1
00:00:02,400 --> 00:00:09,440
Speaker 1: Bloomberg Audio Studios, podcasts, radio news.

2
00:00:08,400 --> 00:00:13,680
Speaker 2: From Mahard where Innovation, money and power Collie in Silicon Valley, NBN.

3
00:00:14,000 --> 00:00:18,520
Speaker 3: This is Bloomberg Technology with Caroline Hyde and Ed Ludlove.

4
00:00:32,200 --> 00:00:34,959
Speaker 4: Live from New York. This is Bloomberg Technology. Coming up.

5
00:00:35,040 --> 00:00:37,760
Speaker 5: Will Google be forced to sell off its Chrome browser?

6
00:00:38,040 --> 00:00:41,080
Speaker 5: We discussed the latest in the historic Alphabet anti trust

7
00:00:41,080 --> 00:00:45,080
Speaker 5: fight here in the United States, plus roadblocks enhances parental

8
00:00:45,159 --> 00:00:48,839
Speaker 5: controls following criticism over child protection on its platform. We

9
00:00:48,880 --> 00:00:52,400
Speaker 5: sit down with the company's chief safety officer and SpaceX

10
00:00:52,479 --> 00:00:55,800
Speaker 5: ready for another starship launch with President elect Donald Trump

11
00:00:56,200 --> 00:00:57,040
Speaker 5: set to attend.

12
00:00:57,560 --> 00:00:58,120
Speaker 4: At first.

13
00:00:58,440 --> 00:01:02,600
Speaker 5: We check in on these markets despite the geopolitical headwinds Russia, Ukraine.

14
00:01:02,640 --> 00:01:05,399
Speaker 5: We push higher just about on tenth of a percent,

15
00:01:05,480 --> 00:01:07,600
Speaker 5: let's call it on the Nazak, one hundred higher still

16
00:01:07,840 --> 00:01:09,880
Speaker 5: on the Nasdaq. That's two straight days of gains when

17
00:01:09,880 --> 00:01:12,920
Speaker 5: you're looking at big tech, and that more broadly is

18
00:01:12,959 --> 00:01:14,840
Speaker 5: as we shake off what have been a big points

19
00:01:14,920 --> 00:01:17,319
Speaker 5: drag on the day, which was Alphabet. Let's just go

20
00:01:17,360 --> 00:01:19,640
Speaker 5: into the intricacies of what's happening with Google at the

21
00:01:19,640 --> 00:01:22,600
Speaker 5: moment in terms of its price point, because we've started

22
00:01:22,600 --> 00:01:26,080
Speaker 5: off the day trading lower, anxiety around the building, anti

23
00:01:26,120 --> 00:01:29,319
Speaker 5: trust crescendo around this particular company, and then maybe we

24
00:01:29,400 --> 00:01:32,080
Speaker 5: push into the green up four tenths of percent. Is

25
00:01:32,120 --> 00:01:34,240
Speaker 5: there a slight relief rally that it's not being forced

26
00:01:34,240 --> 00:01:35,040
Speaker 5: to sell off Android?

27
00:01:35,040 --> 00:01:37,720
Speaker 4: We get into it now, were Bloomberg's Michael Shepherd in Washington,

28
00:01:37,760 --> 00:01:40,640
Speaker 4: because it's an interesting stock market reaction here.

29
00:01:41,040 --> 00:01:43,679
Speaker 5: Michael Many would say, Look, the clouds of anti trust

30
00:01:43,720 --> 00:01:47,040
Speaker 5: concerns still sit on the valuation of Google, but get

31
00:01:47,080 --> 00:01:49,800
Speaker 5: into what we are learning in terms of the DOJ

32
00:01:50,200 --> 00:01:53,200
Speaker 5: looking to treat an anti trust issue from their perspective.

33
00:01:54,600 --> 00:01:58,000
Speaker 2: Well, first credit to our colleagues Leonilan and Josh Cisco

34
00:01:58,120 --> 00:02:01,640
Speaker 2: for breaking the news about what the Justice Department and

35
00:02:01,840 --> 00:02:06,560
Speaker 2: state anti trust regulators are thinking about asking a judge

36
00:02:06,880 --> 00:02:10,720
Speaker 2: in the way of remedies in this historic case. This

37
00:02:10,880 --> 00:02:14,560
Speaker 2: dates to the August ruling in which judge here in

38
00:02:14,680 --> 00:02:19,720
Speaker 2: Washington found that Google had illegally monopolized the search the

39
00:02:19,760 --> 00:02:23,919
Speaker 2: market for search globally, and now they are looking for

40
00:02:24,000 --> 00:02:27,480
Speaker 2: how to make this right for consumers and for website

41
00:02:27,520 --> 00:02:31,799
Speaker 2: operators around around the world. The one step, of course,

42
00:02:31,919 --> 00:02:35,440
Speaker 2: being considered is forcing them to sell off Chrome the browser,

43
00:02:35,560 --> 00:02:38,440
Speaker 2: it's one of the world's most widely used browsers. But

44
00:02:38,560 --> 00:02:43,000
Speaker 2: other measures being considered also include requiring the company to

45
00:02:43,160 --> 00:02:46,400
Speaker 2: license its search results in data, and then also giving

46
00:02:46,520 --> 00:02:50,440
Speaker 2: websites the option of opting out of having their content

47
00:02:50,680 --> 00:02:54,320
Speaker 2: used to help Google build its artificial intelligence platforms.

48
00:02:54,880 --> 00:02:59,000
Speaker 5: Once again, Google Alphabet, the power company of Google, speaks

49
00:02:59,040 --> 00:03:03,360
Speaker 5: out saying thesis of radical steps being taken by the

50
00:03:03,480 --> 00:03:06,840
Speaker 5: US government and sort of waiting down, putting their thumb

51
00:03:06,840 --> 00:03:09,880
Speaker 5: on the scale is how Google has been articulating it.

52
00:03:09,919 --> 00:03:14,000
Speaker 5: But more broadly, the intricacies around selling off data, about

53
00:03:14,040 --> 00:03:15,760
Speaker 5: sharing data is a tough one.

54
00:03:15,960 --> 00:03:17,040
Speaker 4: Many in the market.

55
00:03:16,800 --> 00:03:19,680
Speaker 5: Say, look, a, this is going to take years to enforce.

56
00:03:19,760 --> 00:03:22,120
Speaker 5: You're going to see an appeals process, but more broadly,

57
00:03:22,440 --> 00:03:25,560
Speaker 5: you're also going to potentially see other remedies come into

58
00:03:25,720 --> 00:03:28,640
Speaker 5: like that. Google could action first, perhaps April and August.

59
00:03:28,960 --> 00:03:31,280
Speaker 5: What do we think Alphabet's own reaction will be.

60
00:03:33,280 --> 00:03:35,000
Speaker 2: This will play out of for a long time, And

61
00:03:35,040 --> 00:03:37,880
Speaker 2: I'm glad you've brought that up, Caroline. In terms of

62
00:03:37,920 --> 00:03:41,840
Speaker 2: the clock here, the judges said a two week hearing

63
00:03:42,000 --> 00:03:45,240
Speaker 2: in April to actually walk through some of the potential

64
00:03:45,280 --> 00:03:48,640
Speaker 2: remedies that the DOJ and anti trust regulators at the

65
00:03:48,680 --> 00:03:51,840
Speaker 2: state level are offering, and that will give the time

66
00:03:52,280 --> 00:03:54,960
Speaker 2: that we'll give time to the company to think about, Okay,

67
00:03:55,240 --> 00:03:58,240
Speaker 2: what could we offer perhaps as a settlement, what else

68
00:03:58,320 --> 00:04:01,400
Speaker 2: can we do? What is is maybe a middle ground

69
00:04:01,480 --> 00:04:04,360
Speaker 2: that could satisfy some of the concerns being or raised

70
00:04:04,360 --> 00:04:08,880
Speaker 2: by government authorities. Likewise, the judge himself is not expected

71
00:04:08,920 --> 00:04:12,440
Speaker 2: to rule until August twenty twenty five, that is almost

72
00:04:12,480 --> 00:04:15,280
Speaker 2: a year from his original ruling, and then after that

73
00:04:15,360 --> 00:04:19,200
Speaker 2: we'll go through an appeals process, and given the stakes

74
00:04:19,279 --> 00:04:23,280
Speaker 2: for Alphabet in this case, they're likely to take this

75
00:04:23,440 --> 00:04:26,159
Speaker 2: appeal all the way to the US Supreme Court eventually,

76
00:04:26,600 --> 00:04:27,080
Speaker 2: because this.

77
00:04:27,080 --> 00:04:31,039
Speaker 5: Would have a major business impact the way in which

78
00:04:31,120 --> 00:04:33,760
Speaker 5: they can see who I am when I use Chrome,

79
00:04:33,839 --> 00:04:35,680
Speaker 5: which has I think about sixty percent of the market

80
00:04:35,760 --> 00:04:38,400
Speaker 5: from a browser perspective, and then go across to use

81
00:04:38,480 --> 00:04:40,880
Speaker 5: the search function within it that's about ninety percent of

82
00:04:40,920 --> 00:04:43,160
Speaker 5: the market and all the data they're in and they

83
00:04:43,160 --> 00:04:44,920
Speaker 5: can target me better with ads. This will be a

84
00:04:44,960 --> 00:04:48,880
Speaker 5: big business model change, and ultimately many out there are wondering.

85
00:04:48,560 --> 00:04:50,960
Speaker 4: Who on earth would even buy Chrome if it was forced?

86
00:04:52,240 --> 00:04:54,320
Speaker 2: Well, it's a great question. Again, and who would buy

87
00:04:54,400 --> 00:04:59,400
Speaker 2: Chrome because the price for it would preclude many from

88
00:04:59,440 --> 00:05:02,120
Speaker 2: even to pick it up, and a few buyers out

89
00:05:02,120 --> 00:05:05,080
Speaker 2: there with the resources to pull it off, perhaps Amazon

90
00:05:05,160 --> 00:05:09,200
Speaker 2: dot Com Inc. Or even Open Ai. An attempt by

91
00:05:09,200 --> 00:05:12,960
Speaker 2: them to purchase this could also raise separate antitrust questions

92
00:05:13,320 --> 00:05:16,960
Speaker 2: that the government may be eager to avoid. Likewise, as

93
00:05:17,000 --> 00:05:20,400
Speaker 2: you pointed out, this is central to Alphabet's business. They

94
00:05:20,680 --> 00:05:23,480
Speaker 2: use that user data, as you pointed out, to target

95
00:05:23,520 --> 00:05:27,440
Speaker 2: advertising and promotions better, and they're also using it, as

96
00:05:27,480 --> 00:05:30,239
Speaker 2: I mentioned before, to try to craft a better version

97
00:05:30,240 --> 00:05:34,520
Speaker 2: of this Gemini AI platform that they're developing, and we

98
00:05:34,600 --> 00:05:37,400
Speaker 2: saw it in their most recent results. They really pointed

99
00:05:37,480 --> 00:05:40,920
Speaker 2: to search as something that is driving this bet on AI,

100
00:05:41,400 --> 00:05:43,480
Speaker 2: and they really would like to keep this going if

101
00:05:43,480 --> 00:05:44,039
Speaker 2: they can.

