1
00:00:02,759 --> 00:00:10,600
Speaker 1: Bloomberg Audio Studios, podcasts, radio news. You're listening to the

2
00:00:10,640 --> 00:00:14,600
Speaker 1: Bloomberg Intelligence Podcast. Catch us live weekdays at ten am

3
00:00:14,640 --> 00:00:17,880
Speaker 1: Eastern on Apple, Cocklay and Android Auto with the Bloomberg

4
00:00:17,960 --> 00:00:21,080
Speaker 1: Business App. Listen on demand wherever you get your podcasts,

5
00:00:21,400 --> 00:00:23,120
Speaker 1: or watch us live on YouTube.

6
00:00:24,200 --> 00:00:27,520
Speaker 2: Dell reported fewer sales of AI servers, lower profit margins

7
00:00:27,560 --> 00:00:30,360
Speaker 2: and expected that's really taking a toll on shares. They're

8
00:00:30,360 --> 00:00:32,959
Speaker 2: down about nine percent. So you want someone who could

9
00:00:32,960 --> 00:00:35,320
Speaker 2: break it all down for us. Let's bring in Woojin Ho,

10
00:00:35,440 --> 00:00:39,839
Speaker 2: Bloomberg Intelligence senior technology analysts joining us from Princeton. Woojin,

11
00:00:39,960 --> 00:00:42,960
Speaker 2: welcome back to the show. I want to start broader.

12
00:00:43,040 --> 00:00:46,239
Speaker 2: So taking these results into consideration, I mean, is it

13
00:00:46,360 --> 00:00:49,239
Speaker 2: showing that the AI rally has has more room to

14
00:00:49,360 --> 00:00:51,400
Speaker 2: run or not? I mean, what are you hearing?

15
00:00:52,600 --> 00:00:56,440
Speaker 3: And in terms of the AI rally in general, I

16
00:00:56,480 --> 00:00:59,200
Speaker 3: mean we saw in videos numbers and video is actually

17
00:00:59,280 --> 00:01:01,240
Speaker 3: doing quiet well. I mean if you ex out to

18
00:01:01,320 --> 00:01:07,720
Speaker 3: China outlook right, the demand is still robust and this

19
00:01:07,760 --> 00:01:13,080
Speaker 3: actually flows through to Dell. If anything, the AI numbers

20
00:01:13,120 --> 00:01:15,600
Speaker 3: for a server numbers were better than expected in the quarter,

21
00:01:15,880 --> 00:01:18,360
Speaker 3: and they also raised their guidance by another full year

22
00:01:18,360 --> 00:01:20,640
Speaker 3: guidance by another five billion, So instead of fifteen billion

23
00:01:20,640 --> 00:01:23,479
Speaker 3: for the year, they're expecting twenty billion for the year,

24
00:01:23,760 --> 00:01:26,880
Speaker 3: and add more of these sovereign and neocloud deployment start

25
00:01:26,959 --> 00:01:29,920
Speaker 3: ramping up. You know, twenty twenty six AI server revenue

26
00:01:29,959 --> 00:01:33,040
Speaker 3: should be better for Dell given how they're positioned.

27
00:01:33,680 --> 00:01:36,119
Speaker 4: So then can you try to explain why this stock

28
00:01:36,160 --> 00:01:38,680
Speaker 4: is down nine percent? I mean, Dell just said it

29
00:01:38,680 --> 00:01:42,319
Speaker 4: now plans to ship twenty billion dollars of artificial intelligence

30
00:01:42,319 --> 00:01:46,319
Speaker 4: servers in fiscal twenty twenty six, double what it sold

31
00:01:46,400 --> 00:01:48,800
Speaker 4: last year. What is upsetting investors?

32
00:01:49,560 --> 00:01:52,280
Speaker 3: Yeah, So, so there's a couple of things. Let's start

33
00:01:52,360 --> 00:01:55,640
Speaker 3: on the AI server piece first. I've always said and

34
00:01:56,360 --> 00:02:00,320
Speaker 3: Dell is we'll contend to it. Having Aisells is double

35
00:02:00,400 --> 00:02:03,760
Speaker 3: edged sword. What you're getting in sales, you kind of

36
00:02:03,800 --> 00:02:06,040
Speaker 3: lose out on gross margin. So let me give you

37
00:02:06,040 --> 00:02:10,239
Speaker 3: an example. From a gross margin standpoint, traditional servers as

38
00:02:10,280 --> 00:02:14,360
Speaker 3: well as storage combined are roughly in the you know,

39
00:02:14,400 --> 00:02:19,359
Speaker 3: the twenties to thirties range, right. AI servers they're they're

40
00:02:19,639 --> 00:02:23,480
Speaker 3: closer to the mid teens range, right, So from an

41
00:02:23,480 --> 00:02:28,200
Speaker 3: operating margin standpoint, it actually is somewhat earnings deluded to

42
00:02:28,240 --> 00:02:32,040
Speaker 3: some degree if you sell way too much AI servers.

43
00:02:32,240 --> 00:02:35,600
Speaker 3: So you know, the one point two billion AI server

44
00:02:35,760 --> 00:02:39,800
Speaker 3: beat that only equated to two cents in an EPs.

45
00:02:39,840 --> 00:02:42,280
Speaker 3: So what Della has been saying for quite some time now,

46
00:02:42,560 --> 00:02:45,880
Speaker 3: don't focus on operating margin because that's that's what they missed.

47
00:02:46,320 --> 00:02:51,359
Speaker 3: Focus on the operating margin dollars. That's fantastic, But you know,

48
00:02:51,800 --> 00:02:55,000
Speaker 3: the you know, the the incremental five billion for the

49
00:02:55,080 --> 00:02:57,640
Speaker 3: year isn't going to be the big boost the EPs

50
00:02:57,639 --> 00:02:58,880
Speaker 3: that people are hoping for.

51
00:03:00,080 --> 00:03:02,960
Speaker 2: A big concern for investors is at the profitability of

52
00:03:03,000 --> 00:03:05,680
Speaker 2: AI servers. I mean they run on processors from companies

53
00:03:05,680 --> 00:03:07,760
Speaker 2: like Nvideo, A m D right, which are which are

54
00:03:07,840 --> 00:03:10,200
Speaker 2: pretty pricey. I mean, should they be worried about that?

