1
00:00:02,520 --> 00:00:20,640
Speaker 1: Bloomberg Audio Studios, podcasts, radio news.

2
00:00:22,800 --> 00:00:26,480
Speaker 2: This is Bloomberg Tech coming up. SpaceX continues to rocket,

3
00:00:26,640 --> 00:00:30,080
Speaker 2: eclipsing Amazon in value this as it lands it's sixty

4
00:00:30,120 --> 00:00:34,160
Speaker 2: billion dollar takeover of Cursor, plus Anthropic meets with US

5
00:00:34,200 --> 00:00:37,320
Speaker 2: officials to resolve a national security dispute with the Trump

6
00:00:37,360 --> 00:00:41,879
Speaker 2: administration over its most advanced AI models and Sequoia partner

7
00:00:41,960 --> 00:00:45,800
Speaker 2: SpaceX investor and AI thinker Sean maguire, on set in

8
00:00:45,840 --> 00:00:48,120
Speaker 2: San Francisco, is to go deep on the Cursor deal

9
00:00:48,440 --> 00:00:52,480
Speaker 2: and Orbital Data centers. Tuesday, June sixteenth, twenty twenty six,

10
00:00:52,800 --> 00:00:56,120
Speaker 2: SpaceX its third day of trading in the public markets,

11
00:00:56,160 --> 00:00:59,800
Speaker 2: and it continues to soar. It's all about valuation in

12
00:00:59,840 --> 00:01:02,760
Speaker 2: the here and now. We are currently on track for

13
00:01:02,800 --> 00:01:06,960
Speaker 2: SpaceX to eclipse Amazon in market value. It's been a

14
00:01:07,040 --> 00:01:11,160
Speaker 2: volatile session. One point, the stock up seventeen percent and

15
00:01:11,360 --> 00:01:16,319
Speaker 2: in just a brief moment, eclipse Microsoft in terms of

16
00:01:16,360 --> 00:01:20,080
Speaker 2: its market value in the moment. Indeed, but here's the thing.

17
00:01:20,200 --> 00:01:24,000
Speaker 2: When you think about SpaceX's relative value too Microsoft or Amazon,

18
00:01:24,040 --> 00:01:26,720
Speaker 2: there's a catalyst in the news cycle, and that is

19
00:01:27,080 --> 00:01:30,080
Speaker 2: it will move forward with a sixty billion dollar deal

20
00:01:30,440 --> 00:01:33,160
Speaker 2: for Cursor in the AI coding space. Let's get the

21
00:01:33,160 --> 00:01:36,520
Speaker 2: details on that deal with Blimbox and Attasha Mascarinus, who

22
00:01:36,520 --> 00:01:39,280
Speaker 2: covers AI and venture and here with us in SF

23
00:01:39,360 --> 00:01:42,720
Speaker 2: Let's start with the details the structure. There's also the

24
00:01:42,800 --> 00:01:45,240
Speaker 2: idea that in April they had the right to acquire

25
00:01:45,240 --> 00:01:48,400
Speaker 2: a Cursor just days after the going public they.

26
00:01:48,200 --> 00:01:51,120
Speaker 3: Are exactly, I mean, in April weys our first of

27
00:01:51,160 --> 00:01:54,600
Speaker 3: its kind announcement an agreement to have the option to acquire.

28
00:01:55,040 --> 00:01:57,360
Speaker 3: Now we have the news that they are moving ahead

29
00:01:57,680 --> 00:02:01,040
Speaker 3: with the agreement and they will plan to acquire Cursor.

30
00:02:01,080 --> 00:02:03,240
Speaker 3: It's going to be for that price tag that we expected,

31
00:02:03,560 --> 00:02:07,920
Speaker 3: a sixty billion dollar implied valuation, and the Cursor investors

32
00:02:08,080 --> 00:02:09,760
Speaker 3: will be getting stock as a.

33
00:02:09,680 --> 00:02:11,120
Speaker 4: Result of it, SpaceX stock.

34
00:02:11,000 --> 00:02:12,560
Speaker 3: Sorry, SpaceX stock as a result of it.

35
00:02:12,639 --> 00:02:13,119
Speaker 4: Exactly.

36
00:02:13,560 --> 00:02:15,600
Speaker 2: So there's a big question here about like, well, who

37
00:02:15,639 --> 00:02:19,760
Speaker 2: wins on the Cursor side. Cursor has been active in

38
00:02:19,840 --> 00:02:23,440
Speaker 2: raising money in recent years. They had a lot of momentum.

39
00:02:23,760 --> 00:02:25,360
Speaker 2: You have a pretty good understanding of who's on the

40
00:02:25,360 --> 00:02:26,079
Speaker 2: cat table there.

41
00:02:26,360 --> 00:02:27,679
Speaker 3: Yeah, I mean, this has been one of the most

42
00:02:27,720 --> 00:02:31,720
Speaker 3: explosive AI companies outside of open an Anthropic. If you

43
00:02:31,760 --> 00:02:34,320
Speaker 3: ask me this is the number one other company I've

44
00:02:34,320 --> 00:02:36,679
Speaker 3: been focused on for the past year. We think about

45
00:02:36,680 --> 00:02:39,200
Speaker 3: the biggest winners from this deal when the deal does close,

46
00:02:39,240 --> 00:02:41,920
Speaker 3: which is expecting in a few months, it's actually some

47
00:02:41,960 --> 00:02:45,480
Speaker 3: of the same investors for SpaceX. Unsurprisingly, two of the

48
00:02:45,480 --> 00:02:49,960
Speaker 3: biggest investors in Cursor are Thrive Capital and Injuries and

49
00:02:50,000 --> 00:02:54,080
Speaker 3: Horowitz and you know last week we heard them celebrate

50
00:02:54,120 --> 00:02:56,720
Speaker 3: as well when it came to the SpaceX IPO. And

51
00:02:56,760 --> 00:02:59,880
Speaker 3: so Thrive is expected to have a stake worth north

52
00:03:00,080 --> 00:03:02,880
Speaker 3: of ten billion according to sources. When we think about

53
00:03:02,880 --> 00:03:06,080
Speaker 3: the SpaceX and Cursor combination, and Andresa, you know, heading

54
00:03:06,120 --> 00:03:09,760
Speaker 3: into this, had about six billion worth of stake in Cursor,

55
00:03:09,800 --> 00:03:11,440
Speaker 3: and so another kind.

56
00:03:11,280 --> 00:03:13,760
Speaker 2: Of they already reported that, you know, for Indreeson on

57
00:03:13,760 --> 00:03:16,360
Speaker 2: the SPACEXIPO, this would be the biggest payoff in the

58
00:03:16,480 --> 00:03:18,480
Speaker 2: in the firm's history. Let's go to the AI bit

59
00:03:18,520 --> 00:03:20,720
Speaker 2: of it. I think Elon Musk has been pretty honest

60
00:03:21,120 --> 00:03:25,600
Speaker 2: that in the coding space XAI had been behind your anthropics,

61
00:03:25,600 --> 00:03:29,440
Speaker 2: your open AI is even what is SpaceX doing with Cursor?

62
00:03:29,520 --> 00:03:30,400
Speaker 2: What's the plan here?

63
00:03:30,639 --> 00:03:34,040
Speaker 3: Yeah, I mean people forget that SpaceX actually merging with

64
00:03:34,160 --> 00:03:37,800
Speaker 3: XAI is still a recent move. Yeah, I mean the

65
00:03:37,840 --> 00:03:40,800
Speaker 3: ink is still you know, drying on that one. But yeah,

66
00:03:40,800 --> 00:03:43,400
Speaker 3: I mean, listen, it's the same goal. Elon wants to

67
00:03:43,520 --> 00:03:47,120
Speaker 3: strengthen their AI talent Cursor. I mean, when I was

68
00:03:47,120 --> 00:03:49,920
Speaker 3: first writing about Cursor, they were bringing on people to

69
00:03:50,040 --> 00:03:53,160
Speaker 3: their office for two weeks before they made an acquisition,

70
00:03:53,480 --> 00:03:56,240
Speaker 3: before they made a hiring offer, and so they really

71
00:03:56,280 --> 00:03:59,000
Speaker 3: spent a huge amount of time in recruiting the best

72
00:03:59,000 --> 00:04:01,920
Speaker 3: AI talent in samp and Cisco, building a very unique

73
00:04:01,960 --> 00:04:05,320
Speaker 3: kind of culture. And now Elon Musk is getting you know,

74
00:04:05,360 --> 00:04:08,480
Speaker 3: that highly vetted talent as he looks to make SpaceX

75
00:04:08,520 --> 00:04:10,960
Speaker 3: ANAI company, and so there's a lot of interest in

76
00:04:11,000 --> 00:04:14,240
Speaker 3: the talent. We'll see if all people do come on

77
00:04:14,400 --> 00:04:17,480
Speaker 3: as a result of the acquisition, but it's hundreds of

78
00:04:17,480 --> 00:04:19,520
Speaker 3: employees joining the spacexta.

79
00:04:19,040 --> 00:04:22,640
Speaker 2: And even before spacexxai closed there was high turnover at XAI.

80
00:04:22,760 --> 00:04:25,240
Speaker 2: They had people departing from the founding group, et cetera.

81
00:04:25,240 --> 00:04:27,760
Speaker 2: Bloe miss and attached Miss Garinus on the Cursor deal.

82
00:04:28,000 --> 00:04:31,039
Speaker 2: Let's get more in SpaceX's surging valuation in the public markets.

83
00:04:31,080 --> 00:04:34,320
Speaker 2: Bloomberg Equities reported common Rhinikey joins us with the latest,

84
00:04:34,320 --> 00:04:36,719
Speaker 2: so right now on track to eclipse Amazon. You and

85
00:04:36,760 --> 00:04:39,760
Speaker 2: I have been talking all morning about price to sales ratios,

86
00:04:40,040 --> 00:04:43,040
Speaker 2: and then there's the volatility in the session itself. Explain

87
00:04:43,080 --> 00:04:43,360
Speaker 2: it all.

88
00:04:43,760 --> 00:04:46,120
Speaker 5: Yeah, So there are two things that are really important here.

89
00:04:46,120 --> 00:04:48,360
Speaker 5: The first one is that we have really low float

90
00:04:48,440 --> 00:04:50,960
Speaker 5: actually trading for SpaceX. We know it is about five

91
00:04:50,960 --> 00:04:54,000
Speaker 5: percent or less than five percent of total shares became

92
00:04:54,040 --> 00:04:57,320
Speaker 5: available in the io. That can make trading very volatile interday,

93
00:04:57,400 --> 00:05:00,200
Speaker 5: and it can elevate stock prices because there's just not

94
00:05:00,360 --> 00:05:02,560
Speaker 5: that much float on the market.

95
00:05:02,440 --> 00:05:04,520
Speaker 6: So that can elevate the stock price.

96
00:05:04,680 --> 00:05:07,760
Speaker 5: But then the total shares outstanding, even ones that are

97
00:05:07,760 --> 00:05:10,200
Speaker 5: locked up right now, is what's used to calculate the

98
00:05:10,200 --> 00:05:12,720
Speaker 5: market cap, so it can sort of elevate things there.

99
00:05:13,000 --> 00:05:15,240
Speaker 5: The second thing you brought up price to sales. This

100
00:05:15,320 --> 00:05:18,160
Speaker 5: is so important to think about. So SpaceX at about

101
00:05:18,160 --> 00:05:23,159
Speaker 5: two point seven trillion valuation right now, it's alongside Amazon

102
00:05:23,200 --> 00:05:26,960
Speaker 5: and Microsoft. But the total amount of revenue that SpaceX

103
00:05:27,000 --> 00:05:29,440
Speaker 5: is bringing in tells a very different story. So last

104
00:05:29,520 --> 00:05:32,960
Speaker 5: year in twenty twenty five, SpaceX had about nineteen billion

105
00:05:33,000 --> 00:05:37,560
Speaker 5: dollars in revenue and Microsoft, in comparison, had two hundred

106
00:05:37,560 --> 00:05:39,960
Speaker 5: and eighty one billion dollars in sales, So that's a

107
00:05:40,000 --> 00:05:43,880
Speaker 5: huge difference. It's about fifteen times more than SpaceX, even

108
00:05:43,920 --> 00:05:46,800
Speaker 5: though they're valued right now based on market cap at

109
00:05:46,839 --> 00:05:48,080
Speaker 5: a roughly the same.