102
00:05:44,200 --> 00:05:47,120
Speaker 5: AI overviews all anyone sees now when they're typing in

103
00:05:47,120 --> 00:05:47,480
Speaker 5: for search.

104
00:05:47,520 --> 00:05:48,520
Speaker 4: Michael Shepherd, we.

105
00:05:48,480 --> 00:05:51,599
Speaker 5: So appreciate the breakdown from Washington for broader tech market

106
00:05:51,600 --> 00:05:55,520
Speaker 5: prospective and indeed, well what the legal realities are here.

107
00:05:55,640 --> 00:05:58,839
Speaker 5: Sarah Lamb is a senior fellow at the Tech Institute,

108
00:05:58,920 --> 00:06:02,400
Speaker 5: and Sarah just put your thumb on the scales here.

109
00:06:02,520 --> 00:06:05,679
Speaker 5: As Google is pushing back on the government for so doing.

110
00:06:06,160 --> 00:06:08,520
Speaker 5: Do you think this has any chance of getting through

111
00:06:08,560 --> 00:06:09,800
Speaker 5: that something is seismic?

112
00:06:09,960 --> 00:06:11,760
Speaker 4: Is Chrome being sold off would happen?

113
00:06:13,400 --> 00:06:18,440
Speaker 3: I think personally it's unlikely. It was unchallenged conduct in

114
00:06:18,600 --> 00:06:21,520
Speaker 3: the decision. So if you go back and read the

115
00:06:21,560 --> 00:06:24,960
Speaker 3: two hundred and eighty six page decision from August fifth.

116
00:06:25,240 --> 00:06:28,240
Speaker 4: There is good for us because we know how to talk.

117
00:06:29,720 --> 00:06:34,040
Speaker 3: Yeah, there's discussion about the contracts and the pre loading

118
00:06:34,160 --> 00:06:37,320
Speaker 3: of the Google Suite, and a little bit on the

119
00:06:37,440 --> 00:06:43,159
Speaker 3: Chrome browser. But unchallenged conduct was the search bar in

120
00:06:43,279 --> 00:06:47,840
Speaker 3: the Chrome window, and so that wasn't challenged at trial,

121
00:06:48,080 --> 00:06:51,160
Speaker 3: and so you know, there would have to be fact

122
00:06:51,240 --> 00:06:54,600
Speaker 3: finding and a discussion of the browser market. In the

123
00:06:54,640 --> 00:07:00,000
Speaker 3: April hearing, like the prior speaker said, August twenty two,

124
00:07:00,000 --> 00:07:03,000
Speaker 3: twenty five, is when a decision will come out on remedies.

125
00:07:03,720 --> 00:07:07,359
Speaker 3: And also of note, Google has an opportunity to submit

126
00:07:07,480 --> 00:07:11,680
Speaker 3: their proposed remedies December twentieth, So it really is an

127
00:07:11,760 --> 00:07:16,600
Speaker 3: ongoing process. The last two big cases, Microsoft and AT

128
00:07:16,680 --> 00:07:19,720
Speaker 3: and T took nearly a decade to resolve.

129
00:07:19,880 --> 00:07:24,880
Speaker 5: And Microsoft appealed successfully at and T less. So, Sarah,

130
00:07:25,400 --> 00:07:28,920
Speaker 5: do you therefore think some other changes could be made

131
00:07:29,000 --> 00:07:33,120
Speaker 5: to fight the idea that Google is monopoly or monopolistic

132
00:07:33,640 --> 00:07:36,200
Speaker 5: in search. What could Google enact here itself.

133
00:07:38,240 --> 00:07:41,880
Speaker 3: Well, I think at issue are the contracts, so the

134
00:07:41,920 --> 00:07:47,760
Speaker 3: way they contract with Samsung or the other OEM providers,

135
00:07:48,160 --> 00:07:52,760
Speaker 3: that could change. But even so, there's so much dynamism

136
00:07:52,840 --> 00:07:56,560
Speaker 3: in the market. Samsung uploads its own s browser and

137
00:07:56,640 --> 00:08:01,560
Speaker 3: I read that in the decision. Other Android platforms they

138
00:08:01,640 --> 00:08:05,200
Speaker 3: have their own browsers, and so I think right now

139
00:08:05,240 --> 00:08:11,000
Speaker 3: the market is so dynamic too with AI search windows

140
00:08:11,120 --> 00:08:16,360
Speaker 3: that you know, the way Google responds will have to

141
00:08:17,320 --> 00:08:19,560
Speaker 3: reflect what's happening in the AI market.

142
00:08:20,080 --> 00:08:24,080
Speaker 5: Let's just talk about AI and the data that maybe

143
00:08:24,120 --> 00:08:27,160
Speaker 5: they'd be forced to license in some way. Would that

144
00:08:27,200 --> 00:08:30,400
Speaker 5: be positive or negative for the rest of the market, not.

145
00:08:30,280 --> 00:08:34,680
Speaker 3: Just Google, right, I think operationally that would be very

146
00:08:34,720 --> 00:08:37,960
Speaker 3: difficult to work through, so that would also be a

147
00:08:38,040 --> 00:08:43,079
Speaker 3: year's long process. I don't know how great lawyers are

148
00:08:43,200 --> 00:08:47,160
Speaker 3: at product design. So to the extent that I mean,

149
00:08:47,480 --> 00:08:50,800
Speaker 3: there is some talk that there might have frand f

150
00:08:50,960 --> 00:08:55,119
Speaker 3: r A and D licensing terms in the proposed remedies,

151
00:08:55,160 --> 00:08:59,280
Speaker 3: and that is something that is familiar in the intellectual

152
00:08:59,360 --> 00:09:04,080
Speaker 3: property and standards development world. But to the level of

153
00:09:04,280 --> 00:09:08,640
Speaker 3: data sharing and you know, opening that pipeline of AI

154
00:09:09,160 --> 00:09:12,760
Speaker 3: data that would be a very new type of remedy

155
00:09:13,160 --> 00:09:13,559
Speaker 3: that we.

156
00:09:13,520 --> 00:09:15,160
Speaker 5: Can see from the share price that the market isn't

157
00:09:15,160 --> 00:09:17,320
Speaker 5: freaking out about this, but there does seem to be

158
00:09:17,320 --> 00:09:19,920
Speaker 5: a cloud over Google for the next few years, maybe

159
00:09:19,920 --> 00:09:22,880
Speaker 5: even decade. As you were saying, Sarah, what about an

160
00:09:22,920 --> 00:09:26,480
Speaker 5: administration change in twenty twenty five, does that bear any

161
00:09:26,720 --> 00:09:27,360
Speaker 5: thought for you?

162
00:09:28,640 --> 00:09:28,880
Speaker 1: Yes.

163
00:09:28,960 --> 00:09:32,880
Speaker 3: What's interesting though, is the first Trump administration brought the suit,

164
00:09:33,360 --> 00:09:37,439
Speaker 3: so you have that, although Assistant ag make and del

165
00:09:37,520 --> 00:09:43,040
Speaker 3: Rahim recused himself, so this case has a long history.

166
00:09:43,080 --> 00:09:46,440
Speaker 3: So far, you hear kind of murmurs about the next

167
00:09:46,480 --> 00:09:50,480
Speaker 3: administration wanting to continue activity.

168
00:09:50,559 --> 00:09:51,559
Speaker 4: But at the.

169
00:09:51,520 --> 00:09:55,120
Speaker 3: Same time, you know, they seem to be more pro

170
00:09:56,520 --> 00:10:02,560
Speaker 3: innovation and technology. So we'll see what whoever is Attorney

171
00:10:02,559 --> 00:10:06,240
Speaker 3: General and then assistant Attorney General, at least inside the Beltway.

172
00:10:06,360 --> 00:10:08,679
Speaker 3: We're all kind of guessing who it could be, but

173
00:10:08,840 --> 00:10:09,480
Speaker 3: who knows.

174
00:10:09,760 --> 00:10:13,360
Speaker 5: Yeah, Sarah Alam, Senior Fellow at Tech Institute. We appreciate

175
00:10:13,360 --> 00:10:17,000
Speaker 5: the time and analysis. Meanwhile, well, talking of regulatory landscape

176
00:10:17,040 --> 00:10:21,200
Speaker 5: in the UK, Google's partnership with Aifomanthropic has actually avoided

177
00:10:21,240 --> 00:10:24,480
Speaker 5: further scrutiny after the UK Antitrust watchdog concluded it doesn't

178
00:10:24,559 --> 00:10:27,360
Speaker 5: qualify for a full blown investigation under merger rules. The

179
00:10:27,360 --> 00:10:31,400
Speaker 5: CMA found that Google has not gained quote material influence

180
00:10:31,480 --> 00:10:33,840
Speaker 5: of Anthropic as a result of the deal, which included

181
00:10:33,880 --> 00:10:37,559
Speaker 5: a two billion dollar investment and a cloud agreement coming

182
00:10:37,640 --> 00:10:40,560
Speaker 5: up scaling safe AI governance. We're going to be joined

183
00:10:40,559 --> 00:10:43,679
Speaker 5: by Cudo AI CEO sa Rena Singh on a new

184
00:10:43,720 --> 00:10:55,720
Speaker 5: partnership with Microsoft. This simply meg technology. Microsoft wants to

185
00:10:55,720 --> 00:10:58,720
Speaker 5: make it easier to switch between large language models with

186
00:10:58,840 --> 00:11:02,000
Speaker 5: its new tool as You AI Foundry, which can help

187
00:11:02,040 --> 00:11:05,000
Speaker 5: cloud customers build and deploy AI applications, and the company

188
00:11:05,040 --> 00:11:07,520
Speaker 5: is giving away the software in the hopes of persuading

189
00:11:07,520 --> 00:11:08,719
Speaker 5: corporate customers to.

190
00:11:08,640 --> 00:11:10,160
Speaker 4: Buy more of its cloud services.

191
00:11:10,320 --> 00:11:13,800
Speaker 5: Greenberg's Diena Bas has more on what is being unveiled

192
00:11:13,840 --> 00:11:17,520
Speaker 5: at the annual event in Chicago Ignite Dina and really

193
00:11:17,960 --> 00:11:20,760
Speaker 5: the underlying process here is what what is the blocker here?

194
00:11:20,760 --> 00:11:23,240
Speaker 5: People don't feel they can flip from one great model

195
00:11:23,240 --> 00:11:23,640
Speaker 5: to another.

