1
00:00:00,160 --> 00:00:04,557
If I was critical of USV, I would be
critical of the fact that we don't spend

2
00:00:04,640 --> 00:00:08,237
enough time on the founders.
I got really lucky. I backed these two two

3
00:00:08,320 --> 00:00:10,477
founders, Ron Shriber and Jordan Levy.

4
00:00:10,560 --> 00:00:13,277
They were already on their second
company and I showed up at the first

5
00:00:13,360 --> 00:00:15,997
board meeting and I started telling them,
"Well, I think you should do this

6
00:00:16,080 --> 00:00:19,517
and I think you should do this."
And Jordan just like stopped me dead in my

7
00:00:19,600 --> 00:00:23,917
tracks and he said, "If you come up here
and you do every job in the company for

8
00:00:24,000 --> 00:00:27,117
a week, then we'll let you give
this advice, but not until you do that."

9
00:00:27,200 --> 00:00:30,477
>> An investment we kind of we made
recently in FOMO. We first met that

10
00:00:30,560 --> 00:00:34,717
company in September 2025. We decided
to pass. I remember sending them an email

11
00:00:34,800 --> 00:00:38,797
begging them to angel. I think I they
had offered and then I said no cuz I was

12
00:00:38,880 --> 00:00:41,677
a bigger check than I was willing
to write. It it's a hard thing in the

13
00:00:41,760 --> 00:00:44,877
venture business when you miss something
and then you realize after the fact that

14
00:00:44,960 --> 00:00:48,797
you made a mistake trying to get back
into the investment and the company

15
00:00:48,880 --> 00:00:52,397
tripled or quadrupled in size. when
we realized we had made the mistake and

16
00:00:52,480 --> 00:00:56,637
when we actually got to make the
investment being in venture capital for

17
00:00:56,720 --> 00:01:00,877
40 years is I'm used to the industry
working a certain way and when somebody

18
00:01:00,960 --> 00:01:05,037
comes in and disrupts it like Paul
Graham comes in and creates Y Combinator

19
00:01:05,120 --> 00:01:08,957
and you know that's a that was a massive
innovation in the venture capital

20
00:01:09,040 --> 00:01:12,317
industry if you're kind of stuck in your
ways you're going to look at that and

21
00:01:12,400 --> 00:01:16,717
say that's dumb. I grew up in a business
that was a lot more collaborative than

22
00:01:16,800 --> 00:01:20,557
the one we're in today. Now it's not
really like that anymore. And I think

23
00:01:20,640 --> 00:01:25,437
you got to scrap and claw your way into
rounds now. And you can't expect other

24
00:01:25,520 --> 00:01:29,757
firms to help you. I think you have
to treat every single situation kind of

25
00:01:29,840 --> 00:01:33,277
like a jump ball or a rebound.
And you're going to go get it.

26
00:01:33,360 --> 00:01:36,557
>> Yeah. >> Someone's going to get it,
so you're going to get it.

27
00:01:36,640 --> 00:01:39,760
>> Do you have enough material?

28
00:01:46,400 --> 00:01:49,277
So, Cinnamon sprung this on me, I don't

29
00:01:49,360 --> 00:01:53,757
know, 24 hours ago, and he said, "Let's
just chat with Fred." I said, "Sounds

30
00:01:53,840 --> 00:01:57,677
good. Where do I start?" I sat for myself
with myself for 10 minutes, and I

31
00:01:57,760 --> 00:02:01,597
thought, "Well, you know what?
I always find myself in conversations with

32
00:02:01,680 --> 00:02:06,717
friends talking about a topic. It could
be equity option pools for a new term

33
00:02:06,800 --> 00:02:10,317
sheet. could be topics around where
technology is going. And in the course

34
00:02:10,400 --> 00:02:15,277
of just researching that topic, there's
always an ABC blog post that comes up

35
00:02:15,360 --> 00:02:18,557
and and I'm like, well, I feel like
I'm always sending something. So, I went

36
00:02:18,640 --> 00:02:21,437
through my texts. I literally searched
ABC, looked for the last three things

37
00:02:21,520 --> 00:02:22,797
I've sent in the last week to somebody.

38
00:02:22,880 --> 00:02:27,037
I think the one I sent last week was I
was having a debate with Matt Mandel

39
00:02:27,120 --> 00:02:30,877
about you cited I think this you wrote
this post maybe 13 years ago. It was a

40
00:02:30,960 --> 00:02:33,757
quote of a Don Valentine 2 by two.

41
00:02:33,840 --> 00:02:36,157
>> Yeah. where he had he had difficult

42
00:02:36,240 --> 00:02:41,197
founders, easy founders, and then like
easy to work with, hard to work with,

43
00:02:41,280 --> 00:02:45,117
and then kind of like good businesses,
bad businesses. And the debate we were

44
00:02:45,200 --> 00:02:48,557
having was if you look at like
entrepreneurs today, it almost feels

45
00:02:48,640 --> 00:02:50,717
like tech has softened like it almost

46
00:02:50,800 --> 00:02:52,077
feels like everyone is nice.

47
00:02:52,160 --> 00:02:55,837
>> Yeah. This is you you're referring to a

48
00:02:55,920 --> 00:02:58,237
story that Roloff Boa told me. and

49
00:02:58,320 --> 00:03:00,877
Roloff who ran Sequoia for many years

50
00:03:00,960 --> 00:03:03,197
recently passed the baton there to the

51
00:03:03,280 --> 00:03:06,317
next generation. When he arrived at

52
00:03:06,400 --> 00:03:09,597
Seoia in the mid 2000s, so-called 20

53
00:03:09,680 --> 00:03:12,557
years ago, Don Valentine was still there

54
00:03:12,640 --> 00:03:16,397
and Don walked him into his office and

55
00:03:16,480 --> 00:03:19,277
showed him a 2x two matrix and on one

56
00:03:19,360 --> 00:03:23,517
side it was good founders, great
founders and then easy to get along

57
00:03:23,600 --> 00:03:27,357
with, hard to get along with. And then
he pointed to the upper quadrant which

58
00:03:27,440 --> 00:03:31,517
was great founders, hard to get along
with. And he said, "We make all our

59
00:03:31,600 --> 00:03:35,517
money here." >> Yes. Um,
and and Roloff told me that

60
00:03:35,600 --> 00:03:39,357
story, and I can't even remember why
Roloff told me that story, but I'm sure

61
00:03:39,440 --> 00:03:42,717
we were having a conversation about
working with challenging founders. I

62
00:03:42,800 --> 00:03:45,517
don't think it's really so much about

63
00:03:45,600 --> 00:03:48,397
nice or not nice. It's just that some

64
00:03:48,480 --> 00:03:51,837
founders are stubborn and they kind of

65
00:03:51,920 --> 00:03:53,197
want to do things the way they want to

66
00:03:53,280 --> 00:03:56,877
do do things and they're not inclined to

67
00:03:56,960 --> 00:04:02,077
explain themselves. and you just wake up
one day and they've pivoted the company

68
00:04:02,160 --> 00:04:05,677
and they didn't even bother to tell
you or whatever and you're like well you

69
00:04:05,760 --> 00:04:08,077
know we do own you know 10 or 15% of

70
00:04:08,160 --> 00:04:12,797
this company you know but I think Don's
point was that it's those founders who

71
00:04:12,880 --> 00:04:17,197
often you'll make the most money with
and just last week in preparation for

72
00:04:17,280 --> 00:04:21,437
our CEO summit which we're at right
now for two days we're going to be in a

73
00:04:21,520 --> 00:04:25,117
bunch of different sessions with with
our entire portfolio I was talking to

74
00:04:25,200 --> 00:04:28,637
Mark Pinkis who I'm going to interview
tomorrow And one of the three things

75
00:04:28,720 --> 00:04:33,037
that Mark and I are going to talk about
is what he thinks founder mode means.

76
00:04:33,120 --> 00:04:35,437
And in his mind, founder mode means when

77
00:04:35,520 --> 00:04:38,477
a founder can stay true to their

78
00:04:38,560 --> 00:04:43,757
instincts and not let their board push
them in a direction or not let their

79
00:04:43,840 --> 00:04:47,197
management team push them in a direction,
but they can will the company

80
00:04:47,280 --> 00:04:51,117
to do the thing that they want. And I
think that's kind of the same thing that

81
00:04:51,200 --> 00:04:56,077
Don was trying to to explain to Roloff,
which is that it's those sorts of people

82
00:04:56,160 --> 00:04:57,917
that you make all the money with.

83
00:04:58,000 --> 00:04:59,437
>> Do you think that still holds today?

84
00:04:59,520 --> 00:05:01,597
>> I do. I do. In fact,
I mean, if you look

85
00:05:01,680 --> 00:05:05,997
at the the biggest characters in tech,
the ones who are making all the news

86
00:05:06,080 --> 00:05:08,237
right now, they're all like that. I do I

87
00:05:08,320 --> 00:05:11,357
do think that great founders um have

88
00:05:11,440 --> 00:05:16,317
great conviction and great courage
and maybe are not all that great at

89
00:05:16,400 --> 00:05:19,757
communicating. they know where they
want to go and they they go there.

90
00:05:19,840 --> 00:05:23,037
>> Literally, I literally did an iMessage
search and I searched ABC. The second

91
00:05:23,120 --> 00:05:24,397
article I'd sent, this is like maybe

92
00:05:24,480 --> 00:05:27,437
eight days ago, was on types of

93
00:05:27,520 --> 00:05:30,157
investors. It was like career investors

94
00:05:30,240 --> 00:05:34,957
versus founder investors. It's another
decade old blog post, I think. And the

95
00:05:35,040 --> 00:05:38,397
gist of your point was that there
was more and more operator and founder

96
00:05:38,480 --> 00:05:39,917
investors. This is maybe 10 years ago.

97
00:05:40,000 --> 00:05:42,877
Despite that being true, you yourself
were more of a career investor, and

98
00:05:42,960 --> 00:05:43,997
there's still a lot of them out there.