110
00:05:49,279 --> 00:05:51,600
Speaker 2: It's come Rhini case SpaceX up for a third day

111
00:05:51,680 --> 00:05:55,120
Speaker 2: a lot to factor in. Thank you, Franklin Templeton BACKSPACEX

112
00:05:55,600 --> 00:05:59,880
Speaker 2: long before it's IPO. Really interesting private public market and

113
00:06:00,160 --> 00:06:02,599
Speaker 2: esther in the tech space here in the Bay Aria

114
00:06:03,040 --> 00:06:06,000
Speaker 2: Sarah Rahi Franklin Timmeton, portfolio managers with us. It's so

115
00:06:06,040 --> 00:06:07,159
Speaker 2: great to have you on the program.

116
00:06:07,240 --> 00:06:08,360
Speaker 6: Great, thank you for having me.

117
00:06:08,560 --> 00:06:11,800
Speaker 2: Timing is everything right. Let's start with the Cursor deal,

118
00:06:11,839 --> 00:06:14,040
Speaker 2: if we may. So, we knew that this could happen

119
00:06:14,120 --> 00:06:16,719
Speaker 2: because they had the right to acquire Cursor. They stayed

120
00:06:16,760 --> 00:06:19,680
Speaker 2: in April, but they've moved very quickly, you know since

121
00:06:19,839 --> 00:06:23,279
Speaker 2: since the IPO on Thursday Friday. What is it that

122
00:06:23,320 --> 00:06:26,000
Speaker 2: you think SpaceX gets with Cursor? What do you think

123
00:06:26,040 --> 00:06:26,800
Speaker 2: they intend to do?

124
00:06:27,160 --> 00:06:30,440
Speaker 7: Yeah, you know, the company has been very clear that

125
00:06:30,480 --> 00:06:32,880
Speaker 7: they want to have a role in the AI space

126
00:06:32,960 --> 00:06:36,039
Speaker 7: and that they want to be one of the competitors

127
00:06:36,040 --> 00:06:38,640
Speaker 7: within that space. And so with GROC they have the

128
00:06:38,680 --> 00:06:41,080
Speaker 7: opportunity here and now with the Cursor team, as you

129
00:06:41,080 --> 00:06:44,839
Speaker 7: were just talking about earlier, they acquihire a team here

130
00:06:44,880 --> 00:06:47,960
Speaker 7: that they can ultimately come in and potentially give them

131
00:06:47,960 --> 00:06:50,359
Speaker 7: a chance with the competitors in the space. So I

132
00:06:50,400 --> 00:06:53,359
Speaker 7: think it's brilliant. You have the successful IPO and the

133
00:06:53,400 --> 00:06:55,800
Speaker 7: ability to come in and do this deal right on

134
00:06:55,920 --> 00:06:57,120
Speaker 7: day three now.

135
00:06:57,040 --> 00:06:59,400
Speaker 6: So pretty big aquire, Pretty big acquire.

136
00:06:59,520 --> 00:07:01,640
Speaker 2: I mean the mechanics of that I find really interesting

137
00:07:01,680 --> 00:07:06,600
Speaker 2: because SpaceX's valuation, actually it's valuation today relative to Friday,

138
00:07:06,640 --> 00:07:08,680
Speaker 2: it's a bit of an advantage right in an all

139
00:07:08,720 --> 00:07:11,520
Speaker 2: stock deal. How do you feel about it? You know,

140
00:07:11,560 --> 00:07:16,360
Speaker 2: the mechanism a sixty billion dollar deal right now in

141
00:07:16,400 --> 00:07:18,800
Speaker 2: how that impacts the stock as an existing investor.

142
00:07:18,960 --> 00:07:21,200
Speaker 7: Yeah, look, we were already we knew this was going

143
00:07:21,280 --> 00:07:23,440
Speaker 7: to come, right, and so the fact that it's coming

144
00:07:23,480 --> 00:07:26,520
Speaker 7: now it makes sense. With the fact that this IPO

145
00:07:26,720 --> 00:07:29,560
Speaker 7: was very successful. I mean, they've done a fantastic job.

146
00:07:29,560 --> 00:07:32,000
Speaker 7: The Bankersterer, Goldman, Morgan Stanley did a great job bringing

147
00:07:32,000 --> 00:07:34,560
Speaker 7: this to the public markets. It actually opens up the

148
00:07:34,560 --> 00:07:36,720
Speaker 7: public markets for other deals to come and so it

149
00:07:36,720 --> 00:07:38,680
Speaker 7: makes a lot of sense for them to do this

150
00:07:38,760 --> 00:07:42,640
Speaker 7: cursor deal here now and ultimately take advantage of their

151
00:07:42,680 --> 00:07:43,239
Speaker 7: stock price.

152
00:07:43,720 --> 00:07:45,240
Speaker 2: Do you mind if we talk a little bit about

153
00:07:45,280 --> 00:07:50,760
Speaker 2: Franklin Tumbledton's history with SpaceX. You know, maybe viewers of

154
00:07:50,760 --> 00:07:52,680
Speaker 2: Bloomberg Tech will be a little bit more familiar of

155
00:07:52,920 --> 00:07:56,080
Speaker 2: kind of like the mutual funds and private growth teamams

156
00:07:56,120 --> 00:07:58,800
Speaker 2: participating in a lot of the themes we cover. Franklin

157
00:07:58,800 --> 00:08:01,640
Speaker 2: Tumbleton's name that comes up, but just how you got

158
00:08:01,640 --> 00:08:04,800
Speaker 2: in and the journey now to being a public holding

159
00:08:04,800 --> 00:08:05,040
Speaker 2: in the.

160
00:08:05,000 --> 00:08:07,600
Speaker 6: Funds, Yeah, definitely. So I am.

161
00:08:08,040 --> 00:08:10,680
Speaker 7: I work on a team, the Franklin Equity team at

162
00:08:10,760 --> 00:08:14,120
Speaker 7: Franklin Templeton and here. Over a decade ago, we decided

163
00:08:14,160 --> 00:08:16,800
Speaker 7: that it was time for us to invest in private companies,

164
00:08:17,120 --> 00:08:20,440
Speaker 7: not only through our mutual funds, but also we have

165
00:08:20,520 --> 00:08:21,680
Speaker 7: dedicated venture.

166
00:08:21,480 --> 00:08:22,960
Speaker 6: Vehicles within our group.

167
00:08:23,080 --> 00:08:25,680
Speaker 7: And the reasons we did this was, as you've been

168
00:08:25,680 --> 00:08:27,960
Speaker 7: talking about, the companies have been staying These companies have

169
00:08:27,960 --> 00:08:30,240
Speaker 7: been staying private longer, so a lot of the value

170
00:08:31,520 --> 00:08:34,800
Speaker 7: creation was happening within the private markets. Well, we have

171
00:08:35,240 --> 00:08:37,120
Speaker 7: we have, we are based in Silicon Valley, we have

172
00:08:37,160 --> 00:08:39,560
Speaker 7: the same capabilities, and so we decided we should not

173
00:08:39,600 --> 00:08:42,280
Speaker 7: only take advantage of that and generate the returns and

174
00:08:42,320 --> 00:08:45,400
Speaker 7: bring those to our shareholders, but also the insights that

175
00:08:45,440 --> 00:08:48,600
Speaker 7: you get from investing in these companies, both from just

176
00:08:48,679 --> 00:08:51,680
Speaker 7: investing in the companies and themselves and a successful IPO

177
00:08:51,800 --> 00:08:54,960
Speaker 7: here you saw with SpaceX, or just the insights and

178
00:08:55,600 --> 00:08:59,000
Speaker 7: is disruption coming to some of our existing holdings, and

179
00:08:59,040 --> 00:09:01,360
Speaker 7: how do we think about the market leaders in the

180
00:09:01,400 --> 00:09:04,679
Speaker 7: space and whether or not whether or not these private

181
00:09:04,720 --> 00:09:07,959
Speaker 7: companies will overtake them. So those insights our analysts benefit

182
00:09:08,040 --> 00:09:10,520
Speaker 7: from that. So this is something you know with SpaceX.

183
00:09:10,760 --> 00:09:13,200
Speaker 7: We knew them when they were doing reusable rockets for

184
00:09:13,400 --> 00:09:16,800
Speaker 7: cargo ships, but knowing them over time. Our team met

185
00:09:16,840 --> 00:09:19,720
Speaker 7: with Elon in twenty sixteen, I mean even prior to

186
00:09:19,760 --> 00:09:23,000
Speaker 7: that with Tesla, right, but with Elon on SpaceX and

187
00:09:23,440 --> 00:09:26,680
Speaker 7: understanding what he was going to do here, having that faith,

188
00:09:26,800 --> 00:09:31,160
Speaker 7: and then learning from him about AI, infrastructure, defense, space,

189
00:09:31,520 --> 00:09:32,440
Speaker 7: the space economy.

190
00:09:32,480 --> 00:09:34,120
Speaker 6: It's been an incredible journey for us.

191
00:09:34,240 --> 00:09:36,560
Speaker 2: You make a lot of interesting points about how things

192
00:09:36,559 --> 00:09:41,520
Speaker 2: have changed. The story started with reusability and the economics

193
00:09:41,520 --> 00:09:43,599
Speaker 2: of reusability, and now it's something very different. If you

194
00:09:43,720 --> 00:09:46,080
Speaker 2: just got on the TAM and the perspectives, but also

195
00:09:46,120 --> 00:09:48,240
Speaker 2: like the public market investor view, like what is the

196
00:09:48,280 --> 00:09:50,640
Speaker 2: proxy that you look at? I think we're in a

197
00:09:50,720 --> 00:09:53,160
Speaker 2: space now where we're moving outside the mag seven Yes,

198
00:09:53,440 --> 00:09:55,800
Speaker 2: but like if it was in the software space, you go, okay,

199
00:09:55,840 --> 00:09:58,280
Speaker 2: this is a private company where I see a proxy

200
00:09:58,280 --> 00:10:00,440
Speaker 2: in the public market SpaceX.

201
00:10:00,280 --> 00:10:01,440
Speaker 6: We don't have anything like it.

202
00:10:01,480 --> 00:10:04,840
Speaker 7: And look, it's been a huge discussion on what industry,

203
00:10:04,840 --> 00:10:07,320
Speaker 7: how is gigs going to classify this? Is it going

204
00:10:07,320 --> 00:10:09,040
Speaker 7: to be in the telecom industry, is it going to

205
00:10:09,040 --> 00:10:11,280
Speaker 7: be in industrials. We started out thinking it would be

206
00:10:11,320 --> 00:10:14,600
Speaker 7: an industrial entity and what our share investor is going

207
00:10:14,640 --> 00:10:16,920
Speaker 7: to be selling to buy this, but ultimately you ended

208
00:10:17,000 --> 00:10:21,280
Speaker 7: up it's now classified under communications. And look, our analyst

209
00:10:21,360 --> 00:10:24,480
Speaker 7: covering it has to understand AI has been working very

210
00:10:24,480 --> 00:10:27,680
Speaker 7: closely with our tech team, has to understand defense, has

211
00:10:27,720 --> 00:10:31,560
Speaker 7: to understand space. A new economy here, and we're all

212
00:10:31,600 --> 00:10:33,760
Speaker 7: still just learning more and more about that. I mean,

213
00:10:33,760 --> 00:10:35,000
Speaker 7: it's a new frontier.

214
00:10:35,280 --> 00:10:36,760
Speaker 2: Let's bring it back to the news of the day

215
00:10:36,760 --> 00:10:39,000
Speaker 2: in the Cursor deal. You know a lot of people

216
00:10:39,040 --> 00:10:41,920
Speaker 2: will focus on this the very sort of specific upfront

217
00:10:42,120 --> 00:10:45,559
Speaker 2: coding story, but there is a story with the models themselves,

218
00:10:45,559 --> 00:10:48,800
Speaker 2: and I just wonder how much you assign value to

219
00:10:48,880 --> 00:10:52,080
Speaker 2: GROC and the future opportunity of GROC over At Franklin.

220
00:10:51,760 --> 00:10:53,720
Speaker 6: Templeton, Yeah, look, you don't.