196
00:11:24,880 --> 00:11:27,520
Speaker 6: The whole product is meant to help make it easier

197
00:11:27,559 --> 00:11:30,679
Speaker 6: to get your models programmed up, running, deployed, but one

198
00:11:30,720 --> 00:11:33,240
Speaker 6: of the biggest blockers is that, you know, we're seeing

199
00:11:33,280 --> 00:11:36,160
Speaker 6: this constant drumbeat of new models coming out, whether it's

200
00:11:36,200 --> 00:11:39,559
Speaker 6: open AI updating their models, and you know, one of

201
00:11:39,600 --> 00:11:41,600
Speaker 6: them is good for better for reasoning, one of them

202
00:11:41,640 --> 00:11:44,760
Speaker 6: is better for answering questions that are more conversational. And

203
00:11:44,800 --> 00:11:48,480
Speaker 6: then there are you know, numerous competitors of Microsoft has,

204
00:11:48,559 --> 00:11:51,400
Speaker 6: you know, seventeen hundred and eighteen hundred different models that

205
00:11:51,400 --> 00:11:55,400
Speaker 6: are available through this Azure service. And so the issue

206
00:11:55,440 --> 00:11:56,840
Speaker 6: is how do you figure out which is the right

207
00:11:56,880 --> 00:11:59,520
Speaker 6: one and how do you switch if there's an update

208
00:11:59,679 --> 00:12:02,160
Speaker 6: or or you want to try something different. So this

209
00:12:02,480 --> 00:12:05,120
Speaker 6: foundry software basically has a bunch of different tools that

210
00:12:05,160 --> 00:12:07,680
Speaker 6: allow you to experiment with different models, see which one

211
00:12:07,720 --> 00:12:12,160
Speaker 6: works better, switch more easily track what's working and what isn't. So,

212
00:12:12,240 --> 00:12:14,400
Speaker 6: for example, if you want to put your own corporate

213
00:12:14,480 --> 00:12:17,440
Speaker 6: data into someone else's model in order to find tune it,

214
00:12:17,640 --> 00:12:19,320
Speaker 6: and you want to know is that data really making

215
00:12:19,360 --> 00:12:21,439
Speaker 6: a difference, Is it making the answers better or worse?

216
00:12:21,720 --> 00:12:23,840
Speaker 6: Things like that are really hard to test right now,

217
00:12:23,920 --> 00:12:26,480
Speaker 6: and it's imperative for Microsoft to make this easier in

218
00:12:26,600 --> 00:12:29,600
Speaker 6: order to convince people to spend more money on Azure

219
00:12:29,640 --> 00:12:34,240
Speaker 6: Services to build and deploy these AI applications.

220
00:12:33,559 --> 00:12:35,320
Speaker 5: And they want people to spend more money because they're

221
00:12:35,320 --> 00:12:37,679
Speaker 5: having to spend a lot of money invest certainly on

222
00:12:37,800 --> 00:12:40,720
Speaker 5: chips for example, they're building their own. They're also showing

223
00:12:40,720 --> 00:12:43,360
Speaker 5: off how in videos new aichips are going to be

224
00:12:43,360 --> 00:12:44,280
Speaker 5: working within the systems.

225
00:12:44,320 --> 00:12:46,280
Speaker 4: Can you talk us through their own offering?

226
00:12:47,320 --> 00:12:47,640
Speaker 6: Sure?

227
00:12:47,720 --> 00:12:47,800
Speaker 2: So.

228
00:12:47,920 --> 00:12:50,040
Speaker 6: A year ago at the same conference, they announced their

229
00:12:50,080 --> 00:12:52,240
Speaker 6: first AI chip, and today they've said that it is

230
00:12:52,320 --> 00:12:56,320
Speaker 6: now handling customer workloads in the Azure Open AI service.

231
00:12:56,360 --> 00:12:58,520
Speaker 6: They also announced two new chips today and one of

232
00:12:58,520 --> 00:13:00,680
Speaker 6: them is for security. It's going to go in Microsoft's

233
00:13:00,720 --> 00:13:03,520
Speaker 6: own data center servers to better secure them. And one

234
00:13:03,520 --> 00:13:06,200
Speaker 6: of them is a data process and unit, which is

235
00:13:06,280 --> 00:13:08,599
Speaker 6: basically a networking chip that takes a bunch of the

236
00:13:08,920 --> 00:13:11,800
Speaker 6: data crunching off of the other chips in order to

237
00:13:11,800 --> 00:13:15,160
Speaker 6: make things faster. So they are now increasingly getting into

238
00:13:15,200 --> 00:13:17,880
Speaker 6: a position that Amazon has been Amazon Web Services has

239
00:13:17,880 --> 00:13:20,080
Speaker 6: been in for a while of really coming up with

240
00:13:20,120 --> 00:13:22,520
Speaker 6: a bunch of their own chips for different types of

241
00:13:22,559 --> 00:13:26,440
Speaker 6: tasks and you know, things inside the data center in

242
00:13:26,600 --> 00:13:28,920
Speaker 6: order to make sure that Microsoft can do a much

243
00:13:28,920 --> 00:13:34,400
Speaker 6: better job of training AI, you know, inferencing, serving applications,

244
00:13:34,559 --> 00:13:36,440
Speaker 6: and a bunch of other things that you would use

245
00:13:36,480 --> 00:13:40,319
Speaker 6: Microsoft Cloud for. They obviously still work very closely within Vidia.

246
00:13:40,360 --> 00:13:42,520
Speaker 6: It's going to take a very long time, if ever,

247
00:13:42,720 --> 00:13:44,720
Speaker 6: for as you're to be able to replace in VideA

248
00:13:45,160 --> 00:13:47,880
Speaker 6: in its data center, even if they're putting some of

249
00:13:47,920 --> 00:13:51,560
Speaker 6: the AI traffic onto this Azumaya AI chip. But at

250
00:13:51,600 --> 00:13:53,760
Speaker 6: the same time they you know, I spoke to Ronnie

251
00:13:53,760 --> 00:13:56,640
Speaker 6: Borcard their kind of chip design and development, and you know,

252
00:13:56,720 --> 00:13:59,439
Speaker 6: she made the point that everything needs to be optimized

253
00:13:59,440 --> 00:14:01,560
Speaker 6: at every leg in order to get the best possible

254
00:14:01,600 --> 00:14:03,959
Speaker 6: performance because models are growing so rapidly.

255
00:14:04,880 --> 00:14:08,320
Speaker 5: Dana bas aholl things Microsoftware, thank you for more on

256
00:14:08,320 --> 00:14:11,439
Speaker 5: what's coming out of Microsoft and AI. We have CREDOAI

257
00:14:11,480 --> 00:14:14,079
Speaker 5: announcing they are partnering with the tech Gient to power

258
00:14:14,520 --> 00:14:18,560
Speaker 5: ENDAI governance for enterprise applications, which will help companies basically

259
00:14:18,559 --> 00:14:21,800
Speaker 5: safely adopt these cutting JI solutions, but also ensuring that

260
00:14:21,920 --> 00:14:25,200
Speaker 5: compliance is there with global regulations. Navrina sing joins our

261
00:14:25,240 --> 00:14:28,680
Speaker 5: CEO of CREDOAI and how much of a blocker is

262
00:14:29,200 --> 00:14:30,960
Speaker 5: the cloud around governance of AI?

263
00:14:32,920 --> 00:14:35,640
Speaker 1: Caroline, great to be here today, you know, one of

264
00:14:35,640 --> 00:14:37,200
Speaker 1: the things that we have found that there is a

265
00:14:37,240 --> 00:14:41,880
Speaker 1: massive gap between innovative AI solutions and actually practical business applications,

266
00:14:42,280 --> 00:14:45,400
Speaker 1: and the only way to accelerate that gap and bring

267
00:14:45,440 --> 00:14:48,720
Speaker 1: AI innovations to it is through AI governance. So this

268
00:14:48,800 --> 00:14:51,960
Speaker 1: is where we are very excited to announce our partnership

269
00:14:51,960 --> 00:14:56,640
Speaker 1: today with Microsoft, Azure AI Foundry and Preto AI, as

270
00:14:56,640 --> 00:14:58,960
Speaker 1: you know, is the leading provided a writer of.

271
00:14:58,920 --> 00:15:00,239
Speaker 4: AI governance solutions.

272
00:15:00,560 --> 00:15:02,760
Speaker 1: So I do want to just emphasize the three things

273
00:15:02,800 --> 00:15:05,480
Speaker 1: that this partnership is going to bring to life. One

274
00:15:05,600 --> 00:15:09,320
Speaker 1: is end to end governance of Azure AI apps, including

275
00:15:09,400 --> 00:15:13,120
Speaker 1: generative AI and multimodal systems as well as RAG based

276
00:15:13,240 --> 00:15:18,120
Speaker 1: use cases. Secondly, is this comprehensive unified control that Dina

277
00:15:18,160 --> 00:15:22,000
Speaker 1: was just talking about, which is needed for both THEI developers.

278
00:15:21,440 --> 00:15:22,640
Speaker 4: And governance leaders.

279
00:15:23,040 --> 00:15:26,520
Speaker 1: And lastly, the support for governing especially these cutting edge

280
00:15:26,520 --> 00:15:30,200
Speaker 1: systems that are really getting brought to life each and

281
00:15:30,240 --> 00:15:33,960
Speaker 1: every day like GPT four, GPT four H and smallify models.

282
00:15:34,200 --> 00:15:37,040
Speaker 1: So PREDAI is excited to partner with Microsoft to bring

283
00:15:37,240 --> 00:15:41,560
Speaker 1: all these capabilities at scale through an independent governance layer.

284
00:15:42,320 --> 00:15:44,800
Speaker 5: Just talk us through that governance layer, what is the

285
00:15:44,840 --> 00:15:49,080
Speaker 5: technology that enables you to be monitoring consistently to ensuring

286
00:15:49,280 --> 00:15:51,600
Speaker 5: that these models are doing what they say they're doing

287
00:15:51,640 --> 00:15:56,000
Speaker 5: and doing them well with fair use case and ultimately

288
00:15:56,080 --> 00:15:56,600
Speaker 5: no bias.

289
00:15:56,640 --> 00:15:57,200
Speaker 4: For example.

290
00:15:58,640 --> 00:16:01,920
Speaker 1: Yeah, absolutely so, Caroline. One of the core things that

291
00:16:01,960 --> 00:16:06,440
Speaker 1: we are finding with accelerating AI transformation and innovation is

292
00:16:06,760 --> 00:16:10,240
Speaker 1: something called this AI alignment. How do you actually ensure

293
00:16:10,280 --> 00:16:14,400
Speaker 1: that you're measuring the help of these AI systems against

294
00:16:14,440 --> 00:16:18,560
Speaker 1: the policies, the principles, the regulations, AI standards that you

295
00:16:18,640 --> 00:16:21,080
Speaker 1: as a company deeply care about. So this is where

296
00:16:21,120 --> 00:16:25,560
Speaker 1: CREATOAI is pioneering solution in policy intelligence makes that really

297
00:16:25,640 --> 00:16:29,120
Speaker 1: easy for your business experts as well as your AI

298
00:16:29,160 --> 00:16:32,200
Speaker 1: experts to align on what that good looks like.