99
00:05:44,080 --> 00:05:46,717
You're basically saying, you know,
either can work. The reason I I sent

100
00:05:46,800 --> 00:05:50,477
that post to someone was I think it
was in the context of myself. I was an

101
00:05:50,560 --> 00:05:54,637
engineer before I joined USV but uh I've
never been an entrepreneur and you know

102
00:05:54,720 --> 00:05:58,397
as I plan to continue doing venture
capital for a long time. The thing that

103
00:05:58,480 --> 00:06:01,677
I it's been on my mind a lot is okay
well what are the things you need to

104
00:06:01,760 --> 00:06:05,517
really work on if you're coming from an
investor angle and one of them is like

105
00:06:05,600 --> 00:06:07,997
having that natural both empathy for the

106
00:06:08,080 --> 00:06:12,717
founder but also uh ability to help
in some way. I I I don't think most

107
00:06:12,800 --> 00:06:15,197
founders necessarily expect to get
an immense amount of help from their

108
00:06:15,280 --> 00:06:18,237
investors, but it's nice to be able
to work with an entrepreneur deepway in

109
00:06:18,320 --> 00:06:22,237
sport. I think that comes very naturally
like when I watch when I observe Mike or

110
00:06:22,320 --> 00:06:25,277
or Jared, you know, work with an
entrepreneur, it comes from a very

111
00:06:25,360 --> 00:06:27,997
different angle and there's weight
to what they're saying. Well, over time, I

112
00:06:28,080 --> 00:06:31,517
imagine like you've been able
to cultivate weight and and good helpful

113
00:06:31,600 --> 00:06:34,477
advice just because you've been doing
it for so long. You've seen a lot. But

114
00:06:34,560 --> 00:06:38,397
early on, how did you think about that
relationship? It's a really hard thing

115
00:06:38,480 --> 00:06:42,237
to do young early in your career when
you're a young venture capitalist and

116
00:06:42,320 --> 00:06:46,237
you want to give a founder advice and they
look at you and they're like,"Well,

117
00:06:46,320 --> 00:06:49,997
what the hell do you know?" Right?
And that's where somebody like Mike or

118
00:06:50,080 --> 00:06:54,797
Jared, our partners who who have been
founders, um, have a lot of credibility

119
00:06:54,880 --> 00:06:59,197
because when they tell a founder, well,
you know, I had that problem at XYZ

120
00:06:59,280 --> 00:07:01,837
company and this is what we did.
You know, that's really useful for a

121
00:07:01,920 --> 00:07:05,757
founder. And if you start in the venture
business, I started in the venture

122
00:07:05,840 --> 00:07:08,877
business when I was 25. How old
were you when you showed up at USV?

123
00:07:08,960 --> 00:07:11,037
>> 26. >> There you go.
So, and by the way, I was

124
00:07:11,120 --> 00:07:14,237
a software engineer before I became a VC.
So, we have the exact same

125
00:07:14,320 --> 00:07:18,077
trajectory, more or less. Yeah, it's
really, really challenging. I think that

126
00:07:18,160 --> 00:07:22,077
you have to work a lot harder
to understand the business and gain the

127
00:07:22,160 --> 00:07:26,717
founders's trust. If you do the work and
the founder sees you doing the work, to

128
00:07:26,800 --> 00:07:30,557
get to know their team and to get to know
their business and understand their

129
00:07:30,640 --> 00:07:33,277
products, really use their products, you

130
00:07:33,360 --> 00:07:36,077
can earn their trust. But it doesn't it

131
00:07:36,160 --> 00:07:40,557
it's not a given particularly you know
early early on in your career. So I

132
00:07:40,640 --> 00:07:41,517
think you have to work harder.

133
00:07:41,600 --> 00:07:43,677
>> That makes sense. Is that is that
what you felt like you did early on?

134
00:07:43,760 --> 00:07:45,597
>> I got really lucky. I backed these two

135
00:07:45,680 --> 00:07:48,237
two founders Ron Shriber and Jordan Levy

136
00:07:48,320 --> 00:07:53,117
when they were uh when I was 29 and they
were maybe 35 or 36 and they were

137
00:07:53,200 --> 00:07:56,397
already on their second company. And I
showed up at the first board meeting and

138
00:07:56,480 --> 00:07:58,957
I started telling them, "Well,
I think you should do this and I think you

139
00:07:59,040 --> 00:08:02,557
should do this." And Jordan just like
stopped me dead in my tracks and he

140
00:08:02,640 --> 00:08:04,637
said, "Listen,

141
00:08:04,720 --> 00:08:07,357
if you come up here for a week, this was

142
00:08:07,440 --> 00:08:09,917
Buffalo, New York, and you do every job

143
00:08:10,000 --> 00:08:12,877
in the company for a week, then we'll

144
00:08:12,960 --> 00:08:17,117
let you give this advice, but not until
you do that." So, I got in a plane like,

145
00:08:17,200 --> 00:08:21,357
I don't know, three or four weeks later,
and I went up there and I I did customer

146
00:08:21,440 --> 00:08:24,717
support and I sat with next
to the engineers and I sat next to the

147
00:08:24,800 --> 00:08:28,877
marketing team and, you know, I I I
quote unquote did the work. And then

148
00:08:28,960 --> 00:08:32,557
after that they uh they were at least
willing to put up with me a little bit

149
00:08:32,640 --> 00:08:36,157
more. >> You know, as we've kind of worked
on a couple things together, you're not

150
00:08:36,240 --> 00:08:40,397
always when you're giving advice,
you know, looking to give the answer. But

151
00:08:40,480 --> 00:08:41,837
what one thing I find that you do really

152
00:08:41,920 --> 00:08:45,117
well is clarify what people are working

153
00:08:45,200 --> 00:08:50,317
on or maybe put it in terms that are are
very understandable. So I think one of

154
00:08:50,400 --> 00:08:52,877
my takeaways from observing that is,
you know, it doesn't always have to be the

155
00:08:52,960 --> 00:08:56,557
advice. It can always just be, you know,
if if you are doing the work, if you

156
00:08:56,640 --> 00:09:00,317
understand the business and and you're
just helping talk it through, I think

157
00:09:00,400 --> 00:09:01,837
you're going to land
in interesting places.

158
00:09:01,920 --> 00:09:04,397
>> I do think that I have a gift um which

159
00:09:04,480 --> 00:09:07,917
is I can simplify things. And one of the

160
00:09:08,000 --> 00:09:09,997
things I can help founders do is is

161
00:09:10,080 --> 00:09:14,317
craft sort of a a very simple story that

162
00:09:14,400 --> 00:09:17,997
they can tell when they go out and raise
money. Businesses are complicated and

163
00:09:18,080 --> 00:09:21,917
founders sometimes are caught up in the
weeds of their business and investors

164
00:09:22,000 --> 00:09:24,877
really need a simplified version. And so

165
00:09:24,960 --> 00:09:27,197
I'm pretty good at that. And and as part

166
00:09:27,280 --> 00:09:31,997
of that skill, I've also got quite
good at really trying to frame what the

167
00:09:32,080 --> 00:09:34,877
question is that we're trying to ask,
whether it's in a board meeting or just

168
00:09:34,960 --> 00:09:39,037
a conversation. Sometimes reframing
a little bit for the founder. Oh, are you

169
00:09:39,120 --> 00:09:42,397
trying to decide whether to do this
versus this? And the founder's like,

170
00:09:42,480 --> 00:09:45,757
yeah, that's what I've been telling you
for 20 minutes. And I'm like, well, I

171
00:09:45,840 --> 00:09:48,397
don't think we understood that, but now
that we understand that, we can, you

172
00:09:48,480 --> 00:09:51,757
know, give you that advice.
>> There's another thing that's been on my

173
00:09:51,840 --> 00:09:55,517
mind a lot, which is like this whole
concept of of conviction, right? I think

174
00:09:55,600 --> 00:09:58,957
it's it's it's core to the top
of the funnel, the work we do, which is

175
00:09:59,040 --> 00:10:02,317
actually making the investment. And you
said something to me the like maybe a

176
00:10:02,400 --> 00:10:05,757
half a year ago where I was saying,
well, you know, like I've always liked

177
00:10:05,840 --> 00:10:09,277
investing. I've always been like
a trader and investor. I've made angel

178
00:10:09,360 --> 00:10:12,317
investments. Um, but you know, that's
like my personal capital just it's it's

179
00:10:12,400 --> 00:10:16,477
it's different. I'm able to like I'm
more comfortable taking risk. And you

180
00:10:16,560 --> 00:10:18,957
basically said, well, you can't really
treat it that way. You have to kind of

181
00:10:19,040 --> 00:10:20,317
just think about it as one and the same.

182
00:10:20,400 --> 00:10:23,277
Do you ever feel like that was
something that you wrestled with?

183
00:10:23,360 --> 00:10:26,397
>> I did. It took me a while to think of

184
00:10:26,480 --> 00:10:29,437
the venture capital firm's money as my

185
00:10:29,520 --> 00:10:32,637
money and approach the investing that

186
00:10:32,720 --> 00:10:37,517
I'm thinking of the first firm I was at,
Uklid Partners, um, approach investing

187
00:10:37,600 --> 00:10:41,517
our firm's money the way I would invest
my own money because I was in the early

188
00:10:41,600 --> 00:10:44,877
on in my career was hung up with
convincing my partners to do something

189
00:10:44,960 --> 00:10:48,317
and you know would they like it
or would they not like it and I I kind of

190
00:10:48,400 --> 00:10:52,397
ultimately realized I have to like it
first and then I have to convince them

191
00:10:52,480 --> 00:10:55,757
to like it and that's what it is and I
can't really worry about whether they

192
00:10:55,840 --> 00:10:58,557
like it or they don't like it. I just
have to decide if I do and then I have

193
00:10:58,640 --> 00:11:03,197
to convince them. And that's the point
about conviction. And I think you can

194
00:11:03,280 --> 00:11:07,677
over complicate the investment business
if you start thinking about I'm a

195
00:11:07,760 --> 00:11:09,917
steward of someone else's capital or I

196
00:11:10,000 --> 00:11:12,477
have partners. I need them to agree with

197
00:11:12,560 --> 00:11:17,997
me. And it's really about do I think
this is a great investment. If I think

198
00:11:18,080 --> 00:11:20,557
it's a great investment, then I'm
going to figure out how to make that

199
00:11:20,640 --> 00:11:24,797
investment. I'm going to figure out how
to get the people I need uh to go along

200
00:11:24,880 --> 00:11:28,877
with me. And that's essentially
the advice I was giving you at the time.

201
00:11:28,960 --> 00:11:30,397
>> There's an investment
we kind of we made

202
00:11:30,480 --> 00:11:32,957
recently in FOMO. And we first met that

203
00:11:33,040 --> 00:11:37,677
company in September 2025 and we had
a discussion around around the deal and uh

204
00:11:37,760 --> 00:11:41,197
it was moving quickly and we decided
to pass because I had met the company and

205
00:11:41,280 --> 00:11:44,797
around that same week I started using
the app. After that, I kept using the

206
00:11:44,880 --> 00:11:47,757
app and I kept using it more and more
and then I was like, "Wait a second. I

207
00:11:47,840 --> 00:11:51,437
think there's really something here that I
missed." And so I I sent uh Paul and

208
00:11:51,520 --> 00:11:54,237
say an email. At this point, we had
just passed as a firm and I known they'd

209
00:11:54,320 --> 00:11:56,477
already chosen another investor.
I didn't know which, but I knew they'

210
00:11:56,560 --> 00:11:59,277
already got the deal done. So I was like,
okay, we missed our chance, but

211
00:11:59,360 --> 00:12:01,997
you know, I should probably angel invest
in this thing. I I love this app. So I

212
00:12:02,080 --> 00:12:05,277
remember sending them an email begging
them to angel invest. Cuz I I think I

213
00:12:05,360 --> 00:12:08,397
they had offered and then I said no
cuz I it was a bigger check than I was

214
00:12:08,480 --> 00:12:11,037
willing to write. But then I was using
that more and I started begging and then

215
00:12:11,120 --> 00:12:15,437
it struck me as I was middle of this beg
email. I was like, wait a second like we

216
00:12:15,520 --> 00:12:18,237
should be doing from USB. like if if I'm
going to get here on an angel investment

217
00:12:18,320 --> 00:12:22,157
if I have this conviction I it took me
a second to like reframe I should treat it

218
00:12:22,240 --> 00:12:27,197
the same way at all times right and and
then so I sent a new email that same day

219
00:12:27,280 --> 00:12:30,397
saying I think there's something here
I think we missed something it's a hard

220
00:12:30,480 --> 00:12:35,357
thing in the venture business when you
miss something and either you pass on it

221
00:12:35,440 --> 00:12:38,877
uh and then you realize after the fact
that you made a mistake or you didn't

222
00:12:38,960 --> 00:12:41,837
even pay enough attention to it and you
missed it and then you realize you made

223
00:12:41,920 --> 00:12:44,797
a mistake and then trying to get back

224
00:12:44,880 --> 00:12:46,237
into the investment

225
00:12:46,320 --> 00:12:50,317
is is very hard. You and I pursued that

226
00:12:50,400 --> 00:12:54,557
opportunity for probably 6 months before
we got the opportunity to invest in it.

