1
00:00:03,040 --> 00:00:14,360
Speaker 1: Bloomberg Audio Studios, Podcasts, radio News.

2
00:00:18,040 --> 00:00:21,520
Speaker 2: Hello and welcome to another episode of the Odd Lots podcast.

3
00:00:21,600 --> 00:00:24,639
Speaker 3: I'm Joe Wasnal and I'm Tracy Alloway. Tracy.

4
00:00:24,680 --> 00:00:29,120
Speaker 2: We're recording this February eleventh, and IGV the software ETF

5
00:00:29,160 --> 00:00:31,280
Speaker 2: down another three percent today.

6
00:00:31,600 --> 00:00:35,720
Speaker 4: It has been ugly in software. Everyone's throwing around the

7
00:00:35,760 --> 00:00:38,560
Speaker 4: term SaaS apocalypse. I mean, the great thing about SaaS

8
00:00:38,680 --> 00:00:40,160
Speaker 4: is there are a lot of things that like rhyme

9
00:00:40,200 --> 00:00:42,120
Speaker 4: with it, a lot of hominems, so you can you

10
00:00:42,159 --> 00:00:46,240
Speaker 4: can make all those puns. Yeah, exactly, SaaS is trash whatever.

11
00:00:46,600 --> 00:00:50,000
Speaker 4: But I'm looking at the share price of Salesforce, oh yeah,

12
00:00:50,000 --> 00:00:52,800
Speaker 4: in particular because I always think of Salesforce as sort

13
00:00:52,840 --> 00:00:56,720
Speaker 4: of like emblematic poster child. Yeah, post child of like

14
00:00:56,760 --> 00:01:00,520
Speaker 4: a software company that I'm not really sure what they do,

15
00:01:01,080 --> 00:01:02,800
Speaker 4: but yeah, it's it's just ugly.

16
00:01:03,000 --> 00:01:05,160
Speaker 2: It's basically been cut in half, hasn't it. Since it's

17
00:01:05,200 --> 00:01:08,080
Speaker 2: peaked like an early twenty twenty five. Right now it's

18
00:01:08,120 --> 00:01:09,560
Speaker 2: one eighty four eighty.

19
00:01:09,319 --> 00:01:12,039
Speaker 4: Four, and it's all your fault, shoe.

20
00:01:12,040 --> 00:01:12,920
Speaker 3: It's all my fault.

21
00:01:13,000 --> 00:01:16,119
Speaker 2: That's right, because earlier in the year, after we got

22
00:01:16,160 --> 00:01:19,520
Speaker 2: back from Christmas, vacation or Christmas break, you know, around

23
00:01:19,520 --> 00:01:22,920
Speaker 2: that I'd seen everyone playing around with claud code, and

24
00:01:22,959 --> 00:01:25,080
Speaker 2: then I had to do it. We did an episode,

25
00:01:25,120 --> 00:01:27,360
Speaker 2: and so people were like, oh, if Joe Wisittal can

26
00:01:27,600 --> 00:01:30,160
Speaker 2: like figure out claud code, that there must not be

27
00:01:30,200 --> 00:01:33,000
Speaker 2: any value to any of these companies at all.

28
00:01:33,080 --> 00:01:33,360
Speaker 5: There.

29
00:01:33,360 --> 00:01:36,680
Speaker 2: You know, you mentioned Salesforce. That's far from the ugliest one.

30
00:01:36,720 --> 00:01:38,959
Speaker 2: I'm looking at Atlassian, which makes a lot of like

31
00:01:39,000 --> 00:01:43,039
Speaker 2: workforce productivity companies, like some Slack competitors and stuff. That

32
00:01:43,120 --> 00:01:44,880
Speaker 2: was a four hundred and fifty dollars stock back in

33
00:01:44,920 --> 00:01:48,040
Speaker 2: twenty twenty one. That's an eighty six dollars stock. So like, yeah,

34
00:01:48,040 --> 00:01:50,559
Speaker 2: it's ugly, and yeah, as you said, everyone is realizing

35
00:01:50,680 --> 00:01:54,720
Speaker 2: that if any old fool can write software, maybe these companies, yeah,

36
00:01:54,800 --> 00:01:55,720
Speaker 2: I don't have much value.

37
00:01:55,840 --> 00:01:58,080
Speaker 4: I mean, I will just say it's not just software

38
00:01:58,360 --> 00:02:02,080
Speaker 4: right now. So we're seeing sort of rolling series of

39
00:02:02,200 --> 00:02:06,920
Speaker 4: concerns where like every time AI does something or create

40
00:02:07,040 --> 00:02:10,000
Speaker 4: some new product, it hits a particular industry. So on

41
00:02:10,080 --> 00:02:14,240
Speaker 4: Monday it was the insurance industry, insurance brokers, and you

42
00:02:14,240 --> 00:02:17,800
Speaker 4: know today Wednesday, February eleventh, I think it's some of

43
00:02:17,840 --> 00:02:20,120
Speaker 4: the stockbroker firms.

44
00:02:19,840 --> 00:02:21,480
Speaker 2: And yeah, all you have to do is just say

45
00:02:21,520 --> 00:02:24,320
Speaker 2: AI industry and there's a you know, it's really there's

46
00:02:24,320 --> 00:02:26,360
Speaker 2: a lot of anxiety. But there's something that doesn't make

47
00:02:26,400 --> 00:02:28,200
Speaker 2: any sense to me about this or the thing that

48
00:02:28,200 --> 00:02:29,200
Speaker 2: I'm wrapping my head around.

49
00:02:29,240 --> 00:02:30,520
Speaker 3: It's like, sure, any of.

50
00:02:30,520 --> 00:02:34,079
Speaker 2: Us could easily like write some software, but like writing

51
00:02:34,120 --> 00:02:36,160
Speaker 2: software is a cost center for these companies.

52
00:02:36,360 --> 00:02:36,520
Speaker 1: Right.

53
00:02:36,720 --> 00:02:40,440
Speaker 2: If you're salesforce and you can trivially reduce the cost

54
00:02:40,480 --> 00:02:43,600
Speaker 2: of building software, that's also a benefit for you. And

55
00:02:43,639 --> 00:02:45,680
Speaker 2: there's a lot more to a software company than just

56
00:02:45,760 --> 00:02:49,080
Speaker 2: code generation, because there's all kinds of you know, network

57
00:02:49,080 --> 00:02:52,079
Speaker 2: effects and links into this. It's like a software company

58
00:02:52,120 --> 00:02:55,000
Speaker 2: is clearly more than just code, and so the fact

59
00:02:55,040 --> 00:02:57,959
Speaker 2: that maybe code can be generated a lot cheaper does

60
00:02:58,040 --> 00:03:00,400
Speaker 2: not screen to me, like, oh, these companies are worth

61
00:03:00,480 --> 00:03:01,040
Speaker 2: lessing views.

62
00:03:01,360 --> 00:03:04,639
Speaker 4: Sure, but at the same time they've been pricing. Their

63
00:03:04,680 --> 00:03:08,040
Speaker 4: pricing is based on that assumption, right, like that there

64
00:03:08,120 --> 00:03:10,480
Speaker 4: is no competitor for what they're doing, and suddenly you

65
00:03:10,560 --> 00:03:12,000
Speaker 4: might have an in house competitor.

66
00:03:12,880 --> 00:03:15,360
Speaker 2: Absolutely, But you know, it's like network effects and do

67
00:03:15,440 --> 00:03:18,280
Speaker 2: companies want to start like building their own like payroll

68
00:03:18,480 --> 00:03:21,760
Speaker 2: software Anyway, I have a lot of questions about this

69
00:03:21,880 --> 00:03:23,200
Speaker 2: sell off, and to your point.

70
00:03:23,120 --> 00:03:27,000
Speaker 4: No, no, no, this is you doing like penance first,

71
00:03:27,240 --> 00:03:29,720
Speaker 4: for causing causing the sell off.

72
00:03:29,760 --> 00:03:31,640
Speaker 2: All right, let's talk to someone who actually might be

73
00:03:31,639 --> 00:03:35,120
Speaker 2: able to answer some of these questions for us. We're

74
00:03:35,160 --> 00:03:37,240
Speaker 2: gonna be speaking to someone who's been in the software space,

75
00:03:37,240 --> 00:03:39,160
Speaker 2: an investor in the software space for a long time,

76
00:03:39,200 --> 00:03:43,880
Speaker 2: recently put out a great deck really diving into SaaS

77
00:03:43,880 --> 00:03:46,800
Speaker 2: of the SaaS apocalypse and what kinds of companies are

78
00:03:46,840 --> 00:03:49,640
Speaker 2: thriving and what kinds of companies were struggling even before

79
00:03:50,000 --> 00:03:52,840
Speaker 2: everyone started talking about AI code generation and all that.

80
00:03:53,080 --> 00:03:55,160
Speaker 2: We're gonna be speaking with Jared Sleeper. He is a

81
00:03:55,160 --> 00:03:58,320
Speaker 2: partner at Avenue, which does a growth investing private company.

82
00:03:58,440 --> 00:03:59,920
Speaker 3: So, Jared, thanks for coming on off.

83
00:04:00,240 --> 00:04:01,720
Speaker 5: Yeah, my pleasure. Excited to be here.

84
00:04:01,720 --> 00:04:03,440
Speaker 2: Why are we talking to you? Just you know, for

85
00:04:03,480 --> 00:04:06,400
Speaker 2: our listeners. Apparently your first time on a podcast, which

86
00:04:06,440 --> 00:04:07,840
Speaker 2: is crazy. But why are we talking you? Give us

87
00:04:07,840 --> 00:04:08,720
Speaker 2: a little bit about.

88
00:04:08,480 --> 00:04:11,520
Speaker 3: Your background investing in software and understanding the space.

89
00:04:11,600 --> 00:04:14,880
Speaker 6: Yeah, my pleasure. So I think one thing that makes

90
00:04:14,920 --> 00:04:17,400
Speaker 6: me a little bit different in the investor world is

91
00:04:17,400 --> 00:04:21,040
Speaker 6: that I've spent time investing in early stage startups, public companies,

92
00:04:21,080 --> 00:04:22,719
Speaker 6: and everything in between. So I spent a chunk of

93
00:04:22,720 --> 00:04:24,920
Speaker 6: my career at an early stage venture fund in Boston

94
00:04:24,960 --> 00:04:28,039
Speaker 6: called Matrix Partners, working with an og SaaS investor named

95
00:04:28,080 --> 00:04:30,279
Speaker 6: David Scock, and then was also at ko TU where

96
00:04:30,279 --> 00:04:32,800
Speaker 6: I ran public software. And so I've kind of have

97
00:04:32,920 --> 00:04:36,560
Speaker 6: this like experience across a spectrum from ground floor startups

98
00:04:36,560 --> 00:04:38,280
Speaker 6: to looking at the big public companies, which I've done

99
00:04:38,279 --> 00:04:39,080
Speaker 6: for the last ten years.

100
00:04:39,080 --> 00:04:40,480
Speaker 3: Perfect guest, Perfect guests.

101
00:04:40,560 --> 00:04:43,480
Speaker 4: So give us some color on the mood in software

102
00:04:43,560 --> 00:04:46,120
Speaker 4: at the moment. Are people like I don't know, hunkering

103
00:04:46,160 --> 00:04:48,000
Speaker 4: down in their bunkers. How bad is it?

104
00:04:48,279 --> 00:04:48,520
Speaker 5: Yeah?

105
00:04:48,560 --> 00:04:51,800
Speaker 6: I get texted constantly from folks on the byside, just

106
00:04:52,240 --> 00:04:55,760
Speaker 6: you know, retrenching. I can't believe this is happening, can't

107
00:04:55,920 --> 00:04:58,120
Speaker 6: go lower. I keep saying that it's one hundredth time

108
00:04:58,160 --> 00:05:00,600
Speaker 6: I bought the DIP. You use the SaaS apocalypse like

109
00:05:00,680 --> 00:05:05,560
Speaker 6: casastrophe is my science. It's definitely one of those moments.

110
00:05:05,560 --> 00:05:07,239
Speaker 6: And we were talking about this a little bit earlier

111
00:05:07,240 --> 00:05:09,720
Speaker 6: before starting, But one of the things about software that's

112
00:05:09,760 --> 00:05:12,240
Speaker 6: really fascinating is there's very few folks, even on the

113
00:05:12,240 --> 00:05:15,960
Speaker 6: buy side who really understand how software works. It's one

114
00:05:15,960 --> 00:05:19,040
Speaker 6: of those roar shot tests kind of sectors where no

115
00:05:19,040 --> 00:05:21,880
Speaker 6: almost no one's logged into Salesforce and clicked around much

116
00:05:21,920 --> 00:05:24,720
Speaker 6: less than a Salesforce admin and understood the full complexity.

117
00:05:25,080 --> 00:05:27,839
Speaker 6: And so when there's panic, there's not a lot of

118
00:05:27,880 --> 00:05:30,719
Speaker 6: support for the stocks, and people, you know, get scared

119
00:05:30,800 --> 00:05:31,280
Speaker 6: very easily.

120
00:05:31,560 --> 00:05:32,760
Speaker 3: Well, explain what this means.

121
00:05:32,800 --> 00:05:37,279
Speaker 2: So, for example, in a lot of companies, it's like

122
00:05:37,560 --> 00:05:40,360
Speaker 2: you're saying that the people who investor trade these stocks,

123
00:05:40,720 --> 00:05:45,120
Speaker 2: they just know them as financial tables basically, and they

124
00:05:45,120 --> 00:05:47,280
Speaker 2: have some idea of their financials and some idea of

125
00:05:47,279 --> 00:05:49,599
Speaker 2: their customer base, et cetera. But they don't have like

126
00:05:49,640 --> 00:05:53,760
Speaker 2: a great intuition for the product onlike say, you know

127
00:05:54,120 --> 00:05:56,360
Speaker 2: people who use Instagram and therefore might have a feel

128
00:05:56,400 --> 00:05:57,320
Speaker 2: about meta for example.