55
00:03:11,200 --> 00:03:13,440
Speaker 3: Well, you know, the hope is is that once we

56
00:03:13,480 --> 00:03:17,760
Speaker 3: start moving away from these hyper scale and neocloud type

57
00:03:17,760 --> 00:03:22,680
Speaker 3: of deals, right because they're selling you know, fifty fifty

58
00:03:22,720 --> 00:03:26,320
Speaker 3: thousand servers plus to each of these each of these companies,

59
00:03:27,400 --> 00:03:29,680
Speaker 3: you know, they're not going to get as much they're

60
00:03:29,840 --> 00:03:32,000
Speaker 3: they're heavily discounted. But if they start selling to the

61
00:03:32,160 --> 00:03:38,680
Speaker 3: enterprise over time, the gross margin profile should get better

62
00:03:38,800 --> 00:03:42,920
Speaker 3: and be more in line with traditional service. Now, that's

63
00:03:42,960 --> 00:03:45,920
Speaker 3: going to take a couple of years to materialize. But

64
00:03:45,960 --> 00:03:47,560
Speaker 3: the more important thing is if we think about it

65
00:03:47,600 --> 00:03:50,520
Speaker 3: from a longer term perspective, as long as Dell is

66
00:03:50,560 --> 00:03:55,640
Speaker 3: ahead of from from a competitive standpoint, they should continue

67
00:03:55,680 --> 00:03:59,360
Speaker 3: to win this AI game, not only on the neocloud side,

68
00:03:59,360 --> 00:04:02,280
Speaker 3: but also on the enterprise side, which should help profitability

69
00:04:02,320 --> 00:04:04,360
Speaker 3: longer term. But you're gonna have to wait.

70
00:04:05,000 --> 00:04:06,760
Speaker 4: And you know, when I did a little more digging

71
00:04:06,840 --> 00:04:11,000
Speaker 4: into this earnings report, I found that Dell spent one

72
00:04:11,000 --> 00:04:15,720
Speaker 4: point three billion dollars on share repurchases and dividends during

73
00:04:15,720 --> 00:04:18,560
Speaker 4: the quarter. That's an enviable position to be in. But

74
00:04:18,600 --> 00:04:21,359
Speaker 4: do you think it's the best use of Dell's money?

75
00:04:22,120 --> 00:04:28,000
Speaker 3: Well, so, I think they've They've they've always been since

76
00:04:28,600 --> 00:04:30,479
Speaker 3: they came back to public and then spun off the

77
00:04:30,560 --> 00:04:35,920
Speaker 3: VMware business. They've been very focused on shareholder returns. Right now,

78
00:04:36,200 --> 00:04:40,040
Speaker 3: if you think about from a valuation standpoint, you know,

79
00:04:40,440 --> 00:04:42,840
Speaker 3: M and A is probably going to be something that's

80
00:04:42,880 --> 00:04:45,320
Speaker 3: going to be tough for them to do. They've already

81
00:04:45,400 --> 00:04:47,880
Speaker 3: learned a lesson from big m and A with a

82
00:04:47,960 --> 00:04:50,200
Speaker 3: VMware deal, and which is one of the reasons why

83
00:04:50,200 --> 00:04:54,960
Speaker 3: they spunted out. So if from a shareholder standpoint, if

84
00:04:55,040 --> 00:04:58,479
Speaker 3: there's nothing that Dell can buy, I'd rather have it

85
00:04:58,560 --> 00:05:02,240
Speaker 3: back in dividends and and buybacks. How is it changing, sure,

86
00:05:02,279 --> 00:05:03,040
Speaker 3: like a day like today?

87
00:05:03,440 --> 00:05:05,880
Speaker 4: Well yeah, oh yeah.

88
00:05:05,920 --> 00:05:09,080
Speaker 2: How is how is Dell shaping up against its competitors,

89
00:05:09,120 --> 00:05:11,800
Speaker 2: Let's say, like an HP or super micro computer.

90
00:05:12,560 --> 00:05:14,320
Speaker 3: Yeah, so so let's let me take it from a

91
00:05:14,720 --> 00:05:18,960
Speaker 3: two standpoints. Right from super Micro, we're thinking about it

92
00:05:18,960 --> 00:05:21,600
Speaker 3: more along the lines from AI servers. Super Micro still

93
00:05:21,680 --> 00:05:24,960
Speaker 3: leads the way in the AI server market. Now, one

94
00:05:24,960 --> 00:05:27,680
Speaker 3: of the things that we can glean from the numbers

95
00:05:27,760 --> 00:05:31,800
Speaker 3: is that Dell has been heavily discounting to win to

96
00:05:31,839 --> 00:05:35,480
Speaker 3: win its market share, and they actually have the financial

97
00:05:35,480 --> 00:05:37,920
Speaker 3: have to do so with a strong balance sheet. Now,

98
00:05:37,920 --> 00:05:41,239
Speaker 3: from an HP standpoint, this is more personal computing side,

99
00:05:41,279 --> 00:05:43,200
Speaker 3: and this is one of the This is another reason

100
00:05:43,240 --> 00:05:46,159
Speaker 3: why the shares may be pulling back a little bit

101
00:05:46,680 --> 00:05:50,800
Speaker 3: their PC growth. That segment only grew one percent, and

102
00:05:51,279 --> 00:05:54,160
Speaker 3: on contrast, HP grew five percent. So there's a lot

103
00:05:54,200 --> 00:05:57,240
Speaker 3: of questions in terms of the competitive nature for Dell

104
00:05:57,400 --> 00:06:00,919
Speaker 3: versus HP on the PC market. Now that being said,

105
00:06:01,040 --> 00:06:03,240
Speaker 3: what we can glean from the guide and it is

106
00:06:03,279 --> 00:06:07,480
Speaker 3: the second half of this calendar year. HP should get

107
00:06:07,480 --> 00:06:09,719
Speaker 3: back on track from back to school spending as well

108
00:06:09,760 --> 00:06:12,960
Speaker 3: as Windows eleven upgrade, So I think they should do

109
00:06:12,960 --> 00:06:15,839
Speaker 3: about five percent growth in the second half of the year.