227
00:12:54,640 --> 00:12:59,197
And the company, I think, tripled
or quadrupled in size between when we

228
00:12:59,280 --> 00:13:02,157
realized we had made the mistake
and when we actually got to make the

229
00:13:02,240 --> 00:13:03,917
investment, but I'm glad we did. And and

230
00:13:04,000 --> 00:13:06,797
I think, you know, credit to you for

231
00:13:06,880 --> 00:13:10,877
realizing that we had messed up and made

232
00:13:10,960 --> 00:13:13,997
the wrong call. I think sometimes it's

233
00:13:14,080 --> 00:13:18,157
like human emotion like, "Oh, I screwed
that one up. I'll go find the next one."

234
00:13:18,240 --> 00:13:22,877
And like you said, like maybe I could
just, you know, get into the next round

235
00:13:22,960 --> 00:13:24,717
or whatever it is. And and we did that.

236
00:13:24,800 --> 00:13:27,517
And I think we are very
pleased that we did that.

237
00:13:27,600 --> 00:13:30,877
>> Yeah, >> definitely. It was also uh

238
00:13:30,960 --> 00:13:35,517
>> a lot of fun uh to try to chase
something down. We I don't know if we we

239
00:13:35,600 --> 00:13:38,637
take that motion with as
as many investments at USV.

240
00:13:38,720 --> 00:13:41,597
>> Well, it's kind of interesting how we
did it, right? Like we went and saw them

241
00:13:41,680 --> 00:13:46,077
and we explained to them why we were
convinced that we had made a mistake and

242
00:13:46,160 --> 00:13:47,677
that we really wanted to get in and

243
00:13:47,760 --> 00:13:51,277
spent the time with them to con at least

244
00:13:51,360 --> 00:13:53,917
convince them that we really did buy

245
00:13:54,000 --> 00:13:58,317
into their vision. We went to their
holiday party. You know, I think that

246
00:13:58,400 --> 00:14:03,117
was a was a lot of fun and you know,
I think they really appreciated that and

247
00:14:03,200 --> 00:14:04,717
you know, it was like a lot of that

248
00:14:04,800 --> 00:14:07,277
right over 3 months, 4 months, 5 months,

249
00:14:07,360 --> 00:14:11,997
just really, you know, continuing
to build the relationship to the point

250
00:14:12,080 --> 00:14:13,917
where, you know, we made a couple of

251
00:14:14,000 --> 00:14:15,517
offers that that they said no thank you

252
00:14:15,600 --> 00:14:19,277
to and didn't get upset, you know, was

253
00:14:19,360 --> 00:14:22,317
were were totally, you know, gracious in

254
00:14:22,400 --> 00:14:24,717
in those those interactions when they

255
00:14:24,800 --> 00:14:29,757
said no thank you. And you know,
over time, you know, it worked. So I I

256
00:14:29,840 --> 00:14:33,917
there's a real lesson there. Like I
I don't think that I was good at that

257
00:14:34,000 --> 00:14:37,597
early in my career cuz I'm pretty
competitive and you know, when I lose

258
00:14:37,680 --> 00:14:41,197
something, I kind of get a little
annoyed about it. But I think in that

259
00:14:41,280 --> 00:14:45,437
moment, we really did a I think
a pretty good job of chasing it down.

260
00:14:45,520 --> 00:14:49,837
>> Definitely. The the competitive comment
reminds me of another thing I've been

261
00:14:49,920 --> 00:14:53,597
thinking about which is it it feels
like and I could be wrong about this. A

262
00:14:53,680 --> 00:14:57,677
venture at the early stages has become
less collaborative than it once was. You

263
00:14:57,760 --> 00:15:00,317
talk we often talk about how we made
this investment unis swap that was a

264
00:15:00,400 --> 00:15:03,277
great investment for the firm but we
were a minority investor there. A lot of

265
00:15:03,360 --> 00:15:05,837
people came into that deal. I don't see
as many deals in our portfolio that have

266
00:15:05,920 --> 00:15:09,517
that structure. And I don't know if that
if you think that should shift how one

267
00:15:09,600 --> 00:15:14,157
pursues venture capital. Well, you know,
I grew up in a business that was a lot

268
00:15:14,240 --> 00:15:18,397
more collaborative than the one we're
in today. And so, my natural instinct has

269
00:15:18,480 --> 00:15:23,037
always been syndicate investments,
share investments, and then the venture

270
00:15:23,120 --> 00:15:24,477
capitalists that I share investments

271
00:15:24,560 --> 00:15:26,397
with will share them back with me. And

272
00:15:26,480 --> 00:15:30,237
that was the business that existed for,

273
00:15:30,320 --> 00:15:34,397
you know, most of the last 40 years.
But now it's not really like that anymore.

274
00:15:34,480 --> 00:15:36,157
You got to kind of scrap and claw your

275
00:15:36,240 --> 00:15:38,397
way into rounds now. And you can't

276
00:15:38,480 --> 00:15:41,517
expect other firms to help you. one, I

277
00:15:41,600 --> 00:15:44,317
don't think they can. Uh it's not like

278
00:15:44,400 --> 00:15:49,197
um a founder is going to, you know,
give an allocation to another venture firm

279
00:15:49,280 --> 00:15:51,837
because some other venture firm tells
them to. They're not going to do that.

280
00:15:51,920 --> 00:15:56,077
They're going to allocate it how they
want it. Um but it's also the case that

281
00:15:56,160 --> 00:16:00,237
venture firms are getting larger
and larger and larger and they just have

282
00:16:00,320 --> 00:16:03,517
their own needs to put large amounts
of money to work and so the business has

283
00:16:03,600 --> 00:16:06,157
just gotten much much harder in that

284
00:16:06,240 --> 00:16:09,757
way. So, you just have to I think you

285
00:16:09,840 --> 00:16:12,797
have to treat every single situation

286
00:16:12,880 --> 00:16:16,957
kind of like a jump ball or a rebound
and you're going to go get it,

287
00:16:17,040 --> 00:16:19,517
>> right? You know, cuz someone's going to

288
00:16:19,600 --> 00:16:24,077
get it. So, you're going to get it.
You have to win with the founder and they

289
00:16:24,160 --> 00:16:28,557
have or founders and they have to decide
that they want you in the investment. We

290
00:16:28,640 --> 00:16:30,877
recently made this investment in news

291
00:16:30,960 --> 00:16:32,717
research, the people who make Hermes.

292
00:16:32,800 --> 00:16:36,157
And we ended up not being able to lead

293
00:16:36,240 --> 00:16:38,477
the most recent round they did because

294
00:16:38,560 --> 00:16:43,517
they were raising so much money that we
just didn't have we couldn't write a big

295
00:16:43,600 --> 00:16:44,957
enough check to lead the round.

296
00:16:45,040 --> 00:16:47,197
>> U we made them an offer to lead a much

297
00:16:47,280 --> 00:16:49,437
smaller round. They said no thank you.

298
00:16:49,520 --> 00:16:52,157
Um and then so then we quickly went into

299
00:16:52,240 --> 00:16:56,957
we got to get an allocation. And so
every chance we got, we introduced them

300
00:16:57,040 --> 00:17:00,717
to other investors. We introduced them
to some corporations that we thought

301
00:17:00,800 --> 00:17:04,957
would be good partners to them. And
ultimately when the round came together,

302
00:17:05,040 --> 00:17:07,277
we got an allocation. So that's the way

303
00:17:07,360 --> 00:17:08,957
the game is played now. And you just got

304
00:17:09,040 --> 00:17:12,397
to you got to prove when the round is

305
00:17:12,480 --> 00:17:16,077
going on that they want you in on their

306
00:17:16,160 --> 00:17:20,957
cap table and they want you around the
company because you're going to increase

307
00:17:21,040 --> 00:17:25,837
their chance of success. And if you can
do that, then you know more times than

308
00:17:25,920 --> 00:17:29,517
not you can get an allocation. So it's
way way more competitive now. So let's

309
00:17:29,600 --> 00:17:33,117
let's talk a little bit about you.
So you um have made the decision to stick

310
00:17:33,200 --> 00:17:39,037
around at USV.
>> I basically three three years nine

311
00:17:39,120 --> 00:17:43,437
months. Yeah. >> And you
came into USB I guess 26 as an

312
00:17:43,520 --> 00:17:45,997
analyst and as part of our standard

313
00:17:46,080 --> 00:17:48,477
analyst program which is a three-year

314
00:17:48,560 --> 00:17:52,637
program. And at the end of that program,

315
00:17:52,720 --> 00:17:55,597
we essentially said, you know, do you

316
00:17:55,680 --> 00:18:00,717
want to stick around and grow into
a partner at USV? And you thought about

317
00:18:00,800 --> 00:18:02,637
that and ultimately
concluded that you did.

318
00:18:02,720 --> 00:18:07,117
>> Yeah. Well, was it's a funny exercise
actually because um you know the USV

319
00:18:07,200 --> 00:18:10,717
program is is known to be a fixedterm
program and it's advertised that way up

320
00:18:10,800 --> 00:18:14,877
front. So I had no qualms about that or
hesitations and and one nice thing about

321
00:18:14,960 --> 00:18:18,717
that is everyone comes in very aligned
you know about you hit the 6 month out

322
00:18:18,800 --> 00:18:21,597
mark or 9month out mark everyone kind of
is talking to you about like what do you

323
00:18:21,680 --> 00:18:25,197
think about next how can we help you
and because of that process I actually got

324
00:18:25,280 --> 00:18:28,637
to go through like a like a very
interesting exploration I I talked to I

325
00:18:28,720 --> 00:18:31,677
think maybe you know 50 other venture
capital firms over the course of 6

326
00:18:31,760 --> 00:18:35,597
months got to really force myself to like
evaluate you know what is out there

327
00:18:35,680 --> 00:18:38,877
and every shape and size of firm I got
to see I got to see how they think I got

328
00:18:38,960 --> 00:18:42,717
to see how they talk and I I learned
a ton. I mean, it just I think I came back

329
00:18:42,800 --> 00:18:45,437
from that process saying, "Well, here's
what I like. Here's what I don't like."