129
00:05:57,360 --> 00:05:59,440
Speaker 6: Yeah, if you're an investor in Lulu Lemon, you have

130
00:05:59,480 --> 00:06:01,680
Speaker 6: a pretty solid conception of what that business is. You

131
00:06:01,680 --> 00:06:04,360
Speaker 6: can go into the scioga pants exactly. You can buy

132
00:06:04,360 --> 00:06:06,400
Speaker 6: the product to ship it to yourself. If you're an

133
00:06:06,400 --> 00:06:10,040
Speaker 6: investor in Viva, which makes CRM software for pharmaceutical reps,

134
00:06:10,279 --> 00:06:12,760
Speaker 6: I bet you there's almost no investors in Viva who

135
00:06:12,800 --> 00:06:15,600
Speaker 6: have ever been inside the product even once, much less

136
00:06:15,720 --> 00:06:17,480
Speaker 6: used it on a day to day basis and understood

137
00:06:17,480 --> 00:06:17,960
Speaker 6: how it works.

138
00:06:18,720 --> 00:06:21,320
Speaker 4: So I'm going to go way back in time and start,

139
00:06:21,480 --> 00:06:23,760
Speaker 4: I guess the very beginning, But why is it that

140
00:06:23,920 --> 00:06:28,239
Speaker 4: software like this, you know, payment management systems, whatever, Why

141
00:06:28,279 --> 00:06:31,440
Speaker 4: were they historically not developed in house?

142
00:06:31,640 --> 00:06:31,760
Speaker 3: Like?

143
00:06:31,800 --> 00:06:33,920
Speaker 4: How did we get this model where we have these

144
00:06:34,000 --> 00:06:37,680
Speaker 4: huge software companies that are really you know, to date

145
00:06:37,920 --> 00:06:40,320
Speaker 4: have been really integral to a lot of businesses.

146
00:06:40,520 --> 00:06:42,760
Speaker 6: Yeah, it's a great question. You know, back in the

147
00:06:42,880 --> 00:06:44,880
Speaker 6: very early days of software, like back in the seventies

148
00:06:44,960 --> 00:06:46,960
Speaker 6: or eighties, there was a lot done in house, and

149
00:06:46,960 --> 00:06:49,800
Speaker 6: we've seen a very clear mixshift over time towards using

150
00:06:49,839 --> 00:06:53,000
Speaker 6: third party software. And what it comes down to is

151
00:06:53,080 --> 00:06:56,720
Speaker 6: the software was expensive to build and maintain, and there's

152
00:06:56,720 --> 00:07:00,200
Speaker 6: this need for an ecosystem of integrations around it, which

153
00:07:00,200 --> 00:07:02,640
Speaker 6: are also expensive to build and maintain. And so if

154
00:07:02,640 --> 00:07:04,440
Speaker 6: you look at a software company, it can afford to

155
00:07:04,440 --> 00:07:08,880
Speaker 6: have one, two, three thousand engineers plus partnership teams, et cetera,

156
00:07:09,080 --> 00:07:11,960
Speaker 6: all working to build the perfect piece of software for

157
00:07:12,040 --> 00:07:15,320
Speaker 6: a given application, and then what's striking And this will

158
00:07:15,320 --> 00:07:17,240
Speaker 6: come up a lot more in this conversation, is not

159
00:07:17,360 --> 00:07:18,640
Speaker 6: selling it for that much money.

160
00:07:18,840 --> 00:07:19,040
Speaker 5: Right.

161
00:07:19,160 --> 00:07:21,760
Speaker 6: A lot of software companies report a stat which is

162
00:07:21,840 --> 00:07:23,560
Speaker 6: the share of our customers that pay us more than

163
00:07:23,600 --> 00:07:26,239
Speaker 6: one hundred thousand dollars a year, And one hundred thousand

164
00:07:26,280 --> 00:07:28,360
Speaker 6: dollars a year is less than half of the fully

165
00:07:28,400 --> 00:07:31,560
Speaker 6: loaded cost of a software engineer, right. And so the

166
00:07:31,640 --> 00:07:34,520
Speaker 6: software model was build a product that can be applied

167
00:07:34,560 --> 00:07:36,920
Speaker 6: to thousands of customers and it's the same product for

168
00:07:37,000 --> 00:07:39,680
Speaker 6: every customer, and then sell it to them for way

169
00:07:39,760 --> 00:07:42,160
Speaker 6: cheaper than they could ever hope to build it themselves,

170
00:07:42,200 --> 00:07:43,680
Speaker 6: even less in the cost of one employee.

171
00:07:44,080 --> 00:07:46,880
Speaker 2: Okay, I'd love to just talk long term software history

172
00:07:46,920 --> 00:07:49,480
Speaker 2: even before you know, we think a lot about SaaS

173
00:07:49,520 --> 00:07:51,760
Speaker 2: and these startups and stuff like that. But like a

174
00:07:51,800 --> 00:07:54,320
Speaker 2: lot of the big companies that we think of in software,

175
00:07:54,520 --> 00:08:00,440
Speaker 2: especially like pre Salesforce, whether it's like SAP Oracle Microsoft obviously,

176
00:08:00,600 --> 00:08:03,480
Speaker 2: aren't there a bunch of third party companies whose job

177
00:08:03,600 --> 00:08:05,680
Speaker 2: is to just like help install it for you, Ye,

178
00:08:05,840 --> 00:08:08,640
Speaker 2: like an SAP installed, And that'll be a totally separate company.

179
00:08:08,640 --> 00:08:11,840
Speaker 2: Because it's so big and it's so unwieldy and complex

180
00:08:12,040 --> 00:08:14,480
Speaker 2: that you actually you can't just like install it yourself,

181
00:08:14,560 --> 00:08:16,280
Speaker 2: or it has to be customized or whatever.

182
00:08:17,000 --> 00:08:19,120
Speaker 6: And there's two parts to that which I think are important.

183
00:08:19,360 --> 00:08:23,080
Speaker 6: One is the integrations into your existing systems. Right, a

184
00:08:23,120 --> 00:08:26,960
Speaker 6: lot of big old companies have old databases, old applications,

185
00:08:26,960 --> 00:08:29,520
Speaker 6: and it's important for everything to be stitched together. So

186
00:08:29,560 --> 00:08:32,640
Speaker 6: you need software engineers and you know, consultants to go

187
00:08:32,679 --> 00:08:35,200
Speaker 6: in and understand those existing systems and kind of get

188
00:08:35,240 --> 00:08:38,000
Speaker 6: them linked up to the new systems. But the other one,

189
00:08:38,000 --> 00:08:41,960
Speaker 6: which is probably bigger, is just people management and change management.

190
00:08:42,200 --> 00:08:44,720
Speaker 6: You know, any software system is the combination of the

191
00:08:44,760 --> 00:08:48,040
Speaker 6: code and all of the individual users who have learned

192
00:08:48,040 --> 00:08:49,640
Speaker 6: how to use it. If you're trying to change out

193
00:08:49,679 --> 00:08:52,880
Speaker 6: your CRM at a company, that means training every single

194
00:08:52,960 --> 00:08:55,840
Speaker 6: sales rep on how to use the new CRM and

195
00:08:55,880 --> 00:08:58,120
Speaker 6: getting it right. And if they get it wrong, then

196
00:08:58,160 --> 00:09:01,040
Speaker 6: you lose deals that quarter. And so, you know, one

197
00:09:01,080 --> 00:09:03,439
Speaker 6: of the kind of tropes in investing is if you

198
00:09:03,480 --> 00:09:06,400
Speaker 6: see a company that's doing an ERP transition. ERP stands

199
00:09:06,400 --> 00:09:09,360
Speaker 6: for Enterprise Resource planning. It's the kind of core software

200
00:09:09,480 --> 00:09:13,000
Speaker 6: accounting you know, supply chain, et cetera. That company's probably

201
00:09:13,000 --> 00:09:14,719
Speaker 6: going to miss its earnings over the next one or

202
00:09:14,720 --> 00:09:18,080
Speaker 6: two quarters because those transitions are so painful and so yes,

203
00:09:18,120 --> 00:09:21,280
Speaker 6: there's a big consulting complex around it that does its

204
00:09:21,320 --> 00:09:23,520
Speaker 6: best to come in and parachute in the talent that's

205
00:09:23,559 --> 00:09:26,200
Speaker 6: required to make those transitions smooth. And that tells you

206
00:09:26,240 --> 00:09:28,800
Speaker 6: something about what makes software so sticky, or at least

207
00:09:28,800 --> 00:09:29,560
Speaker 6: has historically.

208
00:09:30,400 --> 00:09:32,720
Speaker 4: It's third party agents all the way down, I feel.

209
00:09:32,880 --> 00:09:35,840
Speaker 4: But actually, on this note, so we hear the integration

210
00:09:36,000 --> 00:09:38,719
Speaker 4: point brought up a lot, and I think the very

211
00:09:38,760 --> 00:09:41,000
Speaker 4: first episode we did on claud code we talked a

212
00:09:41,000 --> 00:09:43,240
Speaker 4: little bit about it as well. But like, if you

213
00:09:43,320 --> 00:09:45,800
Speaker 4: have something like claud code where you can just give

214
00:09:45,800 --> 00:09:49,520
Speaker 4: it permissions to make changes to your computer, does some

215
00:09:49,600 --> 00:09:53,760
Speaker 4: of that integration expertise actually start to go away, because

216
00:09:53,840 --> 00:09:57,760
Speaker 4: presumably we are going to get AI. I would assume

217
00:09:57,840 --> 00:10:00,960
Speaker 4: at some point, given the rate that it's developed and improving,

218
00:10:01,440 --> 00:10:03,600
Speaker 4: that we'll be able to do this like plug itself

219
00:10:03,640 --> 00:10:04,720
Speaker 4: into various systems.

220
00:10:04,840 --> 00:10:07,840
Speaker 6: Yeah, one hundred percent. I think the challenge of writing

221
00:10:07,920 --> 00:10:12,560
Speaker 6: the code for the integrations is going away. That's not

222
00:10:12,800 --> 00:10:15,560
Speaker 6: the bulk of the challenge for a majority of integrations.

223
00:10:15,559 --> 00:10:19,480
Speaker 6: It's about really deeply understanding the prior system and how

224
00:10:19,480 --> 00:10:21,880
Speaker 6: it maps to the new system. And the reality is,

225
00:10:21,880 --> 00:10:25,760
Speaker 6: within most organizations that's a human problem. It's hey, this

226
00:10:25,880 --> 00:10:30,640
Speaker 6: column says status two thousand and four, what does that mean? Like?

227
00:10:30,679 --> 00:10:32,720
Speaker 6: How does that map to the new system that we're building?

228
00:10:32,720 --> 00:10:34,199
Speaker 6: So you have to go talk to someone and understand it.

229
00:10:34,280 --> 00:10:37,280
Speaker 6: And so there's certain types of integrations where I think

230
00:10:37,320 --> 00:10:40,040
Speaker 6: they're effectively solved problems now because you can write a

231
00:10:40,120 --> 00:10:42,319
Speaker 6: quick write into chat in to clog code and get

232
00:10:42,360 --> 00:10:44,880
Speaker 6: a perfectly written piece of software to make it happen.

233
00:10:45,120 --> 00:10:47,600
Speaker 6: And then there's others that are just fundamentally human problems

234
00:10:47,640 --> 00:10:49,960
Speaker 6: because the data doesn't exist in digital space.

235
00:10:50,480 --> 00:10:52,720
Speaker 2: Let's talk more about that, because really it is pretty

236
00:10:52,720 --> 00:10:57,280
Speaker 2: extraordinary the degree to which I don't know it's the

237
00:10:57,640 --> 00:10:59,840
Speaker 2: working code. I don't know if it's high quality code,

238
00:11:00,160 --> 00:11:04,240
Speaker 2: but certainly these models can generate working code, and it's

239
00:11:04,280 --> 00:11:06,960
Speaker 2: just it blows my mind whenever I use it. But

240
00:11:07,080 --> 00:11:08,880
Speaker 2: talk to us a little bit more about from the

241
00:11:08,920 --> 00:11:13,040
Speaker 2: perspective of various software vendors, and I'm sure there's a

242
00:11:13,160 --> 00:11:15,520
Speaker 2: range about what they're selling and how much is it

243
00:11:15,600 --> 00:11:18,200
Speaker 2: code versus how much is it other stuff, and which

244
00:11:18,200 --> 00:11:22,400
Speaker 2: ones are more exposed to the pure like code generation ability.

245
00:11:22,520 --> 00:11:24,600
Speaker 6: Yeah, it's a great question, and you're one hundred percent right.

246
00:11:24,640 --> 00:11:27,160
Speaker 6: It's producing working code, and frankly it has been for

247
00:11:27,280 --> 00:11:29,560
Speaker 6: the last year or so. I built my first leveable

248
00:11:29,600 --> 00:11:32,800
Speaker 6: app that was working in production about about a year ago,

249
00:11:33,120 --> 00:11:35,040
Speaker 6: and it's even intensified in the last three months.

250
00:11:35,160 --> 00:11:35,319
Speaker 1: Right.

251
00:11:35,800 --> 00:11:38,960
Speaker 6: I think when people buy software, there's a set of

252
00:11:38,960 --> 00:11:40,839
Speaker 6: things that they're buying. One thing that I think is

253
00:11:40,840 --> 00:11:43,920
Speaker 6: important for everyone understand is that open source software has

254
00:11:43,960 --> 00:11:45,920
Speaker 6: been a thing, and there have been free, open source

255
00:11:46,000 --> 00:11:49,040
Speaker 6: versions of almost any software you could buy for all

256
00:11:49,080 --> 00:11:52,480
Speaker 6: of recorded history. There's actually some companies that are public

257
00:11:52,600 --> 00:11:55,480
Speaker 6: that built their businesses packaging that open source software and

258
00:11:55,520 --> 00:11:58,680
Speaker 6: adding a few custom features and then support on top

259
00:11:58,720 --> 00:12:01,760
Speaker 6: of it. Because of the companies reliant on an open

260
00:12:01,800 --> 00:12:04,959
Speaker 6: source database or a company like Elastic with its Elastic

261
00:12:05,000 --> 00:12:08,079
Speaker 6: Search product, which is an infrastructure tool and it breaks,

262
00:12:08,400 --> 00:12:11,000
Speaker 6: they need someone to call, both for COIA reasons and

263
00:12:11,120 --> 00:12:14,040
Speaker 6: because it can be very complex and technical and they

264
00:12:14,040 --> 00:12:17,600
Speaker 6: need to quickly understand it. And so that has been

265
00:12:18,040 --> 00:12:20,520
Speaker 6: a big part of the story historically, is that need

266
00:12:20,559 --> 00:12:23,319
Speaker 6: to you know, have support. Another thing that you sell

267
00:12:23,360 --> 00:12:25,440
Speaker 6: when you sell as a software vendor is what I

268
00:12:25,520 --> 00:12:29,360
Speaker 6: call herd familiarity, which means everyone on Earth knows how

269
00:12:29,400 --> 00:12:31,800
Speaker 6: to use your software, which just simplifies the training and

270
00:12:31,920 --> 00:12:34,000
Speaker 6: onboarding workflow. I'll give a few examples because I'm sure

271
00:12:34,000 --> 00:12:35,400
Speaker 6: it's a new term for listeners since I.

