1
00:00:00,280 --> 00:00:07,200
Speaker 1: Bloomberg, Audio Studios, podcasts, radio News.

2
00:00:08,560 --> 00:00:11,200
Speaker 2: This week, two of the biggest players in the AI

3
00:00:11,320 --> 00:00:15,400
Speaker 2: industry struck a deal. Open Ai, which owns chat GPT,

4
00:00:15,880 --> 00:00:18,599
Speaker 2: agreed to use tens of billions of dollars worth of

5
00:00:18,640 --> 00:00:24,400
Speaker 2: AI chips over multiple years from Advanced micro Devices or AMD.

6
00:00:24,760 --> 00:00:28,000
Speaker 2: An AMD set the stage for open Ai to become

7
00:00:28,160 --> 00:00:30,120
Speaker 2: one of its largest shareholders.

8
00:00:30,680 --> 00:00:34,400
Speaker 1: It speaks to the commitments that some of these AI

9
00:00:34,479 --> 00:00:37,640
Speaker 1: infrastructure companies are making with open Ai to facilitate this

10
00:00:37,760 --> 00:00:38,880
Speaker 1: AI boom that's happening.

11
00:00:39,120 --> 00:00:41,760
Speaker 2: Emily Foregash is a tech reporter at Bloomberg.

12
00:00:41,960 --> 00:00:44,560
Speaker 1: The CEO of AMD and then the president of open

13
00:00:44,600 --> 00:00:47,040
Speaker 1: Ai were both on Bloomberg and they both spoke about

14
00:00:47,040 --> 00:00:48,880
Speaker 1: how this is a win win for each other.

15
00:00:49,159 --> 00:00:50,160
Speaker 2: This deal is.

16
00:00:50,080 --> 00:00:53,000
Speaker 1: A win for AMD, it's a win for open Ai,

17
00:00:53,200 --> 00:00:55,320
Speaker 1: and it's a win for our shareholders. We really think

18
00:00:55,360 --> 00:00:58,680
Speaker 1: of this as an industry wide effort. If open Ai wins,

19
00:00:58,680 --> 00:01:01,200
Speaker 1: that's a win for AMD because AMDs chips will be

20
00:01:01,280 --> 00:01:03,920
Speaker 1: used for part of the inferencing and vice versa when

21
00:01:03,960 --> 00:01:07,040
Speaker 1: AMD does well. Now open Ai has a stake in

22
00:01:07,080 --> 00:01:07,640
Speaker 1: that company.

23
00:01:08,000 --> 00:01:11,800
Speaker 2: If this arrangement sounds a bit circular to you, that's

24
00:01:11,840 --> 00:01:15,119
Speaker 2: because it is an AI chip maker and an AI

25
00:01:15,240 --> 00:01:19,319
Speaker 2: developer making a multi billion dollar bet on each other.

26
00:01:20,280 --> 00:01:23,120
Speaker 2: It's not the only deal open ai has broken recently

27
00:01:23,319 --> 00:01:25,240
Speaker 2: with the major AI chip maker.

28
00:01:25,560 --> 00:01:28,360
Speaker 1: The circular nature has kind of come to attention in

29
00:01:28,520 --> 00:01:32,520
Speaker 1: recent weeks because of the scale, the size, the frequency

30
00:01:32,560 --> 00:01:34,959
Speaker 1: of these deals. Part of it was tipped off in

31
00:01:35,040 --> 00:01:37,319
Speaker 1: large part by in Vidia's up to one hundred billion

32
00:01:37,360 --> 00:01:39,000
Speaker 1: dollar investment in open Ai.

33
00:01:39,160 --> 00:01:43,160
Speaker 2: Open ai and AMD open Ai and in VideA. There's

34
00:01:43,200 --> 00:01:46,640
Speaker 2: a lot of AI on AI investment going on right now.

35
00:01:46,840 --> 00:01:49,800
Speaker 1: Open ai has a three hundred billion dollar deal with

36
00:01:49,880 --> 00:01:53,480
Speaker 1: Oracle to compete on Oracle's cloud network space. In Vidia

37
00:01:54,040 --> 00:01:58,480
Speaker 1: and core Weave, another cloud provider, also has a relationship.