221
00:10:54,160 --> 00:10:56,720
Speaker 7: The way we think about it is just starlink alone

222
00:10:57,000 --> 00:10:59,600
Speaker 7: brings a lot of value for SpaceX and a lot

223
00:10:59,640 --> 00:11:02,400
Speaker 7: of their endeavors with an AI. A big part of

224
00:11:02,440 --> 00:11:05,120
Speaker 7: it is funding some of that and you know cash

225
00:11:05,160 --> 00:11:08,000
Speaker 7: generation from cash generation, and look, Elon has talked about

226
00:11:08,000 --> 00:11:10,960
Speaker 7: this from day one. Gwinn talks about it. They want

227
00:11:10,960 --> 00:11:13,839
Speaker 7: to go to Mars. They're doing everything in their power

228
00:11:13,920 --> 00:11:17,400
Speaker 7: to make us multiplanetary. That Their goal here is to

229
00:11:17,480 --> 00:11:20,960
Speaker 7: get there, whether it's GROCK and AI AI infrastructure. You

230
00:11:21,000 --> 00:11:24,440
Speaker 7: saw them doing deals with anthropic generating revenue here.

231
00:11:24,240 --> 00:11:25,559
Speaker 6: Today to get to that goal.

232
00:11:26,360 --> 00:11:28,240
Speaker 7: Anything that can get them on that path I think

233
00:11:28,360 --> 00:11:29,360
Speaker 7: is the right move for them.

234
00:11:29,600 --> 00:11:32,520
Speaker 2: Sarah Rahi, Franklin, Templeton, Davy On, Bloomberg Tech. It's so

235
00:11:32,559 --> 00:11:34,800
Speaker 2: great to have you here in Essa. Thank you very much.

236
00:11:35,200 --> 00:11:38,000
Speaker 2: An update on the story brought yesterday in Video's bond sale,

237
00:11:38,120 --> 00:11:41,040
Speaker 2: the AI chip Giant sold twenty five billion dollars in

238
00:11:41,040 --> 00:11:43,960
Speaker 2: investment grade debt, making it one of the largest corporate

239
00:11:44,000 --> 00:11:46,560
Speaker 2: bond deals of the year. Investor demand was strong, with

240
00:11:46,720 --> 00:11:50,439
Speaker 2: orders topping eighty five billion dollars according to sources, and

241
00:11:50,559 --> 00:11:54,920
Speaker 2: video joins companies like Alphabet and Amazon flooding debt markets

242
00:11:54,960 --> 00:11:57,720
Speaker 2: with hundreds of billions of dollars of issuance as they

243
00:11:57,720 --> 00:12:03,280
Speaker 2: build data centers, another infrastructure needed for AI's rapid expansion. Okay,

244
00:12:03,320 --> 00:12:06,360
Speaker 2: coming up anthrop it meets with US officials is IT

245
00:12:06,400 --> 00:12:09,520
Speaker 2: tries to resolve a national security dispute with the Trump

246
00:12:09,559 --> 00:12:12,840
Speaker 2: administration over it's most advanced AI models. We have more

247
00:12:12,880 --> 00:12:15,120
Speaker 2: on that next, this is Bloomberg Tech.

248
00:12:22,600 --> 00:12:25,160
Speaker 8: There are some players who are more trustworthy than others.

249
00:12:25,520 --> 00:12:27,720
Speaker 8: What I think needs to happen is that the trustworthy

250
00:12:27,720 --> 00:12:31,600
Speaker 8: actors need to need to get together and put the

251
00:12:31,679 --> 00:12:35,600
Speaker 8: untrustworthy actors in a position where they kind of have

252
00:12:35,679 --> 00:12:38,880
Speaker 8: to adopt the same standards. With a lot of experience,

253
00:12:38,960 --> 00:12:40,760
Speaker 8: I've learned that there are some folks who don't do

254
00:12:40,800 --> 00:12:42,680
Speaker 8: the right thing on their own. But if there's a

255
00:12:42,760 --> 00:12:45,440
Speaker 8: majority of the industry that's doing the right thing, then

256
00:12:45,480 --> 00:12:47,840
Speaker 8: I think the rest of the industry is kind of

257
00:12:48,720 --> 00:12:50,600
Speaker 8: they're left in a position where there's not much they

258
00:12:50,600 --> 00:12:52,760
Speaker 8: can do that then come along.

259
00:12:53,679 --> 00:12:57,360
Speaker 2: That was Anthropic CEO Dario Ammo Day they're in late April,

260
00:12:57,520 --> 00:13:00,760
Speaker 2: weighing in on accountability in the AA industry and now

261
00:13:01,360 --> 00:13:03,360
Speaker 2: and topics, trying to work through a growing dispute with

262
00:13:03,360 --> 00:13:07,920
Speaker 2: the Trump administration about concerns on foreign access to its

263
00:13:08,000 --> 00:13:11,479
Speaker 2: most advanced air models, which led to the company temporarily

264
00:13:11,600 --> 00:13:16,520
Speaker 2: suspending global access. Senior technical staff met with Commerce Department

265
00:13:16,520 --> 00:13:20,080
Speaker 2: officials on Monday, as talks continued as Mike Sheppard joins

266
00:13:20,160 --> 00:13:22,800
Speaker 2: us from DC, Sheep bring us up to speed. What

267
00:13:22,840 --> 00:13:23,440
Speaker 2: do we need to know?

268
00:13:24,200 --> 00:13:26,040
Speaker 9: Well, what we need to know is this that there

269
00:13:26,160 --> 00:13:28,960
Speaker 9: so far is very little public sign of progress and

270
00:13:29,040 --> 00:13:33,319
Speaker 9: discussions between the government and Anthropic. The senior technical step,

271
00:13:33,880 --> 00:13:37,840
Speaker 9: as you said, did sit down with US officials, but

272
00:13:37,920 --> 00:13:40,640
Speaker 9: there has really been no evidence of a breakthrough, no

273
00:13:40,760 --> 00:13:44,560
Speaker 9: evidence of really what remedies the company might need to

274
00:13:44,559 --> 00:13:48,640
Speaker 9: pursue with the government's behest to solve this security issue,

275
00:13:48,679 --> 00:13:51,960
Speaker 9: and we don't even really have a fully clear picture

276
00:13:52,040 --> 00:13:54,400
Speaker 9: from the US government side of what that is. The

277
00:13:54,440 --> 00:13:57,920
Speaker 9: company has pointed to this jail break concern the idea

278
00:13:57,960 --> 00:14:02,679
Speaker 9: that Garbrails on its Feeble five system could somehow be evaded,

279
00:14:02,840 --> 00:14:09,080
Speaker 9: empowering that software into something that approaches the cyber security

280
00:14:09,120 --> 00:14:13,520
Speaker 9: capabilities of Mythos, and all of this is unfolding at

281
00:14:13,520 --> 00:14:15,920
Speaker 9: this time of great urgency in the AI race. For

282
00:14:15,960 --> 00:14:19,480
Speaker 9: the US, of course, the question is will it maintain

283
00:14:19,520 --> 00:14:22,120
Speaker 9: its lead over China, and Anthropic, of course, is one

284
00:14:22,120 --> 00:14:25,280
Speaker 9: of the top developers, plays such a crucial role in that.

285
00:14:25,640 --> 00:14:28,440
Speaker 9: And then just add to it all the billions, if

286
00:14:28,480 --> 00:14:30,920
Speaker 9: not trillions of dollars in capex that we are seeing

287
00:14:31,080 --> 00:14:34,840
Speaker 9: poured toward this sector by different companies, not only Anthropic

288
00:14:34,920 --> 00:14:40,320
Speaker 9: but the public tradedly publicly traded ventures as well, the SPACEXIPO.

289
00:14:40,480 --> 00:14:43,040
Speaker 9: So much of that money is going to support the

290
00:14:43,120 --> 00:14:47,520
Speaker 9: infrastructure needed to run top flight models like Mythos and Fable,

291
00:14:47,560 --> 00:14:49,080
Speaker 9: which right now are off the market.

292
00:14:49,880 --> 00:14:53,120
Speaker 2: The COMMAS Department put in an export control Friday which

293
00:14:53,160 --> 00:14:56,280
Speaker 2: basically blocked foreign national access to Mythos and Fable, and

294
00:14:56,320 --> 00:14:59,120
Speaker 2: Anthropics response was like, well, in that case, we'll take

295
00:14:59,120 --> 00:15:01,920
Speaker 2: it offline for every because we can't on a case

296
00:15:01,960 --> 00:15:04,880
Speaker 2: by case basis check user status and many of their

297
00:15:04,920 --> 00:15:08,920
Speaker 2: own employees are foreign nationals. But there is something broader

298
00:15:08,960 --> 00:15:15,720
Speaker 2: here in Anthropic and industry's relationship with the administration. Where

299
00:15:15,720 --> 00:15:19,920
Speaker 2: do we stand in that the litigation, the designation on

300
00:15:20,000 --> 00:15:20,840
Speaker 2: supply chain risk.

301
00:15:21,440 --> 00:15:24,880
Speaker 9: Well, this is just the latest flare up between Anthropic

302
00:15:25,120 --> 00:15:27,960
Speaker 9: and the administration really since the start of the year.

303
00:15:28,600 --> 00:15:31,000
Speaker 9: You're referring, of course, to the dispute right now with

304
00:15:31,120 --> 00:15:35,080
Speaker 9: the Pentagon over the company's ability to continue providing the

305
00:15:35,120 --> 00:15:38,920
Speaker 9: military with access to its sophisticated AI tools. There's a

306
00:15:38,960 --> 00:15:43,200
Speaker 9: clog gov AI tool that was approved for classified use

307
00:15:43,240 --> 00:15:46,120
Speaker 9: by the military and in fact have been deployed as

308
00:15:46,200 --> 00:15:51,080
Speaker 9: part of the US war effort against Iran. Now, the Pentagon,

309
00:15:51,160 --> 00:15:53,920
Speaker 9: in this dispute with the company, has really fixed on

310
00:15:54,280 --> 00:15:58,040
Speaker 9: what Anthropic has demanded as two conditions of being able

311
00:15:58,080 --> 00:16:01,080
Speaker 9: to continue its Pentagon work, and that is it once

312
00:16:01,160 --> 00:16:06,760
Speaker 9: guarantees that there will be no surveillance supported by CLAUDE

313
00:16:07,040 --> 00:16:10,560
Speaker 9: and that there will be no fully autonomous weaponry. Now

314
00:16:10,600 --> 00:16:13,720
Speaker 9: the government says that, look, we won't do either of

315
00:16:13,720 --> 00:16:16,920
Speaker 9: those things. Those extra conditions are way too much, and

316
00:16:16,960 --> 00:16:21,000
Speaker 9: that just presages a lasting few between these two sides

317
00:16:21,080 --> 00:16:23,320
Speaker 9: unless they can reach some sort of an agreement here.

318
00:16:23,920 --> 00:16:26,560
Speaker 2: And begs Mike Shephard, thank you very much. As Washington

319
00:16:26,600 --> 00:16:30,240
Speaker 2: wrestles with AI security concerns at home, the technology is

320
00:16:30,280 --> 00:16:33,040
Speaker 2: also taking center stage at the g seventh summit in France,

321
00:16:33,360 --> 00:16:36,480
Speaker 2: where global leaders and AI tech executives and meeting this week.

322
00:16:36,760 --> 00:16:40,000
Speaker 2: Manibas Tyler Kendall joins us like from Evion France. And

323
00:16:40,040 --> 00:16:42,720
Speaker 2: there's the story on the Bloomberg this morning about how

324
00:16:43,520 --> 00:16:47,960
Speaker 2: US Western allies are now taking into account this Commerce

325
00:16:48,000 --> 00:16:51,960
Speaker 2: Department decision with Anthropic. How's that playing out on the ground.

326
00:16:54,520 --> 00:16:56,200
Speaker 10: Well, And I actually had the chance to catch up

327
00:16:56,200 --> 00:16:59,360
Speaker 10: with the spokesperson for France's Foreign Ministry who told me

328
00:16:59,400 --> 00:17:01,800
Speaker 10: that the Trump and Administration's moves here that you've been

329
00:17:01,800 --> 00:17:05,399
Speaker 10: outlining just underscore the need for European allies to have

330
00:17:05,480 --> 00:17:09,440
Speaker 10: what he called, quote strategic autonomy when it comes to tech.