272
00:12:35,320 --> 00:12:35,800
Speaker 5: Made it up.

273
00:12:36,040 --> 00:12:40,280
Speaker 6: You know, Zoom is a great business. Microsoft has been

274
00:12:40,280 --> 00:12:43,560
Speaker 6: giving away a free version of the product forever in teams.

275
00:12:43,800 --> 00:12:47,560
Speaker 6: Why do people use Zoom Because in certain industries almost

276
00:12:47,559 --> 00:12:50,240
Speaker 6: everyone knows how to use Zoom. They have their Zoom setup,

277
00:12:50,320 --> 00:12:52,679
Speaker 6: they have their virtual background chosen. They're not going to

278
00:12:52,720 --> 00:12:54,839
Speaker 6: fumble around for the first minute or two on the call,

279
00:12:55,120 --> 00:12:57,680
Speaker 6: and that's well worth the twenty dollars a month to

280
00:12:57,720 --> 00:13:00,240
Speaker 6: have a Zoom plan. But that applies to lots of

281
00:13:00,280 --> 00:13:03,640
Speaker 6: other areas as well. So think about Microsoft Excel for example.

282
00:13:04,440 --> 00:13:06,240
Speaker 6: You might be able to use Google Sheets to do

283
00:13:06,320 --> 00:13:08,719
Speaker 6: the same thing, but you really want to retrain every

284
00:13:08,760 --> 00:13:11,520
Speaker 6: person who comes in on the Google Sheets shortcuts versus

285
00:13:11,640 --> 00:13:14,560
Speaker 6: Excel shortcuts. It's not a good use of time, especially

286
00:13:14,559 --> 00:13:17,000
Speaker 6: when the software is already so cheap. And so that's

287
00:13:17,040 --> 00:13:19,560
Speaker 6: another plank in what people are buying when they buy

288
00:13:19,600 --> 00:13:22,160
Speaker 6: software is the standardization and the knowledge that they'll be

289
00:13:22,200 --> 00:13:24,920
Speaker 6: able to hire employees who have that, and then there's

290
00:13:24,960 --> 00:13:27,800
Speaker 6: things like brand again, the kind of ecosystem that comes

291
00:13:27,800 --> 00:13:29,840
Speaker 6: around it, and so it really is more than just

292
00:13:29,960 --> 00:13:30,720
Speaker 6: the raw code.

293
00:13:30,880 --> 00:13:33,440
Speaker 4: We've been joking about this, but the idea of software

294
00:13:33,520 --> 00:13:37,280
Speaker 4: companies value lying in being a scapegoat essentially for when

295
00:13:37,320 --> 00:13:40,640
Speaker 4: things go wrong is kind of funny and dystopian, I

296
00:13:40,679 --> 00:13:41,680
Speaker 4: think in many ways.

297
00:13:42,000 --> 00:13:45,360
Speaker 6: Yeah, I mean, I think you know, it's a real fear,

298
00:13:45,600 --> 00:13:47,679
Speaker 6: right And the way I think about it is, there

299
00:13:47,679 --> 00:13:51,320
Speaker 6: are two arguments against software right now. One is the

300
00:13:51,320 --> 00:13:53,160
Speaker 6: world is going to stay the same, but software just

301
00:13:53,160 --> 00:13:54,960
Speaker 6: going to get a lot cheaper over time now that

302
00:13:54,960 --> 00:13:56,880
Speaker 6: it's cheaper to build. And I think there's no one

303
00:13:56,920 --> 00:13:59,000
Speaker 6: who would argue that it's not gotten dramatically cheaper to

304
00:13:59,000 --> 00:14:01,320
Speaker 6: build for reasons that we laid out in our deck

305
00:14:01,360 --> 00:14:03,160
Speaker 6: and we can talk through it more. We don't buy

306
00:14:03,200 --> 00:14:06,160
Speaker 6: that argument. I don't buy that argument. But the second

307
00:14:06,280 --> 00:14:08,400
Speaker 6: is the world's about to get really weird and the

308
00:14:08,440 --> 00:14:10,840
Speaker 6: way that knowledge work happens is going to change. And

309
00:14:10,920 --> 00:14:13,320
Speaker 6: if we think out three, four or five years, who

310
00:14:13,320 --> 00:14:15,559
Speaker 6: knows if there will even be customer support reps or

311
00:14:15,600 --> 00:14:19,280
Speaker 6: sales reps or software engineers, and I think that's what's

312
00:14:19,360 --> 00:14:21,440
Speaker 6: causing the kind of hit to the share prices lately,

313
00:14:21,520 --> 00:14:23,360
Speaker 6: is this terminal value concern.

314
00:14:23,760 --> 00:14:24,920
Speaker 3: Yeah, it was interesting.

315
00:14:25,000 --> 00:14:28,040
Speaker 2: So one of the companies that's been associated with the uh,

316
00:14:28,120 --> 00:14:29,600
Speaker 2: what did you say casastrophe?

317
00:14:29,840 --> 00:14:31,400
Speaker 3: One of those companies that's been.

318
00:14:31,280 --> 00:14:34,680
Speaker 2: Caught up with this blue owl, the private investing firm

319
00:14:34,720 --> 00:14:37,400
Speaker 2: Private Credit. I read through their conference call and their

320
00:14:37,480 --> 00:14:40,160
Speaker 2: CEO was like, not only do we not see red lights,

321
00:14:40,320 --> 00:14:42,480
Speaker 2: not only do we not even see yellow lights, we

322
00:14:42,520 --> 00:14:44,560
Speaker 2: actually see a lot of green lights, which I think

323
00:14:44,640 --> 00:14:46,080
Speaker 2: is really interesting.

324
00:14:45,920 --> 00:14:48,160
Speaker 3: Because it can fit with this idea of.

325
00:14:48,560 --> 00:14:51,000
Speaker 2: This year could be fine, next year could be fine,

326
00:14:51,080 --> 00:14:52,960
Speaker 2: year after that could be fine, and then the year

327
00:14:52,960 --> 00:14:55,200
Speaker 2: after that could be zero or at least that's the

328
00:14:55,240 --> 00:14:57,480
Speaker 2: anxiety that there's this terminal value talking.

329
00:14:57,800 --> 00:14:58,720
Speaker 4: That's like a cliff rest.

330
00:14:58,760 --> 00:15:01,480
Speaker 5: Yeah, there's this cliff Yeah. I think it's really helpful.

331
00:15:01,680 --> 00:15:03,560
Speaker 6: You know, this is our second iteration of the deck,

332
00:15:03,600 --> 00:15:05,560
Speaker 6: and so we kind of force ourselves to recenter on

333
00:15:05,720 --> 00:15:09,760
Speaker 6: what actually happened since the last deck, right, and there's

334
00:15:09,760 --> 00:15:12,120
Speaker 6: a very clear pattern in software and what happened over

335
00:15:12,160 --> 00:15:15,640
Speaker 6: the last five years, which is the pandemic. People freaked

336
00:15:15,640 --> 00:15:17,480
Speaker 6: out at the beginning, but it was rapidly clear that

337
00:15:17,480 --> 00:15:20,080
Speaker 6: it was an accelerant for SaaS as everyone tried to

338
00:15:20,120 --> 00:15:23,040
Speaker 6: digitize their companies, and so you had a spike in

339
00:15:23,080 --> 00:15:25,680
Speaker 6: the growth rate and net retention of the businesses. It

340
00:15:25,720 --> 00:15:28,520
Speaker 6: peaked at just over forty percent and twenty twenty one

341
00:15:28,560 --> 00:15:31,920
Speaker 6: for the median software companies. That's really nice annualized growth.

342
00:15:32,600 --> 00:15:35,040
Speaker 6: And then there was a hangover and that slowed down,

343
00:15:35,280 --> 00:15:37,880
Speaker 6: and we wrote eighteen months ago that that reflected the

344
00:15:37,880 --> 00:15:41,360
Speaker 6: sector sort of maturing. The adoption had just slowed down

345
00:15:41,400 --> 00:15:43,680
Speaker 6: because most folks had adopted the software that they needed

346
00:15:43,880 --> 00:15:46,280
Speaker 6: under the pressure of the pandemic. And so for the

347
00:15:46,360 --> 00:15:49,200
Speaker 6: last few years after that, we saw this degradation and

348
00:15:49,240 --> 00:15:52,120
Speaker 6: growth rates across the sector. By the beginning of last year,

349
00:15:52,400 --> 00:15:55,040
Speaker 6: the median company was growing eighteen percent instead of forty percent,

350
00:15:55,040 --> 00:15:58,160
Speaker 6: so you saw a pretty significant draw down. What's fascinating

351
00:15:58,200 --> 00:16:00,760
Speaker 6: is that if you look at the actual financial performance

352
00:16:00,800 --> 00:16:03,000
Speaker 6: of the companies in the last year, it's been pretty good.

353
00:16:03,440 --> 00:16:05,720
Speaker 6: That growth rate has held. It was eighteen percent again

354
00:16:05,800 --> 00:16:08,720
Speaker 6: in Q three. Net retention has also been consistent at

355
00:16:08,720 --> 00:16:11,200
Speaker 6: about one hundred and ten percent. That's revenue from existing

356
00:16:11,400 --> 00:16:14,400
Speaker 6: customers over the same revenue from those customers a prior year.

357
00:16:14,640 --> 00:16:16,880
Speaker 6: So there's not a churn issue developing or a lack

358
00:16:16,920 --> 00:16:19,120
Speaker 6: of expansion within the customer base, and a lot of

359
00:16:19,160 --> 00:16:22,480
Speaker 6: the companies are actually accelerating growth or guiding to accelerating growth.

360
00:16:22,520 --> 00:16:24,200
Speaker 6: We have a chart showing the number of those companies

361
00:16:24,240 --> 00:16:27,120
Speaker 6: has increased each quarter of the last three successive quarters.

362
00:16:27,400 --> 00:16:30,120
Speaker 6: And so there's a lot going on right now with

363
00:16:30,160 --> 00:16:32,800
Speaker 6: the terminal value. But it's very hard to argue that

364
00:16:32,840 --> 00:16:35,040
Speaker 6: this is something that's happening today and showing up in

365
00:16:35,080 --> 00:16:37,720
Speaker 6: the numbers. The thing is investors are sharp, right, and

366
00:16:37,720 --> 00:16:39,160
Speaker 6: they're constantly looking at that booking.

367
00:16:39,320 --> 00:16:39,520
Speaker 5: Yeah.

368
00:16:39,520 --> 00:16:41,280
Speaker 6: I mean look at CHEG right, which went down very

369
00:16:41,320 --> 00:16:44,280
Speaker 6: quickly in the aftermath of chat GPT coming out, and

370
00:16:44,320 --> 00:16:45,480
Speaker 6: that was completely correct.

371
00:16:45,840 --> 00:16:46,040
Speaker 5: Right.

372
00:16:46,120 --> 00:16:48,400
Speaker 6: Investors were ahead of that, and of course for the

373
00:16:48,400 --> 00:16:51,240
Speaker 6: first few quarters the management team of CHEG, you know,

374
00:16:51,560 --> 00:16:53,760
Speaker 6: had their heads in the sand, but then it became

375
00:16:53,840 --> 00:16:55,760
Speaker 6: clear that it really was existential to their business.

376
00:16:55,840 --> 00:16:57,040
Speaker 4: That's a fun chart.

377
00:16:58,160 --> 00:17:01,440
Speaker 2: I thought I was looking at a typo because I saw, wow,

378
00:17:01,840 --> 00:17:04,480
Speaker 2: that was a near one hundred dollars stock in February

379
00:17:04,520 --> 00:17:05,280
Speaker 2: twenty twenty one.

380
00:17:05,600 --> 00:17:07,159
Speaker 3: It is now a sixty one cent.

381
00:17:07,000 --> 00:17:09,119
Speaker 6: Stock and you have to give the markets credit. Like

382
00:17:09,160 --> 00:17:11,240
Speaker 6: the second chat of T came out, people were like,

383
00:17:11,320 --> 00:17:13,480
Speaker 6: this company's in big trouble. They didn't wait for it

384
00:17:13,520 --> 00:17:16,199
Speaker 6: to hit the financial results, and so there is signal

385
00:17:16,240 --> 00:17:17,000
Speaker 6: and what people think.

386
00:17:32,800 --> 00:17:35,400
Speaker 4: I have a bunch more questions, but just briefly, where

387
00:17:35,440 --> 00:17:37,879
Speaker 4: does data actually fit into all of this? Because the

388
00:17:37,960 --> 00:17:40,840
Speaker 4: other thing we hear about AI is maybe the models

389
00:17:40,880 --> 00:17:43,719
Speaker 4: don't matter that much, but it's the actual data that

390
00:17:43,800 --> 00:17:47,800
Speaker 4: you have access to. And I imagine the customers themselves

391
00:17:47,920 --> 00:17:50,960
Speaker 4: of SaaS companies, they have their own data. Do the

392
00:17:51,000 --> 00:17:53,000
Speaker 4: SaaS companies have their own data as well? Can they

393
00:17:53,000 --> 00:17:53,600
Speaker 4: build off of that?