38
00:01:58,520 --> 00:02:01,440
Speaker 1: In Vidia bought six point three billion dollars of cloud

39
00:02:01,440 --> 00:02:04,400
Speaker 1: services from open Ai also has a deal with core

40
00:02:04,440 --> 00:02:06,640
Speaker 1: Weave to pay as much as twenty two point four

41
00:02:06,680 --> 00:02:09,800
Speaker 1: billion dollars for that compute space as well. Open Ai

42
00:02:09,919 --> 00:02:12,519
Speaker 1: and AMD had the deal that happened on Monday.

43
00:02:12,560 --> 00:02:17,440
Speaker 2: Obviously, AI companies say these kinds of large, reciprocal and

44
00:02:17,520 --> 00:02:21,400
Speaker 2: sometimes overlapping deals are what's needed to meet growing demand

45
00:02:21,520 --> 00:02:24,359
Speaker 2: for their tech, But as more and more money gets

46
00:02:24,440 --> 00:02:29,000
Speaker 2: invested in the space. These unorthodox business arrangements are also

47
00:02:29,080 --> 00:02:32,760
Speaker 2: raising flags for industry analysts who are starting to wonder

48
00:02:33,320 --> 00:02:36,639
Speaker 2: is the trillion dollar AI market being propped up by

49
00:02:36,680 --> 00:02:37,720
Speaker 2: the industry itself.

50
00:02:38,520 --> 00:02:41,720
Speaker 1: The circular financing concern comes hand in hand with concerns

51
00:02:41,760 --> 00:02:44,160
Speaker 1: over an AI bubble, and so when you're putting that

52
00:02:44,240 --> 00:02:46,959
Speaker 1: much money into an industry that hasn't returned on those

53
00:02:47,400 --> 00:02:50,919
Speaker 1: investments at this point in time, you could worry that

54
00:02:51,000 --> 00:02:53,480
Speaker 1: if it never returns on those investments, or if there's

55
00:02:53,760 --> 00:02:56,919
Speaker 1: never turning a profit, then the whole thing will kind

56
00:02:56,919 --> 00:02:57,880
Speaker 1: of combust.

57
00:03:00,639 --> 00:03:03,040
Speaker 2: I'm Sarah Holder, and this is the big take from

58
00:03:03,040 --> 00:03:06,200
Speaker 2: Bloomberg News today on the show why a network of

59
00:03:06,280 --> 00:03:10,080
Speaker 2: circular AI deals is fueling concerns about the possibility of

60
00:03:10,080 --> 00:03:18,399
Speaker 2: a bubble and what could happen if it pops. Over

61
00:03:18,400 --> 00:03:21,160
Speaker 2: the last few months, there's been a frenzy of deal

62
00:03:21,200 --> 00:03:22,960
Speaker 2: making in the AI space.

63
00:03:22,840 --> 00:03:25,760
Speaker 1: And everyone's talking about AI. Everyone's talking about also this

64
00:03:25,840 --> 00:03:28,760
Speaker 1: deal between open Ai and AMD adding.

65
00:03:28,520 --> 00:03:30,320
Speaker 2: Fuel to the AI frenzy.

66
00:03:30,639 --> 00:03:33,480
Speaker 1: The deal expected to generate tens of billions of dollars

67
00:03:33,480 --> 00:03:36,120
Speaker 1: in new revenue, a historic collaboration.

68
00:03:36,360 --> 00:03:38,440
Speaker 2: As part of the deal, Intel is going to use

69
00:03:38,480 --> 00:03:43,119
Speaker 2: in videographic technology and upcoming PC chips. Emily Forgash, who

70
00:03:43,120 --> 00:03:46,440
Speaker 2: covers tech for Bloomberg, has been covering these deals closely,

71
00:03:46,880 --> 00:03:49,480
Speaker 2: and she says a lot of them have involved two

72
00:03:49,720 --> 00:03:51,280
Speaker 2: dominant AI companies.

73
00:03:51,640 --> 00:03:53,839
Speaker 1: The center of the web is definitely in Vidian Open

74
00:03:53,880 --> 00:03:54,840
Speaker 1: Ai in.