331
00:17:09,560 --> 00:17:12,399
Speaker 10: It is no doubt going to be a focus here

332
00:17:12,440 --> 00:17:16,000
Speaker 10: on the ground as these world leaders, including President Trump,

333
00:17:16,000 --> 00:17:17,520
Speaker 10: are going to be in the same room with some

334
00:17:17,560 --> 00:17:21,360
Speaker 10: of these top tech CEOs, including from Anthropic as well

335
00:17:21,359 --> 00:17:24,200
Speaker 10: as Open AI. And in fact, earlier today Bloomberg News

336
00:17:24,240 --> 00:17:27,200
Speaker 10: was able to obtain a draft statement related to this

337
00:17:27,560 --> 00:17:30,359
Speaker 10: meeting that is set to take place tomorrow, which says,

338
00:17:30,400 --> 00:17:33,280
Speaker 10: in part that the leaders are expected to pledge to

339
00:17:33,400 --> 00:17:38,200
Speaker 10: quote further discuss emerging opportunities and potential risks arising from AI,

340
00:17:38,640 --> 00:17:41,720
Speaker 10: notably in the financial sector. It is definitely going to

341
00:17:41,720 --> 00:17:44,000
Speaker 10: be a theme here going forward. At G seven official

342
00:17:44,240 --> 00:17:46,960
Speaker 10: tells me that that meeting tomorrow will also focus on

343
00:17:47,040 --> 00:17:50,600
Speaker 10: potential regulations that all of these G seven allies could

344
00:17:50,640 --> 00:17:54,640
Speaker 10: come together on, as well as respecting individual countries regulations

345
00:17:54,680 --> 00:17:57,560
Speaker 10: when it comes to the tech and ed Another thread

346
00:17:57,560 --> 00:18:00,000
Speaker 10: that we're watching very closely here is that just yesterday

347
00:18:00,040 --> 00:18:03,800
Speaker 10: President Trump threatened a one hundred percent tieriff on France

348
00:18:03,880 --> 00:18:07,080
Speaker 10: over the use of its digital services tax. This has

349
00:18:07,119 --> 00:18:10,439
Speaker 10: been something that the Trump administration has long lobbied criticism

350
00:18:10,640 --> 00:18:13,240
Speaker 10: against European allies about. I had the chance to ask

351
00:18:13,280 --> 00:18:15,640
Speaker 10: the French spokesperson about this. He did confirm that there

352
00:18:15,640 --> 00:18:18,840
Speaker 10: are ongoing discussions about the DST, but he said that

353
00:18:18,880 --> 00:18:22,800
Speaker 10: this is not a moment for quote antagonism.

354
00:18:22,800 --> 00:18:25,400
Speaker 2: Most Tyler Kendall and Avian France the G seven, thank

355
00:18:25,400 --> 00:18:28,240
Speaker 2: you very much. Another story we're watching in AI open

356
00:18:28,280 --> 00:18:31,600
Speaker 2: AI's losses ballooned last year as the company poured money

357
00:18:31,800 --> 00:18:34,720
Speaker 2: into the AI race. The Financial Times reports the chat

358
00:18:34,800 --> 00:18:38,240
Speaker 2: GBT maker spent more than thirty four billion dollars on research,

359
00:18:38,600 --> 00:18:41,800
Speaker 2: marketing and other expenses. In twenty twenty five, driving an

360
00:18:41,800 --> 00:18:45,280
Speaker 2: eightfold increase in net losses. Open Ai did not comment

361
00:18:45,760 --> 00:18:50,040
Speaker 2: on the Financial Times report. Data briggs Is Data Summit

362
00:18:50,119 --> 00:18:52,840
Speaker 2: twenty twenty six is underway in San Francisco, as the

363
00:18:52,840 --> 00:18:56,399
Speaker 2: software maker competes with Snowflake and Google. It's launching new

364
00:18:56,520 --> 00:19:01,439
Speaker 2: enterprise AI products including Genie one, Gentic co worker for

365
00:19:01,560 --> 00:19:05,439
Speaker 2: business teams, and Unity AI gateway that helps businesses control

366
00:19:05,520 --> 00:19:08,680
Speaker 2: their costs. Talent is proving the biggest expense for them,

367
00:19:08,760 --> 00:19:11,520
Speaker 2: but the companies yet to tap the public markets this

368
00:19:11,680 --> 00:19:15,280
Speaker 2: year amid other blockbuster tech listings. Is what CEO Ali

369
00:19:15,280 --> 00:19:16,520
Speaker 2: Gazi told us yesterday.

370
00:19:17,280 --> 00:19:19,320
Speaker 11: So well, you know, we're not We've not said that

371
00:19:19,400 --> 00:19:21,399
Speaker 11: we're doing any fundraise. I know that there's you know,

372
00:19:22,080 --> 00:19:26,200
Speaker 11: rumors about this, but we're always talking to investors. Why well,

373
00:19:26,359 --> 00:19:29,960
Speaker 11: we're investing heavily into these new product categories. We just

374
00:19:30,080 --> 00:19:34,200
Speaker 11: launching a completely new product for ours that's called Customer

375
00:19:34,280 --> 00:19:37,119
Speaker 11: Lake which is an area of marketing, so we're you know,

376
00:19:37,160 --> 00:19:39,920
Speaker 11: targeting marketing folks, which is not an area where we.

377
00:19:39,920 --> 00:19:41,199
Speaker 12: Used to be active.

378
00:19:41,480 --> 00:19:41,679
Speaker 2: You know.

379
00:19:41,800 --> 00:19:45,200
Speaker 11: Two months ago we launched a security product called lake Watch.

380
00:19:45,200 --> 00:19:46,639
Speaker 12: And we're getting into a security space.

381
00:19:46,880 --> 00:19:50,120
Speaker 11: You know, it takes resources to get into these spaces

382
00:19:50,200 --> 00:19:52,240
Speaker 11: and all of these products that I mentioned, we're building

383
00:19:52,280 --> 00:19:55,360
Speaker 11: it agentic first, so it's like they start with agents

384
00:19:55,400 --> 00:19:56,359
Speaker 11: from the ground up.

385
00:19:56,400 --> 00:19:58,400
Speaker 12: We're not boulting it on. That requires that.

386
00:19:58,359 --> 00:20:01,960
Speaker 11: You have AI researchers, as we all know, AI researchers

387
00:20:02,040 --> 00:20:03,040
Speaker 11: you know, don't come cheap.

388
00:20:03,600 --> 00:20:05,000
Speaker 12: So all this requires funding.

389
00:20:05,119 --> 00:20:07,639
Speaker 11: So we're always in talk with investers, you know, to

390
00:20:07,680 --> 00:20:10,240
Speaker 11: be able to just be aggressive and expand and launch

391
00:20:10,280 --> 00:20:13,040
Speaker 11: new products, do research, you know, go to more.

392
00:20:12,920 --> 00:20:15,399
Speaker 12: Regions, be available in every part of the world.

393
00:20:16,040 --> 00:20:17,960
Speaker 11: So you know, that's that's that's why we're you know,

394
00:20:18,040 --> 00:20:18,800
Speaker 11: require capital.

395
00:20:19,840 --> 00:20:22,119
Speaker 2: Allie, let's let's end it here. Earlier this month you

396
00:20:22,200 --> 00:20:24,880
Speaker 2: told me, you know, Data Bricks in the end will

397
00:20:24,920 --> 00:20:27,280
Speaker 2: be a public company. You want to be a public company,

398
00:20:27,720 --> 00:20:29,840
Speaker 2: but that you think this is just a terrible year

399
00:20:29,880 --> 00:20:33,439
Speaker 2: to go public. Do you stand by that. Has anything

400
00:20:33,560 --> 00:20:35,960
Speaker 2: changed in your mind in the last two weeks.

401
00:20:37,119 --> 00:20:37,840
Speaker 12: No, I think that.

402
00:20:38,119 --> 00:20:40,960
Speaker 11: You know, it's just there's three Mega I pos that

403
00:20:40,960 --> 00:20:43,880
Speaker 11: have been rumored. One of them happened, you know, so

404
00:20:44,240 --> 00:20:46,360
Speaker 11: letting those kind of go through the system and see

405
00:20:46,400 --> 00:20:48,480
Speaker 11: how that sort of plays out and things stabilize a

406
00:20:48,480 --> 00:20:51,440
Speaker 11: little bit, and I think that you know, getting a

407
00:20:51,480 --> 00:20:54,000
Speaker 11: little bit further in the sort of AI revolution that

408
00:20:54,040 --> 00:20:54,720
Speaker 11: we're going through.

409
00:20:55,440 --> 00:20:56,280
Speaker 12: I prefer for the.

410
00:20:56,280 --> 00:20:58,560
Speaker 11: Company to go public when you know, water is a

411
00:20:58,600 --> 00:21:00,960
Speaker 11: little bit more palm and there's a little bit more predictability.

412
00:21:01,840 --> 00:21:03,600
Speaker 11: So you know, I stand by what I said. I

413
00:21:03,600 --> 00:21:06,080
Speaker 11: think for most of the companies, you know, it's probably

414
00:21:06,080 --> 00:21:07,720
Speaker 11: way best to wait it out. And that's also what

415
00:21:07,880 --> 00:21:09,920
Speaker 11: CEOs that I know tell me that I'm thinking about

416
00:21:09,920 --> 00:21:12,560
Speaker 11: the IPO. They're saying, you know, I don't want to

417
00:21:12,600 --> 00:21:14,919
Speaker 11: you know, go between you know, two mega IPOs like that.

418
00:21:16,960 --> 00:21:19,240
Speaker 2: It was data break CEO Ali GODSI let's go out

419
00:21:19,240 --> 00:21:21,600
Speaker 2: to bloom Bag, to Jahara and in New York for

420
00:21:21,760 --> 00:21:24,280
Speaker 2: everything else while we track in your horror, Hi.

421
00:21:24,320 --> 00:21:27,000
Speaker 13: Ed, Well, it's time for talking tech. Fir stup prediction

422
00:21:27,080 --> 00:21:30,960
Speaker 13: market platform Caushi has developed a new AI agent designed

423
00:21:30,960 --> 00:21:34,080
Speaker 13: to stress test the wording of its contracts. The company

424
00:21:34,119 --> 00:21:38,119
Speaker 13: says the tool, known internally as Harrison, helps identify potential

425
00:21:38,160 --> 00:21:41,520
Speaker 13: issues before they affect the millions of wagers it handles

426
00:21:41,600 --> 00:21:44,520
Speaker 13: each day on events ranging, of course, from elections to

427
00:21:44,600 --> 00:21:48,760
Speaker 13: sports games and award ceremonies. Plus robin Hood is the

428
00:21:48,920 --> 00:21:52,920
Speaker 13: latest fintech to downsize. The online brokerage is eliminating three

429
00:21:53,040 --> 00:21:57,040
Speaker 13: hundred jobs, adding to a deepening wave of layoffs across

430
00:21:57,080 --> 00:21:59,639
Speaker 13: the industry. The company says the latest cut is to

431
00:21:59,760 --> 00:22:04,560
Speaker 13: quote further accelerate parts velocity and remain lean and disciplined.

432
00:22:04,800 --> 00:22:09,960
Speaker 13: No mention of AI and several studios inside Microsoft's Xbox

433
00:22:10,000 --> 00:22:13,440
Speaker 13: gaming division are an active negotiations to spin off as

434
00:22:13,480 --> 00:22:16,840
Speaker 13: they try to avoid closure. That's according to sources. The

435
00:22:16,920 --> 00:22:21,040
Speaker 13: talks come as newly appointed CEO Asha Sharma accelerates a

436
00:22:21,160 --> 00:22:23,320
Speaker 13: broader restructuring of the business.

437
00:22:23,440 --> 00:22:24,840
Speaker 6: And I'll leave the gaming.

438
00:22:24,680 --> 00:22:25,120
Speaker 4: Up to you.

439
00:22:26,200 --> 00:22:28,640
Speaker 2: Thank you very much, Yahara, and a coming up. Sean maguire,

440
00:22:28,640 --> 00:22:31,600
Speaker 2: who LEDs to COIs SpaceX Investment, joins us to talk

441
00:22:31,640 --> 00:22:35,040
Speaker 2: about the Cursor deal, orbital data centers, and so much more.