110
00:06:16,720 --> 00:06:19,080
Speaker 4: In the seconds we have left, can you tell us

111
00:06:19,120 --> 00:06:22,520
Speaker 4: how Dell is stacking up against the competition, namely HP.

112
00:06:23,839 --> 00:06:28,080
Speaker 3: Yeah, so from a PC stamp from HP, you know,

113
00:06:29,279 --> 00:06:31,640
Speaker 3: HP is actually winning the game over the last couple

114
00:06:31,720 --> 00:06:33,880
Speaker 3: of quarters, but it's going to be neck and neck

115
00:06:33,880 --> 00:06:36,479
Speaker 3: over the next couple of years. But from an AI

116
00:06:36,680 --> 00:06:39,240
Speaker 3: and a server standpoint, Dell should be fine.

117
00:06:39,440 --> 00:06:41,279
Speaker 2: And in the last minute we have where does Dell

118
00:06:41,400 --> 00:06:43,599
Speaker 2: go from here? I mean, what should their focus be?

119
00:06:44,680 --> 00:06:46,760
Speaker 3: Oh, I really do think it should be on the

120
00:06:46,800 --> 00:06:50,559
Speaker 3: traditional server side as well as source side with AI,

121
00:06:50,720 --> 00:06:53,000
Speaker 3: because they're all going to be melded into one type

122
00:06:53,000 --> 00:06:55,560
Speaker 3: of business going forward, and the Dell's well positioned their

123
00:06:55,600 --> 00:06:56,279
Speaker 3: longer term.

124
00:06:56,560 --> 00:06:59,040
Speaker 2: Stay with us for more Bloomberg Intelligence coming.

125
00:06:58,880 --> 00:07:04,799
Speaker 1: Up after this, you're listening to the Bloomberg Intelligence podcast.

126
00:07:05,160 --> 00:07:08,080
Speaker 1: Catch us live weekdays at ten am. He's Dene on Apple,

127
00:07:08,120 --> 00:07:11,360
Speaker 1: Cocklay and Android Auto with the Bloomberg Business app, listen

128
00:07:11,440 --> 00:07:14,560
Speaker 1: on demand wherever you get your podcasts, or watch us

129
00:07:14,600 --> 00:07:15,560
Speaker 1: live on YouTube.

130
00:07:16,560 --> 00:07:18,880
Speaker 4: One of the big movers today is Caterpillar. Mover to

131
00:07:18,920 --> 00:07:22,080
Speaker 4: the downside, that is off nearly four percent right now.

132
00:07:22,120 --> 00:07:24,720
Speaker 4: It's warning investors that tariffs are expected to have an

133
00:07:24,720 --> 00:07:27,960
Speaker 4: even greater impact, take an even bigger bite out of

134
00:07:28,000 --> 00:07:30,880
Speaker 4: its bottom line, expected to cost CAT as much as

135
00:07:30,880 --> 00:07:34,600
Speaker 4: one point eight billion dollars this year. Let's dig into

136
00:07:34,600 --> 00:07:37,080
Speaker 4: it a little more with Christopher Ciolino. He is Bloomberg

137
00:07:37,120 --> 00:07:41,760
Speaker 4: Intelligence Senior US Machinery analyst. So what is it about

138
00:07:41,760 --> 00:07:45,920
Speaker 4: this report that's got investors in a tizzy?

139
00:07:46,040 --> 00:07:48,600
Speaker 5: Well, you're right, I mean, tariff's will be a bigger

140
00:07:48,680 --> 00:07:52,840
Speaker 5: cost headwind this year, more than KAT previously expected. And

141
00:07:52,880 --> 00:07:56,800
Speaker 5: the revision really seems to be driven by additional clarifications

142
00:07:56,880 --> 00:08:00,120
Speaker 5: around the Section two thirty two steal in aluminum tariffs

143
00:08:00,320 --> 00:08:03,160
Speaker 5: and some of the reciprocal rates on India as you

144
00:08:03,200 --> 00:08:05,600
Speaker 5: alluded to. You know, KAT thinks these TIFFs could be

145
00:08:05,760 --> 00:08:07,560
Speaker 5: you know, up to a one point eight billion dollar

146
00:08:07,640 --> 00:08:10,760
Speaker 5: impact in twenty twenty five. That's up you know, two

147
00:08:10,760 --> 00:08:13,240
Speaker 5: to three hundred million versus their prior view, so a

148
00:08:13,280 --> 00:08:16,800
Speaker 5: little bit higher. So you know, even though that we're

149
00:08:16,840 --> 00:08:19,240
Speaker 5: looking at these incremental costs this year, we think it's

150
00:08:19,280 --> 00:08:23,800
Speaker 5: pretty easily digestible for CAT and probably represents, you know,

151
00:08:23,880 --> 00:08:27,280
Speaker 5: the peak headwind that we'll likely see. They really haven't

152
00:08:27,280 --> 00:08:31,960
Speaker 5: implemented significant cost mitigation or pricing actions yet. I suspect

153
00:08:32,000 --> 00:08:34,280
Speaker 5: you'll see more of those measured than twenty twenty six,

154
00:08:34,800 --> 00:08:37,000
Speaker 5: which could ultimately offset these impacts.

155
00:08:37,240 --> 00:08:41,040
Speaker 2: So is it steal an aluminum? Is that the main focus?

156
00:08:42,440 --> 00:08:44,680
Speaker 5: Yeah, we think so. So we did get an update

157
00:08:44,720 --> 00:08:47,599
Speaker 5: on the section two thirty two taraffs for steel and

158
00:08:47,679 --> 00:08:50,280
Speaker 5: aluminum that expanded the coverage to you know, I think

159
00:08:50,280 --> 00:08:53,560
Speaker 5: an additional four hundred different products. So you know, we

160
00:08:53,600 --> 00:08:57,160
Speaker 5: had CAT report earlier this month and we had that

161
00:08:57,240 --> 00:09:00,240
Speaker 5: subsequent update, So I think that's probably the big driver

162
00:09:00,360 --> 00:09:03,600
Speaker 5: behind it. Obviously, heavy machinery companies use a lot of

163
00:09:03,640 --> 00:09:06,960
Speaker 5: steel and aluminum. It's their largest input costs, so that

164
00:09:07,080 --> 00:09:09,000
Speaker 5: certainly does move the needle.