299
00:16:32,600 --> 00:16:34,280
Speaker 4: And once you've aligned on what that good.

300
00:16:34,320 --> 00:16:38,040
Speaker 1: Like looks like, really measuring your AI systems, your models,

301
00:16:38,040 --> 00:16:42,760
Speaker 1: your data sets against those requirements and then lastly really

302
00:16:42,800 --> 00:16:46,960
Speaker 1: putting that out in a continuous monitoring, risk management, and

303
00:16:47,000 --> 00:16:49,640
Speaker 1: compliance engine so that you can make sure that you're

304
00:16:49,680 --> 00:16:53,200
Speaker 1: always in alignment alignment. And as you can imagine, Caroline,

305
00:16:53,200 --> 00:16:57,920
Speaker 1: this becomes really critical with the AI agents energenic systems

306
00:16:58,280 --> 00:17:01,040
Speaker 1: that have started to take a lot of excitement in

307
00:17:01,120 --> 00:17:02,280
Speaker 1: business organizations.

308
00:17:02,440 --> 00:17:05,760
Speaker 5: Microsoft talking a lot about AI agents today in Avrena like.

309
00:17:06,040 --> 00:17:09,080
Speaker 5: The real world impact is that we hear that people

310
00:17:09,080 --> 00:17:12,760
Speaker 5: are actually kind of plateauing in their own use employees

311
00:17:13,000 --> 00:17:17,200
Speaker 5: use of agents AI general to AI in the workplace

312
00:17:17,280 --> 00:17:19,639
Speaker 5: largely because they're worried that they could be deemed lazy

313
00:17:19,760 --> 00:17:22,240
Speaker 5: or they don't really understand what the guardrails are from

314
00:17:22,520 --> 00:17:25,080
Speaker 5: those that are above them. Is management on top of this,

315
00:17:27,720 --> 00:17:28,000
Speaker 5: you know.

316
00:17:28,000 --> 00:17:30,639
Speaker 1: The reason for the AI platu is because most of

317
00:17:30,680 --> 00:17:33,879
Speaker 1: the organizations, more than seventy percent of them, don't have

318
00:17:33,920 --> 00:17:36,960
Speaker 1: guardrails in place. And this is where predo AI like

319
00:17:37,040 --> 00:17:41,479
Speaker 1: solutions become really critical, because we are converting you know,

320
00:17:41,600 --> 00:17:45,119
Speaker 1: governance from a safety ness to really being a launch

321
00:17:45,160 --> 00:17:47,919
Speaker 1: pad from a four AI innovations. So what we are

322
00:17:47,960 --> 00:17:50,240
Speaker 1: finding in the companies that we work with, whether it

323
00:17:50,320 --> 00:17:55,200
Speaker 1: is MasterCard, Regender and you know Booz Allen and others,

324
00:17:55,720 --> 00:18:00,960
Speaker 1: is that when you have the clearly defined guardrails, employees

325
00:18:01,000 --> 00:18:03,320
Speaker 1: actually are much more excited to use these.

326
00:18:03,200 --> 00:18:04,400
Speaker 4: Technologies at scale.

327
00:18:04,520 --> 00:18:06,560
Speaker 1: So with governance, what we are going to see is

328
00:18:06,880 --> 00:18:10,360
Speaker 1: much more excited as well as rampant use of artificial

329
00:18:10,359 --> 00:18:13,040
Speaker 1: intelligence and AI transformation within organizations.

330
00:18:13,160 --> 00:18:16,520
Speaker 5: Do you think any of that regulatory outlook, governance, desires,

331
00:18:16,560 --> 00:18:18,720
Speaker 5: and desires is going to change out of the next

332
00:18:18,720 --> 00:18:19,919
Speaker 5: administration in the US.

333
00:18:22,640 --> 00:18:25,119
Speaker 1: You know, one of the exciting things is that AI

334
00:18:25,160 --> 00:18:29,119
Speaker 1: is actually central and core to the strategy irrespective of

335
00:18:29,160 --> 00:18:32,119
Speaker 1: the administration. We are ready to actually engage with the

336
00:18:32,160 --> 00:18:36,800
Speaker 1: new administration who is really looking at AI innovation much

337
00:18:36,800 --> 00:18:37,600
Speaker 1: more holistically.

338
00:18:37,960 --> 00:18:38,720
Speaker 4: So more to.

339
00:18:38,880 --> 00:18:41,040
Speaker 1: Come on that, but right now, I think there's a

340
00:18:41,119 --> 00:18:43,879
Speaker 1: lot of excitement on how do you actually use AI

341
00:18:43,920 --> 00:18:47,320
Speaker 1: for transforming government as well as AI for making sure

342
00:18:47,600 --> 00:18:51,840
Speaker 1: United States competitiveness in the world continues to stay at forefront.

343
00:18:52,359 --> 00:18:56,159
Speaker 5: A finnessing, creative AI CEO on new partnership with Microsoftware.

344
00:18:56,200 --> 00:18:57,479
Speaker 4: Appreciate you joining. Thank you.

345
00:19:05,080 --> 00:19:07,800
Speaker 5: Let's check in on the shares of super Micro jumping

346
00:19:07,920 --> 00:19:10,760
Speaker 5: after the company hired a new auditor filed a plan

347
00:19:10,880 --> 00:19:14,160
Speaker 5: to come into compliance with the NASDAT listing requirements. Bloomberg's

348
00:19:14,200 --> 00:19:16,480
Speaker 5: Brody Ford joins us small We're up thirty three percent,

349
00:19:16,520 --> 00:19:17,440
Speaker 5: but we are still.

350
00:19:17,240 --> 00:19:19,160
Speaker 4: Way off of our hives.

351
00:19:19,000 --> 00:19:21,639
Speaker 5: From when the market cap was spiraling higher because of

352
00:19:21,680 --> 00:19:25,680
Speaker 5: general to AI. Brody talk US through Hybdo for example,

353
00:19:25,840 --> 00:19:28,240
Speaker 5: is now willing to work with super Micro computer.

354
00:19:29,040 --> 00:19:32,600
Speaker 7: Yeah, it's quite a dramatic episode. Backing up a little bit,

355
00:19:32,720 --> 00:19:37,000
Speaker 7: super Micro missed a deadline to file financials back in August.

356
00:19:38,040 --> 00:19:41,240
Speaker 7: Then their auditor ey resign saying we're not willing to

357
00:19:41,240 --> 00:19:44,840
Speaker 7: be associated with these numbers. This is scary stuff. You

358
00:19:44,880 --> 00:19:47,080
Speaker 7: don't see this too often, and so it was up

359
00:19:47,160 --> 00:19:49,719
Speaker 7: to the wire here. Super Micro had to find a

360
00:19:49,760 --> 00:19:53,159
Speaker 7: new auditor to get them in line with NASDAC rules

361
00:19:53,200 --> 00:19:57,080
Speaker 7: to avoid delisting, and on the very last day, yesterday evening,

362
00:19:57,560 --> 00:20:01,639
Speaker 7: they found BDO. It was a fairly major auditor. But

363
00:20:01,680 --> 00:20:04,280
Speaker 7: it's worth noting they are not one of the Big four.

364
00:20:04,480 --> 00:20:06,879
Speaker 7: They're the Big fifth or sixth, depending on who you

365
00:20:06,960 --> 00:20:09,439
Speaker 7: listen to. But what matters is they are an auditor

366
00:20:09,960 --> 00:20:12,240
Speaker 7: and they're willing to, you know, try to help super

367
00:20:12,280 --> 00:20:14,920
Speaker 7: Micro get back into compliance, which buys them a couple

368
00:20:14,960 --> 00:20:18,440
Speaker 7: more months before you know, another deadline.

369
00:20:18,960 --> 00:20:22,000
Speaker 5: February is likely when the document deadline is going to

370
00:20:22,040 --> 00:20:24,720
Speaker 5: be pushed to for this ten K filing. The risk

371
00:20:24,800 --> 00:20:27,639
Speaker 5: here is when they're ousted out of the Nasdaq and

372
00:20:27,640 --> 00:20:29,879
Speaker 5: therefore liquidity is going to be far more difficult.

373
00:20:29,880 --> 00:20:31,600
Speaker 4: People will be forced to sell their holdings.

374
00:20:31,640 --> 00:20:35,359
Speaker 7: Right yeah, February, three months from now and the news

375
00:20:35,400 --> 00:20:38,280
Speaker 7: business sounds like forever and the accounting world it is,

376
00:20:38,920 --> 00:20:41,919
Speaker 7: you know, very short, because they essentially have to go

377
00:20:42,040 --> 00:20:45,040
Speaker 7: back through all of their financials and figure out there

378
00:20:45,160 --> 00:20:48,119
Speaker 7: was some issue. Maybe it was around just processes. Maybe

379
00:20:48,200 --> 00:20:51,440
Speaker 7: no restatement is needed. Even in that best case scenario,

380
00:20:51,560 --> 00:20:55,400
Speaker 7: it's quite tight. Right. Let's say that they were delisted.

381
00:20:55,720 --> 00:20:59,480
Speaker 7: What would happen is a significant amount of their shareholders

382
00:20:59,520 --> 00:21:03,359
Speaker 7: would are passive, meaning it would be automatic selling. And

383
00:21:03,440 --> 00:21:05,480
Speaker 7: it's possible that they would have to repay some of

384
00:21:05,480 --> 00:21:09,080
Speaker 7: their bonds early. That's you know, about one point seven

385
00:21:09,160 --> 00:21:12,040
Speaker 7: billion or so. Essentially. Super Micro is going to do

386
00:21:12,119 --> 00:21:15,560
Speaker 7: everything within its power to avoid that scenario, and its

387
00:21:15,600 --> 00:21:18,679
Speaker 7: odds just got a whole lot better after yesterday's news.

388
00:21:18,880 --> 00:21:22,080
Speaker 5: Mean, while its still faces a DOJ probe, we understand, lady,

389
00:21:22,119 --> 00:21:23,680
Speaker 5: for all across this company.

390
00:21:24,080 --> 00:21:34,600
Speaker 4: Thanks, thank you. Welcome back to Bluemote Technology. I'm Caroin

391
00:21:34,640 --> 00:21:35,280
Speaker 4: Hyde in New York.

392
00:21:35,280 --> 00:21:37,640
Speaker 5: It's time now for talking tech and first up shares.