442
00:22:35,080 --> 00:22:38,280
Speaker 2: This is what markets look like. Reminder Kevin Walsh FED

443
00:22:38,320 --> 00:22:50,840
Speaker 2: decision debut Wednesday. This is Bloomberg Tech. Welcome back to

444
00:22:50,840 --> 00:22:54,520
Speaker 2: Bloomberg Tech. It's just the third trading day in SpaceX's

445
00:22:54,560 --> 00:22:57,320
Speaker 2: life as a public company, and my goodness, so many headlines.

446
00:22:57,400 --> 00:22:59,560
Speaker 2: It's kind of been an interesting, volatile session. And Carmen

447
00:22:59,640 --> 00:23:01,919
Speaker 2: Ryne made the point earlier in the show that you

448
00:23:01,960 --> 00:23:03,240
Speaker 2: need to kind of look at the flow and the

449
00:23:03,320 --> 00:23:05,439
Speaker 2: volumes that are going on. But at one point in

450
00:23:05,480 --> 00:23:07,840
Speaker 2: the session we were really out of this world. Sorry,

451
00:23:08,200 --> 00:23:10,040
Speaker 2: not sorry. There is a lot of numbers to go

452
00:23:10,119 --> 00:23:12,160
Speaker 2: through and the lot to pass here, So let's check

453
00:23:12,160 --> 00:23:14,840
Speaker 2: in on these SpaceX shares with Bluebots. Cross Asset reporter

454
00:23:14,880 --> 00:23:15,600
Speaker 2: Isabel Lee.

455
00:23:15,520 --> 00:23:17,800
Speaker 14: Is about hi and out of this world. Indeed, you

456
00:23:17,800 --> 00:23:20,000
Speaker 14: see the stock rallying for the third straight day, and

457
00:23:20,040 --> 00:23:22,879
Speaker 14: it's contractors are past the market cap of Amazon and

458
00:23:22,960 --> 00:23:26,320
Speaker 14: Microsoft to become the fourth largest company in the world.

459
00:23:26,359 --> 00:23:28,400
Speaker 14: So you see the rally today pushing the market cap

460
00:23:28,440 --> 00:23:32,280
Speaker 14: of SpaceX to nearly three trillion dollars. So obviously this

461
00:23:32,400 --> 00:23:34,600
Speaker 14: comes ferce that the market will not be able to

462
00:23:34,600 --> 00:23:37,400
Speaker 14: adjust an IPO this big, and obviously it bodes well

463
00:23:37,400 --> 00:23:39,520
Speaker 14: for the likes of a DROPIC and open Ai which

464
00:23:39,560 --> 00:23:42,080
Speaker 14: is waiting in the wings, all both valued at almost

465
00:23:42,080 --> 00:23:44,080
Speaker 14: one trillion dollars. And to be sure, the price action

466
00:23:44,160 --> 00:23:46,280
Speaker 14: is driven by just a small percentage of shares, but.

467
00:23:46,200 --> 00:23:47,400
Speaker 4: Still it's so interesting.

468
00:23:47,440 --> 00:23:48,840
Speaker 14: And you look at the next page, and this is

469
00:23:48,880 --> 00:23:51,440
Speaker 14: the ETF world, which I really love to look at

470
00:23:51,440 --> 00:23:55,000
Speaker 14: So all of these issuers filed to launch nearly a

471
00:23:55,040 --> 00:23:59,160
Speaker 14: dozen SpaceX link ETFs, so they're both leveraged. They're all

472
00:23:59,240 --> 00:24:02,000
Speaker 14: leverage rather, so they track SpaceX two times to day

473
00:24:02,080 --> 00:24:04,480
Speaker 14: upside and two times to the downside, and in total

474
00:24:04,560 --> 00:24:07,840
Speaker 14: on Monday all of them combined, so one billion dollars

475
00:24:07,880 --> 00:24:10,800
Speaker 14: in trading volume, which is easily a record for any

476
00:24:10,800 --> 00:24:13,720
Speaker 14: single stock etr for any new launch, and a cohort

477
00:24:13,800 --> 00:24:16,000
Speaker 14: like this. So the friends is just really continues, and

478
00:24:16,040 --> 00:24:18,600
Speaker 14: it's really hard not to imagine that because the windfall

479
00:24:18,880 --> 00:24:21,040
Speaker 14: is just so hard to ignore. And you look at

480
00:24:21,040 --> 00:24:23,960
Speaker 14: these companies, these are the ones that are gaining billions

481
00:24:24,000 --> 00:24:26,040
Speaker 14: and billions in net windfall. Take a look of Balor

482
00:24:26,080 --> 00:24:29,160
Speaker 14: Equity Partners. This is founded by muss Ally Antonio Gracia,

483
00:24:29,240 --> 00:24:32,000
Speaker 14: so nearly seventy billion dollars at stake that they have

484
00:24:32,160 --> 00:24:34,639
Speaker 14: founders fund. This is a venture capital firm led by

485
00:24:34,640 --> 00:24:36,879
Speaker 14: Peter Theil. It now has a fifty billion dollars stake.

486
00:24:37,119 --> 00:24:38,480
Speaker 6: Sequoya has twenty billion.

487
00:24:38,480 --> 00:24:40,560
Speaker 14: They're one of the first companies to back SpaceX in

488
00:24:40,600 --> 00:24:42,920
Speaker 14: twenty nineteen, so first move and move there, and they

489
00:24:42,960 --> 00:24:44,800
Speaker 14: were the first to put the trust in Elon Musk

490
00:24:44,840 --> 00:24:47,600
Speaker 14: among the first and Anderson Horowitz a ten billion dollars,

491
00:24:47,600 --> 00:24:50,160
Speaker 14: so really lots of excitement and billions at stake.

492
00:24:50,320 --> 00:24:53,760
Speaker 2: Ed Isabelli at New York City, thank you very much.

493
00:24:53,840 --> 00:24:56,679
Speaker 2: Conversation been looking forward to for some time now. We're

494
00:24:56,720 --> 00:24:59,159
Speaker 2: joined by Sean MacGuire, a partner as Core Capital. With

495
00:24:59,200 --> 00:25:02,480
Speaker 2: the firm, it has invested around two billion dollars across

496
00:25:02,520 --> 00:25:06,359
Speaker 2: funds into SpaceX. That includes an investment into x the

497
00:25:06,400 --> 00:25:09,240
Speaker 2: platform formerly known as Twitter, and all told, that amounts

498
00:25:09,240 --> 00:25:11,600
Speaker 2: to a nearly one and a half percent stake in

499
00:25:11,720 --> 00:25:16,840
Speaker 2: SpaceX SPACEXAI. There's a much bigger story than those two parts,

500
00:25:16,880 --> 00:25:19,960
Speaker 2: the rocket part and the AI part. Welcome to Tech.

501
00:25:20,160 --> 00:25:20,879
Speaker 4: Thank you for having me.

502
00:25:20,960 --> 00:25:23,840
Speaker 15: And we're both football fans, so maybe we even get

503
00:25:23,880 --> 00:25:25,320
Speaker 15: some time for some football commentary.

504
00:25:26,320 --> 00:25:28,240
Speaker 2: There is a lot to discuss. I think should we

505
00:25:28,320 --> 00:25:30,600
Speaker 2: just start with cursor like it's right in front of us.

506
00:25:31,600 --> 00:25:34,880
Speaker 2: The focus of this morning was, oh well, Elon Musk

507
00:25:34,880 --> 00:25:37,919
Speaker 2: has been quite honest about XAIS need to catch up

508
00:25:37,920 --> 00:25:41,760
Speaker 2: in coding visa vanthropic OpenAI. What I'm trying to understand

509
00:25:41,840 --> 00:25:44,600
Speaker 2: is what this means for the models themselves. Do you

510
00:25:44,680 --> 00:25:46,600
Speaker 2: do you have a sort of thesis on that.

511
00:25:47,000 --> 00:25:47,200
Speaker 4: Yeah.

512
00:25:47,240 --> 00:25:49,919
Speaker 15: Look, I've said this publicly, but I think about SpaceX

513
00:25:49,960 --> 00:25:53,680
Speaker 15: as there's five layers of the company or business. Layer

514
00:25:53,760 --> 00:25:56,679
Speaker 15: one is launch. This is foundational, it's kind of the

515
00:25:56,720 --> 00:26:00,560
Speaker 15: future of the company. Layer two is connectivity, which is

516
00:26:00,960 --> 00:26:03,679
Speaker 15: you know, in its early stages with stallink and direct

517
00:26:03,680 --> 00:26:07,639
Speaker 15: to sell, but kind of proven and I think the

518
00:26:07,680 --> 00:26:11,880
Speaker 15: compounding is pretty easy to underwrite. Layer three is compute,

519
00:26:12,320 --> 00:26:14,360
Speaker 15: which today is terrestrial and in the.

520
00:26:14,320 --> 00:26:15,960
Speaker 4: Future will be orbital as well.

521
00:26:16,960 --> 00:26:20,800
Speaker 15: I think that the Google ananthropic deals kind of show

522
00:26:22,240 --> 00:26:26,080
Speaker 15: show that that is very likely to work. And I

523
00:26:26,240 --> 00:26:27,960
Speaker 15: actually want to spend more time on trust you Compute

524
00:26:28,000 --> 00:26:30,520
Speaker 15: later because I think it's a very very exciting and

525
00:26:30,600 --> 00:26:32,960
Speaker 15: underappreciated part of we will of.

526
00:26:32,960 --> 00:26:36,280
Speaker 4: The company today. Layer four is the model layer.

527
00:26:36,600 --> 00:26:41,879
Speaker 15: I think that's where probably rightfully so, most investors don't

528
00:26:41,880 --> 00:26:47,040
Speaker 15: know how to assess where SpaceX SPACEXAI is today, and

529
00:26:47,480 --> 00:26:48,600
Speaker 15: so I think we should talk about that the more.

530
00:26:48,600 --> 00:26:50,640
Speaker 15: And that's where Cursor fits in. And then layer five

531
00:26:50,720 --> 00:26:53,359
Speaker 15: is all the other bets. It's sayings that you know,

532
00:26:53,840 --> 00:26:55,800
Speaker 15: people today don't even know how to think about it.

533
00:26:55,880 --> 00:26:59,560
Speaker 15: That's terrify, that's a moon base someday that has a

534
00:26:59,600 --> 00:27:03,159
Speaker 15: rail gun launching you know, orbital satellites in space.

535
00:27:04,480 --> 00:27:06,200
Speaker 4: But look on the on the Cursor deal.

536
00:27:07,320 --> 00:27:10,639
Speaker 15: I think people are underappreciating just how good that team

537
00:27:10,720 --> 00:27:14,080
Speaker 15: is and also how well they work with Elon and

538
00:27:14,119 --> 00:27:14,879
Speaker 15: the company A.

539
00:27:14,960 --> 00:27:17,520
Speaker 2: Lost guests cooled to Inaquahar. That's just it's a sixty

540
00:27:17,560 --> 00:27:19,000
Speaker 2: billion dollar aquah It's.

541
00:27:18,960 --> 00:27:21,399
Speaker 15: It's I mean, I don't think when a company has

542
00:27:21,440 --> 00:27:24,640
Speaker 15: a three billion plus run rate is the reported number

543
00:27:24,760 --> 00:27:27,120
Speaker 15: in you know, three is a three year old company.

544
00:27:27,520 --> 00:27:31,280
Speaker 15: I don't think it's only an aquahier, but I do.

545
00:27:31,440 --> 00:27:33,480
Speaker 15: I do think that this is one of the best

546
00:27:33,480 --> 00:27:36,800
Speaker 15: teams has ever been assembled in the Cursor team. And

547
00:27:37,600 --> 00:27:40,240
Speaker 15: I also think they've proven over the last few months

548
00:27:40,480 --> 00:27:42,240
Speaker 15: that they can keep pace with Elon.

549
00:27:42,560 --> 00:27:44,720
Speaker 4: Like it is, unless.