330
00:18:45,520 --> 00:18:48,637
A couple things, you know, stood out is,
you know, a lot of firms have fantastic

331
00:18:48,720 --> 00:18:51,837
investors. Not all of them are
necessarily people you you want to work

332
00:18:51,920 --> 00:18:53,277
with in work with day in day out.

333
00:18:53,360 --> 00:18:57,517
Another thing is a lot of teams are more
collection collections of individuals

334
00:18:57,600 --> 00:19:01,517
than real collaborative team environments,
which is uh not to say one

335
00:19:01,600 --> 00:19:04,877
is better or worse than the other.
You know, there's firms out there that are

336
00:19:04,960 --> 00:19:08,877
fantastic and they're they run very
much as solo entities and there's al also

337
00:19:08,960 --> 00:19:11,757
firms have great returns in the other
direction. So, I don't think one is

338
00:19:11,840 --> 00:19:14,317
better than the other, but I think it
it took me it was a it was a productive

339
00:19:14,400 --> 00:19:16,717
exploration to figure out what I like
and what I don't like. And yeah, I'm

340
00:19:16,800 --> 00:19:18,317
really excited by the opportunity. When

341
00:19:18,400 --> 00:19:21,757
you first came into USV, you really

342
00:19:21,840 --> 00:19:24,317
understood crypto, blockchain, stable

343
00:19:24,400 --> 00:19:29,117
coins, DeFi, that whole world. And we

344
00:19:29,200 --> 00:19:31,357
really leaned on you to be sort of a

345
00:19:31,440 --> 00:19:33,277
subject matter expert there in the first

346
00:19:33,360 --> 00:19:37,277
few years. Um, but we also encouraged

347
00:19:37,360 --> 00:19:39,997
you to spread your wings so that you

348
00:19:40,080 --> 00:19:42,317
could be a generalist because at at at

349
00:19:42,400 --> 00:19:44,877
its core, USV is a generalist firm. What

350
00:19:44,960 --> 00:19:47,837
would you say are the areas that you

351
00:19:47,920 --> 00:19:52,557
were able to come up to a level
of expertise in over the last couple years

352
00:19:52,640 --> 00:19:56,237
to kind of meet meet you where you
already were with crypto? It's a great

353
00:19:56,320 --> 00:19:59,997
question. The areas that come to mind
the most are honestly areas where USV

354
00:20:00,080 --> 00:20:03,197
spends time because that's where I've
been able to learn which is one is

355
00:20:03,280 --> 00:20:07,357
energy. I feel like like my my
understanding of of how a lot of our

356
00:20:07,440 --> 00:20:10,317
portfolio companies work and how that
space works has leveled up an immense

357
00:20:10,400 --> 00:20:14,557
amount while being at USV. I was an
engineer before USV still like to play a

358
00:20:14,640 --> 00:20:18,317
lot with with with uh all the tools
out there. So I think I gravitate a lot

359
00:20:18,400 --> 00:20:20,557
towards like developerf facing AI tools.

360
00:20:20,640 --> 00:20:25,037
And the third area which is a kind of a
newer exploration for USV is is biology.

361
00:20:25,120 --> 00:20:28,077
I've increasingly spent a lot of time
there. That one's a really fun one

362
00:20:28,160 --> 00:20:32,717
because um I think we all are you know
convinced and convicted that it's an

363
00:20:32,800 --> 00:20:35,117
important place for us to spend time
but we haven't made that many investments

364
00:20:35,200 --> 00:20:37,757
there. But you really pushed an insight

365
00:20:37,840 --> 00:20:40,797
on us that has been helpful, which is

366
00:20:40,880 --> 00:20:44,797
that biology is really moving from like

367
00:20:44,880 --> 00:20:47,437
a wet lab type of process where

368
00:20:47,520 --> 00:20:50,397
innovation happens in in a traditional

369
00:20:50,480 --> 00:20:53,437
lab to one where it's really more like

370
00:20:53,520 --> 00:20:57,277
software. There's large models and bi

371
00:20:57,360 --> 00:21:00,557
biology is effectively
becoming more programmable. Yep.

372
00:21:00,640 --> 00:21:05,437
>> And so programmable biology would be
maybe the term for the things that we're

373
00:21:05,520 --> 00:21:07,677
most interested in. >> Yeah,
I think that's right. It's

374
00:21:07,760 --> 00:21:10,717
becoming much more of an engineering
discipline. And I think we're seeing

375
00:21:10,800 --> 00:21:13,837
that in in a lot of the entrepreneurs
we're meeting. We're meeting a lot of

376
00:21:13,920 --> 00:21:16,957
entrepreneurs who don't necessarily come
from traditional bio backgrounds. And so

377
00:21:17,040 --> 00:21:19,997
I I think it's really exciting. I think
it it in many ways it open up it opens

378
00:21:20,080 --> 00:21:23,197
up the domain. one, we're just going
to see way more top off ofunnel drug

379
00:21:23,280 --> 00:21:26,637
exploration as we're moving in into more
of an engineering discipline, but also

380
00:21:26,720 --> 00:21:30,797
just like way more I think entrepreneurs
will take a stab at building because I

381
00:21:30,880 --> 00:21:33,757
think it'll be more approachable to them.
Um, so I'm really excited about

382
00:21:33,840 --> 00:21:37,437
that. For me, you know, I came in with a
lot of knowledge on crypto and financial

383
00:21:37,520 --> 00:21:40,797
services. Took me a while to build up
that knowledge and, you know, spreading

384
00:21:40,880 --> 00:21:44,397
my wings into other areas I think
will be a long exploration and a long

385
00:21:44,480 --> 00:21:49,037
journey. I'm curious how you have
handled that over the years. Well, the I

386
00:21:49,120 --> 00:21:50,957
learn the most from making investments.

387
00:21:51,040 --> 00:21:53,677
I don't make investments to learn.

388
00:21:53,760 --> 00:21:55,757
>> Um I think that's a bad reason to make

389
00:21:55,840 --> 00:22:00,717
an investment. But if we make
an investment in a company in a relatively

390
00:22:00,800 --> 00:22:02,637
new area for us, like when when I

391
00:22:02,720 --> 00:22:05,917
invested in Coinbase back in 2012 or

392
00:22:06,000 --> 00:22:08,077
whenever it was, you know, that was

393
00:22:08,160 --> 00:22:11,117
>> basically introductory class in crypto

394
00:22:11,200 --> 00:22:13,677
for me. and you know at at a time when

395
00:22:13,760 --> 00:22:18,237
the industry was still pretty nent.
And so I I got to know a lot of the people

396
00:22:18,320 --> 00:22:20,637
who were sort of the the leading

397
00:22:20,720 --> 00:22:23,277
entrepreneurs in the sector and all of

398
00:22:23,360 --> 00:22:25,917
the the key sort of technologies and

399
00:22:26,000 --> 00:22:29,197
market opportunities. And so that was

400
00:22:29,280 --> 00:22:31,757
very helpful to me. And so like if we

401
00:22:31,840 --> 00:22:34,077
wanted to get deeper in programmable

402
00:22:34,160 --> 00:22:37,437
biology, I think one thing that I feel

403
00:22:37,520 --> 00:22:41,597
like we need to do is to to make
some investments and then that'll just

404
00:22:41,680 --> 00:22:45,277
naturally lead to more or more knowledge.
We'll just get we'll be we'll

405
00:22:45,360 --> 00:22:49,357
just we'll just start to learn more
by working with those companies and those

406
00:22:49,440 --> 00:22:53,917
founders. I guess there's sort of like
a you know a bit of like a catch 22 which

407
00:22:54,000 --> 00:22:58,237
is chicken and egg a bit when there's
areas that you know you knew when you

408
00:22:58,320 --> 00:23:01,597
made for example like some of your early
investments in in energy. Is there like

409
00:23:01,680 --> 00:23:03,917
an extra leap of faith there where
you're like I don't necessarily

410
00:23:04,000 --> 00:23:06,717
understand all the aspects of this
domain but I'm I'm willing to take I'm

411
00:23:06,800 --> 00:23:09,197
willing to like uh you know be
comfortable with those unknowns.

412
00:23:09,280 --> 00:23:13,197
>> Yeah I I I I think
sometimes it's better

413
00:23:13,280 --> 00:23:17,917
when you know a little bit less because
you can take some leaps of faith. Like

414
00:23:18,000 --> 00:23:21,837
if you think about radiant which is a
nuclear reactor company that you know I

415
00:23:21,920 --> 00:23:25,357
led our I led the first round in that
company the first venture round in that

416
00:23:25,440 --> 00:23:28,797
company and the idea was to make really

417
00:23:28,880 --> 00:23:33,437
really small nuclear reactors 1 megawatt
nuclear reactors and to massproduce them

418
00:23:33,520 --> 00:23:38,157
literally like make a factory that's
just spitting out nuclear reactors and

419
00:23:38,240 --> 00:23:39,997
you know if you need 5 megawws well then

420
00:23:40,080 --> 00:23:42,877
you just stack five 1 megawatt reactors

421
00:23:42,960 --> 00:23:46,077
and and it was just a a very radical

422
00:23:46,160 --> 00:23:51,357
idea um And you know when I talk to a
lot of sort of people who had spent 20

423
00:23:51,440 --> 00:23:54,797
30 years in the nuclear industry they're
like that's but that's crazy. That's not

424
00:23:54,880 --> 00:23:59,437
how reactors are built. You know we
reactors are these huge construction

425
00:23:59,520 --> 00:24:01,677
projects that last you know a decade

426
00:24:01,760 --> 00:24:04,637
long and they're regulated and it's it's

427
00:24:04,720 --> 00:24:08,157
a very like this is just crazy idea.

428
00:24:08,240 --> 00:24:12,637
But, you know, my my career is a career
where I watched, you know, computing go

429
00:24:12,720 --> 00:24:15,037
from these huge mainframes down to like

430
00:24:15,120 --> 00:24:17,837
phones in our pocket or, you know,

431
00:24:17,920 --> 00:24:20,157
networks go from super large networks

432
00:24:20,240 --> 00:24:22,717
to, you know, super micro networks. And

433
00:24:22,800 --> 00:24:25,437
I just seen again and again and again

434
00:24:25,520 --> 00:24:27,757
that these large centralized systems

435
00:24:27,840 --> 00:24:30,797
eventually get broken down into really,

436
00:24:30,880 --> 00:24:34,317
really small, much less centralized

437
00:24:34,400 --> 00:24:38,077
systems. And I just thought, well, why

438
00:24:38,160 --> 00:24:39,917
wouldn't that also be true of nuclear

439
00:24:40,000 --> 00:24:42,797
reactors? And so I was drawn to this

440
00:24:42,880 --> 00:24:45,997
idea that I think maybe had I known more

441
00:24:46,080 --> 00:24:48,637
about nuclear energy, I would have

442
00:24:48,720 --> 00:24:50,237
talked myself out of. So I don't think

443
00:24:50,320 --> 00:24:53,837
it's necessarily a bad idea to come into

444
00:24:53,920 --> 00:24:57,917
a sector a little um naive. I think too

445
00:24:58,000 --> 00:25:01,597
naive is a bad idea.
>> Um but a little naive sometimes I think

446
00:25:01,680 --> 00:25:03,117
can be a good idea. And I think some of

447
00:25:03,200 --> 00:25:06,797
the computer scientists are coming into

448
00:25:06,880 --> 00:25:09,837
biology now who are not necessarily

449
00:25:09,920 --> 00:25:13,757
biology majors or biology PhDs but are

450
00:25:13,840 --> 00:25:16,477
bringing a knowledge of, you know,

451
00:25:16,560 --> 00:25:21,117
artificial intelligence and how you
create models and how you train models.