393
00:21:37,680 --> 00:21:40,840
Speaker 5: A delivery Heroes Middle East unit tell About sold out

394
00:21:40,920 --> 00:21:44,600
Speaker 5: min minutes after its subscriptions opened. The IPO value tell

395
00:21:44,600 --> 00:21:47,119
Speaker 5: About and up to ten point two billion dollars in

396
00:21:47,160 --> 00:21:49,280
Speaker 5: the final price of Talabas offer will be announced on

397
00:21:49,280 --> 00:21:51,879
Speaker 5: November the twenty ninth, with shares set to start trading

398
00:21:51,920 --> 00:21:55,159
Speaker 5: on December the tenth plus. Apple is offering Indonesia one

399
00:21:55,240 --> 00:21:57,800
Speaker 5: hundred million dollars in investment as the company hopes to

400
00:21:57,800 --> 00:22:00,359
Speaker 5: persuade the government in lifting a sales band on its

401
00:22:00,400 --> 00:22:03,600
Speaker 5: iPhone sixteen. That's according to sources, the proposal is an

402
00:22:03,720 --> 00:22:06,960
Speaker 5: increase of a prior offer of ten million dollars made

403
00:22:07,000 --> 00:22:10,280
Speaker 5: last month. Indonesia's Ministry of Industry was not made a

404
00:22:10,280 --> 00:22:13,600
Speaker 5: final decision on the revised offer, and Sony is exploring

405
00:22:13,640 --> 00:22:16,800
Speaker 5: a takeover of publisher Karakawa, maker of the hit game

406
00:22:16,920 --> 00:22:20,200
Speaker 5: elden Ring. According to sources, the two companies have held talks,

407
00:22:20,200 --> 00:22:24,080
Speaker 5: with deliberations still ongoing. The purchase would allow Sony access

408
00:22:24,119 --> 00:22:28,919
Speaker 5: to Kadakawa's expansive library of manga, anime, film, and gaming content.

409
00:22:30,000 --> 00:22:31,400
Speaker 4: Let's stick with gaming now.

410
00:22:31,280 --> 00:22:34,639
Speaker 5: Because roblocks has announced more protections to its platform for

411
00:22:34,680 --> 00:22:38,880
Speaker 5: its youngest users, with new parental controls, restrictions on communication

412
00:22:39,280 --> 00:22:42,560
Speaker 5: and the content that is accessible, as well as clearer

413
00:22:42,560 --> 00:22:46,200
Speaker 5: content labels. This is all after Roblocks's child's safety policies

414
00:22:46,320 --> 00:22:49,000
Speaker 5: have come under increasing scrutiny in recent months and following

415
00:22:49,080 --> 00:22:50,280
Speaker 5: arrests of alleged child.

416
00:22:50,160 --> 00:22:51,440
Speaker 4: Abuses using the service.

417
00:22:51,760 --> 00:22:54,840
Speaker 5: Now, Matt Kaufman, who's chief safety officer at Roblocks and

418
00:22:54,880 --> 00:22:57,159
Speaker 5: peaces Say, joins us now to discuss these changes and

419
00:22:57,240 --> 00:22:59,040
Speaker 5: just go into some of the measures in particular that

420
00:22:59,080 --> 00:23:00,520
Speaker 5: you think will make a real differference here.

421
00:23:01,520 --> 00:23:05,000
Speaker 8: Sure, First, thank you for having me. We've waunched three things.

422
00:23:05,119 --> 00:23:07,920
Speaker 8: The first is new parental controls that make it easier

423
00:23:07,920 --> 00:23:12,320
Speaker 8: for parents to monitor and manage their kids accounts even

424
00:23:12,359 --> 00:23:15,119
Speaker 8: when they're not sitting next to their kids. We've also

425
00:23:15,200 --> 00:23:18,840
Speaker 8: updated how our younger users under thirteen can communicate on

426
00:23:18,880 --> 00:23:22,600
Speaker 8: the platform, and finally, we've built in additional protections for

427
00:23:22,680 --> 00:23:23,760
Speaker 8: our youngest users.

428
00:23:24,320 --> 00:23:27,520
Speaker 5: Let's just talk about ultimately defining and working out who

429
00:23:27,640 --> 00:23:30,800
Speaker 5: is thirteen and under. It's a really difficult lift if

430
00:23:30,800 --> 00:23:33,119
Speaker 5: people aren't going to be always telling the truth. Are

431
00:23:33,240 --> 00:23:35,879
Speaker 5: using technology here or is it really depending on the

432
00:23:35,920 --> 00:23:37,320
Speaker 5: truth of the userbase.

433
00:23:38,400 --> 00:23:40,439
Speaker 8: You know, if you're under thirteen and you sign up

434
00:23:40,440 --> 00:23:43,399
Speaker 8: for an account online, we actually cannot ask you for

435
00:23:43,440 --> 00:23:48,520
Speaker 8: additional information because of privacy reasons. So from a Roadblocks perspective,

436
00:23:48,560 --> 00:23:50,800
Speaker 8: we work to make sure that everybody is on the

437
00:23:50,800 --> 00:23:53,960
Speaker 8: platform is safe no matter what age they sign up at.

438
00:23:54,840 --> 00:23:58,479
Speaker 5: So how therefore are you looking to if you kind

439
00:23:58,520 --> 00:24:02,400
Speaker 5: of ask if you cannot get the data and their clarity,

440
00:24:03,200 --> 00:24:04,560
Speaker 5: is there a fix?

441
00:24:04,720 --> 00:24:05,800
Speaker 4: Is there something you're working on?

442
00:24:06,720 --> 00:24:09,959
Speaker 8: Well, so what we do is we take extraordinary steps

443
00:24:10,000 --> 00:24:12,840
Speaker 8: to keep everybody safe on the platform. For example, when

444
00:24:12,840 --> 00:24:17,880
Speaker 8: you're communicating on roadblocks, we don't encrypt anything. We monitor

445
00:24:17,960 --> 00:24:21,399
Speaker 8: all communication and we have an investigation team that's constantly

446
00:24:21,440 --> 00:24:25,000
Speaker 8: looking into issues where our system detects that something may

447
00:24:25,040 --> 00:24:29,200
Speaker 8: be violative to our process, our policies, and we have

448
00:24:29,359 --> 00:24:32,199
Speaker 8: a lot of machine learning and AI systems that are

449
00:24:32,280 --> 00:24:35,480
Speaker 8: running to keep everybody safe. And really that's part of

450
00:24:35,480 --> 00:24:36,720
Speaker 8: our go ahead.

451
00:24:36,920 --> 00:24:40,800
Speaker 5: Well talk through that AI, the machine learning how sopheificated

452
00:24:40,880 --> 00:24:45,360
Speaker 5: is it becoming because the subtleties of grooming are so hard.

453
00:24:45,160 --> 00:24:45,720
Speaker 4: To pick up on.

454
00:24:47,160 --> 00:24:50,000
Speaker 8: That's true, and AI is advancing very fast, and we've

455
00:24:50,040 --> 00:24:54,320
Speaker 8: made significant leaps in the technology. For example, our automated

456
00:24:54,359 --> 00:24:59,080
Speaker 8: moderation systems for voice communication. Just earlier this year, we

457
00:24:59,160 --> 00:25:03,240
Speaker 8: open source some of the models behind that technology and

458
00:25:03,280 --> 00:25:05,760
Speaker 8: they're some of the most popular models in open source

459
00:25:05,760 --> 00:25:08,919
Speaker 8: today for safety. So we feel that roadblocks is really

460
00:25:09,040 --> 00:25:12,080
Speaker 8: at the leading edge of developing AI to keep people

461
00:25:12,119 --> 00:25:13,240
Speaker 8: safe on the platform.

462
00:25:13,400 --> 00:25:16,080
Speaker 5: Can you give that balance of AI and then the

463
00:25:16,119 --> 00:25:19,280
Speaker 5: team that does the deeper investigation. Is it that you

464
00:25:19,320 --> 00:25:22,680
Speaker 5: can ever more focus on technology being the fix here?

465
00:25:22,960 --> 00:25:24,679
Speaker 4: How big is the team going to remain?

466
00:25:26,040 --> 00:25:29,240
Speaker 8: Really everything is almost like a multi layered approach. You

467
00:25:29,320 --> 00:25:32,720
Speaker 8: have AI systems that make initial decisions and act very

468
00:25:32,760 --> 00:25:36,320
Speaker 8: fast and at super scale. But behind those AIS, you

469
00:25:36,440 --> 00:25:40,080
Speaker 8: have a team of professional safety experts that are doing

470
00:25:40,160 --> 00:25:42,040
Speaker 8: all sorts of things. They're the ones who are of

471
00:25:42,119 --> 00:25:44,680
Speaker 8: course who are training the AI. They're also the ones

472
00:25:44,720 --> 00:25:47,880
Speaker 8: that are handling appeals because sometimes the AI doesn't get

473
00:25:47,920 --> 00:25:51,119
Speaker 8: things right. So if somebody says, hey, I feel like

474
00:25:51,160 --> 00:25:53,880
Speaker 8: it made a mistake, we have our team of professionals

475
00:25:53,920 --> 00:25:56,320
Speaker 8: who step in to make decisions, and of course that

476
00:25:56,359 --> 00:25:59,280
Speaker 8: trains the AI. And then we also have that team

477
00:25:59,320 --> 00:26:03,320
Speaker 8: of professionals handling the most complicated cases. So, for example,

478
00:26:03,320 --> 00:26:06,320
Speaker 8: we have a team of investigators who are constantly looking

479
00:26:06,359 --> 00:26:08,720
Speaker 8: at the platform trying to understand if there are bad

480
00:26:08,760 --> 00:26:11,840
Speaker 8: actors and if there are, immediately removing them from the

481
00:26:11,880 --> 00:26:15,120
Speaker 8: platform and working with our law enforcement partners to keep

482
00:26:15,119 --> 00:26:16,040
Speaker 8: the platform safe.

483
00:26:16,119 --> 00:26:18,920
Speaker 5: Yeah, focus in on the bad actor side. You've outlined

484
00:26:18,960 --> 00:26:21,880
Speaker 5: how you're going to be protecting the users, the children,

485
00:26:22,280 --> 00:26:24,720
Speaker 5: but the people who come on who aren't who they

486
00:26:24,720 --> 00:26:27,200
Speaker 5: say they are, who are indeed the predators, what are

487
00:26:27,200 --> 00:26:28,200
Speaker 5: you doing to innovate there?