550
00:27:44,320 --> 00:27:47,760
Speaker 15: You've seen it, it is impossible to explain the intensity

551
00:27:47,840 --> 00:27:52,520
Speaker 15: and work ethic of Elon and Michael and the team

552
00:27:52,960 --> 00:27:57,880
Speaker 15: have kind of shown that they can keep pace and

553
00:27:58,200 --> 00:28:02,880
Speaker 15: fit into this much broad ecosystem adding value, you know,

554
00:28:02,960 --> 00:28:07,640
Speaker 15: on the on terrifab, planning their orbital satellites model layer,

555
00:28:08,280 --> 00:28:12,639
Speaker 15: you know, cleaning up data to train models, you know,

556
00:28:12,720 --> 00:28:15,280
Speaker 15: helping plan infrastructure.

557
00:28:15,080 --> 00:28:15,639
Speaker 4: For the future.

558
00:28:15,680 --> 00:28:18,480
Speaker 15: It's just like it is such a strong team that

559
00:28:18,680 --> 00:28:20,600
Speaker 15: can go all the way to the bar metal in

560
00:28:20,680 --> 00:28:23,600
Speaker 15: terms of understanding everything needed to build a model of

561
00:28:23,640 --> 00:28:26,240
Speaker 15: the future. And so I like, no, it is not

562
00:28:26,400 --> 00:28:30,680
Speaker 15: an aquhier, but I cannot stress how good that team is.

563
00:28:30,720 --> 00:28:34,560
Speaker 15: And for me personally, it makes me much more optimistic

564
00:28:34,920 --> 00:28:37,720
Speaker 15: that SPACEXAI will win or at least be one of

565
00:28:37,720 --> 00:28:39,720
Speaker 15: the winning players on the Model air.

566
00:28:40,680 --> 00:28:43,320
Speaker 2: You're someone that's embedded in the teams. When I say embedded,

567
00:28:43,320 --> 00:28:45,040
Speaker 2: I mean you were able to go in work with

568
00:28:45,080 --> 00:28:47,160
Speaker 2: the team, see what they're working on across space x

569
00:28:47,360 --> 00:28:50,160
Speaker 2: xai X. And when I said that you were coming

570
00:28:50,200 --> 00:28:51,960
Speaker 2: on the show, that was pretty much one of the

571
00:28:52,040 --> 00:28:53,840
Speaker 2: questions that we got for you is how do you

572
00:28:53,880 --> 00:28:58,320
Speaker 2: think Elon must takes that cursor team and integrates it

573
00:28:58,400 --> 00:29:02,200
Speaker 2: and moves quickly. You know, there's a particular musk way

574
00:29:02,200 --> 00:29:05,120
Speaker 2: of doing things. But but you know, this has happened

575
00:29:05,600 --> 00:29:08,440
Speaker 2: three days since an IPO. It's pretty fast.

576
00:29:08,880 --> 00:29:11,040
Speaker 4: Look, I think there was a couple months.

577
00:29:12,280 --> 00:29:15,040
Speaker 15: You know, I don't know what the right technical word is,

578
00:29:15,080 --> 00:29:17,120
Speaker 15: and you know, I'm on live TV and there's a

579
00:29:17,160 --> 00:29:19,440
Speaker 15: recent IPO, But I think there was a roughly two

580
00:29:19,480 --> 00:29:24,800
Speaker 15: month trial period and I think that only one percent

581
00:29:24,960 --> 00:29:29,320
Speaker 15: or less entrepreneurs could survive that trial period of working

582
00:29:29,440 --> 00:29:31,440
Speaker 15: very closely with like a pass the test.

583
00:29:31,600 --> 00:29:32,280
Speaker 4: Yes, I think I.

584
00:29:32,280 --> 00:29:36,000
Speaker 15: Think Cursor passed the test, showing that they can keep pace,

585
00:29:36,400 --> 00:29:39,200
Speaker 15: that they can operate without making mistakes, that they can

586
00:29:39,320 --> 00:29:43,040
Speaker 15: actually elevate what's already one of the.

587
00:29:42,960 --> 00:29:44,560
Speaker 4: Most talented teams in the world.

588
00:29:45,160 --> 00:29:47,840
Speaker 15: And so I don't view this as like three days

589
00:29:47,840 --> 00:29:51,440
Speaker 15: after IPO, there's, you know, something that came out of nowhere.

590
00:29:51,480 --> 00:29:56,280
Speaker 15: I view this as there was you know, many months

591
00:29:56,400 --> 00:29:58,640
Speaker 15: leading up to that two month period and then two

592
00:29:58,720 --> 00:30:02,719
Speaker 15: months of like deep deep partnership and so like, I

593
00:30:02,760 --> 00:30:05,800
Speaker 15: personally think the probability that this isn't a creative acquisition

594
00:30:05,920 --> 00:30:08,440
Speaker 15: is like ninety nine point ninety nine nine percent because

595
00:30:08,480 --> 00:30:09,280
Speaker 15: the trial.

596
00:30:09,440 --> 00:30:10,200
Speaker 4: Was so intense.

597
00:30:10,720 --> 00:30:12,720
Speaker 2: Sean before we get to the near cloud business, I

598
00:30:12,720 --> 00:30:15,880
Speaker 2: think that's like really important and timely. We're in day

599
00:30:15,880 --> 00:30:18,760
Speaker 2: three of trading as a public company, and a lot

600
00:30:18,760 --> 00:30:22,000
Speaker 2: of people will make a lot of the milestone of

601
00:30:22,040 --> 00:30:24,320
Speaker 2: being worth more than Amazon. Maybe that happens at the

602
00:30:24,480 --> 00:30:27,720
Speaker 2: Clothes maybe it doesn't. And I'm guessing you'll tell me

603
00:30:27,760 --> 00:30:30,640
Speaker 2: you'll that you'll distribute to the LPs when the time

604
00:30:30,680 --> 00:30:33,360
Speaker 2: comes and you expect them to hold the shares. You know,

605
00:30:33,400 --> 00:30:37,320
Speaker 2: what is your kind of longer term position on SpaceX.

606
00:30:39,440 --> 00:30:42,240
Speaker 15: I mean, look for me, on a personal level, and

607
00:30:42,840 --> 00:30:44,640
Speaker 15: I say this to someone, I think I understand this

608
00:30:44,680 --> 00:30:49,000
Speaker 15: business very well, but better than you know, most investors,

609
00:30:49,000 --> 00:30:52,400
Speaker 15: better than almost any investor. I am personally in a

610
00:30:52,440 --> 00:30:56,000
Speaker 15: hold my shares in this company forever, like quite literally,

611
00:30:56,040 --> 00:30:58,680
Speaker 15: I think that. And sure there's some things that could

612
00:30:58,720 --> 00:31:04,440
Speaker 15: change or you know, you know, in the macro ten

613
00:31:04,520 --> 00:31:07,760
Speaker 15: years from now. But this is the biggest vision, the

614
00:31:07,760 --> 00:31:10,200
Speaker 15: biggest mission of any company in history, going after the

615
00:31:10,200 --> 00:31:13,320
Speaker 15: biggest markets of any company history, with the biggest moat,

616
00:31:13,560 --> 00:31:15,400
Speaker 15: and so I just think the compounding is going to

617
00:31:15,440 --> 00:31:18,440
Speaker 15: be like it's it's hard for people to even imagine

618
00:31:18,440 --> 00:31:18,960
Speaker 15: the competitions.

619
00:31:19,080 --> 00:31:21,560
Speaker 2: Part is the moat. I'm sorry to interrupt you, but

620
00:31:21,960 --> 00:31:25,920
Speaker 2: the five layer Kate that you've outlined, which one.

621
00:31:25,840 --> 00:31:27,400
Speaker 4: So obviously a great question.

622
00:31:27,880 --> 00:31:32,280
Speaker 15: By far, the biggest moat is launch, which you know,

623
00:31:32,360 --> 00:31:34,960
Speaker 15: we call people rocket scientists for a reason, or at

624
00:31:35,040 --> 00:31:37,200
Speaker 15: least in the past. You know, that was an expression

625
00:31:37,680 --> 00:31:41,280
Speaker 15: like the engineering that has gone into building Starship.

626
00:31:41,720 --> 00:31:42,880
Speaker 4: I believe it's the.

627
00:31:44,360 --> 00:31:49,040
Speaker 15: I believe it's the hardest like single engineering project that any.

628
00:31:49,000 --> 00:31:50,080
Speaker 4: Company has ever done.

629
00:31:50,120 --> 00:31:53,440
Speaker 15: I think it's to me, it's meaningly beyond like ASML,

630
00:31:54,000 --> 00:31:56,560
Speaker 15: like intrinting, complexity of building ev machine.

631
00:31:56,560 --> 00:31:57,479
Speaker 4: I'm happy to go into this.

632
00:31:57,960 --> 00:32:01,120
Speaker 15: So I would say Mote number one is the launch business,

633
00:32:01,280 --> 00:32:04,960
Speaker 15: which is just so far ahead of everyone else. Remote

634
00:32:05,040 --> 00:32:07,600
Speaker 15: number two and for SORT, I think there's this kind

635
00:32:07,600 --> 00:32:10,320
Speaker 15: of weird thing where I do think that people right

636
00:32:10,320 --> 00:32:16,640
Speaker 15: now are really underestimating the thrust you'll compute past. Yeah,

637
00:32:16,840 --> 00:32:19,240
Speaker 15: I'll go there in one second. And I think that

638
00:32:19,320 --> 00:32:23,960
Speaker 15: the moat forterrestrial compute is not launch. It's the quality

639
00:32:23,960 --> 00:32:26,480
Speaker 15: of the team that's been assembled. And you know, in

640
00:32:26,560 --> 00:32:33,320
Speaker 15: kind of this very heavy distillation process over twenty years

641
00:32:33,800 --> 00:32:38,400
Speaker 15: of finding just bringing in the most talented people from Stanford,

642
00:32:38,400 --> 00:32:41,560
Speaker 15: from MIT, et cetera, and only one percent where zero

643
00:32:41,600 --> 00:32:45,080
Speaker 15: point one percent actually survive this distillation process. What they

644
00:32:45,120 --> 00:32:47,719
Speaker 15: prove that they can work with the level of intensity,

645
00:32:47,840 --> 00:32:52,440
Speaker 15: well simultaneous level of team orientation, without making mistakes, et cetera.

646
00:32:52,880 --> 00:32:55,880
Speaker 15: So that team has figured out how to solve basically

647
00:32:55,880 --> 00:32:58,720
Speaker 15: every Harbord problem in the world, and when you apply

648
00:32:58,800 --> 00:33:01,760
Speaker 15: them to something like Treshell data centers, they're now like

649
00:33:01,840 --> 00:33:05,240
Speaker 15: they're building data centers faster and better than anyone else.

650
00:33:05,800 --> 00:33:08,200
Speaker 15: And so the team is a second mote that I

651
00:33:08,240 --> 00:33:10,520
Speaker 15: know that seems like, oh, you can just go create

652
00:33:10,520 --> 00:33:13,240
Speaker 15: a team. No, you cannot, because it took twenty plus

653
00:33:13,360 --> 00:33:16,720
Speaker 15: years of distillation. But I think that we're in this

654
00:33:16,800 --> 00:33:20,280
Speaker 15: weird metastable period where when we go five years from now,

655
00:33:20,680 --> 00:33:23,520
Speaker 15: I think that launch will again be the mote because

656
00:33:23,520 --> 00:33:26,920
Speaker 15: I think that orbital compute will be dominating net new

657
00:33:27,080 --> 00:33:29,440
Speaker 15: inference compute, like just absolutely dominating.

658
00:33:29,800 --> 00:33:31,000
Speaker 4: But I think for the next five.

659
00:33:30,840 --> 00:33:35,040
Speaker 15: Years, like the team feeds into dominating trust for compute,

660
00:33:35,200 --> 00:33:36,360
Speaker 15: so as now you want to go to the trust.

661
00:33:36,840 --> 00:33:39,800
Speaker 2: So the state of play is this the XAI team

662
00:33:39,840 --> 00:33:42,600
Speaker 2: space Xai build data center quickly. And I think your

663
00:33:42,680 --> 00:33:44,920
Speaker 2: argument is that on a dollar a token basis highly

664
00:33:44,920 --> 00:33:48,480
Speaker 2: competitively operate them. If you go on to a map

665
00:33:48,520 --> 00:33:52,480
Speaker 2: of Tennessee, they've Clossus one, Clusus two, and Clusus three.