394
00:17:53,840 --> 00:17:55,399
Speaker 6: Yeah, it's a great question. And we're here at one

395
00:17:55,440 --> 00:17:58,840
Speaker 6: of the world's biggest data companies, so very apt. Well,

396
00:17:58,880 --> 00:18:03,560
Speaker 6: disclosure data is definitely something that gets more valuable in

397
00:18:03,640 --> 00:18:07,040
Speaker 6: this world. If you think about a stylized AI model,

398
00:18:07,400 --> 00:18:10,040
Speaker 6: it could have PhD level intelligence in a domain. But

399
00:18:10,119 --> 00:18:12,440
Speaker 6: if you hired a PhD in your company and sat

400
00:18:12,480 --> 00:18:15,360
Speaker 6: her down on her first day, she wouldn't be very useful, right.

401
00:18:15,440 --> 00:18:18,160
Speaker 6: She would have to understand how the organization functions, where

402
00:18:18,280 --> 00:18:20,760
Speaker 6: things live. Do I trust this chart or that chart.

403
00:18:20,760 --> 00:18:22,760
Speaker 6: I need access to the Google drive, I need access

404
00:18:22,800 --> 00:18:24,639
Speaker 6: to Slack, I need to spend some time reading up

405
00:18:24,720 --> 00:18:26,879
Speaker 6: and so we call this kind of context.

406
00:18:27,000 --> 00:18:27,200
Speaker 5: Right.

407
00:18:27,240 --> 00:18:29,879
Speaker 6: It's all the extra information that an AI needs to

408
00:18:30,040 --> 00:18:33,680
Speaker 6: get something done, no matter how intelligent it is. And

409
00:18:33,800 --> 00:18:35,520
Speaker 6: we wrote about this in the chart in the deck.

410
00:18:35,560 --> 00:18:38,959
Speaker 6: But there's a real question of who becomes that system

411
00:18:39,000 --> 00:18:41,520
Speaker 6: of context. And you're right, a lot of the software

412
00:18:41,560 --> 00:18:44,320
Speaker 6: companies do sit on a pool of very important data.

413
00:18:44,440 --> 00:18:46,320
Speaker 6: Let's talk about Salesforce for example.

414
00:18:46,400 --> 00:18:48,400
Speaker 5: Right. CRM is where.

415
00:18:48,240 --> 00:18:51,480
Speaker 6: You track the records of every customer. You have every

416
00:18:51,640 --> 00:18:55,160
Speaker 6: prospect in your pipeline, all of your historic interactions with them,

417
00:18:55,320 --> 00:18:58,080
Speaker 6: notes from sales reps on what's going on the status

418
00:18:58,080 --> 00:19:00,919
Speaker 6: of their account, their customer support requests. It's an incredibly

419
00:19:01,000 --> 00:19:04,920
Speaker 6: complex piece of software for a large enterprise, and obviously

420
00:19:04,960 --> 00:19:07,879
Speaker 6: if you are an AI agent working within a company,

421
00:19:07,880 --> 00:19:09,760
Speaker 6: you would need access to that in order to get

422
00:19:09,760 --> 00:19:14,120
Speaker 6: almost anything done. Right, But you need more than is there.

423
00:19:14,600 --> 00:19:17,040
Speaker 6: You don't know what happened at the sales dinner last

424
00:19:17,119 --> 00:19:19,560
Speaker 6: night unless the rep took really detailed notes. And I

425
00:19:19,600 --> 00:19:21,760
Speaker 6: can tell you one comment learning and software is they

426
00:19:21,800 --> 00:19:23,360
Speaker 6: do not take very detailed notes.

427
00:19:23,320 --> 00:19:25,680
Speaker 5: Especially have a sales party, right, yeah, exactly.

428
00:19:25,760 --> 00:19:28,320
Speaker 6: Can people assume that software management teams know exactly what's

429
00:19:28,359 --> 00:19:30,880
Speaker 6: going on, but they're looking through really messy salesforce data

430
00:19:31,040 --> 00:19:32,000
Speaker 6: and doing their very best.

431
00:19:32,359 --> 00:19:35,240
Speaker 4: Now I'm imagining a sales agent being like the Cabernet

432
00:19:35,480 --> 00:19:38,399
Speaker 4: was exquisite at last night's party, just putting in all

433
00:19:38,400 --> 00:19:41,199
Speaker 4: these irrelevant like diary entries, exactly.

434
00:19:41,280 --> 00:19:44,360
Speaker 6: But a lot of that context does live in human brains.

435
00:19:44,440 --> 00:19:46,680
Speaker 6: You know, a sales rep meets a person at dinner,

436
00:19:46,720 --> 00:19:48,880
Speaker 6: gets to know their kids, figures out what sports team

437
00:19:48,880 --> 00:19:51,320
Speaker 6: they root for, and they're not automatically pumping all that

438
00:19:51,359 --> 00:19:54,560
Speaker 6: into the CRM. And so there's this race to collect

439
00:19:54,720 --> 00:19:57,240
Speaker 6: the information that an AI agent would need in order

440
00:19:57,240 --> 00:20:00,919
Speaker 6: to actually take proactive action, and the software companies have

441
00:20:01,000 --> 00:20:03,320
Speaker 6: a position there. But there's also this set of AI

442
00:20:03,400 --> 00:20:07,360
Speaker 6: native startups that are coming in building actual agents who

443
00:20:07,359 --> 00:20:09,480
Speaker 6: are doing their own work to collect that context.

444
00:20:09,480 --> 00:20:10,200
Speaker 5: And that's one of the.

445
00:20:10,200 --> 00:20:13,040
Speaker 6: Battles that we saw, you know, kind of highlighted in

446
00:20:13,080 --> 00:20:15,600
Speaker 6: our deck is whoever wins that has a chance to

447
00:20:15,640 --> 00:20:17,400
Speaker 6: be a really valuable company.

448
00:20:17,720 --> 00:20:19,240
Speaker 2: You know what I think about and I think you

449
00:20:19,359 --> 00:20:21,080
Speaker 2: talk about this in your deck, But when I think

450
00:20:21,119 --> 00:20:25,000
Speaker 2: about software, I sort of have like if there's a spectrum,

451
00:20:25,240 --> 00:20:28,600
Speaker 2: you know, I think about Salesforce dot com, which is

452
00:20:28,640 --> 00:20:31,560
Speaker 2: a platform, and there's third party developers that build on

453
00:20:31,600 --> 00:20:34,240
Speaker 2: top of Salesforce and they sort of offer any everything,

454
00:20:34,520 --> 00:20:37,399
Speaker 2: And then I think about something niche like this is

455
00:20:37,440 --> 00:20:41,679
Speaker 2: the company that makes point of sales software for dentist's office,

456
00:20:42,000 --> 00:20:45,439
Speaker 2: and they went around by giving them free payment terminals,

457
00:20:45,520 --> 00:20:48,520
Speaker 2: and they joined y Combinator, and you know, they signed

458
00:20:48,560 --> 00:20:50,560
Speaker 2: up ten thousand dentist office and then they pay those

459
00:20:50,680 --> 00:20:54,199
Speaker 2: offices pay them ten dollars a month forever for access

460
00:20:54,240 --> 00:20:56,280
Speaker 2: to that. You know what, I'm just making it up,

461
00:20:56,280 --> 00:20:58,600
Speaker 2: but things like that. Is there a side of the

462
00:20:58,640 --> 00:21:02,920
Speaker 2: spectrum that's more at risk here? Is that spectrum legitimate

463
00:21:02,920 --> 00:21:06,320
Speaker 2: way to think about the industry or is there threats

464
00:21:06,359 --> 00:21:07,960
Speaker 2: on sort of wherever you look?

465
00:21:08,480 --> 00:21:10,600
Speaker 6: Yeah, it's a great question. I mean, certainly in the

466
00:21:10,680 --> 00:21:13,680
Speaker 6: world gets really weird scenario, it's not clear there's anywhere

467
00:21:13,680 --> 00:21:16,800
Speaker 6: immune from threats, but it's important to think through what

468
00:21:16,840 --> 00:21:19,399
Speaker 6: it looks like. I think what's most at threat is

469
00:21:19,760 --> 00:21:25,480
Speaker 6: companies that serve enterprises with very customized software already or

470
00:21:25,520 --> 00:21:28,520
Speaker 6: software that takes a very heavy implementation. And the reason

471
00:21:28,640 --> 00:21:32,000
Speaker 6: is if anyone's going to take advantage of this wave

472
00:21:32,000 --> 00:21:35,600
Speaker 6: of technology to really you know, advance and replace a

473
00:21:35,640 --> 00:21:38,400
Speaker 6: core system of software, it's going to be the enterprises

474
00:21:38,440 --> 00:21:41,600
Speaker 6: that have the resources and the customization needs. If you

475
00:21:41,600 --> 00:21:45,199
Speaker 6: think about SMBs. You know, my dad runs our family's

476
00:21:45,240 --> 00:21:47,199
Speaker 6: grocery store in the family for one hundred years, and

477
00:21:47,240 --> 00:21:49,359
Speaker 6: he just changed his point of sale for the first

478
00:21:49,359 --> 00:21:51,879
Speaker 6: time in a few decades, and it was a really

479
00:21:51,960 --> 00:21:52,960
Speaker 6: messy process.

480
00:21:53,080 --> 00:21:55,880
Speaker 5: It took a long time. Your dad, come on all alls, Yeah, sure,

481
00:21:56,160 --> 00:21:57,679
Speaker 5: we love, we love to.

482
00:21:58,480 --> 00:22:01,800
Speaker 6: All about independent grocery. And but you know he's certainly

483
00:22:01,840 --> 00:22:04,320
Speaker 6: not going to sit down and vibe code himself a

484
00:22:04,359 --> 00:22:06,359
Speaker 6: point of sale system and put the store on it.

485
00:22:06,440 --> 00:22:08,359
Speaker 6: I can guarantee you that, nor will any dentist.

486
00:22:08,520 --> 00:22:08,680
Speaker 5: Right.

487
00:22:09,160 --> 00:22:12,760
Speaker 6: There's a chance that someone comes along with a cheaper version,

488
00:22:13,000 --> 00:22:16,520
Speaker 6: but you know, I think that's not something he's going

489
00:22:16,560 --> 00:22:18,440
Speaker 6: to switch to anytime soon. He's going to go through

490
00:22:18,440 --> 00:22:21,359
Speaker 6: that pain for another few decades to come right. And

491
00:22:21,440 --> 00:22:24,800
Speaker 6: so it really is, you know, kind of company by company,

492
00:22:24,840 --> 00:22:27,560
Speaker 6: Like I'm doing this exercise right now on X where

493
00:22:27,600 --> 00:22:29,480
Speaker 6: every day I look at a different software company and

494
00:22:29,560 --> 00:22:32,199
Speaker 6: just think hard about what will AI look like for

495
00:22:32,240 --> 00:22:34,720
Speaker 6: this company, And it's really interesting when you press I'll

496
00:22:34,720 --> 00:22:37,680
Speaker 6: give an example like DocuSign, which I think to most

497
00:22:37,720 --> 00:22:42,080
Speaker 6: investors would seem like an incredibly simple, easy piece of software. Right,

498
00:22:42,160 --> 00:22:45,360
Speaker 6: it's an eating software. We've all experienced it. DocuSign has

499
00:22:45,400 --> 00:22:49,760
Speaker 6: more employees today than open ai and andthrop it combined god,

500
00:22:50,080 --> 00:22:53,800
Speaker 6: which is a crazy stat and probably reflects that labor

501
00:22:53,880 --> 00:22:57,200
Speaker 6: is inefficiently allocated across the market. But when you actually

502
00:22:57,240 --> 00:23:00,200
Speaker 6: double click into what DocuSign does, there are ways in

503
00:23:00,240 --> 00:23:04,520
Speaker 6: which it's very complicated. Right understanding the signature regulations in

504
00:23:04,560 --> 00:23:06,720
Speaker 6: every country around the world, what does it take for

505
00:23:06,760 --> 00:23:09,399
Speaker 6: signature to be legally valid. Most of its signatures are

506
00:23:09,440 --> 00:23:11,840
Speaker 6: done as an API, so folks are integrating it into

507
00:23:11,880 --> 00:23:15,280
Speaker 6: their own applications. And there's a benefit to using docu

508
00:23:15,320 --> 00:23:17,400
Speaker 6: sign which is the brand people have been giving away

509
00:23:17,400 --> 00:23:20,159
Speaker 6: free e signature software for a very long time. But

510
00:23:20,200 --> 00:23:22,520
Speaker 6: if you're a company of a certain esteem, you want

511
00:23:22,560 --> 00:23:25,240
Speaker 6: to make sure your customers trust what they're signing, and

512
00:23:25,240 --> 00:23:27,360
Speaker 6: if they're getting a contract from you, you'd much rather

513
00:23:27,400 --> 00:23:31,000
Speaker 6: say DocuSign than xyz sign that someone vibe coded, right,

514
00:23:31,320 --> 00:23:34,040
Speaker 6: And so I think it's really important to look a

515
00:23:34,080 --> 00:23:38,000
Speaker 6: company by company. It's definitely a stock pickers market where

516
00:23:38,160 --> 00:23:41,119
Speaker 6: there's some that are either relatively immune or have a

517
00:23:41,200 --> 00:23:42,960
Speaker 6: chance to benefit, and tho's others that could be in

518
00:23:42,960 --> 00:23:43,480
Speaker 6: real trouble.

519
00:23:44,160 --> 00:23:47,159
Speaker 4: So is the argument the bold case for software, or

520
00:23:47,200 --> 00:23:50,679
Speaker 4: at least the non sudden death case for software. This

521
00:23:50,800 --> 00:23:53,560
Speaker 4: idea that like, Okay, if you have a software company

522
00:23:53,600 --> 00:23:57,000
Speaker 4: that's producing I don't know, like DocuSign, you're able to

523
00:23:57,240 --> 00:24:00,159
Speaker 4: sign documents digitally and track them and share them and

524
00:24:00,200 --> 00:24:03,840
Speaker 4: all of that. You can build more quickly and more

525
00:24:03,920 --> 00:24:09,159
Speaker 4: efficiently off of that base model and provide new versions,

526
00:24:09,320 --> 00:24:14,680
Speaker 4: new customizations for customers. So I could do DocuSign for dentists,

527
00:24:14,760 --> 00:24:16,600
Speaker 4: just to stick with that example. I don't know what

528
00:24:16,640 --> 00:24:19,880
Speaker 4: specific needs dentists would have. I don't know, maybe marking

529
00:24:20,000 --> 00:24:22,520
Speaker 4: up like teeth or something. Yeah, and then I can

530
00:24:22,600 --> 00:24:26,840
Speaker 4: do like DocuSign for doctors and DocuSign for sales agents

531
00:24:26,920 --> 00:24:28,040
Speaker 4: or whatever and just keep going.