488
00:26:29,359 --> 00:26:31,240
Speaker 8: Sure, So let me step back for a minute and say,

489
00:26:31,280 --> 00:26:34,239
Speaker 8: you know, roblocks has nearly ninety million users from one

490
00:26:34,320 --> 00:26:37,200
Speaker 8: hundred and ninety countries around the world, and these users

491
00:26:37,200 --> 00:26:40,440
Speaker 8: are enjoying experiences with their friends and with their families,

492
00:26:40,720 --> 00:26:43,840
Speaker 8: and it's a really important part of their life. There's

493
00:26:43,920 --> 00:26:46,120
Speaker 8: a very very small number of people in the world

494
00:26:46,200 --> 00:26:48,919
Speaker 8: who are just they're bad actors, and so we have

495
00:26:49,080 --> 00:26:53,879
Speaker 8: systems that automatically detect behavioral anomalies to determine when something

496
00:26:53,960 --> 00:26:57,320
Speaker 8: might be violating our policy, and we try and act

497
00:26:57,320 --> 00:27:00,119
Speaker 8: as quickly as we can. That notion of acting as

498
00:27:00,160 --> 00:27:03,439
Speaker 8: quickly as we can actually applies to how people upload

499
00:27:03,480 --> 00:27:06,159
Speaker 8: content to Roadblocks. First of all, when you're building a

500
00:27:06,160 --> 00:27:09,640
Speaker 8: game on Roadblocks, we review every piece of content that's

501
00:27:09,720 --> 00:27:12,760
Speaker 8: uploaded to the platform before anybody sees it. And when

502
00:27:12,760 --> 00:27:16,439
Speaker 8: you're communicating using text on the platform, we actually block

503
00:27:16,520 --> 00:27:19,760
Speaker 8: out and filter a lot of that communication before it

504
00:27:19,800 --> 00:27:22,320
Speaker 8: ever transmits. To other people if we find that it

505
00:27:22,359 --> 00:27:23,600
Speaker 8: is violated for our.

506
00:27:23,400 --> 00:27:27,879
Speaker 5: Policies, So those sorts of changes that you just outline,

507
00:27:28,040 --> 00:27:31,280
Speaker 5: are they new and they novel because I mean, very sadly,

508
00:27:31,320 --> 00:27:33,320
Speaker 5: as you say, there are very few bad actors and

509
00:27:33,400 --> 00:27:36,760
Speaker 5: a percentage perspective, but raw numbers of at least twenty

510
00:27:36,760 --> 00:27:39,520
Speaker 5: four people being arrested since twenty eighteen because of abusing

511
00:27:39,640 --> 00:27:42,400
Speaker 5: or abducting victims met through the platform, it's twenty four

512
00:27:42,440 --> 00:27:43,840
Speaker 5: too many. As I'm sure you.

513
00:27:43,800 --> 00:27:49,320
Speaker 8: Agree, I think anybody obviously who comes to any harm

514
00:27:49,359 --> 00:27:52,000
Speaker 8: on roadblocks is one too many. And we take these

515
00:27:52,040 --> 00:27:55,080
Speaker 8: situations incredibly seriously, and we do everything we can to

516
00:27:55,160 --> 00:27:58,040
Speaker 8: keep the platform safe. It is our number one priority

517
00:27:59,080 --> 00:28:02,000
Speaker 8: when it comes to the bad actors on roadblocks. We

518
00:28:02,040 --> 00:28:05,679
Speaker 8: are constantly innovating, and we've actually this year alone have

519
00:28:05,720 --> 00:28:09,359
Speaker 8: already had thirty major updates to our safety systems. So

520
00:28:09,480 --> 00:28:11,920
Speaker 8: the release that we're talking about today and we announced

521
00:28:11,960 --> 00:28:15,760
Speaker 8: this week is really about limiting exposure to some of

522
00:28:15,760 --> 00:28:19,200
Speaker 8: our youngest audiences to content that's more appropriate to teenagers

523
00:28:19,680 --> 00:28:22,520
Speaker 8: when we're talking about the actual, the really bad actors.

524
00:28:22,760 --> 00:28:26,120
Speaker 8: We are constantly innovating on that front and constantly updating

525
00:28:26,119 --> 00:28:26,760
Speaker 8: our systems.

526
00:28:27,040 --> 00:28:30,199
Speaker 5: You talk about safety being your number one priority, and

527
00:28:30,240 --> 00:28:33,120
Speaker 5: there have of course been accusations that there was growth

528
00:28:33,520 --> 00:28:36,879
Speaker 5: that came first over safety. Do you still feel in

529
00:28:36,960 --> 00:28:40,520
Speaker 5: any way that that's the case, or do you ultimately

530
00:28:40,560 --> 00:28:43,160
Speaker 5: have to slow down sign ups in some way to

531
00:28:43,360 --> 00:28:45,200
Speaker 5: ensure the number one priority is safety.

532
00:28:45,960 --> 00:28:47,800
Speaker 8: You know, our number one priority is safety. I mean,

533
00:28:47,800 --> 00:28:51,240
Speaker 8: I think that's just our first principle. And if you

534
00:28:51,240 --> 00:28:54,400
Speaker 8: think about it from a company perspective, Roadblocks was founded

535
00:28:54,840 --> 00:28:57,680
Speaker 8: nearly twenty years ago, where the audience was primarily kids

536
00:28:57,680 --> 00:29:01,600
Speaker 8: in the beginning, and really since it's found, protecting the

537
00:29:01,680 --> 00:29:04,800
Speaker 8: users on the platform has been our first priority, and

538
00:29:05,200 --> 00:29:07,840
Speaker 8: it's good for our business, it's good for growth, it's

539
00:29:07,880 --> 00:29:13,560
Speaker 8: obviously good for our users. Everyone at Roadblocks understands that, yes,

540
00:29:13,640 --> 00:29:18,040
Speaker 8: sometimes we slow things down to account for safety issues.

541
00:29:18,520 --> 00:29:22,120
Speaker 8: For example, when we first start developing new features, we'll

542
00:29:22,160 --> 00:29:24,720
Speaker 8: sit down with those feature teams and talk about risks

543
00:29:24,720 --> 00:29:27,440
Speaker 8: associated with the feature and make sure we address those

544
00:29:27,480 --> 00:29:29,840
Speaker 8: at the very beginning, rather than waiting to tell the

545
00:29:29,920 --> 00:29:32,720
Speaker 8: very end and trying to fix something we build safety

546
00:29:32,720 --> 00:29:34,160
Speaker 8: in at the beginning.

547
00:29:35,080 --> 00:29:38,800
Speaker 5: You're not the only company that, in many ways both

548
00:29:38,880 --> 00:29:41,080
Speaker 5: has led a genie out of a bottle here because

549
00:29:41,120 --> 00:29:44,840
Speaker 5: of those few bad actors. And I'm interested is ultimately

550
00:29:46,040 --> 00:29:49,680
Speaker 5: how the focus shifts from being one where companies are

551
00:29:49,720 --> 00:29:51,959
Speaker 5: the ones we turn to to fix this, or how

552
00:29:52,000 --> 00:29:54,600
Speaker 5: much the onus is on the parents, on the children,

553
00:29:54,840 --> 00:29:59,000
Speaker 5: on the users. Where do you think that balance lies, Matt, I.

554
00:29:58,960 --> 00:30:02,680
Speaker 8: Think companies need to take a proactive stance to keeping

555
00:30:03,080 --> 00:30:06,120
Speaker 8: their users safe on the platform. Obviously, we want to

556
00:30:06,160 --> 00:30:09,120
Speaker 8: provide as many systems and controls and insights to parents

557
00:30:09,160 --> 00:30:12,480
Speaker 8: as possible, but from our perspective, we need to make

558
00:30:12,520 --> 00:30:15,080
Speaker 8: the assumption that parents aren't always in the room and

559
00:30:15,160 --> 00:30:18,400
Speaker 8: aren't always involved, and that's okay. Parents are really busy.

560
00:30:18,760 --> 00:30:22,080
Speaker 8: They're spread between their jobs and all kinds of different

561
00:30:22,120 --> 00:30:24,920
Speaker 8: things going on, and we need to play a proactive role.

562
00:30:25,280 --> 00:30:27,640
Speaker 8: I think when you're talking about like the larger problem

563
00:30:27,720 --> 00:30:30,880
Speaker 8: that we face, I think it's almost like it's a

564
00:30:30,880 --> 00:30:33,440
Speaker 8: problem across the Internet. And so the other thing that

565
00:30:33,480 --> 00:30:35,800
Speaker 8: we think a lot about on roadblocks is how do

566
00:30:35,840 --> 00:30:39,040
Speaker 8: we keep people safe on roadblocks? Knowing that our users

567
00:30:39,120 --> 00:30:42,480
Speaker 8: also have accounts on other platforms, and they have phone

568
00:30:42,560 --> 00:30:45,640
Speaker 8: numbers and cell phones and things like that, and we

569
00:30:45,720 --> 00:30:49,000
Speaker 8: do our best to keep people on roadblocks, so we

570
00:30:49,040 --> 00:30:52,959
Speaker 8: try and block people from sharing personal identifiable information. We

571
00:30:53,000 --> 00:30:56,760
Speaker 8: try and block people from directing users to other platforms

572
00:30:57,040 --> 00:31:00,120
Speaker 8: because we know that we can keep roadblocks safe and

573
00:31:00,160 --> 00:31:03,400
Speaker 8: we know that we go through great ends to keep

574
00:31:03,440 --> 00:31:06,560
Speaker 8: the platform safe, but other platforms don't operate at the

575
00:31:06,560 --> 00:31:09,080
Speaker 8: same standards that we do, so it's very important for

576
00:31:09,200 --> 00:31:11,200
Speaker 8: us to do what we can to make sure that

577
00:31:11,240 --> 00:31:13,440
Speaker 8: people stay safe on roadblocks while they're there.

578
00:31:13,840 --> 00:31:17,160
Speaker 5: Web Rocks Chief safety Officer Matt Kaufman, thanks for your

579
00:31:17,200 --> 00:31:21,000
Speaker 5: time today on their latest protections being unveiled. Meanwhile, we

580
00:31:21,200 --> 00:31:24,360
Speaker 5: shift from particular corporate story over to what's happening in

581
00:31:24,400 --> 00:31:25,560
Speaker 5: the broader markets right now.

582
00:31:25,640 --> 00:31:27,160
Speaker 4: We actually want to shine like what's happening in.

583
00:31:27,160 --> 00:31:31,040
Speaker 5: Crypto because it is continuing to sustain it's rally. Bitcoin

584
00:31:31,160 --> 00:31:33,400
Speaker 5: up to now ninety two thousand. We're off of that

585
00:31:33,760 --> 00:31:35,880
Speaker 5: key high reached in the previous week, but we're still

586
00:31:35,960 --> 00:31:38,640
Speaker 5: up a percentage point on the day. Micro Strategy extraordinary

587
00:31:38,720 --> 00:31:40,760
Speaker 5: run more than four hundred percent higher over the last

588
00:31:41,240 --> 00:31:42,040
Speaker 5: year trading.

589
00:31:42,360 --> 00:31:43,719
Speaker 4: We're up another six percent.

590
00:31:43,880 --> 00:31:47,040
Speaker 5: The company's going to be selling yet more convertible debt

591
00:31:47,040 --> 00:31:48,520
Speaker 5: to be able to buy you guessed.

592
00:31:48,240 --> 00:31:49,240
Speaker 4: It yet more bitcoin.

593
00:31:49,520 --> 00:31:52,080
Speaker 5: They've just added more than fifty six thousand already, they

594
00:31:52,120 --> 00:31:54,680
Speaker 5: have more than thirty billion dollars worth on their balance sheet.