666
00:33:53,000 --> 00:33:54,720
Speaker 2: And a big question a lot of people had was,

667
00:33:54,760 --> 00:33:59,240
Speaker 2: We'll hold on, it's amazing to get two billion dollars

668
00:33:59,280 --> 00:34:02,400
Speaker 2: per month for Anthropic and Google, but why is that

669
00:34:02,440 --> 00:34:06,960
Speaker 2: capacity not being used to train the latest greatest models

670
00:34:06,960 --> 00:34:09,439
Speaker 2: at the model air or run inference on what Rock's

671
00:34:09,440 --> 00:34:13,440
Speaker 2: already doing seemed like a blinder of a business decision,

672
00:34:13,760 --> 00:34:17,040
Speaker 2: but longer term, like it's hard to reconcile.

673
00:34:17,640 --> 00:34:21,040
Speaker 15: This is my personal opinion, kind of my assessment of

674
00:34:21,040 --> 00:34:25,319
Speaker 15: the situation. First of all, one the way, I think

675
00:34:25,360 --> 00:34:27,400
Speaker 15: that SpaceX in terms of.

676
00:34:28,400 --> 00:34:32,600
Speaker 4: In terms of treastial compute and you know, you.

677
00:34:32,560 --> 00:34:37,640
Speaker 15: Mentioned Amazon, Amazon and Microsoft are incredible businesses. I have

678
00:34:37,760 --> 00:34:42,680
Speaker 15: massive respect for those founders, for the for the businesses,

679
00:34:43,560 --> 00:34:49,200
Speaker 15: for Satia, running, running Microsoft, but a lot of their

680
00:34:49,239 --> 00:34:53,319
Speaker 15: market cap comes from their cloud businesses. And you know,

681
00:34:53,760 --> 00:34:56,719
Speaker 15: also for their legacy businesses in the case of you know,

682
00:34:57,080 --> 00:34:58,040
Speaker 15: delivery or.

683
00:34:59,600 --> 00:35:02,239
Speaker 4: Or the whole.

684
00:35:02,040 --> 00:35:06,000
Speaker 15: Microsoft Office bundle and operating system, but a lot of

685
00:35:06,040 --> 00:35:10,600
Speaker 15: it is their cloud businesses. We're in the generational transition

686
00:35:11,000 --> 00:35:14,040
Speaker 15: of cloud from kind of legacy CPU based jobs to

687
00:35:14,320 --> 00:35:21,640
Speaker 15: AI cloud. I strongly believe that SPACEXAI is the best

688
00:35:21,680 --> 00:35:26,560
Speaker 15: in the world at building this next generation of AI clouds, like,

689
00:35:27,000 --> 00:35:30,880
Speaker 15: not even close an order of magnitude, better at building

690
00:35:31,200 --> 00:35:33,879
Speaker 15: threstrial compute in this AI era than anyone else.

691
00:35:33,920 --> 00:35:36,040
Speaker 4: And it's not to say that other companies are bad.

692
00:35:36,400 --> 00:35:39,160
Speaker 15: It's to say you have the best rocket scientists in

693
00:35:39,200 --> 00:35:44,520
Speaker 15: the world are now using their skills to build threstal compute.

694
00:35:45,000 --> 00:35:47,960
Speaker 15: And what that leads to, in my opinion is, I

695
00:35:47,960 --> 00:35:51,440
Speaker 15: think people are really underestimating how much more compute is

696
00:35:51,480 --> 00:35:53,760
Speaker 15: going to come online next year and the year after

697
00:35:54,080 --> 00:35:59,760
Speaker 15: for SpaceX, and that, you know, means I really believe

698
00:35:59,760 --> 00:36:02,560
Speaker 15: space going to have a surplus of compute, and I

699
00:36:02,560 --> 00:36:05,480
Speaker 15: think it's very rational and I'm as an investor, I

700
00:36:05,560 --> 00:36:08,800
Speaker 15: love that they are offering compute to Anthropic and Google

701
00:36:08,920 --> 00:36:11,400
Speaker 15: because one, it brings in a ton of revenue that

702
00:36:11,560 --> 00:36:12,160
Speaker 15: I think.

703
00:36:13,480 --> 00:36:17,279
Speaker 4: In an IPO shows the world that all of.

704
00:36:17,320 --> 00:36:20,040
Speaker 15: This CAPEX build out is not for nothing, and it's

705
00:36:20,080 --> 00:36:23,160
Speaker 15: not all or nothing based on the model layer. There's

706
00:36:23,200 --> 00:36:27,560
Speaker 15: this you know, massive value that the company has a

707
00:36:27,600 --> 00:36:30,279
Speaker 15: dial of how they choose what to do with it.

708
00:36:31,400 --> 00:36:35,759
Speaker 15: Kind of next point, I think that people are underestimating

709
00:36:35,760 --> 00:36:39,399
Speaker 15: that most of that compute is kind of CLASSUS one,

710
00:36:39,840 --> 00:36:42,640
Speaker 15: which is roughly ten miles away from classes two and

711
00:36:42,640 --> 00:36:43,440
Speaker 15: CLASSUS three.

712
00:36:43,560 --> 00:36:44,920
Speaker 4: It's heterogeneous GPS.

713
00:36:45,560 --> 00:36:48,080
Speaker 15: Anthropic in particular has done an amazing job at being

714
00:36:48,080 --> 00:36:50,800
Speaker 15: able to run inference on heterogeneous compute.

715
00:36:50,800 --> 00:36:52,879
Speaker 4: Like there's a lot of nuances here anyways.

716
00:36:53,280 --> 00:36:54,799
Speaker 2: And by that, I think you know the expert. They

717
00:36:54,840 --> 00:36:57,960
Speaker 2: have different generations of GPU hop per black Quote. We're

718
00:36:57,960 --> 00:36:59,279
Speaker 2: going to take a quick break and we're going to

719
00:36:59,440 --> 00:37:01,440
Speaker 2: keep the covers going. We have plenty of time coming up.

720
00:37:01,520 --> 00:37:04,960
Speaker 2: SpaceX has big plans for orbital data centers and Squoil

721
00:37:04,960 --> 00:37:07,120
Speaker 2: Sew MacGuire is going to stick with us and break

722
00:37:07,200 --> 00:37:11,399
Speaker 2: down how and when that space compute becomes reality. That's next.

723
00:37:11,400 --> 00:37:12,239
Speaker 2: This is bloom bo tech.

724
00:37:18,800 --> 00:37:22,400
Speaker 16: Data centers just require an enormous amount of power in

725
00:37:22,480 --> 00:37:25,680
Speaker 16: order to run them, also to cool them, and so

726
00:37:25,760 --> 00:37:29,640
Speaker 16: they've essentially hit what is an energy wall, and.

727
00:37:29,640 --> 00:37:31,959
Speaker 17: Yet demand keeps accelerating.

728
00:37:32,960 --> 00:37:33,520
Speaker 4: Globally.

729
00:37:33,680 --> 00:37:36,960
Speaker 17: Electricity use for data centers is expected to double by

730
00:37:37,000 --> 00:37:41,680
Speaker 17: twenty thirty. By twenty fifty, it will likely represent a

731
00:37:41,880 --> 00:37:47,279
Speaker 17: tenth of all electricity consumed on Earth. So what happens

732
00:37:47,400 --> 00:37:52,319
Speaker 17: when artificial intelligence outpaces the power and land needed to

733
00:37:52,360 --> 00:37:53,000
Speaker 17: sustain it?

734
00:37:55,960 --> 00:37:56,200
Speaker 10: Why?

735
00:37:57,160 --> 00:38:02,799
Speaker 17: The answer may be above Earth. In late twenty twenty five,

736
00:38:03,080 --> 00:38:06,800
Speaker 17: space startups star Cloud sent a satellite into orbit carrying

737
00:38:06,840 --> 00:38:11,400
Speaker 17: an Nvidia H one hundred chip, the most powerful processor

738
00:38:11,480 --> 00:38:15,759
Speaker 17: ever deployed in space. Star Cloud one is not a

739
00:38:15,840 --> 00:38:19,640
Speaker 17: data center, not yet. It's a single prototype in orbit.

740
00:38:20,800 --> 00:38:24,440
Speaker 17: The future that Star Cloud imagines is tens of thousands

741
00:38:24,480 --> 00:38:27,840
Speaker 17: of satellites, each carrying chips like this one, working together

742
00:38:28,080 --> 00:38:32,399
Speaker 17: to do complex large scale computing in space, also known

743
00:38:32,719 --> 00:38:37,200
Speaker 17: as orbital compute. For many, The vision begins with a

744
00:38:37,320 --> 00:38:41,200
Speaker 17: prompt from Earth beamed into space through laser links. The

745
00:38:41,320 --> 00:38:45,480
Speaker 17: request is then processed by AI chips inside a satellite bus,

746
00:38:45,840 --> 00:38:49,359
Speaker 17: which is powered by massive solar arrays stretching as much

747
00:38:49,400 --> 00:38:54,120
Speaker 17: as four square kilometers. Together, these satellites form a linked

748
00:38:54,120 --> 00:38:58,240
Speaker 17: computing network that then sends results back to Earth in milliseconds.

749
00:39:00,040 --> 00:39:02,400
Speaker 9: Two years ago, when we first started talking about building

750
00:39:02,520 --> 00:39:05,200
Speaker 9: data centers in space, people you know, frankly thought we

751
00:39:05,200 --> 00:39:05,400
Speaker 9: were a.

752
00:39:05,400 --> 00:39:07,640
Speaker 2: Bit crazy, and it does sound a little bit sci fi.

753
00:39:08,920 --> 00:39:12,000
Speaker 17: What sounds like sci fi is a response to a

754
00:39:12,120 --> 00:39:17,440
Speaker 17: very real problem on Earth. Data centers require vast amounts

755
00:39:17,480 --> 00:39:21,920
Speaker 17: of land, water, and electricity, and as the AI systems

756
00:39:22,000 --> 00:39:25,440
Speaker 17: inside them grow more powerful, the heat they generate is

757
00:39:25,440 --> 00:39:30,640
Speaker 17: becoming harder and costlier to contain. In space. On the

758
00:39:30,680 --> 00:39:34,240
Speaker 17: other hand, there are no land constraints or permitting battles

759
00:39:34,440 --> 00:39:38,560
Speaker 17: not to mention access to almost unlimited solar energy if

760
00:39:38,960 --> 00:39:40,160
Speaker 17: it can be made to work.

761
00:39:42,760 --> 00:39:45,040
Speaker 2: So that was a segment from Bloomberger Regionals looking at

762
00:39:45,120 --> 00:39:47,720
Speaker 2: orbital data centers, and here on the show, Shan MacGuire

763
00:39:47,920 --> 00:39:51,279
Speaker 2: Square Capital still with us the firm stake of around

764
00:39:51,320 --> 00:39:54,400
Speaker 2: one point five percent in SpaceX, and this is a

765
00:39:54,440 --> 00:39:57,479
Speaker 2: big part of the plan in the prospectus. The team

766
00:39:57,560 --> 00:40:01,279
Speaker 2: is saying at SpaceX as early as twenty twenty eight,

767
00:40:02,040 --> 00:40:03,840
Speaker 2: in a moment, we're going to bring up a Bloomberg

768
00:40:03,920 --> 00:40:07,560
Speaker 2: rendering of the design of and that's actually the SpaceX one.

769
00:40:07,680 --> 00:40:07,839
Speaker 4: Right.

770
00:40:08,840 --> 00:40:10,560
Speaker 2: I think you're probably in the camp of people that

771
00:40:11,080 --> 00:40:14,960
Speaker 2: think they can do this quickly. Where do you stand

772
00:40:15,120 --> 00:40:19,120
Speaker 2: on your data orbital data center outlook for SpaceX?

773
00:40:20,160 --> 00:40:23,760
Speaker 4: I stand incredibly bullish, which just incredibly bullish.

774
00:40:24,200 --> 00:40:25,840
Speaker 2: Yeah, I'm not surprised by that.