75
00:03:54,960 --> 00:03:59,240
Speaker 2: Video designs chips that power AI systems, and open Ai

76
00:03:59,520 --> 00:04:05,040
Speaker 2: developed AI and AI driven products like chat GPT. In September,

77
00:04:05,360 --> 00:04:08,600
Speaker 2: just two weeks before open Ai announced its deal with AMD,

78
00:04:09,280 --> 00:04:12,600
Speaker 2: open Ai and Nvidia announced a partnership of their own.

79
00:04:13,320 --> 00:04:15,840
Speaker 2: Nvidia agreed to invest as much as one hundred billion

80
00:04:15,880 --> 00:04:19,359
Speaker 2: dollars in open Ai to build data centers, and open

81
00:04:19,400 --> 00:04:23,280
Speaker 2: Ai agreed to fill those sites with millions of Nvidia chips.

82
00:04:24,240 --> 00:04:27,560
Speaker 2: Emily says, in some ways, the deal just made good

83
00:04:27,680 --> 00:04:28,400
Speaker 2: business sense.

84
00:04:29,000 --> 00:04:31,280
Speaker 1: Part of that is just the position that Nvidia has

85
00:04:31,279 --> 00:04:34,159
Speaker 1: found itself in. It's the wealthiest company in the entire world,

86
00:04:34,200 --> 00:04:36,120
Speaker 1: the biggest company in the entire world, and it has

87
00:04:36,160 --> 00:04:37,080
Speaker 1: dominance in chips.

88
00:04:37,480 --> 00:04:41,359
Speaker 2: But it also raised eyebrows among industry analysts.

89
00:04:41,279 --> 00:04:45,240
Speaker 1: The circular nature of investing in a company like open ai,

90
00:04:45,680 --> 00:04:49,240
Speaker 1: which then needs compute capacity which could ultimately use in

91
00:04:49,360 --> 00:04:51,839
Speaker 1: Vidia's chips in the end and give back to Nvidia

92
00:04:51,880 --> 00:04:55,120
Speaker 1: in that way. That's kind of more recently becoming a concern.

93
00:04:55,440 --> 00:04:59,560
Speaker 1: It's not necessarily planned to be circular, but analysts have

94
00:04:59,560 --> 00:05:02,359
Speaker 1: pointed out the fact that when you froth up a

95
00:05:02,400 --> 00:05:04,240
Speaker 1: market like this, it could lead to a bubble, and

96
00:05:04,839 --> 00:05:07,599
Speaker 1: the original sin of this bubble could have been like

97
00:05:07,640 --> 00:05:08,599
Speaker 1: circular financing.

98
00:05:08,960 --> 00:05:12,680
Speaker 2: The industry has seen a number of other circular financing

99
00:05:12,720 --> 00:05:16,760
Speaker 2: deals recently between open ai and Oracle in Nvidia and

100
00:05:16,839 --> 00:05:20,840
Speaker 2: core Weave, Open Ai and core Weave, Open Ai and AMD,

101
00:05:21,600 --> 00:05:25,640
Speaker 2: and in Vidia has also made investments in smaller AI companies.

102
00:05:27,240 --> 00:05:30,000
Speaker 1: In Vidia doesn't just invest hundreds of billions of dollars

103
00:05:30,080 --> 00:05:32,440
Speaker 1: into the biggest companies like open Ai. It also has

104
00:05:32,440 --> 00:05:35,039
Speaker 1: been funding a lot of AI startups, in large part

105
00:05:35,120 --> 00:05:38,320
Speaker 1: to push AI innovation forward, with the thinking that if

106
00:05:38,360 --> 00:05:40,640
Speaker 1: you invest in the smaller AI startups and kind of

107
00:05:40,640 --> 00:05:43,560
Speaker 1: give them a boost, then they'll ultimately get somewhere, maybe

108
00:05:43,640 --> 00:05:45,839
Speaker 1: not somewhere as big as open ai, but maybe somewhere close.

109
00:05:46,279 --> 00:05:49,080
Speaker 2: Why are open Ai and Nvidia at the center of

110
00:05:49,120 --> 00:05:50,200
Speaker 2: so many of these deals.