165
00:09:09,880 --> 00:09:12,560
Speaker 4: We're talking about headwinds, but what about tailwinds any of

166
00:09:12,559 --> 00:09:13,640
Speaker 4: those for Caterpillar.

167
00:09:14,880 --> 00:09:18,280
Speaker 5: Yeah, you know, despite some of these near term costs headwinds,

168
00:09:18,840 --> 00:09:21,600
Speaker 5: I don't think it really ultimately changes the narrative here.

169
00:09:21,640 --> 00:09:24,120
Speaker 5: You know, we're still quite optimistic that earnings are going

170
00:09:24,200 --> 00:09:27,360
Speaker 5: to bottom this year. There's a number of cyclical and

171
00:09:27,400 --> 00:09:29,880
Speaker 5: secular tailwinds that we think are going to drive higher

172
00:09:29,880 --> 00:09:33,640
Speaker 5: earnings in twenty six and twenty seven. The backlocks is

173
00:09:33,679 --> 00:09:36,679
Speaker 5: sitting at a record high. We've seen very solid order

174
00:09:36,760 --> 00:09:40,319
Speaker 5: trends these past you know, four quarters, and dealer inventories

175
00:09:40,800 --> 00:09:44,680
Speaker 5: remain at you know, pretty healthy levels considering softness in

176
00:09:44,760 --> 00:09:48,280
Speaker 5: the broader economy. So all of these are really positive

177
00:09:48,360 --> 00:09:49,959
Speaker 5: leading indicators for Caterpillar.

178
00:09:50,400 --> 00:09:52,760
Speaker 2: And Chris, you follow this company very closely. So when

179
00:09:52,760 --> 00:09:57,120
Speaker 2: we have an economics flowdown, what segments of Caterpillar's business

180
00:09:57,200 --> 00:09:58,720
Speaker 2: tend to hold up the best.

181
00:10:00,160 --> 00:10:02,520
Speaker 5: So right now, the segment holding up the best would

182
00:10:02,559 --> 00:10:05,480
Speaker 5: be their E and T or energy and transportation business,

183
00:10:06,160 --> 00:10:08,640
Speaker 5: and really I think the big driver behind that is

184
00:10:08,720 --> 00:10:12,760
Speaker 5: power generation, particularly on the data center side. Those are

185
00:10:12,880 --> 00:10:17,319
Speaker 5: much longer cycle products have you know, very extended backlogs

186
00:10:17,320 --> 00:10:21,480
Speaker 5: for many years. Caterpillars adding capacity there, so we've seen

187
00:10:21,520 --> 00:10:24,760
Speaker 5: that business hold up very well. On the flip side,

188
00:10:25,160 --> 00:10:28,839
Speaker 5: the more cyclical construction industry's business is probably the one

189
00:10:29,160 --> 00:10:33,520
Speaker 5: more harder hit, and also their resource industry's business, which

190
00:10:33,640 --> 00:10:36,199
Speaker 5: is a lot of mining equipment that's been a little

191
00:10:36,200 --> 00:10:38,680
Speaker 5: softer as well. But you know, I think, you know,

192
00:10:38,679 --> 00:10:41,120
Speaker 5: if we're looking at a lower rate environment moving forward,

193
00:10:42,280 --> 00:10:44,480
Speaker 5: we're certainly a little bit more optimistic on some of

194
00:10:44,480 --> 00:10:47,640
Speaker 5: the growth prospects and those two businesses moving forward.

195
00:10:48,160 --> 00:10:51,280
Speaker 2: Stay with us for more Bloomberg Intelligence coming up after this.

196
00:10:53,840 --> 00:10:57,520
Speaker 1: You're listening to the Bloomberg Intelligence Podcast. Catch us live

197
00:10:57,600 --> 00:11:00,720
Speaker 1: weekdays at ten am Eastern on apple Cock and Android

198
00:11:00,720 --> 00:11:03,520
Speaker 1: Auto with the Blue Work Business app. Listen on demand

199
00:11:03,640 --> 00:11:07,200
Speaker 1: wherever you get your podcasts, or watch us live on YouTube.

200
00:11:08,120 --> 00:11:11,200
Speaker 4: In Video and AI, it was out with earnings this

201
00:11:11,280 --> 00:11:15,120
Speaker 4: week and it impressed with more proof that that AI

202
00:11:15,200 --> 00:11:17,480
Speaker 4: spending continues to be strong. I want to talk more

203
00:11:17,480 --> 00:11:21,400
Speaker 4: about AI in the real world with Binran. He is

204
00:11:21,480 --> 00:11:24,760
Speaker 4: founder and CEO at the software company sig Tech. He

205
00:11:24,880 --> 00:11:28,400
Speaker 4: joins us from London. Ben, thanks so much for being

206
00:11:28,480 --> 00:11:31,040
Speaker 4: with us. Before we get into Sigtech and what you

207
00:11:31,080 --> 00:11:33,240
Speaker 4: do with AI agents, we'll get into what all of

208
00:11:33,280 --> 00:11:35,800
Speaker 4: that is. I just want your take your thoughts on

209
00:11:36,160 --> 00:11:37,760
Speaker 4: In Video's earnings report this week.

210
00:11:38,280 --> 00:11:42,040
Speaker 6: I think we are seeing unprecedented demand for AI services,

211
00:11:42,080 --> 00:11:46,040
Speaker 6: AI inference. If you look at the you know, not

212
00:11:46,160 --> 00:11:49,960
Speaker 6: only the hyperscaters, purchasing GPUs and building data centers.

213
00:11:50,000 --> 00:11:51,000
Speaker 7: It's keep going up.