775
00:40:26,000 --> 00:40:32,440
Speaker 15: Yeah, but like I think it's rationally rationally bullish. I

776
00:40:32,480 --> 00:40:34,239
Speaker 15: think something that a lot of people are struggling with

777
00:40:34,360 --> 00:40:38,080
Speaker 15: here is most people just never thought about what is

778
00:40:38,080 --> 00:40:39,719
Speaker 15: the satellite? How do you build a satellite, What are

779
00:40:39,719 --> 00:40:42,040
Speaker 15: the components of a satellite? What are the constraints around

780
00:40:42,120 --> 00:40:45,360
Speaker 15: launching these satellites? And I think that people that have

781
00:40:45,360 --> 00:40:49,520
Speaker 15: been following starlink closely are very well positioned.

782
00:40:49,040 --> 00:40:51,919
Speaker 2: To assess because you can extrap lay out on both

783
00:40:52,000 --> 00:40:55,640
Speaker 2: the economics but also the engineering and physics exactly.

784
00:40:55,760 --> 00:40:59,960
Speaker 15: The orbital compute satellites are very similar in many many

785
00:41:00,239 --> 00:41:02,080
Speaker 15: ways to starling satellites.

786
00:41:03,880 --> 00:41:04,799
Speaker 4: I truly believe in.

787
00:41:04,920 --> 00:41:07,760
Speaker 15: Elon said this last week in a video on orbital

788
00:41:07,800 --> 00:41:13,040
Speaker 15: compute that the intrinsic complexity of making a stalling satellite

789
00:41:13,400 --> 00:41:17,640
Speaker 15: is probably a little bit greater than an orbital compute satellite. So,

790
00:41:17,960 --> 00:41:20,080
Speaker 15: I know this sounds kind of crazy, but starlink is

791
00:41:20,120 --> 00:41:24,480
Speaker 15: on the satellites themselves. Stalling satellites are more complicated than

792
00:41:25,080 --> 00:41:28,800
Speaker 15: orbital compute satellites. That said, orbital compute satellites are bigger.

793
00:41:29,080 --> 00:41:30,880
Speaker 15: For them to make sense, they have to be much bigger.

794
00:41:30,920 --> 00:41:33,280
Speaker 15: You basically want to put a whole server rack in space.

795
00:41:33,520 --> 00:41:35,680
Speaker 15: Just in terms of doing inference on a model, you

796
00:41:35,760 --> 00:41:38,440
Speaker 15: kind of need a minimum size for it to be useful.

797
00:41:39,160 --> 00:41:43,600
Speaker 15: And you can't launch these bigger satellites unless you have starship.

798
00:41:44,320 --> 00:41:45,959
Speaker 4: So almost all of.

799
00:41:45,880 --> 00:41:50,120
Speaker 15: The difficulty of orbital compute satellites is in starship.

800
00:41:50,280 --> 00:41:51,480
Speaker 4: It's in the launch vehicle.

801
00:41:51,880 --> 00:41:55,720
Speaker 15: And so if you want to estimate when orbital compute

802
00:41:55,840 --> 00:41:59,880
Speaker 15: is going to be scaling, it's really a question around

803
00:42:00,440 --> 00:42:02,319
Speaker 15: like when is Starship going.

804
00:42:02,239 --> 00:42:03,960
Speaker 2: To be still we're showing the videos. This is this

805
00:42:04,000 --> 00:42:07,000
Speaker 2: is file right of Starship, and you know the basics

806
00:42:07,040 --> 00:42:09,760
Speaker 2: of it are that on paper, you know, one hundred

807
00:42:09,800 --> 00:42:12,280
Speaker 2: and thirty five one hundred and fifty metric tons payload

808
00:42:12,360 --> 00:42:14,360
Speaker 2: to orbit. And I think there's some idea that in

809
00:42:14,400 --> 00:42:16,680
Speaker 2: the future in iterations four hundred metric tons.

810
00:42:18,200 --> 00:42:20,960
Speaker 15: I think the company said, you know, two hundred metric

811
00:42:21,000 --> 00:42:25,160
Speaker 15: tons is pretty like pretty achievable. Four hundred metric metric

812
00:42:25,200 --> 00:42:28,800
Speaker 15: guns maybe I don't know. It seems likely, but I

813
00:42:28,840 --> 00:42:31,120
Speaker 15: don't know. You just make it. There's something with rockets

814
00:42:31,120 --> 00:42:33,800
Speaker 15: where the taller you make it, kind of the easier

815
00:42:33,840 --> 00:42:37,680
Speaker 15: things become the surface area of volume ratio and that

816
00:42:38,000 --> 00:42:43,000
Speaker 15: benefits you with drag. So it seems very likely, but

817
00:42:43,680 --> 00:42:45,080
Speaker 15: manufacturing it that skills harder.

818
00:42:45,200 --> 00:42:46,200
Speaker 2: Can you compare and contrast?

819
00:42:46,239 --> 00:42:46,359
Speaker 6: Then?

820
00:42:46,440 --> 00:42:49,480
Speaker 2: So you outlined in great detail why you're bullish on

821
00:42:49,560 --> 00:42:53,480
Speaker 2: the terrestrial data center neocloud business. Where does the orbital

822
00:42:53,560 --> 00:42:57,320
Speaker 2: data center business compare over? It's just like reasonable timeline

823
00:42:57,320 --> 00:42:57,560
Speaker 2: for you.

824
00:42:57,680 --> 00:43:02,640
Speaker 4: Yep again, And I love the question. I appreciate the question.

825
00:43:03,480 --> 00:43:06,040
Speaker 15: I think people are looking at orbital compute and they think,

826
00:43:06,080 --> 00:43:08,520
Speaker 15: oh wow, this sounds so crazy, and I wasn't even

827
00:43:08,920 --> 00:43:11,440
Speaker 15: I wasn't thinking about it two years ago. So it

828
00:43:11,560 --> 00:43:13,600
Speaker 15: must take a long time. And a lot of Elon's

829
00:43:13,600 --> 00:43:16,120
Speaker 15: timelines in the past have have been wrong, which you know,

830
00:43:16,160 --> 00:43:20,440
Speaker 15: he says himself. But if you look at what an

831
00:43:20,520 --> 00:43:24,959
Speaker 15: orbital compute satellite is, it's basically some compute element, whether

832
00:43:25,000 --> 00:43:25,359
Speaker 15: that's a.

833
00:43:25,280 --> 00:43:28,919
Speaker 4: GPU or a TPU or tranium or you know, proprietary.

834
00:43:29,000 --> 00:43:30,839
Speaker 2: You said you have reference designs for all of them.

835
00:43:30,920 --> 00:43:34,239
Speaker 15: Yeah, but so there's the compute, and you want a

836
00:43:34,280 --> 00:43:37,840
Speaker 15: bout a server racks amount of compute. There are solar

837
00:43:37,880 --> 00:43:42,880
Speaker 15: panels like that's easy, fully kind of proven, uh, spaces

838
00:43:42,920 --> 00:43:44,600
Speaker 15: has huge experience building solar panels.

839
00:43:44,920 --> 00:43:47,319
Speaker 4: And the third main piece are radiators.

840
00:43:47,600 --> 00:43:49,920
Speaker 15: So this is what takes all the heat from the

841
00:43:49,960 --> 00:43:54,600
Speaker 15: compute chips and radiates it into space via you know,

842
00:43:55,880 --> 00:43:57,920
Speaker 15: via black body radiation.

843
00:43:59,520 --> 00:44:00,920
Speaker 4: Those are the three main things.

844
00:44:00,960 --> 00:44:04,480
Speaker 15: And then you also have laser links to connect to

845
00:44:04,520 --> 00:44:07,200
Speaker 15: other satellites or you know, connect to the ground, and

846
00:44:07,480 --> 00:44:13,160
Speaker 15: radio connect to the ground. Of these things, every individual

847
00:44:13,200 --> 00:44:18,040
Speaker 15: component here is kind of fully fully proven by SpaceX,

848
00:44:18,040 --> 00:44:23,160
Speaker 15: except for the compute side, but that is not it's

849
00:44:23,200 --> 00:44:26,439
Speaker 15: just not that heart and so the satellites themselves are

850
00:44:26,840 --> 00:44:27,879
Speaker 15: pretty easy to make.

851
00:44:28,000 --> 00:44:31,480
Speaker 2: That's rendering, again, a bloomberg rendering of what you're describing.

852
00:44:31,680 --> 00:44:35,040
Speaker 15: Yeah, So to me, it really is a question of

853
00:44:35,040 --> 00:44:39,160
Speaker 15: when Starship is flying. I don't I don't understand why

854
00:44:39,320 --> 00:44:42,120
Speaker 15: SpaceX wouldn't be able to have an orbital compute satellite

855
00:44:42,160 --> 00:44:44,759
Speaker 15: up within you know, six months or so of the

856
00:44:44,800 --> 00:44:47,319
Speaker 15: first storage Starship Payload play.

857
00:44:47,440 --> 00:44:49,200
Speaker 2: We have ninety seconds left in the show. I can't

858
00:44:49,239 --> 00:44:52,279
Speaker 2: believe it. We need to address the open AI and

859
00:44:52,320 --> 00:44:55,799
Speaker 2: anthropic elephant in the room, which is on paper, they

860
00:44:55,840 --> 00:44:58,080
Speaker 2: are very similar if you read the prospectus and the

861
00:44:58,120 --> 00:45:02,000
Speaker 2: tam that SpaceX outlines their room on Earth and outside

862
00:45:02,040 --> 00:45:03,399
Speaker 2: of the Earth for all three.

863
00:45:04,480 --> 00:45:09,719
Speaker 15: I think intelligence is you know, the most important. It's

864
00:45:09,719 --> 00:45:13,719
Speaker 15: the defining capability of our time. And I think there's

865
00:45:14,200 --> 00:45:18,080
Speaker 15: unlimited demand for intelligence. I think it's very clear that

866
00:45:18,120 --> 00:45:21,760
Speaker 15: SpaceX will be, in my opinion, will be the dominant

867
00:45:21,760 --> 00:45:24,960
Speaker 15: provider of compute for intelligence. And I don't know what's

868
00:45:24,960 --> 00:45:27,280
Speaker 15: going to happen on the model layer, but there's absolute

869
00:45:27,320 --> 00:45:29,080
Speaker 15: space for all of these companies.

870
00:45:29,920 --> 00:45:32,200
Speaker 2: Fifteen seconds. If you had like a kind of black

871
00:45:32,239 --> 00:45:35,040
Speaker 2: Swan memo right now for the firm Sequoia. What would

872
00:45:35,040 --> 00:45:38,560
Speaker 2: the headline be. What's the biggest focus for you?

873
00:45:40,320 --> 00:45:44,520
Speaker 15: I think it's probably regulatory risk regulatory.

874
00:45:44,719 --> 00:45:46,720
Speaker 2: Okay, let's leave it as a cliffhanger, and you're always

875
00:45:46,760 --> 00:45:49,120
Speaker 2: welcome here back on the show, Bomberg Tech, thank you

876
00:45:49,239 --> 00:45:51,239
Speaker 2: very much for coming in. That does do it for

877
00:45:51,239 --> 00:45:53,279
Speaker 2: this addition of Bloomberg Tech. I want to check on

878
00:45:53,320 --> 00:45:55,280
Speaker 2: the market. It's actually in the course of that conversation

879
00:45:55,400 --> 00:45:57,960
Speaker 2: then as that one hundred headed south, it's paired some

880
00:45:58,000 --> 00:46:00,960
Speaker 2: of its declines, but with down one point two, SpaceX

881
00:46:01,000 --> 00:46:03,600
Speaker 2: is not at session highs, but it is up ten percent.

882
00:46:03,680 --> 00:46:08,120
Speaker 2: It is past Amazon on market value as it stands

883
00:46:08,120 --> 00:46:10,840
Speaker 2: where we close who knows long time away and maybe

884
00:46:10,880 --> 00:46:13,400
Speaker 2: it will eclipse Microsoft and market cap by the closer

885
00:46:13,400 --> 00:46:15,960
Speaker 2: as well. Stay with us on Bloomberg TV throughout the day.

886
00:46:16,000 --> 00:46:18,160
Speaker 2: What a show it's been. Recap on the podcast. Really

887
00:46:18,200 --> 00:46:20,600
Speaker 2: appreciate all of you that do listen to the show

888
00:46:20,600 --> 00:46:23,000
Speaker 2: as a podcast. You know where to find it on Bloomberg, Spotify,

889
00:46:23,080 --> 00:46:26,919
Speaker 2: Apple and iHeart from San Francisco. This is Bloomberg Tech.