111
00:05:50,320 --> 00:05:52,400
Speaker 1: Part of the reason in Nvidia can put its hands

112
00:05:52,440 --> 00:05:54,760
Speaker 1: in so many places is because it has so much

113
00:05:54,800 --> 00:05:57,560
Speaker 1: money to do so. But also, as I mentioned before,

114
00:05:57,600 --> 00:06:00,919
Speaker 1: in Vidia is the dominant chip make in the US,

115
00:06:01,000 --> 00:06:03,760
Speaker 1: and it's competing for that title globally as well. So

116
00:06:04,000 --> 00:06:08,280
Speaker 1: most of the high level AI capacity comes from in

117
00:06:08,440 --> 00:06:13,480
Speaker 1: videos advanced chips. Open ai was just very much ahead

118
00:06:13,480 --> 00:06:15,960
Speaker 1: of the curve. It has more than seven hundred million

119
00:06:16,360 --> 00:06:20,160
Speaker 1: weekly users now of chat gibt. It's just becoming so

120
00:06:20,440 --> 00:06:23,799
Speaker 1: well known and so used. It's almost gotten like verb status,

121
00:06:24,200 --> 00:06:25,760
Speaker 1: the way that we say, oh, I'm just going to

122
00:06:25,839 --> 00:06:27,479
Speaker 1: Google that, and by that we mean we're going to

123
00:06:27,480 --> 00:06:29,600
Speaker 1: search on the Internet. Now I'm kind of hearing people

124
00:06:29,640 --> 00:06:31,640
Speaker 1: say I'm just going to chat gbt that rather than

125
00:06:31,680 --> 00:06:33,800
Speaker 1: being like I'm going to search on an LM or

126
00:06:33,839 --> 00:06:36,000
Speaker 1: I'm going to search through one of these chatbots. It's

127
00:06:36,080 --> 00:06:38,719
Speaker 1: chat gbt that they go to. And so that kind

128
00:06:38,760 --> 00:06:42,440
Speaker 1: of global dominance that both companies have and the amount

129
00:06:42,440 --> 00:06:44,800
Speaker 1: of money that they both are dealing with. At this

130
00:06:44,839 --> 00:06:47,440
Speaker 1: point in time, open ai just reached a five hundred

131
00:06:47,440 --> 00:06:51,000
Speaker 1: billion dollar valuation. Those two are making it kind of

132
00:06:51,720 --> 00:06:53,240
Speaker 1: ultimately the center of this web.

133
00:06:53,600 --> 00:06:55,799
Speaker 2: But does open a I have the revenue to support

134
00:06:55,839 --> 00:06:58,720
Speaker 2: these kinds of deals. From my understanding, when I chat

135
00:06:58,800 --> 00:07:00,440
Speaker 2: GPT something that's free.

136
00:07:00,600 --> 00:07:03,640
Speaker 1: Open Ai does have high revenue, but it also burns cash.

137
00:07:03,640 --> 00:07:06,200
Speaker 1: It's not a profitable company at this point in time.

138
00:07:06,640 --> 00:07:09,400
Speaker 1: That is one of the key concerns of this circular

139
00:07:09,440 --> 00:07:13,520
Speaker 1: financing leading to an AI bubble, because companies are just

140
00:07:13,520 --> 00:07:16,720
Speaker 1: putting so much money into not just open ai, but

141
00:07:16,760 --> 00:07:19,720
Speaker 1: the other companies that open ai surrounds itself with, and

142
00:07:19,760 --> 00:07:22,480
Speaker 1: it's just kind of propping up the entire AI sector.

143
00:07:22,800 --> 00:07:25,400
Speaker 1: There's one hundred billion dollars from Nvidian open Ai, There's

144
00:07:25,440 --> 00:07:28,520
Speaker 1: three hundred billion dollars with open ai and Oracle. There's

145
00:07:28,640 --> 00:07:30,800
Speaker 1: as much as twenty two billion dollars between open ai

146
00:07:30,800 --> 00:07:32,560
Speaker 1: and core Weave, and those continuing to grow. And then

147
00:07:32,600 --> 00:07:34,840
Speaker 1: if you think about how much those companies are also

148
00:07:34,880 --> 00:07:39,120
Speaker 1: then spending on compute, spending on chips, the numbers are

149
00:07:39,160 --> 00:07:41,880
Speaker 1: just staggering in terms of how much money they're giving

150
00:07:41,880 --> 00:07:44,560
Speaker 1: to each other. And so the concern is that all

151
00:07:44,600 --> 00:07:46,880
Speaker 1: this money is going in, but none of it will

152
00:07:46,920 --> 00:07:49,000
Speaker 1: ever come back out because there will never be a profit.