532
00:24:28,359 --> 00:24:28,560
Speaker 5: Yeah.

533
00:24:28,600 --> 00:24:30,399
Speaker 6: I think that's right. I actually kind of think of

534
00:24:30,400 --> 00:24:33,600
Speaker 6: it as there's three cases. There's the software gets wiped

535
00:24:33,640 --> 00:24:37,520
Speaker 6: out case, there's the not much happens to software case,

536
00:24:37,680 --> 00:24:39,920
Speaker 6: and then there's the bual case where the software companies

537
00:24:39,960 --> 00:24:42,040
Speaker 6: capture a lot of value. I think it's a little

538
00:24:42,080 --> 00:24:45,040
Speaker 6: different than them adding a lot of features and functionality. Frankly,

539
00:24:45,040 --> 00:24:47,399
Speaker 6: I think a lot of software products today are pretty mature.

540
00:24:47,840 --> 00:24:50,439
Speaker 6: There than thousand engineers working on them for ten years,

541
00:24:50,560 --> 00:24:52,800
Speaker 6: and they've built not all, but most of the things

542
00:24:52,840 --> 00:24:56,000
Speaker 6: that you'd want to build with today's technology. But with agents,

543
00:24:56,119 --> 00:24:59,160
Speaker 6: there's ways to automate a big chunk of the work.

544
00:24:59,320 --> 00:25:02,480
Speaker 6: So software company that's done this very well is Intercom.

545
00:25:02,720 --> 00:25:05,960
Speaker 6: Intercom sells customer support software. It's those little widgets on

546
00:25:06,000 --> 00:25:07,800
Speaker 6: the bottom right hand corner of websites. They were the

547
00:25:07,840 --> 00:25:10,280
Speaker 6: creators of that. They had a nice business, but then

548
00:25:10,320 --> 00:25:13,400
Speaker 6: they got very aggressive about building out an AI product

549
00:25:13,440 --> 00:25:16,560
Speaker 6: called Finn, which answers customer support queries on its own.

550
00:25:17,080 --> 00:25:19,480
Speaker 6: And they've i think they've mentioned that it's almost one

551
00:25:19,560 --> 00:25:21,800
Speaker 6: hundred million of ARR now on a base that was

552
00:25:22,000 --> 00:25:24,640
Speaker 6: you know, like three hundred million of ARR or something

553
00:25:24,680 --> 00:25:27,840
Speaker 6: like that. And so they've really reaccelerated their business by

554
00:25:27,880 --> 00:25:30,719
Speaker 6: building an AI native tool that actually solves the work,

555
00:25:31,040 --> 00:25:33,960
Speaker 6: not just a tool that not just kind of exists

556
00:25:33,960 --> 00:25:37,439
Speaker 6: as a tool that humans use. And so yeah, I

557
00:25:37,440 --> 00:25:39,800
Speaker 6: think that's like the mega bowlcase, right. I think about

558
00:25:39,800 --> 00:25:43,080
Speaker 6: it like almost a transition from brick and mortar retail

559
00:25:43,119 --> 00:25:45,600
Speaker 6: to e commerce, where you have a brand new way

560
00:25:45,800 --> 00:25:48,200
Speaker 6: of doing business and you have a bunch of legacy

561
00:25:48,240 --> 00:25:52,000
Speaker 6: companies and some of them will probably just exist as

562
00:25:52,040 --> 00:25:56,560
Speaker 6: they always have. Others can benefit from the change and

563
00:25:56,600 --> 00:25:58,840
Speaker 6: add new business lines. You look at Walmart's share price,

564
00:25:58,920 --> 00:26:01,639
Speaker 6: it's done amazingly well at incorporateing e commerce into its business.

565
00:26:02,080 --> 00:26:03,560
Speaker 6: And then there's going to be some that are like

566
00:26:03,640 --> 00:26:04,840
Speaker 6: Seers and go away.

567
00:26:05,160 --> 00:26:07,800
Speaker 4: That's funny. Sears always reminds me of my dad loves

568
00:26:07,800 --> 00:26:10,359
Speaker 4: Sears because he always said the parking lot was empty

569
00:26:10,760 --> 00:26:12,879
Speaker 4: when he goes to the shopping mall, so he always

570
00:26:12,920 --> 00:26:16,320
Speaker 4: went through Sears anyway. So I understand like the cost

571
00:26:16,520 --> 00:26:19,080
Speaker 4: argument and brings down the cost of code. Maybe you

572
00:26:19,160 --> 00:26:22,680
Speaker 4: have fewer employees or whatever, But where does growth actually

573
00:26:22,720 --> 00:26:25,640
Speaker 4: come from in that world? How are you expanding your

574
00:26:25,680 --> 00:26:26,440
Speaker 4: customer base?

575
00:26:26,800 --> 00:26:29,520
Speaker 6: Yeah, you're really going to them and saying we are

576
00:26:29,760 --> 00:26:34,520
Speaker 6: replacing human labor. And there's a different pricing paradigm.

577
00:26:34,600 --> 00:26:34,800
Speaker 7: Now.

578
00:26:34,920 --> 00:26:37,160
Speaker 6: You used to think of us as something you paid,

579
00:26:37,440 --> 00:26:40,080
Speaker 6: you know, twenty thirty forty fifty dollars per seat, per

580
00:26:40,080 --> 00:26:43,280
Speaker 6: month for as a tool for your employees, almost as

581
00:26:43,280 --> 00:26:45,680
Speaker 6: if you know your employees are artisans and they're getting

582
00:26:45,680 --> 00:26:48,439
Speaker 6: a toolkit to work with. And now we're just selling

583
00:26:48,440 --> 00:26:51,960
Speaker 6: you an employee or the results of an employee. So

584
00:26:52,520 --> 00:26:55,520
Speaker 6: you know, we will sell you customer support tickets getting

585
00:26:55,520 --> 00:26:59,280
Speaker 6: closed out for fifty cents or a dollar per ticket,

586
00:26:59,400 --> 00:27:01,360
Speaker 6: and you can do math of what it would cost

587
00:27:01,400 --> 00:27:02,840
Speaker 6: you for the human to do that, or what it

588
00:27:02,880 --> 00:27:05,359
Speaker 6: would cost you for AI to do that, and we'll

589
00:27:05,400 --> 00:27:08,120
Speaker 6: be cheaper. But we're also dramatically increasing what you pay

590
00:27:08,200 --> 00:27:11,320
Speaker 6: us because you know, we're cutting into a completely different stream.

591
00:27:11,359 --> 00:27:13,240
Speaker 6: And so that's what I think it looks like. We

592
00:27:13,240 --> 00:27:16,119
Speaker 6: see a lot of exciting examples in the startup space

593
00:27:16,280 --> 00:27:19,120
Speaker 6: of companies that are getting much much higher pricing.

594
00:27:19,680 --> 00:27:23,240
Speaker 2: This is a totally new pricing model for software. If

595
00:27:23,240 --> 00:27:26,480
Speaker 2: you just recorded another episode and they guess teased that,

596
00:27:26,560 --> 00:27:29,440
Speaker 2: but talk to us about like results based pricing, talk

597
00:27:29,480 --> 00:27:29,760
Speaker 2: to us.

598
00:27:29,960 --> 00:27:32,240
Speaker 6: Yeah, it's results based pricing. There's a lot of questions

599
00:27:32,280 --> 00:27:35,879
Speaker 6: on how it'll ultimately shake out. Fundamentally, what these companies

600
00:27:35,880 --> 00:27:38,760
Speaker 6: are doing is they are reselling intelligence.

601
00:27:39,000 --> 00:27:40,080
Speaker 5: Right the core model.

602
00:27:40,160 --> 00:27:43,720
Speaker 6: Vendors Open Eye and Thropict, Google have created a way

603
00:27:43,760 --> 00:27:47,280
Speaker 6: to get elastic intelligence. And if you have the right

604
00:27:47,359 --> 00:27:49,520
Speaker 6: data and you can put the right harness around it,

605
00:27:50,160 --> 00:27:52,479
Speaker 6: you can now sell that to your customers. What's an

606
00:27:52,480 --> 00:27:55,879
Speaker 6: open question is how do you price that relative to

607
00:27:55,920 --> 00:27:58,400
Speaker 6: the intelligence. So I was talking to someone this morning.

608
00:27:58,400 --> 00:28:01,159
Speaker 6: You said, I think fifty percent growth smart on intelligence

609
00:28:01,200 --> 00:28:03,479
Speaker 6: are about right. But we see a lot of variants

610
00:28:03,480 --> 00:28:05,240
Speaker 6: in how startups are doing it today. Some are getting

611
00:28:05,280 --> 00:28:07,600
Speaker 6: eighty percent gross margins on top of the model vendors.

612
00:28:07,880 --> 00:28:11,399
Speaker 6: Others are getting twenty percent. But what's absolutely true in

613
00:28:11,440 --> 00:28:14,320
Speaker 6: any case is if you're able to do that, you

614
00:28:14,359 --> 00:28:17,480
Speaker 6: get much much higher pricing in total dollars than you

615
00:28:17,520 --> 00:28:19,880
Speaker 6: did before orders a magnitude in some cases.

616
00:28:19,960 --> 00:28:22,919
Speaker 4: But just to be clear, like the cost savings, it

617
00:28:22,960 --> 00:28:25,880
Speaker 4: can't be priced so high that the company that's using

618
00:28:25,960 --> 00:28:30,520
Speaker 4: the software to produce these outcomes like isn't saving money, right,

619
00:28:30,560 --> 00:28:32,120
Speaker 4: that's the balancing act.

620
00:28:32,000 --> 00:28:33,800
Speaker 6: One hundred percent. But I think if you think about

621
00:28:33,840 --> 00:28:35,160
Speaker 6: when we talk about this a little bit earlier, think

622
00:28:35,160 --> 00:28:39,560
Speaker 6: about where software pricing was already right, you know, think

623
00:28:39,560 --> 00:28:42,520
Speaker 6: about salesforce. You know, at the elite tier, you know,

624
00:28:42,840 --> 00:28:45,320
Speaker 6: eighty ninety one hundred dollars per user per month. So

625
00:28:45,440 --> 00:28:48,240
Speaker 6: for round numbers, say one thousand dollars per user per year,

626
00:28:48,760 --> 00:28:50,959
Speaker 6: for sales reps who could be making on average two

627
00:28:51,000 --> 00:28:53,720
Speaker 6: hundred and fifty three hundred thousand dollars per year, if

628
00:28:53,760 --> 00:28:55,760
Speaker 6: you have a technology that can come in and replace

629
00:28:55,840 --> 00:28:59,280
Speaker 6: a sales rep, you can charge fifty thousand dollars still

630
00:28:59,280 --> 00:29:02,040
Speaker 6: give the customer or a five x ROI, and then

631
00:29:02,200 --> 00:29:05,400
Speaker 6: you have effectively fifty x your take rate on that revenue.

632
00:29:05,400 --> 00:29:08,040
Speaker 6: And so that's the exciting opportunity. That's that's what has

633
00:29:08,080 --> 00:29:10,360
Speaker 6: people excited in startup land for sure. If you talk

634
00:29:10,400 --> 00:29:13,120
Speaker 6: to folks from Silicon Valley, they are foaming at the

635
00:29:13,160 --> 00:29:16,480
Speaker 6: mouth about the opportunity to really expand tech spend in

636
00:29:16,520 --> 00:29:19,080
Speaker 6: this way. And that's also the opportunity for the software

637
00:29:19,080 --> 00:29:19,920
Speaker 6: companies that get it right.

638
00:29:20,120 --> 00:29:22,560
Speaker 2: There must be another risk too, which is that if

639
00:29:22,560 --> 00:29:25,520
Speaker 2: you can sort of resell intelligence it'd say an eighty

640
00:29:25,560 --> 00:29:30,239
Speaker 2: percent gross margin, then for the model makers themselves, they're like, well,

641
00:29:30,240 --> 00:29:32,000
Speaker 2: why do we just want to be this is gonna

642
00:29:32,000 --> 00:29:34,320
Speaker 2: sound weird. Why do we want to be the dumb intelligence?

643
00:29:34,440 --> 00:29:34,640
Speaker 5: Right?

644
00:29:34,680 --> 00:29:36,240
Speaker 2: And that's sort of like we don't they use it.

645
00:29:36,280 --> 00:29:37,600
Speaker 2: We don't want to be the dumb pipe. If we

646
00:29:37,640 --> 00:29:41,360
Speaker 2: saw that like in the cloud era, right, the Asias

647
00:29:41,400 --> 00:29:44,480
Speaker 2: and the Google Cloud and the Amazon One, they didn't

648
00:29:44,480 --> 00:29:47,400
Speaker 2: want to just be commodity cloud and they started building

649
00:29:47,480 --> 00:29:51,080
Speaker 2: like medical features. They wanted to differentiate themselves. So it

650
00:29:51,160 --> 00:29:54,080
Speaker 2: must be a risk for the company's reselling intelligence that

651
00:29:54,160 --> 00:29:56,840
Speaker 2: it's so lucrative. And then like, how are you thinking

652
00:29:56,880 --> 00:29:59,920
Speaker 2: about the core model makers themselves? Yeah, and how they're

653
00:30:00,080 --> 00:30:03,800
Speaker 2: thinking about expanding into some of these fields rather than

654
00:30:03,920 --> 00:30:05,600
Speaker 2: just piping in intelligence for them.