452
00:25:21,200 --> 00:25:24,557
They're going to bring some
breakthroughs into the area of biology

453
00:25:24,640 --> 00:25:29,597
that biologists couldn't have
gotten gotten on their own. Because

454
00:25:29,680 --> 00:25:32,957
when you worked in an industry for 20
or 30 years, you kind of get caught up in

455
00:25:33,040 --> 00:25:37,517
the way that the business works and you
it's hard for you to kind of imagine

456
00:25:37,600 --> 00:25:41,597
these radical things. I mean, venture
capital for example, you know, one of

457
00:25:41,680 --> 00:25:45,597
the disadvantages I have of being in
venture capital for 40 years is I'm used

458
00:25:45,680 --> 00:25:47,597
to the industry working a certain way.

459
00:25:47,680 --> 00:25:50,077
>> Yeah. And when somebody comes in and

460
00:25:50,160 --> 00:25:52,717
disrupts it like Paul Graham comes in

461
00:25:52,800 --> 00:25:57,517
and creates Y Combinator and you
know that's a that was a massive

462
00:25:57,600 --> 00:25:59,197
innovation in the venture capital

463
00:25:59,280 --> 00:26:01,917
industry that happened in like 2006 or

464
00:26:02,000 --> 00:26:06,397
2007. Um and it really kind of reshaped

465
00:26:06,480 --> 00:26:10,957
how the earliest stages of company
formation happens. You know if you're

466
00:26:11,040 --> 00:26:14,157
kind of stuck in your ways you're going
to look at that and say that's dumb.

467
00:26:14,240 --> 00:26:19,117
Like >> we don't want to touch that. Yeah.
and and a lot of people that were in the

468
00:26:19,200 --> 00:26:22,877
venture industry for a long time
looked at that and felt that way. But

469
00:26:22,960 --> 00:26:26,957
fortunately, you know, we didn't and we
embraced that. Um, and now, you know,

470
00:26:27,040 --> 00:26:29,437
there's a lot of new stuff happening in

471
00:26:29,520 --> 00:26:34,637
venture as well. It's important
to be um open-minded, I think.

472
00:26:34,720 --> 00:26:37,917
>> I mean, speaking of that, I would say
there's a couple major shifts that have

473
00:26:38,000 --> 00:26:42,557
happened in the last 10 years. Um I'd
say andre was probably the first to do

474
00:26:42,640 --> 00:26:46,717
it which was really pursue venture
capital as a business really scaling to

475
00:26:46,800 --> 00:26:50,317
touch every part part of the stack
of business and uh many people would say

476
00:26:50,400 --> 00:26:54,557
that you know that trend of larger
firms more capital has created sort of a

477
00:26:54,640 --> 00:26:59,197
bifurcation where it's really hard
to play in the middle in maybe the middle

478
00:26:59,280 --> 00:27:02,637
looks like not to put a number on it
but like you know you're not a small fund

479
00:27:02,720 --> 00:27:05,597
but you're not a mega fund and the mega
fund can kind of throw their weight

480
00:27:05,680 --> 00:27:09,997
around in in unique ways and small funds
can stay small and and USV you we're

481
00:27:10,080 --> 00:27:13,277
we're, you know, gravitating not exactly
to the middle. I'd say we're still on

482
00:27:13,360 --> 00:27:15,677
the smaller side for a lot of the
industry, but I'm curious kind of how

483
00:27:15,760 --> 00:27:18,557
you you think about when these
innovations come. You're trying to stay

484
00:27:18,640 --> 00:27:21,517
open-minded. Uh but you also don't
want to play everyone else's game.

485
00:27:21,600 --> 00:27:24,157
>> The institutionalization of venture

486
00:27:24,240 --> 00:27:27,597
capital that Andre Horowitz, you know,

487
00:27:27,680 --> 00:27:30,797
really led um is a fundamental change to

488
00:27:30,880 --> 00:27:33,357
the business and one that I think, you

489
00:27:33,440 --> 00:27:36,397
know, is going to be long lasting. And

490
00:27:36,480 --> 00:27:39,117
I'm not convinced that any

491
00:27:39,200 --> 00:27:42,237
there's any size venture firm that isn't

492
00:27:42,320 --> 00:27:44,637
at some level at risk because there's

493
00:27:44,720 --> 00:27:47,597
nothing stopping, you know, a firm that

494
00:27:47,680 --> 00:27:50,477
manages, you know, 50 billion or 100

495
00:27:50,560 --> 00:27:52,637
billion from making a seed investment.

496
00:27:52,720 --> 00:27:55,917
They they may talk themselves out of it
because they're like, well, how could a

497
00:27:56,000 --> 00:27:57,917
million dollar seed move the needle for

498
00:27:58,000 --> 00:28:01,277
us? But, you know, they could have scout

499
00:28:01,360 --> 00:28:03,277
programs. they could they could do seed

500
00:28:03,360 --> 00:28:08,317
in ways that maybe is a more scalable
way to do it. I think it's it's the new

501
00:28:08,400 --> 00:28:12,317
normal and that's how the venture
business is organized and the largest

502
00:28:12,400 --> 00:28:14,077
players and the largest platforms have

503
00:28:14,160 --> 00:28:17,917
the most market power and I think that

504
00:28:18,000 --> 00:28:22,477
for a smaller firm like USV I think we
have to be honest with ourselves that

505
00:28:22,560 --> 00:28:24,317
we're at a competitive disadvantage with

506
00:28:24,400 --> 00:28:26,797
them in many ways getting bigger

507
00:28:26,880 --> 00:28:30,717
probably is a logical move but it's hard

508
00:28:30,800 --> 00:28:34,157
to get a little bigger right cuz they

509
00:28:34,240 --> 00:28:39,037
raise capital at a scale that we don't
raise capital at and they, you know, are

510
00:28:39,120 --> 00:28:40,557
organized at a scale that we're not

511
00:28:40,640 --> 00:28:43,277
organized. So, you know, I don't think

512
00:28:43,360 --> 00:28:46,557
it would be simple for us to to to copy

513
00:28:46,640 --> 00:28:51,037
what they've done. And I'm I'm not sure
it would be healthy for us to do that,

514
00:28:51,120 --> 00:28:55,277
but I do think it's it it changes
the competitive dynamic in a way that

515
00:28:55,360 --> 00:28:58,877
doesn't necessarily favor us. And do
you think like if you were, you know, if

516
00:28:58,960 --> 00:29:02,237
you're USV or advising a manager and you
take this dynamic and you want to create

517
00:29:02,320 --> 00:29:05,437
an advantage position for yourself,
you know, one thing that we've found, I

518
00:29:05,520 --> 00:29:07,997
think at USV, we we find
ourselves going earlier.

519
00:29:08,080 --> 00:29:10,557
>> Right. >> Right. Do you think
that's that's probably the main answer?

520
00:29:10,640 --> 00:29:13,357
>> I think go earlier,

521
00:29:13,440 --> 00:29:16,397
domain knowledge, become a recognized a

522
00:29:16,480 --> 00:29:20,557
recognized expert in a domain. Um,
so that founders look at you and like, oh,

523
00:29:20,640 --> 00:29:24,637
I want that person. They really know
my market really well. And I think also

524
00:29:24,720 --> 00:29:27,197
maybe willing to deal with some things

525
00:29:27,280 --> 00:29:30,077
that you know maybe have a few

526
00:29:30,160 --> 00:29:34,237
complications on them. You know that could
be a regulatory complication or it

527
00:29:34,320 --> 00:29:36,637
could be you know a geographic

528
00:29:36,720 --> 00:29:41,357
complication or something where a big
platform might struggle a little bit uh

529
00:29:41,440 --> 00:29:44,237
cuz it's not you know
a strike right down the middle.

530
00:29:44,320 --> 00:29:48,877
>> Um so we you know you mentioned APD
>> uh they've rebranded to Panandina.

531
00:29:48,960 --> 00:29:50,877
>> Panandina they
changed their name. Well,

532
00:29:50,960 --> 00:29:56,637
the the P uh the P in Pandino APD was a
Torid Pandina. They dropped they dropped

533
00:29:56,720 --> 00:29:57,197
the A.

534
00:29:57,280 --> 00:29:59,837
>> So, it's Pandina.
Y even I can say that.

535
00:29:59,920 --> 00:30:03,757
Um so, that's a company in Colombia. I I

536
00:30:03,840 --> 00:30:07,117
think the there's plenty of large
platforms that would be comfortable

537
00:30:07,200 --> 00:30:11,357
investing in Colombia, but you know,
maybe not, you know, in the trenches

538
00:30:11,440 --> 00:30:13,677
kind of work in the market quite the way

539
00:30:13,760 --> 00:30:18,717
that you know, we we have now in that
company we have Audi, we have Somos, and

540
00:30:18,800 --> 00:30:22,477
we have Penadino. We had three companies
in Colombia. Now, you know, I don't

541
00:30:22,560 --> 00:30:24,157
think we're going to be opening a a a

542
00:30:24,240 --> 00:30:27,837
Medaheen office, but um you know, uh I

543
00:30:27,920 --> 00:30:32,957
think geography of that sort can be
a little bit of an advantage, too. The

544
00:30:33,040 --> 00:30:36,077
geography piece is really interesting
to me, you know, because we're going

545
00:30:36,160 --> 00:30:39,517
through a period of immense uh price

546
00:30:39,600 --> 00:30:42,557
inflation uh valid or not in the venture

547
00:30:42,640 --> 00:30:45,677
industry. And uh it's largely centered

548
00:30:45,760 --> 00:30:50,557
around very known pockets of talent. Um
I think you know people will cite if you

549
00:30:50,640 --> 00:30:52,077
just look year over year and just if you

550
00:30:52,160 --> 00:30:54,637
use YC as kind of a you know a market

551
00:30:54,720 --> 00:30:57,997
benchmark we're like 3x the seed prices

552
00:30:58,080 --> 00:31:02,957
we were about 2 and a half years ago
maybe even 4x now. And so the geography

553
00:31:03,040 --> 00:31:05,117
is interesting because there's I think
a lot of people are looking around

554
00:31:05,200 --> 00:31:09,997
thinking okay this math pencils if if
you know if your business is a trillion

555
00:31:10,080 --> 00:31:12,477
dollar outcome. >> I I I I
know where you're going with

556
00:31:12,560 --> 00:31:16,477
this. I'm going to interrupt you because
um I wanted to share with you this thing

557
00:31:16,560 --> 00:31:18,157
that I was thinking about. So, we have

558
00:31:18,240 --> 00:31:20,717
two companies that are existing USV

559
00:31:20,800 --> 00:31:23,917
portfolio companies that recently gave

560
00:31:24,000 --> 00:31:26,797
us updates and they have businesses that

561
00:31:26,880 --> 00:31:30,237
are growing very quickly now. And when I

562
00:31:30,320 --> 00:31:34,797
asked them, you know, what kinds
of margins were they making, you know, on

563
00:31:34,880 --> 00:31:38,797
like 30 million of revenue, they were
producing like $25 million worth of

564
00:31:38,880 --> 00:31:40,957
profit. Like I cannot remember a time

565
00:31:41,040 --> 00:31:42,557
when we had companies who were

566
00:31:42,640 --> 00:31:46,637
generating like EBIDA profits not not

567
00:31:46,720 --> 00:31:49,597
gross margins not even like contribution

568
00:31:49,680 --> 00:31:51,837
but like Ebida margins of like 80 or

569
00:31:51,920 --> 00:31:55,277
90%. Now part of it is that they don't

570
00:31:55,360 --> 00:31:57,197
have to build large software engineering

571
00:31:57,280 --> 00:31:59,837
teams anymore because they're using uh

572
00:31:59,920 --> 00:32:02,557
all the gentic coding techniques and you

573
00:32:02,640 --> 00:32:04,717
know marketing has become maybe a little

574
00:32:04,800 --> 00:32:09,597
bit more automated. you know,
they're not don't have large sales forces.