214
00:11:51,440 --> 00:11:55,680
Speaker 6: And if you look at the consumptions of tokens by

215
00:11:56,240 --> 00:11:59,880
Speaker 6: large language model providers such as Opening Ai and Google

216
00:12:00,080 --> 00:12:04,880
Speaker 6: and the Anthropic, it's growing at one hundred percent every

217
00:12:04,880 --> 00:12:10,360
Speaker 6: six months. So the adoption, the usage, the consumption is

218
00:12:10,400 --> 00:12:15,280
Speaker 6: all going up the trend. There's no no reason why

219
00:12:15,280 --> 00:12:18,079
Speaker 6: it's going to slow down or certainly all stopping.

220
00:12:19,200 --> 00:12:21,920
Speaker 2: We've heard a lot about different types of AI. Jen

221
00:12:21,960 --> 00:12:24,439
Speaker 2: AI is a huge one. I mean, I know everyone's

222
00:12:24,480 --> 00:12:28,319
Speaker 2: on the CHADGBT or whatever whatever else you have, but

223
00:12:28,520 --> 00:12:31,360
Speaker 2: do generic AI tools They can help out in a

224
00:12:31,400 --> 00:12:34,679
Speaker 2: lot of fields. But what about financial institutions, I mean

225
00:12:35,160 --> 00:12:37,440
Speaker 2: you focus more on on custom AI?

226
00:12:37,520 --> 00:12:38,120
Speaker 6: Is that correct?

227
00:12:39,600 --> 00:12:39,800
Speaker 7: Yeah?

228
00:12:39,800 --> 00:12:43,920
Speaker 6: We focused on building entirely bespoke and custom AI agents

229
00:12:43,920 --> 00:12:47,480
Speaker 6: for customers in capital markets. I think people now one

230
00:12:47,559 --> 00:12:51,480
Speaker 6: hundreds of millions of people have used CHGBT, so we all

231
00:12:51,520 --> 00:12:54,880
Speaker 6: have a good idea of you know, what a chat

232
00:12:54,960 --> 00:12:58,920
Speaker 6: bolt can do for us, which is extremely useful. But

233
00:12:58,960 --> 00:13:02,840
Speaker 6: when it comes to business is there's an increasing gap

234
00:13:03,400 --> 00:13:07,439
Speaker 6: between what the large language models are capable of and

235
00:13:07,640 --> 00:13:11,840
Speaker 6: what actually value users in businesses are getting out of

236
00:13:13,000 --> 00:13:16,960
Speaker 6: the getting out of the applications using using the same models.

237
00:13:17,840 --> 00:13:20,920
Speaker 6: You know, the large nugree model capability has been improving

238
00:13:22,120 --> 00:13:30,839
Speaker 6: dramatically and frankly the probably the most improved applications unfortunately

239
00:13:30,920 --> 00:13:34,440
Speaker 6: have been restricted to only a few areas, mainly about

240
00:13:35,120 --> 00:13:40,520
Speaker 6: writing code. That that has been the killer app outside chatboard.

241
00:13:41,120 --> 00:13:44,800
Speaker 6: In finance, the difficulty is there actually there are many

242
00:13:44,800 --> 00:13:48,600
Speaker 6: different reasons why the application is not delivering the kind

243
00:13:48,600 --> 00:13:51,480
Speaker 6: of potential we are expecting. One of the reasons is

244
00:13:52,320 --> 00:13:56,000
Speaker 6: in finance we use a lot of private data that

245
00:13:56,440 --> 00:13:59,160
Speaker 6: the large language models are mainly trained on public data.

246
00:14:00,000 --> 00:14:04,400
Speaker 6: And then the question of how to converting the vast

247
00:14:04,480 --> 00:14:09,600
Speaker 6: corpus of financial documents in PDFs, in world documents in

248
00:14:09,679 --> 00:14:15,040
Speaker 6: excels into the right formats that actually can be effectively

249
00:14:15,200 --> 00:14:20,920
Speaker 6: used by large language models. Any error English data conversion

250
00:14:21,320 --> 00:14:27,120
Speaker 6: document processing pipeline will immediately result in very pro qualities

251
00:14:27,280 --> 00:14:32,200
Speaker 6: downstream and in other areas. In finance, we prontion numbers,

252
00:14:32,640 --> 00:14:35,880
Speaker 6: and the large language models are very bad at cremching

253
00:14:35,960 --> 00:14:39,400
Speaker 6: numbers right and it's probably the most expensive calculator we

254
00:14:39,440 --> 00:14:39,920
Speaker 6: can think of.

255
00:14:40,880 --> 00:14:42,240
Speaker 7: So to make large.

256
00:14:42,160 --> 00:14:45,960
Speaker 6: Nuguerte models working in finance we have to provide a

257
00:14:46,280 --> 00:14:51,080
Speaker 6: very wide range of financial tools. These tools can be software,

258
00:14:51,280 --> 00:14:55,400
Speaker 6: can be APIs, can be desktop applications. Then we have

259
00:14:55,480 --> 00:14:58,880
Speaker 6: to train models to use them, and there are other

260
00:14:59,200 --> 00:15:01,880
Speaker 6: challenges to Yeah, that's why everything has to be bespoke.

261
00:15:02,720 --> 00:15:05,400
Speaker 4: Then we keep hearing that AI is going to be

262
00:15:05,480 --> 00:15:08,600
Speaker 4: a job killer, and we're already seeing it. Frankly, a

263
00:15:08,600 --> 00:15:12,920
Speaker 4: lot of you know, entry level positions being taken over

264
00:15:12,960 --> 00:15:15,920
Speaker 4: by AI. That means a lot of recent graduates from

265
00:15:15,960 --> 00:15:18,320
Speaker 4: college or finding it even harder to find a job

266
00:15:18,360 --> 00:15:22,800
Speaker 4: in unemployment during and amidst that demographics about six percent

267
00:15:22,880 --> 00:15:24,680
Speaker 4: right now versus you know, a little more than four

268
00:15:24,720 --> 00:15:28,000
Speaker 4: percent for the rest of us. So what what how

269
00:15:28,040 --> 00:15:30,120
Speaker 4: do you see the future of work looking and how

270
00:15:30,120 --> 00:15:31,120
Speaker 4: does AI play a part?