655
00:30:05,880 --> 00:30:07,960
Speaker 6: Well, look like in any situation, they're gonna have to

656
00:30:08,000 --> 00:30:11,080
Speaker 6: make decisions, right, So in Amazon, you know, built aws,

657
00:30:11,480 --> 00:30:13,680
Speaker 6: they had to decide where are we going to press

658
00:30:13,880 --> 00:30:15,640
Speaker 6: and where are we not? Are we going to sell

659
00:30:15,720 --> 00:30:17,760
Speaker 6: database software or are we going to let other vendors

660
00:30:17,840 --> 00:30:19,360
Speaker 6: do that on top of us? And they kind of

661
00:30:19,400 --> 00:30:22,080
Speaker 6: made those decisions as it went out. What's really interesting

662
00:30:22,120 --> 00:30:24,560
Speaker 6: is if you look at the foundation model vendors, they

663
00:30:24,560 --> 00:30:27,920
Speaker 6: have been racing towards the application layer. Both cloud code

664
00:30:27,960 --> 00:30:31,600
Speaker 6: and cowork and open ai codex are applications that people

665
00:30:31,760 --> 00:30:35,640
Speaker 6: download and use right, And I think that reflects this

666
00:30:35,800 --> 00:30:39,480
Speaker 6: understanding that there is value in getting the users used

667
00:30:39,520 --> 00:30:43,479
Speaker 6: to using your application. Otherwise, as you know, you may

668
00:30:43,600 --> 00:30:46,160
Speaker 6: you risk being an API that's commoditized. People switch back

669
00:30:46,160 --> 00:30:48,560
Speaker 6: and forth between you and that kind of application vendor

670
00:30:48,840 --> 00:30:49,560
Speaker 6: has that control.

671
00:31:05,640 --> 00:31:08,800
Speaker 4: So one of the advantages that software has is like

672
00:31:09,360 --> 00:31:14,880
Speaker 4: this network effect comfort software as a security blanket for management. Right,

673
00:31:15,240 --> 00:31:18,480
Speaker 4: But at the same time, people are getting really comfortable

674
00:31:18,760 --> 00:31:22,600
Speaker 4: with AI, like telling them everything. And I keep thinking

675
00:31:22,680 --> 00:31:26,200
Speaker 4: like if part of the sales pitch for software is

676
00:31:26,280 --> 00:31:30,640
Speaker 4: like this sense of comfort, but then AI is rapidly

677
00:31:30,720 --> 00:31:35,240
Speaker 4: becoming the thing that you talk to for everything. Does

678
00:31:35,240 --> 00:31:38,000
Speaker 4: it eventually just become a portal for doing all these

679
00:31:38,360 --> 00:31:39,120
Speaker 4: different things.

680
00:31:39,880 --> 00:31:42,480
Speaker 6: It's a really interesting question, and this is where there's

681
00:31:42,520 --> 00:31:46,840
Speaker 6: probably the biggest disparity between how enterprise buyers think and

682
00:31:46,880 --> 00:31:49,440
Speaker 6: how humans think. Right, I'm sure you guys have seen

683
00:31:49,600 --> 00:31:52,600
Speaker 6: claude Bot and the kind of rise of these, you know,

684
00:31:52,720 --> 00:31:55,720
Speaker 6: open source agents that people are deploying for themselves, giving

685
00:31:55,720 --> 00:31:57,920
Speaker 6: it and access to everything their whole computer, et cetera.

686
00:31:58,080 --> 00:31:58,560
Speaker 1: That's Joe.

687
00:31:58,840 --> 00:32:01,760
Speaker 3: Yeah, no, I didn't install I didn't install Club.

688
00:32:01,880 --> 00:32:02,640
Speaker 4: Oh you didn't know.

689
00:32:02,680 --> 00:32:04,640
Speaker 5: I'm getting mindset up. I lacked many with a hammer

690
00:32:04,680 --> 00:32:05,040
Speaker 5: next to it.

691
00:32:05,080 --> 00:32:06,240
Speaker 4: Well, I'm really curious why not?

692
00:32:06,800 --> 00:32:07,760
Speaker 6: Because of this issue?

693
00:32:08,040 --> 00:32:11,040
Speaker 2: Yeah, because of that and it just seemed like a

694
00:32:11,080 --> 00:32:12,640
Speaker 2: potential waste of tokens and stuff.

695
00:32:12,680 --> 00:32:14,560
Speaker 6: And yeah, and then then it turned out that for

696
00:32:14,600 --> 00:32:17,200
Speaker 6: a while on Moltbook, which was the social media for

697
00:32:17,440 --> 00:32:20,040
Speaker 6: all of the all the APIs are available in a

698
00:32:20,080 --> 00:32:22,240
Speaker 6: public facing database and one could go read and so

699
00:32:22,280 --> 00:32:24,240
Speaker 6: it was like a completely open system.

700
00:32:23,960 --> 00:32:24,760
Speaker 5: That had to get fixed.

701
00:32:24,760 --> 00:32:29,000
Speaker 6: And so, you know, enterprises really do worry about this stuff,

702
00:32:29,000 --> 00:32:30,920
Speaker 6: and they worry about it for a good reason. You

703
00:32:31,000 --> 00:32:33,000
Speaker 6: have another really interesting example. So there's a bunch of

704
00:32:33,040 --> 00:32:36,400
Speaker 6: startups that help you record zoom calls and transcribe them.

705
00:32:36,600 --> 00:32:40,040
Speaker 6: All of those zoom calls then become legally discoverable because

706
00:32:40,080 --> 00:32:42,720
Speaker 6: they're transcribed somewhere, and so you have vcs in Silicon

707
00:32:42,800 --> 00:32:44,560
Speaker 6: Valley who will refuse to use them, and you have

708
00:32:44,600 --> 00:32:46,680
Speaker 6: other firms that are all in and recording everything that

709
00:32:46,720 --> 00:32:49,400
Speaker 6: happens across the board so that they can upload that

710
00:32:49,440 --> 00:32:52,360
Speaker 6: into AI as context. I think it's a really it's

711
00:32:52,360 --> 00:32:54,320
Speaker 6: a really great point, you know. And one of the

712
00:32:54,320 --> 00:32:57,160
Speaker 6: things that makes me wonder is companies that are willing

713
00:32:57,200 --> 00:33:00,320
Speaker 6: to skirt the rules or you know, play fast and loose,

714
00:33:00,400 --> 00:33:02,640
Speaker 6: will be moving much faster over the next two or

715
00:33:02,680 --> 00:33:05,200
Speaker 6: three years. And One of the reasons big incumbents struggle

716
00:33:05,560 --> 00:33:08,000
Speaker 6: is because they actually do have to care about this stuff.

717
00:33:08,000 --> 00:33:09,320
Speaker 5: They have stuff to protect, they.

718
00:33:09,280 --> 00:33:12,640
Speaker 6: Don't want to be sued. They can't handle a major breach,

719
00:33:12,920 --> 00:33:15,320
Speaker 6: and startups are able to just move faster given that.

720
00:33:16,280 --> 00:33:20,000
Speaker 2: So every time software stocks sell off with this and

721
00:33:20,040 --> 00:33:22,400
Speaker 2: people say, oh, they micro bargain hunting, and they say

722
00:33:22,440 --> 00:33:25,240
Speaker 2: what's cheap and what baby is being thrown out with

723
00:33:25,280 --> 00:33:25,880
Speaker 2: the bath water?

724
00:33:26,000 --> 00:33:27,320
Speaker 3: Someone always a bunch.

725
00:33:27,160 --> 00:33:29,400
Speaker 2: Of people is like, yes, they look cheap, but have

726
00:33:29,520 --> 00:33:32,960
Speaker 2: you considered a stock based compensation? And it turns out

727
00:33:33,040 --> 00:33:35,400
Speaker 2: that these companies are not nearly as profitable once you

728
00:33:35,480 --> 00:33:39,160
Speaker 2: factor this in. Its very interesting note from Barkleys I

729
00:33:39,160 --> 00:33:41,720
Speaker 2: think it was I think it was Barkleys. This is

730
00:33:41,800 --> 00:33:44,800
Speaker 2: very interesting, and it said our European investors are always

731
00:33:44,840 --> 00:33:48,120
Speaker 2: asking about SBC. Our American investors only ask whether there's

732
00:33:48,160 --> 00:33:48,680
Speaker 2: a crisis.

733
00:33:49,000 --> 00:33:50,320
Speaker 3: I think tells you something about.

734
00:33:50,120 --> 00:33:52,200
Speaker 2: The differentiat in Europeans and Americas. I thought that was

735
00:33:52,240 --> 00:33:56,280
Speaker 2: a fascinating sociological observation. Tell us, like, how should we

736
00:33:56,360 --> 00:33:59,880
Speaker 2: think about the cost, because again, if code generation is

737
00:33:59,880 --> 00:34:03,400
Speaker 2: a cost base, presumably these software companies don't need as

738
00:34:03,440 --> 00:34:05,520
Speaker 2: many employees either. And they could pair back on this,

739
00:34:05,600 --> 00:34:07,320
Speaker 2: So talk to us about how we're thinking about the

740
00:34:07,360 --> 00:34:09,480
Speaker 2: costs inside the software company.

741
00:34:09,520 --> 00:34:12,719
Speaker 6: Great question, and yeah, I mean certainly theoretically true, right,

742
00:34:12,760 --> 00:34:17,239
Speaker 6: but aside from Elon cutting eighty percent of twitter X's headcount,

743
00:34:17,520 --> 00:34:20,440
Speaker 6: we really haven't seen any companies take the pill and

744
00:34:20,520 --> 00:34:23,560
Speaker 6: kind of realize the benefits of that. The SBC debate

745
00:34:23,800 --> 00:34:26,719
Speaker 6: has been going on for a long time. I've had

746
00:34:26,719 --> 00:34:29,080
Speaker 6: it at nausimo of the course of my career. It's

747
00:34:29,080 --> 00:34:32,320
Speaker 6: a real expense. You're issuing your employees stock. They value

748
00:34:32,320 --> 00:34:34,640
Speaker 6: it like cash. Many of them auto sell it the

749
00:34:34,719 --> 00:34:37,960
Speaker 6: day at vests for them. And I think what the

750
00:34:38,080 --> 00:34:41,480
Speaker 6: problem that it creates for software companies is they the

751
00:34:41,520 --> 00:34:44,560
Speaker 6: management teams are addicted to reporting non gap, which excludes

752
00:34:44,600 --> 00:34:48,320
Speaker 6: the impact of SBC. And so if you are an

753
00:34:48,480 --> 00:34:51,959
Speaker 6: entrepreneur who founded the software business who's technical, hasn't really

754
00:34:51,960 --> 00:34:54,200
Speaker 6: ever cared more that much about the financial side your

755
00:34:54,280 --> 00:34:56,799
Speaker 6: product person, you may think that you've been doing a

756
00:34:56,840 --> 00:34:59,920
Speaker 6: good job of being a profitable company because your CFO

757
00:35:00,120 --> 00:35:02,000
Speaker 6: telling you, well, we're at a twenty five percent non

758
00:35:02,000 --> 00:35:05,279
Speaker 6: gap operating margin. That's pretty good when the reality is

759
00:35:05,360 --> 00:35:07,879
Speaker 6: you're running break even, which is a very common state

760
00:35:07,920 --> 00:35:10,120
Speaker 6: of affairs. You know, we looked at the whole universe

761
00:35:10,160 --> 00:35:12,680
Speaker 6: and the media and public software company has a five

762
00:35:12,719 --> 00:35:15,320
Speaker 6: percent non gap net income or gap net income margin,

763
00:35:15,680 --> 00:35:18,360
Speaker 6: which is not enough to value the companies on. And

764
00:35:18,400 --> 00:35:21,839
Speaker 6: so it creates this dynamic where you know, yes, there's

765
00:35:21,840 --> 00:35:24,160
Speaker 6: a terminal value concern, which by far the most important thing,

766
00:35:24,160 --> 00:35:25,560
Speaker 6: but there's also no floor.

767
00:35:25,800 --> 00:35:25,960
Speaker 5: Right.

768
00:35:26,000 --> 00:35:28,480
Speaker 6: I was looking at the earnings report from fresh Works,

769
00:35:28,480 --> 00:35:31,560
Speaker 6: which is a mid market seller of customer support and

770
00:35:31,640 --> 00:35:34,479
Speaker 6: IT management software. It trades at one and a half

771
00:35:34,520 --> 00:35:37,759
Speaker 6: times EVY to sales. If it ran at even a

772
00:35:37,800 --> 00:35:41,960
Speaker 6: ten percent gap margin, it'd be trading at fifteen times earnings.

773
00:35:42,000 --> 00:35:44,600
Speaker 6: You know, that's which is a pretty attractive place to be.

774
00:35:44,680 --> 00:35:47,480
Speaker 6: You could get some value investors, maybe some European investors

775
00:35:47,880 --> 00:35:51,320
Speaker 6: interested in buying it there, but it doesn't have material

776
00:35:51,360 --> 00:35:54,440
Speaker 6: gap earnings. And on their earnings call, there was no

777
00:35:54,640 --> 00:35:58,200
Speaker 6: real you know, sense of trajectory towards that, and you

778
00:35:58,239 --> 00:36:01,040
Speaker 6: see the share price down sixteen p exactly, and like,

779
00:36:01,280 --> 00:36:03,360
Speaker 6: the top line results were actually pretty good, and so

780
00:36:04,040 --> 00:36:06,800
Speaker 6: there's a real issue here on the financial side as well.

781
00:36:07,680 --> 00:36:11,520
Speaker 6: It's incredibly disappointing to me that management teams haven't embraced

782
00:36:11,520 --> 00:36:14,920
Speaker 6: this as a way to cut costs themselves, and I

783
00:36:15,000 --> 00:36:15,840
Speaker 6: expect they will.

784
00:36:16,000 --> 00:36:19,640
Speaker 2: Yeah, talk to us about this specifically. Are we going

785
00:36:19,640 --> 00:36:23,000
Speaker 2: to see big layoffs across the SaaS space in the

786
00:36:23,040 --> 00:36:24,960
Speaker 2: near term and what do you think is the timeframe

787
00:36:24,960 --> 00:36:26,000
Speaker 2: for that great question?