595
00:31:54,680 --> 00:31:56,240
Speaker 4: We're looking at backed holdings as well.

596
00:31:56,360 --> 00:32:01,040
Speaker 5: Interesting story that maybe Trump True Social might well be

597
00:32:01,120 --> 00:32:04,840
Speaker 5: purchasing backed of course, a marketplace for crypto coming up

598
00:32:05,160 --> 00:32:07,200
Speaker 5: Trump that we have Trump Media currently off by seven

599
00:32:07,200 --> 00:32:08,600
Speaker 5: percent on the back of that reporting.

600
00:32:08,720 --> 00:32:10,280
Speaker 4: Tech firms, meanwhile, are.

601
00:32:10,280 --> 00:32:13,960
Speaker 5: Racing to develop AI powered agents to help improve efficiency

602
00:32:13,960 --> 00:32:16,120
Speaker 5: across sales industry, and one of the leaders in the

603
00:32:16,120 --> 00:32:19,760
Speaker 5: field is Rocks Company CEO and co founder Shan Mukerje.

604
00:32:19,920 --> 00:32:21,880
Speaker 4: It's going to be joining us. This is Bloomberg Technology.

605
00:32:30,880 --> 00:32:33,320
Speaker 5: There is a wave of startups and bigger companies that

606
00:32:33,360 --> 00:32:36,960
Speaker 5: are focusing in on developing AI powered agents, in particular

607
00:32:36,960 --> 00:32:41,160
Speaker 5: for salespeople, all in an effort to more efficiently complete tasks.

608
00:32:40,720 --> 00:32:43,920
Speaker 4: With minimal human supervision. One of them is Rocks, who.

609
00:32:43,720 --> 00:32:47,600
Speaker 5: Today announces a fifty million dollar raise here with more

610
00:32:47,720 --> 00:32:52,520
Speaker 5: Rock CEO and co founder Ishan Mukerji. Ishan, first things first,

611
00:32:52,560 --> 00:32:54,680
Speaker 5: before we get to the exact details of the numbers,

612
00:32:55,080 --> 00:32:58,160
Speaker 5: is how many companies do you estimate are out there

613
00:32:58,200 --> 00:33:00,400
Speaker 5: building agents for enterprize.

614
00:33:01,880 --> 00:33:05,920
Speaker 9: There's quite a few have a growing incumbent kind of

615
00:33:05,960 --> 00:33:09,760
Speaker 9: market of vendors who are repositioning themselves as the I companies.

616
00:33:10,280 --> 00:33:12,960
Speaker 9: And then because of the barriers of entry being lowered,

617
00:33:13,360 --> 00:33:16,000
Speaker 9: there is always early stage companies entering the market.

618
00:33:16,400 --> 00:33:21,840
Speaker 5: And yet Sequoia picks you, General Catalyst picks you. What

619
00:33:21,920 --> 00:33:25,400
Speaker 5: are you offering that's different that Salesforce isn't already plowing

620
00:33:25,440 --> 00:33:28,239
Speaker 5: money into Microsoft isn't doing or some of the other

621
00:33:28,280 --> 00:33:29,600
Speaker 5: startups that are in the field.

622
00:33:31,000 --> 00:33:35,440
Speaker 9: Yeah, at Rocks, we're building AI that focuses on accelerating

623
00:33:35,520 --> 00:33:39,600
Speaker 9: revenue for the largest enterprises. Caroline, what we're seeing is

624
00:33:40,000 --> 00:33:43,800
Speaker 9: every business owner across every vertical is blowing out the

625
00:33:43,880 --> 00:33:47,800
Speaker 9: revenue targets for next year. Some of our customers are

626
00:33:47,800 --> 00:33:51,840
Speaker 9: doubling the revenue targets when none of them are doubling

627
00:33:51,840 --> 00:33:55,920
Speaker 9: the team size. So the interesting thing about sales is

628
00:33:55,960 --> 00:33:58,960
Speaker 9: it follows the power law. Ninety percent of the revenue

629
00:33:58,960 --> 00:34:03,040
Speaker 9: in a scaled enterprise comes in through your largest enterprise customers,

630
00:34:03,360 --> 00:34:07,440
Speaker 9: which is supported by a top fifteen percent of your employees. Okay,

631
00:34:08,280 --> 00:34:10,359
Speaker 9: we have to w revenue. You have to get more

632
00:34:10,360 --> 00:34:12,480
Speaker 9: out of like the top fifteen percent of the employees.

633
00:34:12,960 --> 00:34:16,120
Speaker 5: Okay, so you're saying with your technology, with your AI agents,

634
00:34:16,200 --> 00:34:21,040
Speaker 5: you're sort of super powering, augmenting the humans that are

635
00:34:21,080 --> 00:34:24,120
Speaker 5: currently being hired. But my question remains sort of why

636
00:34:24,160 --> 00:34:26,479
Speaker 5: wouldn't someone do that through Microsoft? How can you ensure

637
00:34:26,480 --> 00:34:29,080
Speaker 5: you continue to build your pipeline of users and indeed,

638
00:34:29,200 --> 00:34:31,480
Speaker 5: with a fifty million that you currently raised, be able to.

639
00:34:31,440 --> 00:34:31,920
Speaker 4: Tackle that.

640
00:34:33,680 --> 00:34:36,719
Speaker 9: To be essential to an enterprise and essentially securing and

641
00:34:36,760 --> 00:34:42,000
Speaker 9: growing their customers. Like, the core advantage of AI is

642
00:34:42,040 --> 00:34:44,400
Speaker 9: to be able to deeply couple with the data layer.

643
00:34:44,480 --> 00:34:47,640
Speaker 9: So at Rocks we build the first enterprise ready AI

644
00:34:47,680 --> 00:34:51,360
Speaker 9: agent swarm that's powered by a warehouse native system of record.

645
00:34:51,680 --> 00:34:55,320
Speaker 9: So Rocks we've seen amazing kind of early adoption because

646
00:34:55,360 --> 00:34:57,799
Speaker 9: of the part of the data that we aggregate that

647
00:34:57,880 --> 00:34:58,920
Speaker 9: fuels these agents.

648
00:35:00,000 --> 00:35:02,880
Speaker 5: Co founder Shan Muckergee, we appreciate you coming on to

649
00:35:02,880 --> 00:35:06,200
Speaker 5: talk about the rays and the company. Meanwhile, Huawei's ambitions

650
00:35:06,200 --> 00:35:09,879
Speaker 5: to create more powerful chips, FRAI and smartphones has hit

651
00:35:09,960 --> 00:35:12,680
Speaker 5: some madjor snags because of US sanctions. This is the

652
00:35:12,760 --> 00:35:16,719
Speaker 5: Chinese firm is designing its next to ascend processors. It's

653
00:35:16,760 --> 00:35:19,920
Speaker 5: answered basically to Nvidia's dominant accelerators and so all, according

654
00:35:19,920 --> 00:35:22,440
Speaker 5: to sources who say this means that the firm's marquee

655
00:35:22,520 --> 00:35:26,200
Speaker 5: chips will be stuck at aging technology until at least

656
00:35:26,200 --> 00:35:30,000
Speaker 5: twenty twenty six. Meanwhile, let's talk about innovation over it

657
00:35:30,000 --> 00:35:32,680
Speaker 5: in video, it's teaming up with Google to pursue another

658
00:35:32,719 --> 00:35:38,360
Speaker 5: technology once well relegated the science fiction quantum computing. Bloomberg's

659
00:35:38,400 --> 00:35:40,680
Speaker 5: Inking has the details. We were just talking about quantum

660
00:35:40,719 --> 00:35:43,319
Speaker 5: computing with a market analyst yesterday, and here we have

661
00:35:43,440 --> 00:35:44,719
Speaker 5: a new deal being forged.

662
00:35:44,480 --> 00:35:45,640
Speaker 4: Between in Video and Google.

663
00:35:46,000 --> 00:35:48,600
Speaker 5: What are they going to be doing together to progress us?

664
00:35:49,640 --> 00:35:52,040
Speaker 10: Well, this is basically in Video saying hey, this is

665
00:35:52,080 --> 00:35:57,920
Speaker 10: another use of our technology RAAI chips and systems and software,

666
00:35:58,239 --> 00:36:00,960
Speaker 10: and hey look there's a really important company called Google

667
00:36:01,000 --> 00:36:03,600
Speaker 10: that you've heard of that's using this to try to

668
00:36:03,840 --> 00:36:08,719
Speaker 10: push the whole quantum computing world forward early stages. But

669
00:36:08,840 --> 00:36:12,600
Speaker 10: basically Google is going to be using an in video supercomputer.

670
00:36:13,880 --> 00:36:17,040
Speaker 5: Okay, in videos supercomputer will then be able to strip

671
00:36:17,120 --> 00:36:19,400
Speaker 5: out so called noises. You say, what exactly adds to

672
00:36:19,440 --> 00:36:21,120
Speaker 5: the noise? What is it that the innovation we're seeing

673
00:36:21,160 --> 00:36:23,520
Speaker 5: right here right now that can cut that down or

674
00:36:23,520 --> 00:36:26,520
Speaker 5: indeed understand how it impacts the future of quantum computing.

675
00:36:27,000 --> 00:36:29,520
Speaker 10: Yeah, I mean, we're using the phrase quantum computing and

676
00:36:29,560 --> 00:36:33,879
Speaker 10: everybody throws it around, but it doesn't really exist. There

677
00:36:33,880 --> 00:36:38,600
Speaker 10: are components, or the early stage components that are attempts

678
00:36:38,600 --> 00:36:43,319
Speaker 10: to use quantum mechanics to basically represent information and to

679
00:36:43,440 --> 00:36:47,279
Speaker 10: replace standard semiconductors. We're at the really early stages of

680
00:36:47,320 --> 00:36:50,399
Speaker 10: that and actually deciding what is information and what is

681
00:36:50,680 --> 00:36:53,200
Speaker 10: just random noise generated by the environment.

682
00:36:53,440 --> 00:36:54,520
Speaker 4: Is it really difficult?

683
00:36:54,600 --> 00:36:57,080
Speaker 10: What in video is saying is hey, use our technology

684
00:36:57,280 --> 00:37:01,120
Speaker 10: and you'll be able to develop components more quickly that

685
00:37:01,480 --> 00:37:04,240
Speaker 10: can do that by simulating that environment.

686
00:37:04,680 --> 00:37:06,759
Speaker 5: So when video wants to ensure that we still think

687
00:37:06,760 --> 00:37:10,000
Speaker 5: of it going forward past into quantum computing.

688
00:37:10,280 --> 00:37:12,799
Speaker 4: Google meanwhile is in on the app.

689
00:37:12,600 --> 00:37:15,120
Speaker 5: Just because of the sheer force of data processing that

690
00:37:15,160 --> 00:37:18,040
Speaker 5: will be achieved under this what is currently still in

691
00:37:18,080 --> 00:37:18,960
Speaker 5: the bounds of fiction.