271
00:15:33,920 --> 00:15:38,360
Speaker 6: I think whatever, there's you know, unprecedented new technology coming along,

272
00:15:38,440 --> 00:15:39,760
Speaker 6: so transformative and.

273
00:15:39,800 --> 00:15:42,000
Speaker 7: So disruptive like AI.

274
00:15:42,240 --> 00:15:46,240
Speaker 6: But previously there was internet, there was industrial revolution, there

275
00:15:46,360 --> 00:15:50,440
Speaker 6: was the invention of airplanes and cars. I mean, inevitably

276
00:15:50,560 --> 00:15:53,680
Speaker 6: some jobs will be killed, I guess you can say that,

277
00:15:54,080 --> 00:15:57,240
Speaker 6: but there will be new jobs created. I think the

278
00:15:57,320 --> 00:16:01,360
Speaker 6: key here is to understand from them mentally what AI

279
00:16:01,440 --> 00:16:05,520
Speaker 6: can do and what AI cannot do and what makes

280
00:16:05,600 --> 00:16:10,000
Speaker 6: us by us, I mean humans special. AI is very

281
00:16:10,040 --> 00:16:18,320
Speaker 6: good at doing repetitive, non non creative work, non creative

282
00:16:18,560 --> 00:16:24,760
Speaker 6: knowledge work. Humans we are uniquely equipped to be creative

283
00:16:24,920 --> 00:16:28,720
Speaker 6: and imaginative. And also, by the way you you probably

284
00:16:28,760 --> 00:16:31,760
Speaker 6: know that AI is really bad at telling a.

285
00:16:31,760 --> 00:16:34,200
Speaker 7: Joke, a good joke, and there's a reason for.

286
00:16:34,200 --> 00:16:39,960
Speaker 6: It, because AI is trained to predict the next the

287
00:16:40,000 --> 00:16:43,640
Speaker 6: next most likely word, the word by word, and to

288
00:16:43,800 --> 00:16:46,560
Speaker 6: deliver a good joke, and the punchline has to be

289
00:16:46,600 --> 00:16:48,480
Speaker 6: surprising and the witty.

290
00:16:48,640 --> 00:16:50,040
Speaker 7: So it's large number.

291
00:16:50,120 --> 00:16:52,160
Speaker 6: Models are literally not trained to be able to tell

292
00:16:52,200 --> 00:16:58,600
Speaker 6: a good joke. So all the comedians and they have

293
00:16:58,720 --> 00:17:01,480
Speaker 6: right own stuff, and they have to have inspiration, they

294
00:17:01,520 --> 00:17:04,800
Speaker 6: have to have new experiences, imagination and creativity. I think

295
00:17:04,800 --> 00:17:08,440
Speaker 6: that's probably a good room model for the future work.

296
00:17:08,480 --> 00:17:12,400
Speaker 6: Look like we can spend more time on being imaginative

297
00:17:12,440 --> 00:17:17,000
Speaker 6: and creative and frankly more intellectually satisfying, and less time

298
00:17:17,080 --> 00:17:20,080
Speaker 6: on things frankly we don't wake up and be excited about.

299
00:17:20,440 --> 00:17:23,520
Speaker 2: Stay with us for more Bloomberg Intelligence coming up after this.

300
00:17:26,080 --> 00:17:29,760
Speaker 1: You're listening to the Bloomberg Intelligence podcast. Catch us Live

301
00:17:29,840 --> 00:17:32,960
Speaker 1: weekdays at ten am Eastern on Apple, Cocklay and Android

302
00:17:32,960 --> 00:17:36,280
Speaker 1: Auto with the Bloomberg Business app. Listen on demand wherever

303
00:17:36,320 --> 00:17:39,960
Speaker 1: you get your podcasts, or watch us live on YouTube.

304
00:17:40,440 --> 00:17:42,560
Speaker 4: Want to switch gears for a minute and talk about

305
00:17:42,600 --> 00:17:46,840
Speaker 4: the sleepy summer for the housing market would be home buyers.

306
00:17:46,880 --> 00:17:48,680
Speaker 4: We know they've been staying on the sideline, so they're

307
00:17:48,680 --> 00:17:50,800
Speaker 4: waiting for mortgage rates to come down. They're waiting for

308
00:17:51,320 --> 00:17:54,840
Speaker 4: housing prices to come down. So will they And when

309
00:17:55,400 --> 00:17:58,359
Speaker 4: Best Friedman, she's been looking into her crystal ball, we

310
00:17:58,480 --> 00:18:01,720
Speaker 4: called her in for some help. EO at Brown, Harris Stevens,

311
00:18:02,200 --> 00:18:04,239
Speaker 4: that's always good to see you. Thanks so much for

312
00:18:04,359 --> 00:18:07,080
Speaker 4: joining us. So let's just start with a reality check.

313
00:18:07,480 --> 00:18:10,359
Speaker 4: Are things improving? I mean I heard that inventory was

314
00:18:10,400 --> 00:18:12,639
Speaker 4: starting to beef up a little bit, So are you

315
00:18:12,640 --> 00:18:14,360
Speaker 4: seeing some pockets of improvement?

316
00:18:15,359 --> 00:18:17,640
Speaker 8: You know, it's interesting. Good morning, so nice to see

317
00:18:17,640 --> 00:18:19,800
Speaker 8: you as well. Thanks for having me. I think we

318
00:18:19,840 --> 00:18:22,399
Speaker 8: see a step forward and then a step back. And

319
00:18:22,440 --> 00:18:25,800
Speaker 8: I would describe our housing market at least through the summer,

320
00:18:25,840 --> 00:18:30,199
Speaker 8: it's very frustrating people buyers and sellers couldn't get on

321
00:18:30,200 --> 00:18:35,000
Speaker 8: the same page. You know, for buyers things have become unaffordable.

322
00:18:35,080 --> 00:18:37,880
Speaker 8: Rates are too high. Sellers aren't budging on their prices.