788
00:36:26,160 --> 00:36:29,160
Speaker 6: I think we will. I think we've seen that management

789
00:36:29,160 --> 00:36:32,840
Speaker 6: teams do respond to price signals. If you look at

790
00:36:33,080 --> 00:36:35,480
Speaker 6: the history of the sector, it was in twenty twenty

791
00:36:35,520 --> 00:36:37,960
Speaker 6: three when there was a round of layoffs and companies

792
00:36:38,000 --> 00:36:40,360
Speaker 6: showed margin, and then they've kind of resisted it for

793
00:36:40,400 --> 00:36:43,640
Speaker 6: the last two years. The thing about it is layoffs

794
00:36:43,640 --> 00:36:46,319
Speaker 6: can help you move faster, right. I think if you

795
00:36:46,400 --> 00:36:50,160
Speaker 6: look within any company today, unfortunately, there is this spectrum

796
00:36:50,200 --> 00:36:53,480
Speaker 6: of employees and how fast they've adopted AI, whether they're

797
00:36:53,480 --> 00:36:56,000
Speaker 6: still doing things the old way or they're on cloud code,

798
00:36:56,200 --> 00:36:58,759
Speaker 6: cloud cowork, kind of changing the way that they work.

799
00:36:59,160 --> 00:37:02,320
Speaker 6: And employees who are on the lower end are actually

800
00:37:02,360 --> 00:37:05,040
Speaker 6: slowing you down as a company. They're not even zero

801
00:37:05,120 --> 00:37:08,319
Speaker 6: marginal product, they're negative marginal product. There's just been such

802
00:37:08,320 --> 00:37:10,800
Speaker 6: a change in how you work, especially in software development,

803
00:37:11,200 --> 00:37:14,120
Speaker 6: and so I think management teams are going to realize

804
00:37:14,120 --> 00:37:18,440
Speaker 6: that there's two benefits to actually doing layoffs in addition

805
00:37:18,440 --> 00:37:20,120
Speaker 6: to the obvious pain of it and the kind of

806
00:37:20,200 --> 00:37:24,000
Speaker 6: human costs which I never forget to discuss. But one

807
00:37:24,040 --> 00:37:27,480
Speaker 6: is saving money and showing your shareholders you're financially disciplined

808
00:37:27,520 --> 00:37:30,279
Speaker 6: and probably seeing your share price stabilize, especially if you're

809
00:37:30,280 --> 00:37:32,920
Speaker 6: trading at some very low multiple. And the second is

810
00:37:33,560 --> 00:37:36,520
Speaker 6: moving faster and also almost as importantly, being able to

811
00:37:36,560 --> 00:37:39,440
Speaker 6: pay your top performing employees. The war for talent in

812
00:37:39,480 --> 00:37:42,360
Speaker 6: Silicon Valley has never been more intense right now. Is

813
00:37:42,400 --> 00:37:45,359
Speaker 6: talking to a private company you invested in and they're

814
00:37:45,400 --> 00:37:48,000
Speaker 6: losing employees left and right to these high growth AI

815
00:37:48,080 --> 00:37:51,680
Speaker 6: companies who can afford to pay huge compackages in both

816
00:37:51,680 --> 00:37:54,720
Speaker 6: equity and stock, and you want to keep your good people.

817
00:37:54,960 --> 00:37:55,440
Speaker 5: You don't want it.

818
00:37:55,520 --> 00:37:57,839
Speaker 6: You don't want these AI companies that pluck away all

819
00:37:57,920 --> 00:37:59,960
Speaker 6: the best people and leave you with the folks who

820
00:38:00,080 --> 00:38:02,160
Speaker 6: relative leddites. And so I do think we'll see this.

821
00:38:02,360 --> 00:38:05,160
Speaker 6: It's very sad that that will have to happen, but

822
00:38:06,040 --> 00:38:08,319
Speaker 6: it's the obvious path forward for the sector, and I

823
00:38:08,360 --> 00:38:10,360
Speaker 6: think if done right, it accelerates innovation.

824
00:38:10,640 --> 00:38:13,680
Speaker 4: I have a tangential question on that note, which is

825
00:38:13,960 --> 00:38:17,560
Speaker 4: whenever we talk about technological disruption, you know, people bring

826
00:38:17,560 --> 00:38:22,040
Speaker 4: out examples of like remember when Excel was basically actual,

827
00:38:22,120 --> 00:38:25,520
Speaker 4: people sat down with like papers in front of them

828
00:38:25,880 --> 00:38:29,720
Speaker 4: doing the math. And those people didn't disappear when Excel

829
00:38:29,800 --> 00:38:34,280
Speaker 4: got created, but they started doing new things. I imagine

830
00:38:34,320 --> 00:38:36,319
Speaker 4: a lot of people are very interested right now in

831
00:38:36,600 --> 00:38:42,360
Speaker 4: alternative careers for basic commoditized coders. What do those actually

832
00:38:42,400 --> 00:38:45,080
Speaker 4: look like? Yeah, I feel like you might have some

833
00:38:45,160 --> 00:38:46,800
Speaker 4: insight here the alternative.

834
00:38:46,800 --> 00:38:48,880
Speaker 6: Well, so I think there's two ways to answer the question. Right,

835
00:38:48,880 --> 00:38:50,120
Speaker 6: There's like what do you do if you want to

836
00:38:50,160 --> 00:38:52,560
Speaker 6: stay a coder? And then there's what are the careers

837
00:38:52,600 --> 00:38:54,239
Speaker 6: that are going to still exist over time?

838
00:38:54,320 --> 00:38:54,520
Speaker 5: Right?

839
00:38:55,040 --> 00:38:57,120
Speaker 6: I think if you think about what's happening to coding,

840
00:38:57,880 --> 00:39:00,520
Speaker 6: it reminds me a lot of civil engineering, and so

841
00:39:00,840 --> 00:39:02,760
Speaker 6: it's kind of a funky example, but you know, civil

842
00:39:02,760 --> 00:39:05,319
Speaker 6: engineers used to work pen and paper doing calculus will

843
00:39:05,360 --> 00:39:08,799
Speaker 6: this bridge hold up or not? That's been obsolete for

844
00:39:08,840 --> 00:39:11,080
Speaker 6: a very long time. All those calculations are done by

845
00:39:11,080 --> 00:39:13,480
Speaker 6: a computer. They're kind of clicking and moving, and they

846
00:39:13,520 --> 00:39:16,359
Speaker 6: go on site, they collect some data, they talk to stakeholders,

847
00:39:16,600 --> 00:39:19,480
Speaker 6: and they're effectively project managing this computer that can do

848
00:39:19,520 --> 00:39:22,080
Speaker 6: the physics part of their job. For them, it's important

849
00:39:22,080 --> 00:39:24,840
Speaker 6: that they understand the physics in case something looks strange,

850
00:39:25,080 --> 00:39:28,000
Speaker 6: but they're not doing much physics right. That's clearly where

851
00:39:28,040 --> 00:39:30,360
Speaker 6: software engineering is heading in the near term, and a

852
00:39:30,400 --> 00:39:32,719
Speaker 6: lot of companies that's already there. And these companies are

853
00:39:32,760 --> 00:39:35,759
Speaker 6: still hiring software engineers because that job is valuable, and

854
00:39:35,800 --> 00:39:38,640
Speaker 6: in fact, each individual software engineer is way more productive

855
00:39:38,920 --> 00:39:42,440
Speaker 6: than they were before. And there's happily elastic demand for software,

856
00:39:42,480 --> 00:39:44,879
Speaker 6: Like we still are undersupplied with software in the world,

857
00:39:45,160 --> 00:39:46,560
Speaker 6: and so there's quite a bit of room to go

858
00:39:46,600 --> 00:39:49,560
Speaker 6: to add those and so I'm not necessarily bearish on

859
00:39:50,440 --> 00:39:53,360
Speaker 6: the demand for software engineers, at least for the next

860
00:39:53,520 --> 00:39:55,680
Speaker 6: you know, three to five years beyond that if things

861
00:39:55,680 --> 00:39:58,520
Speaker 6: get weird hard to tell. But then for people more broadly,

862
00:39:58,680 --> 00:40:02,759
Speaker 6: I think the the best advice is just adaptability, you know,

863
00:40:03,000 --> 00:40:06,319
Speaker 6: constantly trying and testing these tools, making sure you're staying

864
00:40:06,320 --> 00:40:08,720
Speaker 6: at the cutting edge of them, and then being aware

865
00:40:08,760 --> 00:40:11,600
Speaker 6: of what's human right. I think in like in my

866
00:40:11,719 --> 00:40:14,200
Speaker 6: work and venture investing, you know, there's a lot of

867
00:40:14,239 --> 00:40:16,600
Speaker 6: data that comes out of human relationships that an AI

868
00:40:16,640 --> 00:40:18,799
Speaker 6: wouldn't have access to you know, an AI can't call

869
00:40:18,880 --> 00:40:21,439
Speaker 6: its friend at another fund and ask how company is doing.

870
00:40:21,760 --> 00:40:24,160
Speaker 6: Not yet, at least they have to make some friends first.

871
00:40:24,239 --> 00:40:26,480
Speaker 3: Right, they're talking about they are talking to each other

872
00:40:26,520 --> 00:40:27,000
Speaker 3: on moultbook.

873
00:40:27,080 --> 00:40:28,759
Speaker 6: Right, they're talking about their Moult book. Yeah, so maybe

874
00:40:28,760 --> 00:40:30,120
Speaker 6: if there's an AI agent from Sequi.

875
00:40:31,480 --> 00:40:33,080
Speaker 3: I was intrigued by that.

876
00:40:33,280 --> 00:40:35,759
Speaker 5: Yeah, it was very evocative, but pretty fake.

877
00:40:35,840 --> 00:40:38,040
Speaker 4: Well also there was that Wired article of the guy

878
00:40:38,080 --> 00:40:41,480
Speaker 4: who like infiltrated as a bought and pretended to.

879
00:40:41,440 --> 00:40:44,880
Speaker 2: Be They're like, oh, why are we let's create a

880
00:40:44,920 --> 00:40:46,759
Speaker 2: new language just for us, Like they're not making new.

881
00:40:46,719 --> 00:40:49,120
Speaker 6: Languages right, But I think I think the rough mental

882
00:40:49,120 --> 00:40:51,600
Speaker 6: model is if there was any effort to outsource your

883
00:40:51,680 --> 00:40:54,759
Speaker 6: job to India, that's risk because that tells you that

884
00:40:54,840 --> 00:40:57,320
Speaker 6: job can be done by person, someone who's not physically

885
00:40:57,360 --> 00:41:00,960
Speaker 6: present in a space. And you know, if you like

886
00:41:01,040 --> 00:41:04,920
Speaker 6: working on problems in isolation, not socially with other people,

887
00:41:05,239 --> 00:41:08,120
Speaker 6: you know, grinding out math problems or little coding assignments,

888
00:41:08,520 --> 00:41:11,160
Speaker 6: that's probably also a pretty tough place to be. It's

889
00:41:11,200 --> 00:41:12,239
Speaker 6: going to be more social world.

890
00:41:12,440 --> 00:41:14,960
Speaker 4: This is something we've touched on before, which makes me

891
00:41:15,080 --> 00:41:18,440
Speaker 4: kind of sad, which is the edge in the AI

892
00:41:18,560 --> 00:41:23,080
Speaker 4: world becomes like sociability right, And to some extent we

893
00:41:23,160 --> 00:41:25,440
Speaker 4: talked about this in the context of look smack, I

894
00:41:25,520 --> 00:41:27,200
Speaker 4: know you love it, Joe, I do not.

895
00:41:27,920 --> 00:41:32,160
Speaker 2: Can I say two little observations from my time vibe

896
00:41:32,160 --> 00:41:35,040
Speaker 2: coding in twenty twenty six that are interesting. One is

897
00:41:35,480 --> 00:41:39,080
Speaker 2: I have zero technical background. I've been surprised by the

898
00:41:39,200 --> 00:41:42,799
Speaker 2: speed with which I can build intuitions about when it's

899
00:41:42,840 --> 00:41:45,960
Speaker 2: going off the rails, like when it's doing something that

900
00:41:46,000 --> 00:41:48,759
Speaker 2: doesn't seem right. Like I joke that vibe coding is

901
00:41:48,840 --> 00:41:51,560
Speaker 2: just typing make it better presenter over and over again

902
00:41:51,800 --> 00:41:54,760
Speaker 2: and then hitting yes. When it offers to do something.

903
00:41:55,680 --> 00:41:58,960
Speaker 2: You actually can start to build an intuition fairly quickly,

904
00:41:59,040 --> 00:42:01,160
Speaker 2: like when this doesn't make sense, and then the other thing.

905
00:42:01,160 --> 00:42:04,320
Speaker 2: And this relates to your question of like trusting the AI.

906
00:42:04,760 --> 00:42:06,359
Speaker 2: So one of the things I'm doing is I'm having

907
00:42:06,360 --> 00:42:08,960
Speaker 2: a lot of documents get annotated, and that I do

908
00:42:09,000 --> 00:42:11,560
Speaker 2: that through the Claude API, which actually runs up the

909
00:42:11,600 --> 00:42:13,799
Speaker 2: bills a little bit. And one thing this API run

910
00:42:13,920 --> 00:42:16,839
Speaker 2: was gonna cost like one hundred dollars, and I stupidly

911
00:42:17,080 --> 00:42:18,600
Speaker 2: asked Claude it is like, is this a good thing.

912
00:42:18,600 --> 00:42:21,359
Speaker 2: It's like, well, when you're done with this API API run,

913
00:42:21,520 --> 00:42:24,080
Speaker 2: you're gonna have this annotated asset that no one else

914
00:42:24,120 --> 00:42:25,480
Speaker 2: has done, and that'll be very.

915
00:42:25,640 --> 00:42:28,320
Speaker 3: It was sort of useless what I did, so you shouldn't.

916
00:42:28,719 --> 00:42:29,320
Speaker 5: It's selling it.

917
00:42:29,880 --> 00:42:31,120
Speaker 3: Selling itself is like, oh.