575
00:32:09,680 --> 00:32:11,357
Customer support now is largely

576
00:32:11,440 --> 00:32:13,437
automated. You know, if companies truly

577
00:32:13,520 --> 00:32:17,917
are a multiple of cash flow and we just

578
00:32:18,000 --> 00:32:21,037
saw an order of magnitude growth in how

579
00:32:21,120 --> 00:32:24,237
much margin a company can generate out

580
00:32:24,320 --> 00:32:26,477
of revenue, then maybe an order of

581
00:32:26,560 --> 00:32:29,357
magnitude bump in valuations

582
00:32:29,440 --> 00:32:31,757
does make sense. That that's very fair.

583
00:32:31,840 --> 00:32:35,437
the in one of those examples you gave,
you know, when you have these businesses

584
00:32:35,520 --> 00:32:39,277
that are cash flowing really well out
of the gates, uh, one of them is actually

585
00:32:39,360 --> 00:32:43,357
debating whether or not they should
even raise equity. Um, and I think

586
00:32:43,440 --> 00:32:47,597
historically, you know, we've had
an aversion to companies taking on debt

587
00:32:47,680 --> 00:32:51,517
unnecessarily. It usually, you know,
tends to be tends to cause a lot of pain

588
00:32:51,600 --> 00:32:55,917
down the line. Do you think more
entrepreneurs should or will uh pursue

589
00:32:56,000 --> 00:32:58,397
different equity structures?
>> I think it depends on what you're going

590
00:32:58,480 --> 00:33:01,357
to use the capital for. If you if you're

591
00:33:01,440 --> 00:33:07,117
not losing money and you don't have
any losses to fund, there's there's a

592
00:33:07,200 --> 00:33:08,157
question about why should I raise

593
00:33:08,240 --> 00:33:12,077
capital at all? But companies do want to

594
00:33:12,160 --> 00:33:14,237
um get rerated from a valuation spec

595
00:33:14,320 --> 00:33:17,837
perspective from time to time um because

596
00:33:17,920 --> 00:33:21,357
they might want to do acquisitions
or they might want to make some hires and

597
00:33:21,440 --> 00:33:25,917
they want to be able to, you know, have
valuable stock to recruit to attract the

598
00:33:26,000 --> 00:33:28,077
right people. To me, it's not so much a

599
00:33:28,160 --> 00:33:30,477
question of debt as coming up with new

600
00:33:30,560 --> 00:33:34,077
ways to raise capital that maybe are not

601
00:33:34,160 --> 00:33:38,317
as dilutive. So, you know, maybe more

602
00:33:38,400 --> 00:33:41,437
kind of secondary uh transactions, get

603
00:33:41,520 --> 00:33:46,157
new names on the cap table, get,
you know, a new valuation, but maybe just

604
00:33:46,240 --> 00:33:49,357
raising a bunch of money to just have
it go sit in a bank doesn't make a lot of

605
00:33:49,440 --> 00:33:51,357
sense. So, I I do think we're going to

606
00:33:51,440 --> 00:33:54,077
see that change over time if these

607
00:33:54,160 --> 00:33:58,637
companies can just produce, you know,
a lot of cash flow. Like, they're going to

608
00:33:58,720 --> 00:34:02,797
we're going to have to rethink financing
markets. I was talking to a friend of

609
00:34:02,880 --> 00:34:06,077
mine who's the chairman of a company
that's about to go public. I can't say

610
00:34:06,160 --> 00:34:10,157
the name of the company because it's
in its quiet period. He told me the entire

611
00:34:10,240 --> 00:34:12,877
IPO is going to be secondary. >> Wow.

612
00:34:12,960 --> 00:34:17,037
>> No primary whatsoever. Now, we played
around with direct listings for a while

613
00:34:17,120 --> 00:34:20,157
that were a version of that. But there
was a bunch of things about direct

614
00:34:20,240 --> 00:34:22,637
listings that I think have created some

615
00:34:22,720 --> 00:34:24,397
issues. So, this is going to be a

616
00:34:24,480 --> 00:34:28,637
traditional IPO, but all the capital is

617
00:34:28,720 --> 00:34:33,037
going to go to selling shareholders
who are going to be listed in the S1 as

618
00:34:33,120 --> 00:34:37,197
selling shareholders. And so,
it's it's just going to be pure secondary.

619
00:34:37,280 --> 00:34:41,277
You used to never see that, you know,
and still I think it's kind of rare, but

620
00:34:41,360 --> 00:34:44,877
maybe we'll see that more over time
as we get these kinds of companies.

621
00:34:44,960 --> 00:34:48,957
>> This has popped up another question
in my head which is uh you're on a number

622
00:34:49,040 --> 00:34:51,837
of public company boards and have been
and you also see in late stage private

623
00:34:51,920 --> 00:34:54,317
companies. There's an ongoing debate
really, you know, the last couple

624
00:34:54,400 --> 00:34:57,197
decades around the trend line in terms
of companies going public, staying

625
00:34:57,280 --> 00:34:59,757
private, right? >> Um there's
a side discussion probably to

626
00:34:59,840 --> 00:35:03,197
be had here around everything coming on
chain, but we don't we don't want to get

627
00:35:03,280 --> 00:35:06,477
there right away. question in my head
is where do where do you think the

628
00:35:06,560 --> 00:35:10,077
incentives are going over time?
You know, you have folks like Stripe who I

629
00:35:10,160 --> 00:35:13,277
don't really see them ever going
public or at least not anytime soon. And

630
00:35:13,360 --> 00:35:15,837
they've been able to demonstrate
that it's created a lot of advantage for

631
00:35:15,920 --> 00:35:18,797
them. And then also you have you have
companies like SpaceX coming out and and

632
00:35:18,880 --> 00:35:23,677
able to pay, you know, instantly 60
billion for an asset. Uh and in my

633
00:35:23,760 --> 00:35:25,997
estimation, a lot of that market power
came from the fact that they were public

634
00:35:26,080 --> 00:35:28,397
and had liquid liquid tradable shares.

635
00:35:28,480 --> 00:35:32,397
like the price you can wield for um
acquisitions probably shifts when people

636
00:35:32,480 --> 00:35:36,877
are getting liquid stock. I'm curious,
you know, in your view if you have if

637
00:35:36,960 --> 00:35:41,357
you care at all with the companies you
work with about whether they, you know,

638
00:35:41,440 --> 00:35:44,637
go public or not, when they're at when
they're ready for it. Um if you think

639
00:35:44,720 --> 00:35:48,077
it's healthier to be have more companies
in the public markets. You said an

640
00:35:48,160 --> 00:35:49,117
important thing when they're ready for

641
00:35:49,200 --> 00:35:53,437
it. And ready for it means that they can

642
00:35:53,520 --> 00:35:56,557
>> deal with the burdens of being a public

643
00:35:56,640 --> 00:35:59,517
company, which is um there's a bunch of

644
00:35:59,600 --> 00:36:01,837
legal and financial obligations that you

645
00:36:01,920 --> 00:36:04,877
have to your shareholders that require,

646
00:36:04,960 --> 00:36:09,437
you know, beefed up systems, you know,
certain kinds of people that in the

647
00:36:09,520 --> 00:36:14,477
accounting and legal department and just
a just a way of running the business.

648
00:36:14,560 --> 00:36:18,077
you know, a lot of the companies in our
portfolio who are here this these two

649
00:36:18,160 --> 00:36:21,677
days, they can't do that. But I think
when you get to that point, I think

650
00:36:21,760 --> 00:36:23,597
there's a real benefit to being public.

651
00:36:23,680 --> 00:36:28,397
There's downside, of course. Um,
there's shareholder lawsuits and there's

652
00:36:28,480 --> 00:36:31,917
regulation by the SEC and a lot
of things that, you know, some founders

653
00:36:32,000 --> 00:36:36,797
maybe don't want to deal with. But I
think the benefit of having a very broad

654
00:36:36,880 --> 00:36:42,077
and diverse shareholder base where
anybody could wake up today and say, "I

655
00:36:42,160 --> 00:36:46,077
want to buy SpaceX." and they could
go buy SpaceX. Um, in the same way that

656
00:36:46,160 --> 00:36:50,317
anybody who wanted to buy Hyperlid like
you did convinced us and we went and

657
00:36:50,400 --> 00:36:53,037
bought it, right? Boom. We were
a Hyperlid shareholder. We didn't have to

658
00:36:53,120 --> 00:36:57,357
convince anybody, right? We you don't
have to go to Elon Musk and say, "Let me

659
00:36:57,440 --> 00:37:00,957
buy some shares of SpaceX." No, you just
go to the public markets and buy them.

660
00:37:01,040 --> 00:37:04,957
>> I think that's really powerful and that
way anybody can be a shareholder in your

661
00:37:05,040 --> 00:37:07,117
company. >> I think that's better.

662
00:37:07,200 --> 00:37:11,517
>> I'm biased to think so as well. the you
know it's it's been interesting to see

663
00:37:11,600 --> 00:37:15,597
more and more businesses you know
we've gone we've seen every side of the

664
00:37:15,680 --> 00:37:19,917
experiment in crypto right where people
you know issued tokens which which um

665
00:37:20,000 --> 00:37:23,757
you know were meant to give ownership
to more participants in in a network and

666
00:37:23,840 --> 00:37:26,317
we've seen when that works well
and when doesn't work well and it'll be

667
00:37:26,400 --> 00:37:30,317
interesting to see you know like what
types of businesses are able to wield

668
00:37:30,400 --> 00:37:33,757
the advantage of hey I actually want
to bring more people around the table I

669
00:37:33,840 --> 00:37:36,957
don't know I think a lot of consumer
businesses I think will probably

670
00:37:37,040 --> 00:37:39,597
continue to do so >> I think I
think it's really important as

671
00:37:39,680 --> 00:37:43,917
a consumer business. It's a big branding
moment. It's a big coming out moment.

