 00:18:27,040
Speaker 3: the world. I think we can be a critical technology partner.

327
00:18:27,720 --> 00:18:30,919
Speaker 3: I think we want to participate in important issues that

328
00:18:30,960 --> 00:18:35,200
Speaker 3: affect these countries, be its skilling and educating their workforce.

329
00:18:35,880 --> 00:18:38,439
Speaker 3: Be it bringing access to more knowledge and information, and

330
00:18:38,480 --> 00:18:40,399
Speaker 3: that's the role. And be it helping them build out

331
00:18:40,400 --> 00:18:45,640
Speaker 3: their digital infrastructure, including AI. And I think that's the role.

332
00:18:46,400 --> 00:18:48,639
Speaker 3: You know, we don't see it in a geopolitical context.

333
00:18:49,080 --> 00:18:51,399
Speaker 3: We see it in an enabling context. We want to

334
00:18:51,400 --> 00:18:54,680
Speaker 3: be partners to these companies and you know, and there

335
00:18:54,680 --> 00:18:57,680
Speaker 3: are times information plays an important role in these moments

336
00:18:58,040 --> 00:19:01,159
Speaker 3: and we want to get those moments right. And so

337
00:19:01,240 --> 00:19:01,920
Speaker 3: that's the way I.

338
00:19:01,880 --> 00:19:05,400
Speaker 2: Think about it Google, and so much not every government

339
00:19:05,440 --> 00:19:07,600
Speaker 2: is like minded, right Google, and so much of US

340
00:19:07,680 --> 00:19:11,200
Speaker 2: tech relies on Taiwan for chips. If tensions escalate between

341
00:19:11,200 --> 00:19:14,040
Speaker 2: Taiwan and China, how big a threat is that to

342
00:19:14,080 --> 00:19:17,120
Speaker 2: the US tech industry. It is enough being done to

343
00:19:17,200 --> 00:19:19,359
Speaker 2: mitigate that risk.

344
00:19:19,400 --> 00:19:23,000
Speaker 3: This goes beyond Google. I think in today's world, a

345
00:19:23,040 --> 00:19:27,640
Speaker 3: lot of the semiconductive manufacturing technology, you know, comes out

346
00:19:27,680 --> 00:19:30,399
Speaker 3: of the innovation from Taiwan. And you know that's a

347
00:19:30,440 --> 00:19:34,440
Speaker 3: dipendency which exists. And so you know, but this is

348
00:19:34,760 --> 00:19:37,480
Speaker 3: not a unique Google thing, and you know, my senses

349
00:19:37,520 --> 00:19:39,120
Speaker 3: it will be that way for a long time to come.

350
00:19:40,160 --> 00:19:43,680
Speaker 2: Eight years ago, you changed Google's motto Google changed its

351
00:19:43,680 --> 00:19:45,960
Speaker 2: motto from don't be evil to do the right thing.

352
00:19:46,760 --> 00:19:49,159
Speaker 2: What does it mean to do the right thing in

353
00:19:49,200 --> 00:19:51,320
Speaker 2: an AI powered world?

354
00:19:51,960 --> 00:19:54,960
Speaker 3: Hopefully AI will also help us, you know, give inputs

355
00:19:54,960 --> 00:19:56,520
Speaker 3: to it. But look, I think it has to be

356
00:19:56,600 --> 00:20:02,399
Speaker 3: grounded in the fundament the values of humanity, you know,

357
00:20:02,560 --> 00:20:06,600
Speaker 3: human rights and universal human values we all agree on

358
00:20:06,720 --> 00:20:09,639
Speaker 3: has got to be the foundation for it. And you

359
00:20:09,680 --> 00:20:11,880
Speaker 3: know you have to build it up from there, and

360
00:20:12,040 --> 00:20:13,919
Speaker 3: there'll be a lot of debate about it, but you know,

361
00:20:13,960 --> 00:20:16,760
Speaker 3: that's the foundational thing I would go back to. And

362
00:20:17,040 --> 00:20:20,000
Speaker 3: you know, as long as AI reflects the essence of

363
00:20:20,119 --> 00:20:23,640
Speaker 3: what's good about innately good about humanity, I think will

364
00:20:23,640 --> 00:20:24,480
Speaker 3: be will be Okay.

365
00:20:24,760 --> 00:20:26,880
Speaker 2: Everyone likes to ask you what keeps you up at night?

366
00:20:26,960 --> 00:20:31,040
Speaker 2: And I'm curious right now this year, what what do

367
00:20:31,080 --> 00:20:33,320
Speaker 2: you worry about you about?

368
00:20:33,640 --> 00:20:37,560
Speaker 3: Look, I'm excited because it's an extraordinary moment of innovation.

369
00:20:39,040 --> 00:20:42,080
Speaker 3: You know, when I look at the pace of activity,

370
00:20:42,920 --> 00:20:44,760
Speaker 3: you know, even at Google, it reminds me of Google's

371
00:20:44,760 --> 00:20:48,480
Speaker 3: earliest days. So so there's a lot of energy I

372
00:20:48,480 --> 00:20:52,359
Speaker 3: get from that, you know, innovation, because you know it

373
00:20:52,400 --> 00:20:54,320
Speaker 3: plays out and it will drive benefits in the world.

374
00:20:55,880 --> 00:20:58,199
Speaker 3: I do worry about many things, making sure we are

375
00:20:58,240 --> 00:21:00,760
Speaker 3: meeting the moment and moving fast as a company to

376
00:21:02,359 --> 00:21:06,159
Speaker 3: making sure today election question, we talked about getting it

377
00:21:06,280 --> 00:21:08,920
Speaker 3: right and the responsibility to do so. So it's a

378
00:21:09,000 --> 00:21:10,960
Speaker 3: it's a balance of all of that, all right.

379
00:21:11,080 --> 00:21:14,000
Speaker 2: Sindhar Pachai, CEO of Google, thank you for spending your

380
00:21:14,040 --> 00:21:14,600
Speaker 2: time with us

381
00:21:14,720 --> 00:21:16,280
Speaker 3: To thank you, thank you, thank you,