918
00:42:31,040 --> 00:42:34,160
Speaker 2: Yeah, use the API, Joe run this like annotated all

919
00:42:34,200 --> 00:42:36,680
Speaker 2: these documents. It wasn't actually like a good use of

920
00:42:36,719 --> 00:42:39,600
Speaker 2: my time. So you can't really always they're just gonna

921
00:42:39,840 --> 00:42:42,960
Speaker 2: they're gonna just sell these things. So Tracy asked about data,

922
00:42:43,160 --> 00:42:45,960
Speaker 2: and there's one other sectors that I'm interested in.

923
00:42:46,360 --> 00:42:47,719
Speaker 3: You see these companies like.

924
00:42:47,760 --> 00:42:50,480
Speaker 7: Moody's or fair Isaac.

925
00:42:50,480 --> 00:42:53,399
Speaker 2: Or sp Global they have an index, yeah, and they're

926
00:42:53,400 --> 00:42:55,799
Speaker 2: getting they're selling off too. And it's not like this

927
00:42:55,840 --> 00:42:58,960
Speaker 2: is another area where like it's you know, people were

928
00:42:59,320 --> 00:43:02,320
Speaker 2: fund managers for a long time. Unless things get really weird,

929
00:43:02,920 --> 00:43:05,279
Speaker 2: the super they are gonna be like benchmarking themselves off

930
00:43:05,320 --> 00:43:06,799
Speaker 2: of like that S and P five hundred for a

931
00:43:06,840 --> 00:43:10,120
Speaker 2: long time, or lenders are gonna be using the FICO

932
00:43:10,440 --> 00:43:13,759
Speaker 2: for a very long time, et cetera. Intuitively, it would

933
00:43:13,800 --> 00:43:15,560
Speaker 2: strike be is this would be a very hard thing

934
00:43:15,640 --> 00:43:16,640
Speaker 2: for ad or replace.

935
00:43:16,920 --> 00:43:19,520
Speaker 6: I share your intuition, right, I can't say I fully

936
00:43:19,600 --> 00:43:23,000
Speaker 6: understand the selloff and these companies, I wonder if there's

937
00:43:23,040 --> 00:43:25,440
Speaker 6: not some parts of their businesses that are more services

938
00:43:25,520 --> 00:43:28,799
Speaker 6: or consulting heads that people are there's often like combinations

939
00:43:28,960 --> 00:43:31,440
Speaker 6: right where like I don't think anyone suspecting that, like,

940
00:43:31,520 --> 00:43:33,360
Speaker 6: you know, the S and P five hundred is going

941
00:43:33,440 --> 00:43:37,520
Speaker 6: to be displaced as an index throughout Yeah. Maybe, but look,

942
00:43:37,600 --> 00:43:41,200
Speaker 6: we're in a world where folks are very happy to

943
00:43:41,239 --> 00:43:44,760
Speaker 6: shoot first and ask questions later once AI risk comes out.

944
00:43:45,000 --> 00:43:48,600
Speaker 2: Actually going back to your hedge fund time, like how

945
00:43:48,680 --> 00:43:50,880
Speaker 2: much is it just the sort of the nature of

946
00:43:50,920 --> 00:43:55,200
Speaker 2: hedge fund traders right now where there's very little stomach

947
00:43:55,239 --> 00:43:58,720
Speaker 2: to take any downside risk and appear to look stupid

948
00:43:58,760 --> 00:44:00,840
Speaker 2: for missing you know, yeah the bag here, and how

949
00:44:00,920 --> 00:44:02,480
Speaker 2: much do you think that's contributing to some of these

950
00:44:02,480 --> 00:44:02,959
Speaker 2: market money?

951
00:44:03,000 --> 00:44:03,600
Speaker 5: It's a great question.

952
00:44:03,600 --> 00:44:05,000
Speaker 6: I won't speak to my fund because I think CO

953
00:44:05,120 --> 00:44:07,480
Speaker 6: two and Tiger cubs like it are a fairly small

954
00:44:07,560 --> 00:44:09,520
Speaker 6: share of the overall market and dollars, right, but if

955
00:44:09,560 --> 00:44:12,920
Speaker 6: you look at trading volumes, the pod shops, Citadel, Ballyas

956
00:44:12,920 --> 00:44:15,040
Speaker 6: and the millennium are very large share of the volumes,

957
00:44:15,040 --> 00:44:19,680
Speaker 6: and yeah, those people can't afford draughtouts, right. And the

958
00:44:19,719 --> 00:44:24,399
Speaker 6: scary thing about this for them is because it's not fundamental.

959
00:44:24,719 --> 00:44:28,320
Speaker 6: Because the companies aren't struggling themselves, they have no idea

960
00:44:28,360 --> 00:44:31,520
Speaker 6: when it will stop, right, And so you're left predicting

961
00:44:31,520 --> 00:44:33,320
Speaker 6: this thing. And you're like, well, I bet my career

962
00:44:33,640 --> 00:44:36,440
Speaker 6: that people are going to feel better about software companies

963
00:44:36,480 --> 00:44:38,600
Speaker 6: in three to six months than they do right now.

964
00:44:38,880 --> 00:44:41,160
Speaker 6: And you know, your one open AI model release or

965
00:44:41,200 --> 00:44:44,319
Speaker 6: anthropic model release away from more fear. And so I

966
00:44:44,360 --> 00:44:48,000
Speaker 6: do think there's a lot of short termism right now,

967
00:44:48,520 --> 00:44:51,000
Speaker 6: and there's all but again I think we come back

968
00:44:51,000 --> 00:44:53,480
Speaker 6: to the SBC point. But there's also no valuation support,

969
00:44:53,680 --> 00:44:56,839
Speaker 6: no real valuation support. You know, in normal times, if

970
00:44:56,840 --> 00:44:58,560
Speaker 6: companies were like this, they'd be buying back a bunch

971
00:44:58,600 --> 00:45:01,799
Speaker 6: of their stock, taking their share out, issuing dividends. You know,

972
00:45:01,840 --> 00:45:03,440
Speaker 6: I have a friend who works at a mutual fund

973
00:45:03,480 --> 00:45:05,600
Speaker 6: with is a lot of dividend funds that would love

974
00:45:05,680 --> 00:45:08,719
Speaker 6: to buy dividending companies growing ten to fifteen percent, like

975
00:45:08,760 --> 00:45:11,560
Speaker 6: a lot of software companies. But they're not right, and

976
00:45:11,640 --> 00:45:14,480
Speaker 6: so you've kind of lost the ability to kind of

977
00:45:14,480 --> 00:45:16,800
Speaker 6: put an actual floor in underneath evaluations.

978
00:45:16,800 --> 00:45:19,360
Speaker 2: As a result of that, Jared Sleeper, thank you so

979
00:45:19,440 --> 00:45:22,480
Speaker 2: much for coming on odd Laws and explaining how socrow works.

980
00:45:22,520 --> 00:45:23,799
Speaker 5: My pleasure is super fun.

981
00:45:23,960 --> 00:45:37,240
Speaker 7: Thanks Tracy.

982
00:45:37,280 --> 00:45:40,080
Speaker 2: I thought that was really interesting. I'm very fascinated with

983
00:45:40,160 --> 00:45:43,359
Speaker 2: this idea that in the short term most of these

984
00:45:43,400 --> 00:45:46,799
Speaker 2: businesses are doing fine. In the long term it might

985
00:45:46,840 --> 00:45:49,560
Speaker 2: go to zero. But also with the short term they're

986
00:45:49,560 --> 00:45:53,240
Speaker 2: not really doing fine because actually, you on a gap basis,

987
00:45:53,239 --> 00:45:55,279
Speaker 2: they're not making much money. I guess that sort of

988
00:45:55,280 --> 00:45:57,239
Speaker 2: makes sense that they're all selling off right now.

989
00:45:57,360 --> 00:46:00,680
Speaker 4: Yeah, I keep thinking this is probably a stretch, but

990
00:46:00,719 --> 00:46:04,440
Speaker 4: I keep thinking back to that book Full Jobs. Yeah, remember,

991
00:46:04,719 --> 00:46:07,360
Speaker 4: and like the argument there is a lot of jobs

992
00:46:07,400 --> 00:46:12,200
Speaker 4: exist not because like people are doing anything specific, but

993
00:46:12,280 --> 00:46:15,799
Speaker 4: because they're like providing some sort of social value in

994
00:46:15,840 --> 00:46:18,560
Speaker 4: a way. So, for instance, like you have a person

995
00:46:18,600 --> 00:46:22,920
Speaker 4: who is essentially the designated scapegoat for senior management. And

996
00:46:22,960 --> 00:46:25,640
Speaker 4: I keep thinking of you know, business is basically an

997
00:46:25,640 --> 00:46:28,600
Speaker 4: ecosystem of different players, So it might be that in

998
00:46:28,640 --> 00:46:32,960
Speaker 4: the new AI world the role of software companies just

999
00:46:33,120 --> 00:46:37,400
Speaker 4: kind of changes, like their social role changes. And I

1000
00:46:37,440 --> 00:46:39,640
Speaker 4: don't know what the price or the valuation looks like

1001
00:46:39,719 --> 00:46:40,799
Speaker 4: on that it.

1002
00:46:40,800 --> 00:46:45,000
Speaker 2: Still feels like to me, I have no I was

1003
00:46:45,040 --> 00:46:46,759
Speaker 2: going to say something about how businesses we're going to

1004
00:46:47,000 --> 00:46:49,160
Speaker 2: what do I know? I have no idea how businesses

1005
00:46:49,400 --> 00:46:51,440
Speaker 2: are going to buy software in the future. I did

1006
00:46:51,520 --> 00:46:53,719
Speaker 2: think that was really helpful, Like I really don't know

1007
00:46:53,760 --> 00:46:56,640
Speaker 2: anything about how the software business works generally, so I

1008
00:46:56,719 --> 00:46:59,520
Speaker 2: find that very helpful. One other interesting thing though, and

1009
00:46:59,560 --> 00:47:01,839
Speaker 2: it may sort of speak to think about this risk,

1010
00:47:02,280 --> 00:47:05,080
Speaker 2: is just the idea that, like, even high end software

1011
00:47:05,120 --> 00:47:06,600
Speaker 2: is not that much money, right, So if you have

1012
00:47:06,640 --> 00:47:09,200
Speaker 2: a salesperson who's making two hundred and fifty thousand dollars,

1013
00:47:09,560 --> 00:47:12,799
Speaker 2: what is one thousand dollars a year from salesforce.

1014
00:47:12,280 --> 00:47:12,919
Speaker 7: To do their job?

1015
00:47:13,400 --> 00:47:16,279
Speaker 2: Particularly and then also given the fact that you know,

1016
00:47:16,440 --> 00:47:19,280
Speaker 2: free and open source software has existed for a long time,

1017
00:47:19,600 --> 00:47:21,960
Speaker 2: but still you know you want to pay an implementer

1018
00:47:22,160 --> 00:47:25,680
Speaker 2: for a company that like manages et cetera. Getting from

1019
00:47:25,680 --> 00:47:29,759
Speaker 2: like here to there where okay Ai really changes the

1020
00:47:29,880 --> 00:47:34,120
Speaker 2: nature of software buying does feel like you have to get.

1021
00:47:33,920 --> 00:47:36,920
Speaker 3: Into this is going to be weird toward tear, but maybe.

1022
00:47:36,680 --> 00:47:37,120
Speaker 7: Things are good.

1023
00:47:37,160 --> 00:47:39,480
Speaker 3: I think things probably are going to get really weird.

1024
00:47:39,400 --> 00:47:41,840
Speaker 4: Yeah, I think that's that's a pretty good bet. Right, Like,

1025
00:47:41,960 --> 00:47:45,800
Speaker 4: if you bet on weirdness, if there is a weirdness index.

1026
00:47:45,680 --> 00:47:47,200
Speaker 3: Someone should build that weirdness index.

1027
00:47:47,320 --> 00:47:49,920
Speaker 4: That would be a pretty good investment. Yeah, all right,

1028
00:47:49,960 --> 00:47:51,719
Speaker 4: shall we leave it there, Let's live it there. This

1029
00:47:51,760 --> 00:47:54,319
Speaker 4: has been another episode of the All Thoughts podcast. I'm

1030
00:47:54,360 --> 00:47:57,400
Speaker 4: Tracy Alloway. You can follow me at Tracy Alloway.

1031
00:47:57,080 --> 00:47:58,960
Speaker 2: And I'm Joe Why Isn't Though? You can follow me

1032
00:47:59,080 --> 00:48:02,000
Speaker 2: at the Stalwart. Follow our guest Jared Sleeper He's at

1033
00:48:02,120 --> 00:48:06,000
Speaker 2: Jared's Sleeper. Follow our producers Carmen Rodriguez at Carmen armand

1034
00:48:06,040 --> 00:48:09,319
Speaker 2: dash Ol Bennett at dashbod and Kilbrooks at Kilbrooks. For

1035
00:48:09,400 --> 00:48:11,960
Speaker 2: more odd Laws content, go to Bloomberg dot com slash

1036
00:48:11,960 --> 00:48:14,200
Speaker 2: odd Lots or of a daily newsletter on all of our.

1037
00:48:14,120 --> 00:48:16,000
Speaker 7: Episodes, and you can chat about all.

1038
00:48:15,880 --> 00:48:18,240
Speaker 2: Of these topics twenty four to seven in our discord

1039
00:48:18,600 --> 00:48:20,480
Speaker 2: Discord dot gg slash odd.

1040
00:48:20,280 --> 00:48:22,880
Speaker 4: Lots And if you enjoy odd Lots, if you like

1041
00:48:22,920 --> 00:48:27,840
Speaker 4: it when we talk about the casastrophy Casastrophe, then please

1042
00:48:27,960 --> 00:48:31,840
Speaker 4: leave a positive review on your favorite podcast platform. And remember,

1043
00:48:31,920 --> 00:48:34,400
Speaker 4: if you are a Bloomberg subscriber, you can listen to

1044
00:48:34,560 --> 00:48:37,600
Speaker 4: all of our episodes absolutely ad free. All you need

1045
00:48:37,640 --> 00:48:40,360
Speaker 4: to do is find the Bloomberg channel on Apple Podcasts

1046
00:48:40,400 --> 00:48:43,040
Speaker 4: and follow the instructions there. Thanks for listening.