692
00:37:19,960 --> 00:37:20,920
Speaker 4: Yeah, exactly.

693
00:37:20,960 --> 00:37:23,680
Speaker 10: You said we'll be achieved, and some people would say

694
00:37:23,719 --> 00:37:27,600
Speaker 10: that's an inevitable future. Jensen Wang I believe said, you know,

695
00:37:27,640 --> 00:37:30,840
Speaker 10: a couple of decades, more than a decade. Others have

696
00:37:30,960 --> 00:37:35,200
Speaker 10: said never, So you take your pick. But certainly in

697
00:37:35,280 --> 00:37:38,200
Speaker 10: video is definitely covering its bets and again trying to

698
00:37:38,239 --> 00:37:41,360
Speaker 10: push forward this idea that hey, we have this incredible

699
00:37:41,400 --> 00:37:43,960
Speaker 10: system for accelerating whatever it is you're working on, even

700
00:37:44,000 --> 00:37:46,359
Speaker 10: if it's something that will replace what we do, we

701
00:37:46,400 --> 00:37:48,640
Speaker 10: can get you there use our stuff.

702
00:37:49,120 --> 00:37:51,640
Speaker 5: Or the narrative that builds ahead of their earnings in

703
00:37:52,920 --> 00:37:55,120
Speaker 5: little more than twenty four hours time. In King's going

704
00:37:55,160 --> 00:37:56,520
Speaker 5: to be all across it. We thank you with the

705
00:37:56,600 --> 00:38:07,320
Speaker 5: latest don in video. Later today, SpaceX it slated to launch.

706
00:38:07,400 --> 00:38:11,840
Speaker 5: It's got Gamchuan starship rocket once again out of South Texas,

707
00:38:11,880 --> 00:38:14,240
Speaker 5: and it's a key test that is expected to include

708
00:38:14,280 --> 00:38:15,120
Speaker 5: a guest visit from.

709
00:38:15,040 --> 00:38:17,920
Speaker 4: President elect Donald Trump and begs. Lauren Crush joins.

710
00:38:17,719 --> 00:38:19,560
Speaker 5: Us right here when New York come luckily enough to

711
00:38:19,600 --> 00:38:21,880
Speaker 5: say yes, So what is.

712
00:38:21,880 --> 00:38:23,120
Speaker 4: It that they're testing that? Is it going to be

713
00:38:23,160 --> 00:38:23,799
Speaker 4: a repeat the.

714
00:38:23,800 --> 00:38:28,399
Speaker 5: Chopstick fun incredible viewpoint that the whole internet went well, right, Yeah,

715
00:38:28,400 --> 00:38:31,560
Speaker 5: that's stunning booster ketch that everyone was shocked over.

716
00:38:31,680 --> 00:38:34,080
Speaker 11: They're going to try and do that again and we're

717
00:38:34,120 --> 00:38:37,200
Speaker 11: probably going to see that happen many, many times. The

718
00:38:37,239 --> 00:38:39,040
Speaker 11: goal of these tests is to make it so that

719
00:38:39,160 --> 00:38:41,480
Speaker 11: is perfect right, and so they're going to see if

720
00:38:41,480 --> 00:38:44,600
Speaker 11: they can do it again perfect that booster ketch as

721
00:38:44,640 --> 00:38:47,920
Speaker 11: well as launch Starship around the world and then also

722
00:38:48,000 --> 00:38:50,759
Speaker 11: bring that back to Earth through the atmosphere and hope

723
00:38:50,800 --> 00:38:52,759
Speaker 11: that it survives its fiery plunge.

724
00:38:53,239 --> 00:38:54,839
Speaker 4: Just remind us how difficult this is.

725
00:38:55,000 --> 00:38:58,160
Speaker 5: Just how much bigger a rocket that it currently is

726
00:38:58,200 --> 00:38:59,520
Speaker 5: compared to what's already been tried.

727
00:38:59,600 --> 00:39:03,360
Speaker 11: Right, Starship is by far the most powerful, the largest

728
00:39:03,480 --> 00:39:06,640
Speaker 11: rocket that humans have ever developed. I mean, it's the

729
00:39:06,760 --> 00:39:09,920
Speaker 11: key to what SpaceX was formed for, right, to send

730
00:39:10,160 --> 00:39:13,839
Speaker 11: humans to deep space like the Moon and Mars. It's

731
00:39:14,000 --> 00:39:17,319
Speaker 11: it's a behemoth of a vehicle and it has, you know,

732
00:39:17,480 --> 00:39:19,200
Speaker 11: dozens of engines.

733
00:39:18,800 --> 00:39:19,399
Speaker 4: At its base.

734
00:39:20,000 --> 00:39:22,160
Speaker 11: So you know the fact that they can do this

735
00:39:22,239 --> 00:39:24,399
Speaker 11: alone is a feat. But really they need to get

736
00:39:24,400 --> 00:39:26,600
Speaker 11: to a place where they're comfortable with doing this over

737
00:39:26,600 --> 00:39:29,640
Speaker 11: and over again. I mean just last week, President Gwnschotwell

738
00:39:29,680 --> 00:39:33,080
Speaker 11: said that she anticipates launching Starship up to four hundred

739
00:39:33,120 --> 00:39:35,480
Speaker 11: times over the next four years. You know, So this

740
00:39:35,640 --> 00:39:37,600
Speaker 11: is this is something that they want to make routine,

741
00:39:37,960 --> 00:39:38,839
Speaker 11: four hundred.

742
00:39:38,560 --> 00:39:41,000
Speaker 5: Times to do what exactly like this is where the money,

743
00:39:41,080 --> 00:39:44,120
Speaker 5: the revenue stream comes in we just sort of phenomenal valuation.

744
00:39:43,760 --> 00:39:46,320
Speaker 4: For the company. Part of that is the government contracts

745
00:39:46,320 --> 00:39:47,280
Speaker 4: they can win, right.

746
00:39:47,239 --> 00:39:50,960
Speaker 11: I mean, essentially Starship is going to replace the vehicles

747
00:39:50,960 --> 00:39:53,960
Speaker 11: that they're currently using to launch humans and cargo to orbit.

748
00:39:54,440 --> 00:39:56,040
Speaker 4: That won't happen for a while.

749
00:39:55,760 --> 00:39:58,359
Speaker 11: But they want to use Starship to send their next

750
00:39:58,360 --> 00:40:02,080
Speaker 11: generation Starlink satellites into orbit. For instance, the goal is

751
00:40:02,080 --> 00:40:04,920
Speaker 11: to use this as a lander for NASA's Artemis program

752
00:40:05,000 --> 00:40:07,799
Speaker 11: to land people on the Moon and then eventually to

753
00:40:07,840 --> 00:40:10,960
Speaker 11: send humans to Mars somedays. So it really is going

754
00:40:11,040 --> 00:40:14,120
Speaker 11: to be kind of the crux of SpaceX's business plan

755
00:40:14,239 --> 00:40:15,320
Speaker 11: moving forward.

756
00:40:15,400 --> 00:40:17,120
Speaker 5: On which might well be helped by the fact that,

757
00:40:17,160 --> 00:40:18,920
Speaker 5: you know, Musk is so close to the president of

758
00:40:18,960 --> 00:40:21,840
Speaker 5: that Trump who might well be there to view.

759
00:40:21,680 --> 00:40:22,040
Speaker 4: All of this.

760
00:40:23,120 --> 00:40:26,720
Speaker 5: What is the next technological feat that's necessary? Can't even

761
00:40:26,800 --> 00:40:29,080
Speaker 5: get in this current capacity to someone.

762
00:40:28,920 --> 00:40:30,400
Speaker 4: Light mod so glad you asked. So.

763
00:40:30,600 --> 00:40:34,319
Speaker 11: As impressive as these test flights are, there's still quite

764
00:40:34,320 --> 00:40:36,480
Speaker 11: a long way to go. So to get to deep

765
00:40:36,480 --> 00:40:40,680
Speaker 11: space destinations, for instance, space our starship has to refuel

766
00:40:40,760 --> 00:40:43,600
Speaker 11: in orbit, and that's something that's never really been done

767
00:40:43,680 --> 00:40:46,759
Speaker 11: at this scale before, so Essentially, what they'll have to

768
00:40:46,800 --> 00:40:50,000
Speaker 11: do is launch starships and then launch multiple back to

769
00:40:50,040 --> 00:40:53,799
Speaker 11: back starships that will dock in orbit and refuel kind

770
00:40:53,800 --> 00:40:55,919
Speaker 11: of like you know, filling up at the gas tank

771
00:40:55,960 --> 00:40:58,560
Speaker 11: over and over again, and then once it has enough fuel,

772
00:40:58,640 --> 00:40:59,280
Speaker 11: then they can.

773
00:40:59,160 --> 00:40:59,839
Speaker 4: Go on to the moon.

774
00:41:00,560 --> 00:41:02,759
Speaker 11: But that's still a while away, you know, and.

775
00:41:02,719 --> 00:41:04,800
Speaker 4: They still need to perfect this booster catch.

776
00:41:04,960 --> 00:41:06,879
Speaker 11: They need to make sure that starship can come back

777
00:41:06,880 --> 00:41:09,040
Speaker 11: to Earth in one piece. I mean, they were able

778
00:41:09,080 --> 00:41:11,200
Speaker 11: to do that last time, but we still saw some

779
00:41:11,280 --> 00:41:14,160
Speaker 11: pieces of it burning away, so you know, it's it's

780
00:41:14,200 --> 00:41:14,560
Speaker 11: still a.

781
00:41:14,520 --> 00:41:15,200
Speaker 4: Work in progress.

782
00:41:15,239 --> 00:41:17,080
Speaker 11: But that's what each of these tests are for, right

783
00:41:17,120 --> 00:41:20,279
Speaker 11: It's one step forward into that direction. But yes, it's

784
00:41:20,280 --> 00:41:22,600
Speaker 11: going to be some time before anyone is launching on

785
00:41:22,640 --> 00:41:25,440
Speaker 11: this vehicle, let alone going to the moon of Mars.

786
00:41:25,680 --> 00:41:29,040
Speaker 5: All Eyes around five pm today Eastern Lauren Grush, so

787
00:41:29,120 --> 00:41:29,839
Speaker 5: great to have you here.

788
00:41:30,000 --> 00:41:30,360
Speaker 4: Thank you.

789
00:41:30,560 --> 00:41:32,440
Speaker 5: I Meanwhile, that does it for this edition of BLUEBG

790
00:41:32,440 --> 00:41:35,320
Speaker 5: Technology June. In tomorrow, we've got a great interview with

791
00:41:35,360 --> 00:41:36,719
Speaker 5: the quillcom Cio Cristiano.

792
00:41:36,760 --> 00:41:39,440
Speaker 4: I'm on on all things you guessed it, chips and AI.

793
00:41:40,080 --> 00:41:41,280
Speaker 4: This is Blue meg Technology