153
00:07:49,480 --> 00:07:53,840
Speaker 2: And that's the key concern here that a complex, interconnected

154
00:07:53,880 --> 00:07:58,440
Speaker 2: web of investments are artificially inflating valuations of these AI

155
00:07:58,560 --> 00:08:03,760
Speaker 2: companies before they actually make any money. Is this something

156
00:08:03,800 --> 00:08:07,800
Speaker 2: that would normally concern regulators or is this just how

157
00:08:07,960 --> 00:08:11,320
Speaker 2: the industry works? In other words, does it just make

158
00:08:11,360 --> 00:08:14,760
Speaker 2: sense for AI companies to partner with other AI companies

159
00:08:14,840 --> 00:08:17,640
Speaker 2: because their goals are aligned, or is there something else

160
00:08:17,800 --> 00:08:18,520
Speaker 2: going on here?

161
00:08:18,800 --> 00:08:22,080
Speaker 1: So whether or not regulars get involved is to be seen,

162
00:08:22,200 --> 00:08:24,840
Speaker 1: But at this point in time, it's not necessarily shady

163
00:08:24,880 --> 00:08:27,440
Speaker 1: by any means. It's more so just that these companies

164
00:08:27,480 --> 00:08:30,160
Speaker 1: have so much money. In Video, for example, has so

165
00:08:30,240 --> 00:08:33,480
Speaker 1: much money because of its chip dominance, and it has

166
00:08:33,520 --> 00:08:38,360
Speaker 1: an interest in propelling the AI innovation around the world forward.

167
00:08:38,520 --> 00:08:40,240
Speaker 1: I think that would kind of be the argument for

168
00:08:40,320 --> 00:08:43,079
Speaker 1: why the deals are kind of happening this way, although

169
00:08:43,080 --> 00:08:46,560
Speaker 1: they're super new, this kind of way of financing. These

170
00:08:46,760 --> 00:08:50,400
Speaker 1: AI companies are using equity, debt, investing in one another,

171
00:08:50,480 --> 00:08:52,880
Speaker 1: investing huge tons of money in one another. So it's

172
00:08:52,920 --> 00:08:56,400
Speaker 1: kind of like, since AI is rapidly advancing and there's

173
00:08:56,440 --> 00:08:59,320
Speaker 1: a huge stake in doing so, these companies are just

174
00:08:59,320 --> 00:09:01,640
Speaker 1: pouring in the money as much as they can to

175
00:09:01,640 --> 00:09:02,520
Speaker 1: propel that forward.

176
00:09:02,640 --> 00:09:05,080
Speaker 2: So they're using all the tools in their toolbox, including

177
00:09:05,200 --> 00:09:10,000
Speaker 2: investing in each other. Yes, yes, coming up. What happens

178
00:09:10,040 --> 00:09:23,360
Speaker 2: if the AI boom becomes the AI bust? Frenzied investments

179
00:09:23,400 --> 00:09:28,240
Speaker 2: in new startups, rapid growth, frothy valuations. Bloomberg Tech reporter

180
00:09:28,320 --> 00:09:31,760
Speaker 2: Emily Foregash says, what's happening in the AI space right

181
00:09:31,800 --> 00:09:35,720
Speaker 2: now is reminding some people of another by gone tech era.

182
00:09:36,040 --> 00:09:38,960
Speaker 1: I think a lot of the industry analysts that were

183
00:09:39,000 --> 00:09:42,640
Speaker 1: around during the dot com boom especially, are seeing parallels

184
00:09:42,720 --> 00:09:44,520
Speaker 1: that are concerning.

185
00:09:44,360 --> 00:09:47,600
Speaker 2: The dot com boom, that moment in the late nineties

186
00:09:47,640 --> 00:09:51,439
Speaker 2: when everyone was eager to invest in anything Internet related.