672
00:37:44,000 --> 00:37:48,477
And you got to have a really good reason
not to do to not do it. I mean, Stripe

673
00:37:48,560 --> 00:37:50,797
obviously has some good reason that they

674
00:37:50,880 --> 00:37:53,197
don't want to be public. I I can't

675
00:37:53,280 --> 00:37:58,637
fathom it, but I'm sure they must have
one. If I was the founder and CEO of a

676
00:37:58,720 --> 00:38:01,517
company, I think I would be inclined
to take it public. talking about market

677
00:38:01,600 --> 00:38:02,957
shifts a second ago. The other market

678
00:38:03,040 --> 00:38:06,957
shift I want to touch on was in my view

679
00:38:07,040 --> 00:38:10,077
the pace at which AI has changed and how

680
00:38:10,160 --> 00:38:14,077
much dynamism it's created in the market
has led to a market that I think always

681
00:38:14,160 --> 00:38:17,357
at the early stages of venture capital
majority of venture capital firms that I

682
00:38:17,440 --> 00:38:20,397
interact with have a lens on the
business on any business that is very

683
00:38:20,480 --> 00:38:24,717
very founder centric to the point where
you know when I speak to people about

684
00:38:24,800 --> 00:38:29,437
discussions like amongst firms inside
the firm a lot of air time is spent

685
00:38:29,520 --> 00:38:31,917
dissecting the qualities
of an entrepreneur

686
00:38:32,000 --> 00:38:34,237
It's not something we do a particularly

687
00:38:34,320 --> 00:38:36,797
a whole lot of at USV and we have a lot

688
00:38:36,880 --> 00:38:41,437
of maybe two decades of muscle memory
of really spending immense amounts of time

689
00:38:41,520 --> 00:38:45,597
dissecting businesses, markets, themes
and it's hard for me to say it's a shift

690
00:38:45,680 --> 00:38:48,877
because you probably have a better lens
on that whe that's changed over time but

691
00:38:48,960 --> 00:38:52,317
the three years that I was in venture
capital I felt like it was more

692
00:38:52,400 --> 00:38:55,677
founderentric than ever in terms of how
people approached making investments

693
00:38:55,760 --> 00:38:59,917
thinking about businesses. If I was
critical of USV, I would be critical of

694
00:39:00,000 --> 00:39:01,677
the fact that we don't spend enough time

695
00:39:01,760 --> 00:39:05,517
on the founders. I I actually think that

696
00:39:05,600 --> 00:39:08,797
the magic happens when you get the right

697
00:39:08,880 --> 00:39:11,357
market and the right founders.

698
00:39:11,440 --> 00:39:15,597
I've been in the right market with the
wrong founders, bad place to be. I've

699
00:39:15,680 --> 00:39:19,517
been in the wrong market with the right
founders. That's not a good place to be

700
00:39:19,600 --> 00:39:22,397
either. Some founders can pivot out of

701
00:39:22,480 --> 00:39:25,037
that into a good market and I have seen

702
00:39:25,120 --> 00:39:29,517
that. But many times they they just say,
"You know what? I'm done. This wasn't

703
00:39:29,600 --> 00:39:33,357
the right market. The Farcaster founders
would be an example of that." You know,

704
00:39:33,440 --> 00:39:37,197
I think Dan's a really high quality
founder, but he was working in a market

705
00:39:37,280 --> 00:39:40,237
there was no there there. And he's now

706
00:39:40,320 --> 00:39:45,197
working on Tempo, which maybe is going
to be a better platform for him. So, you

707
00:39:45,280 --> 00:39:49,677
really got to get both right. And when
you get the market right and the founder

708
00:39:49,760 --> 00:39:51,837
or founders right, that's where the

709
00:39:51,920 --> 00:39:54,157
biggest outcomes are. I think you're

710
00:39:54,240 --> 00:39:57,517
right. USV has traditionally been very

711
00:39:57,600 --> 00:39:59,997
good at getting to the right markets and

712
00:40:00,080 --> 00:40:01,917
also the right sides of the right

713
00:40:02,000 --> 00:40:04,877
markets. So, it's not just right to go

714
00:40:04,960 --> 00:40:07,677
into e-commerce, but the best businesses

715
00:40:07,760 --> 00:40:12,237
in e-commerce were marketplaces like
Shopify and Etsy. And so, not only did

716
00:40:12,320 --> 00:40:13,757
we get to e-commerce correctly, we got

717
00:40:13,840 --> 00:40:16,317
to the right side of e-commerce, but I

718
00:40:16,400 --> 00:40:20,237
think that we have

719
00:40:20,320 --> 00:40:23,917
maybe a little bit less great of a

720
00:40:24,000 --> 00:40:26,557
record on um getting to the right

721
00:40:26,640 --> 00:40:29,357
founders. Um, and I think some of it is

722
00:40:29,440 --> 00:40:33,757
that we for a long time would only back
one founder in a market. And I think

723
00:40:33,840 --> 00:40:35,517
that's not good. And we've made that

724
00:40:35,600 --> 00:40:38,237
change now. So that typically if we like

725
00:40:38,320 --> 00:40:43,117
a market we want to make five or 10 bets
that I think frees us up to maybe pick

726
00:40:43,200 --> 00:40:46,477
more founder teams and then you know
more likely that we'll get with the

727
00:40:46,560 --> 00:40:49,277
right one. Yep. >> On that theme,
another thing that's been

728
00:40:49,360 --> 00:40:53,197
in my head is USV sort of became
synonymous in an earlier era with the

729
00:40:53,280 --> 00:40:56,237
concept of network effects, which is sort
of an emerging property of internet

730
00:40:56,320 --> 00:41:00,797
businesses. And it's become
an organizing principle for for us. Not to

731
00:41:00,880 --> 00:41:04,797
say every business we invest in has
network effects. We often use that lens

732
00:41:04,880 --> 00:41:09,517
of how can this be really long-term
defensible and grow organically. Uh and

733
00:41:09,600 --> 00:41:12,557
network effects are a great way to have
both of those things. Eli Gil once made

734
00:41:12,640 --> 00:41:14,957
a comment in some post a long time ago
where he said there's three types of

735
00:41:15,040 --> 00:41:18,477
founders. I think he said like there's
two types who are basically you know

736
00:41:18,560 --> 00:41:22,557
like maybe like a um a missionary um you
know some other great type founder and

737
00:41:22,640 --> 00:41:25,197
then the third founder was a network
effect founder and I don't know if he

738
00:41:25,280 --> 00:41:29,357
meant that in a positive or negative way
but uh basically the implication is you

739
00:41:29,440 --> 00:41:33,677
know certain businesses can support
themselves or maybe you know the the

740
00:41:33,760 --> 00:41:36,957
profile of founders doesn't need
to be the same uh for a business

741
00:41:37,040 --> 00:41:38,797
>> well you don't need
like a hard charging

742
00:41:38,880 --> 00:41:43,837
enterprise salesoriented CEO to build
you know an internet marketplace so I

743
00:41:43,920 --> 00:41:46,877
think he's right about that and
the right kind of person to build an

744
00:41:46,960 --> 00:41:50,317
internet marketplace might be very
different than the hard charging

745
00:41:50,400 --> 00:41:53,037
enterprise sales type person. So I think

746
00:41:53,120 --> 00:41:55,197
that's true. I would challenge you to

747
00:41:55,280 --> 00:41:58,477
find um a great exit that we've had at

748
00:41:58,560 --> 00:42:01,517
USV that didn't have network effects.

749
00:42:01,600 --> 00:42:04,877
That doesn't mean that there aren't
great companies that don't have network

750
00:42:04,960 --> 00:42:10,077
effects, but I think most of our great
successes at USV have some form of

751
00:42:10,160 --> 00:42:13,517
network effects in them. >> I think
that's probably right. Uh, and I

752
00:42:13,600 --> 00:42:17,437
I honestly think they're often more
emergent than we realize. You know, we

753
00:42:17,520 --> 00:42:21,437
have businesses like even like
a Duolingo that looks like a single player

754
00:42:21,520 --> 00:42:24,957
app, but the network effect that ended
up being they have this immense learning

755
00:42:25,040 --> 00:42:28,637
loop that they're on. And I don't know if
that always presented itself on day one.

756
00:42:28,720 --> 00:42:31,197
>> Well, I mean at our CEO summit three or

757
00:42:31,280 --> 00:42:33,117
four years ago, someone asked Luis Vanon

758
00:42:33,200 --> 00:42:36,397
how much AI they use in the product and

759
00:42:36,480 --> 00:42:39,037
he said, "Well, really only one

760
00:42:39,120 --> 00:42:43,117
algorithm. We just look at what is what

761
00:42:43,200 --> 00:42:45,837
is the most likely thing to make the

762
00:42:45,920 --> 00:42:50,237
user get the next question and then
the next question and the next question

763
00:42:50,320 --> 00:42:54,397
right to retain them as long as possible
and whatever that is that's what we do

764
00:42:54,480 --> 00:42:58,397
and the point is the more users they
have using the product the more accurate

765
00:42:58,480 --> 00:43:02,397
they can be in that prediction so it
is it is ultimately a data network effect

766
00:43:02,480 --> 00:43:05,517
as Brad would say >> and and it's
it's likely I mean another

767
00:43:05,600 --> 00:43:08,557
thought on that theme is I've been
thinking okay well network effects are

768
00:43:08,640 --> 00:43:11,437
so powerful right we've seen You know,
Twitter's like a cockroach. You can't

769
00:43:11,520 --> 00:43:15,357
kill that thing, right? That also means
that as we build as we built out the

770
00:43:15,440 --> 00:43:18,317
early internet networks, it's hard for
new networks to form, right? They crowd

771
00:43:18,400 --> 00:43:22,797
out other networks because they're so
strong. And one thought is, okay, have

772
00:43:22,880 --> 00:43:26,637
we built all of the key networks is is
one thought. The other thing is maybe

773
00:43:26,720 --> 00:43:32,157
maybe every business is an AI business
and AI the fuel for AI is data. So it's

774
00:43:32,240 --> 00:43:35,517
just a new form of network effect,
which is every business will have either a

775
00:43:35,600 --> 00:43:37,277
strong or weak data network effect.

776
00:43:37,360 --> 00:43:40,877
>> I don't think so. Because if you look
at if you look outside of AI, what are the

777
00:43:40,960 --> 00:43:44,077
some of the most exciting companies
that have kind of broken out over the last

778
00:43:44,160 --> 00:43:48,957
two or three years? Calcy and Poly
Market network effect businesses. Any

779
00:43:49,040 --> 00:43:52,237
any financial marketplace is a network
effect business. It's a liquidity

780
00:43:52,320 --> 00:43:54,477
business, you know, and so those are

781
00:43:54,560 --> 00:43:56,477
network effect businesses. I don't think

782
00:43:56,560 --> 00:43:59,517
we've seen the last network get created.