323
00:18:37,920 --> 00:18:40,960
Speaker 8: They're taking their homes off the market. So there's been

324
00:18:41,000 --> 00:18:43,480
Speaker 8: a little bit of that, and it's gradual. You know,

325
00:18:43,520 --> 00:18:46,680
Speaker 8: we definitely performed better this year than we did last year.

326
00:18:46,760 --> 00:18:51,640
Speaker 8: It's a better housing market. But having said that, it's

327
00:18:51,760 --> 00:18:53,320
Speaker 8: it's going to take a little bit of time. And

328
00:18:53,359 --> 00:18:56,639
Speaker 8: I know that everybody's expecting or thinking that the FED

329
00:18:56,760 --> 00:19:01,320
Speaker 8: is going to cut the FED fund rate in September.

330
00:19:01,440 --> 00:19:04,200
Speaker 8: Maybe we'll have to see how the jobs report comes out.

331
00:19:04,720 --> 00:19:07,280
Speaker 8: But you know, even that, you know, I don't know

332
00:19:07,320 --> 00:19:08,960
Speaker 8: that's going to have a dramatic impact.

333
00:19:09,000 --> 00:19:11,080
Speaker 7: It's going to take time for us to come.

334
00:19:10,880 --> 00:19:15,120
Speaker 8: To a better place where we see supply and demand intersecting.

335
00:19:15,200 --> 00:19:19,160
Speaker 8: That's a healthy market, and we don't really have that best.

336
00:19:19,200 --> 00:19:21,680
Speaker 2: I have to tell you, I've been kind of caught

337
00:19:21,720 --> 00:19:23,560
Speaker 2: in the middle of all this. I mean, I was

338
00:19:23,560 --> 00:19:25,800
Speaker 2: trying to find a home. I've been out bid by

339
00:19:25,840 --> 00:19:28,280
Speaker 2: like one hundred and seventy thousand dollars in New Jersey.

340
00:19:28,359 --> 00:19:29,680
Speaker 7: I mean, it was rough.

341
00:19:30,520 --> 00:19:32,920
Speaker 2: So can you break it down as far as areas

342
00:19:32,920 --> 00:19:35,439
Speaker 2: of the US right, what areas are are home selling

343
00:19:35,480 --> 00:19:38,000
Speaker 2: at a faster pace or or maybe what areas have

344
00:19:38,160 --> 00:19:39,080
Speaker 2: more inventory.

345
00:19:40,440 --> 00:19:43,360
Speaker 7: I mean, every real estate market is so local. I'm

346
00:19:43,400 --> 00:19:48,040
Speaker 7: sorry that you have that experience. I'm still there are areas.

347
00:19:48,320 --> 00:19:51,320
Speaker 8: I'll give you an example and market that is completely

348
00:19:51,359 --> 00:19:53,960
Speaker 8: on fire. There's no inventory and there's bidding wars. Is

349
00:19:54,080 --> 00:19:58,960
Speaker 8: Upstate New York, Rinebeck. You can't find anything. Whatever comes on,

350
00:19:59,040 --> 00:20:01,040
Speaker 8: there's a bidding war, similar to what happened to you

351
00:20:01,080 --> 00:20:05,000
Speaker 8: in New Jersey. Places like Montclair, New Jersey, very popular,

352
00:20:05,119 --> 00:20:10,119
Speaker 8: a really strong market, and you know, people get discouraged

353
00:20:10,119 --> 00:20:12,480
Speaker 8: when they walk in they see twenty people there waiting,

354
00:20:12,520 --> 00:20:14,480
Speaker 8: and then they know there's going to be multiple bids

355
00:20:14,520 --> 00:20:17,280
Speaker 8: and rates are higher. Price is now going to be higher,

356
00:20:17,320 --> 00:20:19,880
Speaker 8: and so you know, that's very tough. But that's more

357
00:20:19,920 --> 00:20:23,560
Speaker 8: the exception to the rule. Palm Beach is similar, but

358
00:20:23,680 --> 00:20:26,439
Speaker 8: other places like New York is a pretty steady market.

359
00:20:26,440 --> 00:20:29,840
Speaker 8: In New York City right now, the market has been decent,

360
00:20:30,040 --> 00:20:33,919
Speaker 8: not great. The rental market really strong. That's all I

361
00:20:33,960 --> 00:20:36,600
Speaker 8: hear about is people cannot afford to rent in New York.

362
00:20:36,880 --> 00:20:40,960
Speaker 8: There's no vacancies. So it really depends on what's going

363
00:20:41,000 --> 00:20:44,439
Speaker 8: on with supply and demand in the market. Like in Connecticut,

364
00:20:44,480 --> 00:20:48,640
Speaker 8: for example, also really strong demand and lack of supply,

365
00:20:49,320 --> 00:20:51,639
Speaker 8: and so that's starting.

366
00:20:51,680 --> 00:20:54,040
Speaker 7: We're starting to see more inventory come on and so

367
00:20:54,119 --> 00:20:55,000
Speaker 7: that's a good thing.

368
00:20:55,200 --> 00:20:57,480
Speaker 8: And so hopefully if we get a little bit of

369
00:20:57,760 --> 00:20:59,520
Speaker 8: if rates come down a little bit they have come

370
00:20:59,560 --> 00:21:02,600
Speaker 8: down a little bit, that might inspire more buyers to

371
00:21:02,640 --> 00:21:05,800
Speaker 8: get into the market. But it's gonna it's tough. I'm

372
00:21:05,840 --> 00:21:08,920
Speaker 8: sorry that you had that experience. You know, in due time,

373
00:21:09,000 --> 00:21:10,960
Speaker 8: something will come on that you love and you will

374
00:21:11,000 --> 00:21:12,480
Speaker 8: find it and you will bid on it.

375
00:21:13,240 --> 00:21:13,520
Speaker 7: Bess.