187
00:09:51,320 --> 00:09:54,319
Speaker 1: And people just poured in money, so much money into

188
00:09:54,360 --> 00:09:57,400
Speaker 1: these companies, any company that was kind of involved in

189
00:09:57,640 --> 00:10:00,200
Speaker 1: developing the Internet at that point in time, and so

190
00:10:00,679 --> 00:10:03,120
Speaker 1: when the return on investment just wasn't there, it all

191
00:10:03,160 --> 00:10:05,440
Speaker 1: came crashing down. And that's kind of the worry that

192
00:10:05,480 --> 00:10:07,800
Speaker 1: we have here. The difference here, though, is some people

193
00:10:07,800 --> 00:10:10,959
Speaker 1: would argue, like, there are actual products here. They're being

194
00:10:11,000 --> 00:10:15,360
Speaker 1: adopted by so many enterprises and businesses and just consumers

195
00:10:15,360 --> 00:10:18,079
Speaker 1: in their everyday life that the bet that people are

196
00:10:18,080 --> 00:10:20,880
Speaker 1: making on AI might have a bit of a stronger foundation.

197
00:10:21,480 --> 00:10:23,600
Speaker 1: But the question then is, well, where is your profit

198
00:10:23,679 --> 00:10:25,439
Speaker 1: to back up all of these huge investments.

199
00:10:25,800 --> 00:10:28,240
Speaker 2: And the other difference with this moment is that AI

200
00:10:28,320 --> 00:10:31,559
Speaker 2: companies are spending a lot more than dot com startups were.

201
00:10:31,640 --> 00:10:36,120
Speaker 1: Right, absolutely, each chip is tens of thousands of dollars

202
00:10:36,160 --> 00:10:39,800
Speaker 1: each chip, and the compute that the world is using

203
00:10:39,800 --> 00:10:42,000
Speaker 1: at this point in time is not just a few chips,

204
00:10:42,000 --> 00:10:44,560
Speaker 1: It's lots and lots and lots of chips, and so

205
00:10:45,040 --> 00:10:47,560
Speaker 1: this is not a cheap endeavor by any means. Also,

206
00:10:47,840 --> 00:10:51,160
Speaker 1: the data centers are massive. They also coust so much

207
00:10:51,240 --> 00:10:54,200
Speaker 1: money to build and to then fill with the chips.

208
00:10:54,880 --> 00:10:57,720
Speaker 2: That all means that a potential AI bubble could be

209
00:10:57,760 --> 00:11:00,679
Speaker 2: a lot bigger than the dot com bubble, And as

210
00:11:00,720 --> 00:11:03,560
Speaker 2: more money pours into the industry for more corners of

211
00:11:03,600 --> 00:11:07,520
Speaker 2: the economy, the stakes of that bubble bursting get bigger too.

212
00:11:08,040 --> 00:11:11,320
Speaker 1: Blackrock is in advanced talks to buy a data center

213
00:11:11,360 --> 00:11:14,320
Speaker 1: company called Aligned Data Centers with the valuation of forty

214
00:11:14,360 --> 00:11:17,240
Speaker 1: billion dollars. This is kind of a forty billion dollar

215
00:11:17,320 --> 00:11:20,240
Speaker 1: bet that Blackrock is placing on a data center company

216
00:11:20,600 --> 00:11:25,320
Speaker 1: that only has like around six hundred megawatts of live

217
00:11:25,360 --> 00:11:27,520
Speaker 1: capacity at this point. They have a lot that's planned,

218
00:11:27,520 --> 00:11:29,560
Speaker 1: they have a lot in construction. That's all coming, but

219
00:11:29,600 --> 00:11:32,040
Speaker 1: that takes a lot of time. Also, the US government

220
00:11:32,080 --> 00:11:34,720
Speaker 1: is involved to a certain extent. The US took a

221
00:11:34,760 --> 00:11:38,320
Speaker 1: stake in Intel, which is another provider, a ten percent

222
00:11:38,320 --> 00:11:41,880
Speaker 1: stake in Intel using Chips Act funding. Nvidia also now

223
00:11:41,920 --> 00:11:44,360
Speaker 1: invests five billion dollars in Intel, and they're planning to

224
00:11:44,400 --> 00:11:46,679
Speaker 1: code develop chips together. So that's kind of a narrative

225
00:11:46,720 --> 00:11:49,760
Speaker 1: of two rivals becoming partners to kind of boost this

226
00:11:49,840 --> 00:11:53,680
Speaker 1: AI boom. And then also the US is taking a

227
00:11:53,720 --> 00:11:56,840
Speaker 1: fifteen percent of cut of Nvidia and AMD's chip sales

228
00:11:56,840 --> 00:12:00,120
Speaker 1: to China. So just in those three last examples, you

229
00:12:00,120 --> 00:12:02,520
Speaker 1: can see how the US government is like somewhat involved.