783
00:43:59,600 --> 00:44:02,557
I I I think it may be harder. Well, one

784
00:44:02,640 --> 00:44:06,957
thing I think we have learned the hard
way and just to what mistakes have we

785
00:44:07,040 --> 00:44:09,037
made in the last three or four years. I

786
00:44:09,120 --> 00:44:14,317
think we thought that AI or crypto or

787
00:44:14,400 --> 00:44:17,677
some other new innovation would undo the

788
00:44:17,760 --> 00:44:19,597
network effects that were built by these

789
00:44:19,680 --> 00:44:23,037
large web 2 um success stories like

790
00:44:23,120 --> 00:44:25,677
Twitter and Facebook and Tik Tok and

791
00:44:25,760 --> 00:44:28,077
whatever. And I think that has proven to

792
00:44:28,160 --> 00:44:31,277
be incorrect. those businesses are not

793
00:44:31,360 --> 00:44:34,077
going to get killed so easily

794
00:44:34,160 --> 00:44:38,317
>> when you're constructing a new network
like where I'm and this is really kind

795
00:44:38,400 --> 00:44:42,157
of key to how we're thinking about a lot
of AI these days which is like we've had

796
00:44:42,240 --> 00:44:44,157
almost exclusively single player

797
00:44:44,240 --> 00:44:47,277
products to date got to a billion users

798
00:44:47,360 --> 00:44:51,517
being entirely single player and we're
seeing you know maybe new surfaces pop

799
00:44:51,600 --> 00:44:53,997
up that can become more multiplayer and

800
00:44:54,080 --> 00:44:57,197
the question in my mind is like um what

801
00:44:57,280 --> 00:45:01,757
does it look like to either bootstrap of
another network, take on another network

802
00:45:01,840 --> 00:45:04,637
in this new age. >> Well,
if you think about, you know, what

803
00:45:04,720 --> 00:45:06,557
is what is a what does a multiplayer

804
00:45:06,640 --> 00:45:09,757
agent look like? So, it's not I mean, at

805
00:45:09,840 --> 00:45:12,557
USV, we have multiplayer agents. So, you

806
00:45:12,640 --> 00:45:17,197
know, the agents that we use at USV,
Parker and Ruth and Penny, you know, we

807
00:45:17,280 --> 00:45:19,357
we give our agents names and email

808
00:45:19,440 --> 00:45:22,637
addresses. Um, you and I use them in a

809
00:45:22,720 --> 00:45:27,357
multiplayer, you know, uh because we're
all connected to these agents. And so I

810
00:45:27,440 --> 00:45:30,957
could ask Parker a question and when
Parker responds you see the response and

811
00:45:31,040 --> 00:45:32,557
then you could reply to it. So we're

812
00:45:32,640 --> 00:45:35,677
using email or you know messengers

813
00:45:35,760 --> 00:45:39,677
effectively as the way to have
a multiplayer relationship with an agent.

814
00:45:39,760 --> 00:45:41,597
So I actually think that probably does

815
00:45:41,680 --> 00:45:44,237
exist in a lot of enterprises you know

816
00:45:44,320 --> 00:45:47,117
particularly like maybe more cutting

817
00:45:47,200 --> 00:45:52,317
edge techcentric or finance centric uh

818
00:45:52,400 --> 00:45:54,317
small businesses like ours. But on the

819
00:45:54,400 --> 00:45:56,397
consumer side,

820
00:45:56,480 --> 00:45:59,677
it's not obvious to me exactly how you

821
00:45:59,760 --> 00:46:03,837
and I would share an agent and why
and would that actually be be a good idea.

822
00:46:03,920 --> 00:46:07,357
It's pretty obvious to me how my wife
and I would share an agent. That is an

823
00:46:07,440 --> 00:46:11,517
opportunity. Now, the question is,
is that just an opportunity for WhatsApp

824
00:46:11,600 --> 00:46:13,837
and Facebook Messenger and Apple with

825
00:46:13,920 --> 00:46:17,197
iMessage? Um, where most most husbands

826
00:46:17,280 --> 00:46:20,637
and wives have a preferred vehicle to

827
00:46:20,720 --> 00:46:24,637
communicate you and your girlfriend.
Do they do you use a preferred app to

828
00:46:24,720 --> 00:46:27,117
communicate? >> iMessage. Yeah.

829
00:46:27,200 --> 00:46:30,557
>> So, it would seem to me if I owned one

830
00:46:30,640 --> 00:46:32,397
of those things, I would try to figure

831
00:46:32,480 --> 00:46:34,957
that out. That's a little surprising to

832
00:46:35,040 --> 00:46:38,877
me why Facebook and Apple and and, you

833
00:46:38,960 --> 00:46:41,997
know, Telegram haven't figured that out.

834
00:46:42,080 --> 00:46:46,797
That's a big opportunity for them.
Now, someone could come in from some other

835
00:46:46,880 --> 00:46:50,477
angle and get there. It's a little
harder though because my wife and I

836
00:46:50,560 --> 00:46:55,117
already have a place where we communicate.
So, the most obvious place

837
00:46:55,200 --> 00:46:58,237
for this agent that we're going
to share would be right there.

838
00:46:58,320 --> 00:47:00,797
>> You know, you're
a Hermes user, >> right?

839
00:47:00,880 --> 00:47:03,757
>> Um, and you probably interact
with it through Telegram.

840
00:47:03,840 --> 00:47:07,437
>> I do. >> Do you give
it most of your data?

841
00:47:07,520 --> 00:47:10,317
>> No. Um, I don't. I'm a little paranoid

842
00:47:10,400 --> 00:47:13,197
about this. So, I I have I don't have

843
00:47:13,280 --> 00:47:14,797
Hermes running in the cloud. I have

844
00:47:14,880 --> 00:47:17,357
Hermes running on a essentially not

845
00:47:17,440 --> 00:47:20,397
really an airgapped Mac Mini, but let's

846
00:47:20,480 --> 00:47:22,637
just call it a factory reset Mac Mini

847
00:47:22,720 --> 00:47:25,277
that I've been very careful about what

848
00:47:25,360 --> 00:47:28,157
services it has. And that way I know

849
00:47:28,240 --> 00:47:33,437
exactly what of my data my Hermes agent
has and what it doesn't. And I, you

850
00:47:33,520 --> 00:47:34,957
know, I'm just a little paranoid. One

851
00:47:35,040 --> 00:47:36,877
thing that crypto has really taught me

852
00:47:36,960 --> 00:47:41,277
is that bad people can get into your

853
00:47:41,360 --> 00:47:46,157
really easily if you let them. Like
I'm really lucky that I've I mean, you

854
00:47:46,240 --> 00:47:49,677
know, but one time I got hacked
because you helped me get unhacked.

855
00:47:49,760 --> 00:47:52,797
>> But the total amount
of financial losses

856
00:47:52,880 --> 00:47:55,117
I've been hacked like four or five times

857
00:47:55,200 --> 00:47:58,477
for me has been in the singledigit

858
00:47:58,560 --> 00:48:02,957
10,000s, right? So maybe I've lost 10
or $20,000, which is a lot of money. Like

859
00:48:03,040 --> 00:48:06,157
for some people it would be like their
entire life savings. I I I I don't want

860
00:48:06,240 --> 00:48:07,917
to trivialize that, but but it has

861
00:48:08,000 --> 00:48:11,677
really made me kind of I wouldn't call I

862
00:48:11,760 --> 00:48:15,757
wouldn't say I'm paranoid, but I'm
really well aware that like I got to

863
00:48:15,840 --> 00:48:20,157
lock my down. >> Yes. Well, the the reason
I ask is because, you know, I've been

864
00:48:20,240 --> 00:48:23,117
interviewing a bunch of people uh
talking to people who are playing with

865
00:48:23,200 --> 00:48:26,077
all these agents. You know, we've been
playing with Instinct and Tesla in our

866
00:48:26,160 --> 00:48:29,837
portfolio in town and and the thing
that keeps coming up as people border onto

867
00:48:29,920 --> 00:48:34,477
personal use cases is is this data
question, right? And I think people have

868
00:48:34,560 --> 00:48:38,237
a lot of a lot of fear. And someone said
to me yesterday, you know, the the thing

869
00:48:38,320 --> 00:48:41,197
I want really is for the people who
already have my data to make the most of

870
00:48:41,280 --> 00:48:43,917
it. >> Well, this is exactly right.
Right. Like

871
00:48:44,000 --> 00:48:49,277
why do I want to give some random agent
access to my Amazon? Why wouldn't Amazon

872
00:48:49,360 --> 00:48:53,437
just give me an agent that does that for
me? Right. Like and then Coinbase gives

873
00:48:53,520 --> 00:48:56,877
me an agent for my Coinbase and JP
Morgan gives me an agent for my JP

874
00:48:56,960 --> 00:49:00,317
Morgan. And then I don't have
to permission any agents to those things.

875
00:49:00,400 --> 00:49:03,037
Yeah. Yeah. I would feel
a lot safer in that world.

876
00:49:03,120 --> 00:49:06,477
>> And then you you also want there'll
need to be some connective tissue here

877
00:49:06,560 --> 00:49:09,837
because you'll want your Hermes or other
agents to be able to communicate with

878
00:49:09,920 --> 00:49:12,237
those other >> But that
that to me seems like a nice

879
00:49:12,320 --> 00:49:14,557
abstraction layer, right? So my in

880
00:49:14,640 --> 00:49:16,797
theory the Amazons and JP Morgans and

881
00:49:16,880 --> 00:49:19,357
Coinbases the world will

882
00:49:19,440 --> 00:49:22,317
architect their agents that they know

883
00:49:22,400 --> 00:49:25,757
not to give out any of my stuff but they

884
00:49:25,840 --> 00:49:30,637
can at least report up to some
agent and at least give them data.

885
00:49:30,720 --> 00:49:33,117
>> Yeah. >> You know
and then I could have like an

886
00:49:33,200 --> 00:49:36,717
agent that manages all these sub agents
for me but does doesn't actually have

887
00:49:36,800 --> 00:49:38,477
access into my systems.

888
00:49:38,560 --> 00:49:41,037
>> That's right. um that feels a little

889
00:49:41,120 --> 00:49:44,077
complex and like what something someone

890
00:49:44,160 --> 00:49:48,077
like me might want. But I don't know
that that's mainstream. Like there's a

891
00:49:48,160 --> 00:49:51,437
part of me which says like the mainstream
person doesn't have this problem.

892
00:49:51,520 --> 00:49:53,597
>> Yeah. >> Yeah. Yeah.
The main person might not

893
00:49:53,680 --> 00:49:56,397
even have this fear. I mean I think
over and over time and time again I think

894
00:49:56,480 --> 00:50:00,797
people will will jump over any hoop if
they if they have magic at the other end.

895
00:50:00,880 --> 00:50:02,797
>> People will do
the most convenient thing.

896
00:50:02,880 --> 00:50:06,557
>> Yeah. Yep. Yep.
>> Do you have enough material?

897
00:50:06,640 --> 00:50:12,157
>> All right. Uh we got 100 plus CEOs
out there uh wanting to network and schmoo

898
00:50:12,240 --> 00:50:14,637
for the next two days.
I think we got to get going on that.

899
00:50:14,720 --> 00:50:18,440
>> Awesome. Let's wrap.