376
00:21:13,680 --> 00:21:16,720
Speaker 4: I want your thoughts on a trend we're continuing to

377
00:21:16,760 --> 00:21:20,320
Speaker 4: see grow, and that is corporations backed by private equity

378
00:21:20,359 --> 00:21:24,520
Speaker 4: groups now buying up single family homes. So folks are

379
00:21:24,560 --> 00:21:27,840
Speaker 4: not only competing against other would be home buyers, they're

380
00:21:27,880 --> 00:21:30,680
Speaker 4: competing against the big guns, these institutional investors. I saw

381
00:21:30,720 --> 00:21:34,280
Speaker 4: this one report from MetLife Investment Management. Get this, They

382
00:21:34,280 --> 00:21:38,359
Speaker 4: say investors, institutional investors may control forty percent of the

383
00:21:38,440 --> 00:21:41,879
Speaker 4: single family rental home market by twenty thirty. How's the

384
00:21:41,880 --> 00:21:44,320
Speaker 4: little guys supposed to compete with that?

385
00:21:44,320 --> 00:21:45,840
Speaker 7: That's really hard. I don't know.

386
00:21:45,880 --> 00:21:48,800
Speaker 8: And we are and we have housing shortage in the

387
00:21:48,880 --> 00:21:49,720
Speaker 8: United States.

388
00:21:49,760 --> 00:21:51,920
Speaker 7: You know, we're we need to build.

389
00:21:52,000 --> 00:21:54,879
Speaker 8: We need four million homes here, So I don't know,

390
00:21:54,960 --> 00:21:57,879
Speaker 8: that's not a good sign. Because people find that they

391
00:21:57,920 --> 00:22:01,320
Speaker 8: can't afford rent, they can't afford to buy, and so

392
00:22:01,359 --> 00:22:03,720
Speaker 8: we need to figure out ways to work together with

393
00:22:03,800 --> 00:22:07,760
Speaker 8: our state and local officials to make rules and legislation

394
00:22:07,840 --> 00:22:10,240
Speaker 8: that makes sense so that people can afford to buy

395
00:22:10,320 --> 00:22:11,600
Speaker 8: and can afford to rent.

396
00:22:12,040 --> 00:22:14,560
Speaker 7: And so yeah, the little guy matters.

397
00:22:14,600 --> 00:22:16,880
Speaker 8: We have to make sure that everyday people can afford

398
00:22:16,920 --> 00:22:20,000
Speaker 8: to pursue their dream, which is to buy a home

399
00:22:20,119 --> 00:22:22,600
Speaker 8: or to rent the home, whatever it is. It's really

400
00:22:22,600 --> 00:22:25,119
Speaker 8: important and we've lost a little bit of that in

401
00:22:25,160 --> 00:22:28,080
Speaker 8: the last five years, and I would love to see

402
00:22:28,119 --> 00:22:32,280
Speaker 8: that switch direction. I'm happy to be part of those discussions.

403
00:22:31,920 --> 00:22:33,920
Speaker 2: And best in the last minute we have here. How

404
00:22:33,960 --> 00:22:36,320
Speaker 2: are the builders doing? I mean, how is the sentiment

405
00:22:36,359 --> 00:22:38,680
Speaker 2: among the builders or building these homes?

406
00:22:39,800 --> 00:22:43,320
Speaker 8: You know, it's a great question and it depends, you know.

407
00:22:43,400 --> 00:22:45,760
Speaker 8: I'll give you a real example. A few months ago,

408
00:22:45,800 --> 00:22:48,840
Speaker 8: as working friend of mine, builder developer New York City,

409
00:22:48,960 --> 00:22:51,840
Speaker 8: was looking at a site, very excited. He wanted to

410
00:22:51,840 --> 00:22:54,520
Speaker 8: buy the site to develop it. And then the primary

411
00:22:54,600 --> 00:22:58,360
Speaker 8: came out and it looks like and whether your politics

412
00:22:58,400 --> 00:23:01,200
Speaker 8: are I don't care, but it looks so potentially Mom

413
00:23:01,280 --> 00:23:02,640
Speaker 8: Donnie might win in New York.

414
00:23:02,720 --> 00:23:03,840
Speaker 7: It's a possibility.

415
00:23:04,119 --> 00:23:06,560
Speaker 8: And so he got squee mission backed away from it,

416
00:23:07,000 --> 00:23:10,560
Speaker 8: and so but another developer came in and ended up

417
00:23:10,560 --> 00:23:11,359
Speaker 8: buying the site.

418
00:23:11,400 --> 00:23:12,600
Speaker 7: It just really depends.

419
00:23:12,680 --> 00:23:15,600
Speaker 8: People are very emotional when it comes to these things,

420
00:23:15,600 --> 00:23:17,439
Speaker 8: and it depends how they feel. If they think the

421
00:23:17,480 --> 00:23:20,160
Speaker 8: market's going to be good or strong, they'll invest.

422
00:23:20,640 --> 00:23:23,280
Speaker 7: And builders are the same. It's pencils down. If they think.

423
00:23:23,240 --> 00:23:27,280
Speaker 8: Legislation is sloppy and not helpful and not giving them incentives,

424
00:23:27,280 --> 00:23:28,800
Speaker 8: they're not going to continue to build.

425
00:23:28,960 --> 00:23:32,080
Speaker 7: So it's such a nuanced story.

426
00:23:32,280 --> 00:23:34,840
Speaker 8: So we're seeing some builders do things, some stay on

427
00:23:34,880 --> 00:23:36,080
Speaker 8: the sidelines right now.

428
00:23:36,400 --> 00:23:41,080
Speaker 1: This is the Bloomberg Intelligence Podcast, available on Apple, Spotify,

429
00:23:41,280 --> 00:23:44,760
Speaker 1: and anywhere else you get your podcasts. Listen live each

430
00:23:44,760 --> 00:23:48,480
Speaker 1: weekday ten am to noon Eastern on Bloomberg dot Com,

431
00:23:48,640 --> 00:23:52,200
Speaker 1: the iHeartRadio app, tune In, and the Bloomberg Business App.

432
00:23:52,600 --> 00:23:55,480
Speaker 1: You can also watch us live every weekday on YouTube

433
00:23:55,920 --> 00:23:58,160
Speaker 1: and always on the Bloomberg terminal.

434
00:24:01,800 --> 00:24:02,000
Speaker 7: Yeah