230
00:12:02,559 --> 00:12:05,920
Speaker 2: Also, how much of the rest of the economy is

231
00:12:06,080 --> 00:12:09,080
Speaker 2: being propped up by AI right now. If the AI

232
00:12:09,120 --> 00:12:11,240
Speaker 2: bubble bursts, what would it mean for the rest of US.

233
00:12:11,400 --> 00:12:13,520
Speaker 1: Well, the market is propped up by these companies for sure.

234
00:12:13,679 --> 00:12:16,480
Speaker 1: It's propping up the New York Stock Exchange other exchanges,

235
00:12:16,559 --> 00:12:18,679
Speaker 1: and so if those companies were just all of a

236
00:12:18,720 --> 00:12:21,000
Speaker 1: sudden dip even a little bit, you'll see the effects

237
00:12:21,040 --> 00:12:23,199
Speaker 1: of that in the stock market at any time.

238
00:12:23,240 --> 00:12:26,080
Speaker 2: In video dips, that's a huge whole market moving in that.

239
00:12:26,200 --> 00:12:28,520
Speaker 1: Yeah, absolutely, you don't necessarily have to care about AI

240
00:12:29,040 --> 00:12:31,360
Speaker 1: or the future of open ai, or the future of

241
00:12:31,440 --> 00:12:34,800
Speaker 1: Nvidia even to care about this story, because this has

242
00:12:34,840 --> 00:12:38,240
Speaker 1: reached into so many other markets, the financial markets, equity markets,

243
00:12:38,280 --> 00:12:41,520
Speaker 1: debt markets, real estate markets because data centers are real

244
00:12:41,559 --> 00:12:44,280
Speaker 1: places that need to be built. Also, your energy bills

245
00:12:44,320 --> 00:12:46,240
Speaker 1: might be higher if you live next to a data center,

246
00:12:46,559 --> 00:12:50,080
Speaker 1: So this extends far beyond now at this point the

247
00:12:50,120 --> 00:12:50,880
Speaker 1: AI sector.

248
00:12:51,600 --> 00:12:55,120
Speaker 2: If some of this valuation is froth, what metrics can

249
00:12:55,160 --> 00:12:57,360
Speaker 2: we look at to get a true sense of how

250
00:12:57,400 --> 00:13:00,120
Speaker 2: the AI economy is performing right now and how all

251
00:13:00,160 --> 00:13:01,680
Speaker 2: these companies are really doing.

252
00:13:01,880 --> 00:13:04,679
Speaker 1: Open ai is a private company, and it does share

253
00:13:04,880 --> 00:13:08,120
Speaker 1: some information about its financials, but that's kind of where

254
00:13:08,120 --> 00:13:10,560
Speaker 1: it's a little bit tricky, because even its valuation is

255
00:13:10,600 --> 00:13:13,240
Speaker 1: a little bit hard to nail down, because there's valuations

256
00:13:13,320 --> 00:13:16,080
Speaker 1: during funding rounds, and then there's valuations during stock sales.

257
00:13:16,480 --> 00:13:19,200
Speaker 1: A company like in VideA, though public, reports a lot

258
00:13:19,360 --> 00:13:22,680
Speaker 1: about their finances. I think that if in Vidia were

259
00:13:22,720 --> 00:13:25,960
Speaker 1: to post declining revenue for some reason in the coming quarters,

260
00:13:26,160 --> 00:13:28,800
Speaker 1: people might then look back to this circular financing idea

261
00:13:28,840 --> 00:13:31,280
Speaker 1: and be like, is this really paying off in this moment.

262
00:13:31,360 --> 00:13:33,559
Speaker 1: Is this a short term pain that will pay off

263
00:13:33,600 --> 00:13:35,920
Speaker 1: in the future. But these are kind of questions that

264
00:13:36,000 --> 00:13:38,440
Speaker 1: we won't yet know the answers to, and I think

265
00:13:38,480 --> 00:13:41,439
Speaker 1: that's what's a little bit nerve wracking sometimes for investors.

266
00:13:43,040 --> 00:13:46,600
Speaker 2: It's a big question with billions of dollars riding on it,

267
00:13:47,280 --> 00:13:50,720
Speaker 2: will the bet on AI pay off or will it

268
00:13:50,800 --> 00:13:51,760
Speaker 2: break the system.

269
00:13:51,960 --> 00:13:55,560
Speaker 1: Analysts are definitely saying that these big AI companies, their

270
00:13:55,600 --> 00:13:58,679
Speaker 1: capacity for innovation, their capacity to return on investment, they

271
00:13:58,720 --> 00:14:02,559
Speaker 1: could be leading the whole economy towards a huge boom

272
00:14:02,559 --> 00:14:04,600
Speaker 1: in AI, huge boom and money, huge boom and even

273
00:14:04,640 --> 00:14:06,360
Speaker 1: further investments, or could go the other way.

274
00:14:06,679 --> 00:14:11,520
Speaker 2: Yes. Stacey Rasgan with Bernstein Research wrote that Sam Altman,

275
00:14:11,840 --> 00:14:15,320
Speaker 2: the CEO of open Ai, has the power to crash

276
00:14:15,360 --> 00:14:18,000
Speaker 2: the global economy for a decade or take us all

277
00:14:18,080 --> 00:14:20,600
Speaker 2: to the promised land. And right now we don't know

278
00:14:20,640 --> 00:14:21,480
Speaker 2: which is in the cards.

279
00:14:22,000 --> 00:14:24,520
Speaker 1: Everyone's looking towards open Ai at this moment. It's kind

280
00:14:24,520 --> 00:14:26,960
Speaker 1: of like the bell weather for this situation with the

281
00:14:27,000 --> 00:14:30,240
Speaker 1: AI boom. If it continues to innovate and if it

282
00:14:30,320 --> 00:14:33,480
Speaker 1: can turn a profit and become a profitable company, that

283
00:14:33,600 --> 00:14:36,520
Speaker 1: might make all this investment so worth it. But at

284
00:14:36,560 --> 00:14:38,760
Speaker 1: the same time, if it's not able to generate profit

285
00:14:38,920 --> 00:14:42,600
Speaker 1: quickly enough, and if the innovation doesn't continue, then all

286
00:14:42,720 --> 00:14:46,240
Speaker 1: this investment that's kind of circulated around open ai are

287
00:14:46,320 --> 00:14:49,560
Speaker 1: near open ai adjacent to open ai that will all

288
00:14:49,600 --> 00:14:52,480
Speaker 1: see the negative consequences. It's very hard to detect a

289
00:14:52,480 --> 00:14:56,400
Speaker 1: bubble ahead of the bubble, but given that only twenty

290
00:14:56,400 --> 00:14:58,960
Speaker 1: five years ago there was a huge dot com crash,

291
00:14:59,120 --> 00:15:03,000
Speaker 1: people are certainly more aware of how it could ultimately

292
00:15:03,080 --> 00:15:07,160
Speaker 1: come crashing down if the profit returns don't actually manifest.

293
00:15:12,200 --> 00:15:15,240
Speaker 2: This is the Big Take from Bloomberg News. I'm Sarah Holder.

294
00:15:15,360 --> 00:15:18,000
Speaker 2: To get more from The Big Take and unlimited access

295
00:15:18,040 --> 00:15:21,760
Speaker 2: to all of Bloomberg dot com, subscribe today at Bloomberg

296
00:15:21,800 --> 00:15:25,520
Speaker 2: dot com slash podcast offer. If you liked this episode,

297
00:15:25,680 --> 00:15:28,040
Speaker 2: make sure to follow and review The Big Take wherever

298
00:15:28,080 --> 00:15:30,760
Speaker 2: you listen to podcasts. It helps people find the show.

299
00:15:31,680 --> 00:15:33,880
Speaker 2: Thanks for listening. We'll be back tomorrow.