1
00:00:03,320 --> 00:00:06,620
Ejaaz:
Okay, we have an incredibly special episode for you guys today.

2
00:00:06,860 --> 00:00:10,280
Ejaaz:
You're about to hear from one of the most well-researched strategic investors

3
00:00:10,280 --> 00:00:11,560
Ejaaz:
in frontier technologies.

4
00:00:11,780 --> 00:00:15,700
Ejaaz:
If you don't believe me, these guys raised $1 million, just one,

5
00:00:15,860 --> 00:00:20,980
Ejaaz:
back in early 2019 for their first fund and turned that into over 10 figures

6
00:00:20,980 --> 00:00:24,340
Ejaaz:
in value in a year and a half. You want to know what's even crazier?

7
00:00:24,580 --> 00:00:28,620
Ejaaz:
The original $1 million that they raised was on credit card debt and loans.

8
00:00:28,620 --> 00:00:32,980
Ejaaz:
So we know that these guys go all in when they have conviction in something.

9
00:00:33,260 --> 00:00:36,640
Ejaaz:
What I'm really excited about is over the last two years, they've been dialing

10
00:00:36,640 --> 00:00:39,840
Ejaaz:
in to all the stuff going on in AI, and I'm really excited to get into their

11
00:00:39,840 --> 00:00:45,180
Ejaaz:
heads about what trends they think are exciting and what they're excited about investing in.

12
00:00:45,660 --> 00:00:48,900
Ejaaz:
Anil, Jan, Jose, it's great to have you guys on. How are you guys doing today?

13
00:00:49,400 --> 00:00:52,140
Anil:
Yeah, thanks for having us. Amazing intro.

14
00:00:52,620 --> 00:00:53,440
Jose:
Yeah, I appreciate that.

15
00:00:53,560 --> 00:00:59,220
Ejaaz:
Great to be here. Let's go. Okay, so from a lot of our listeners that tune into

16
00:00:59,220 --> 00:01:01,800
Ejaaz:
the show, they've probably never heard of you guys.

17
00:01:02,080 --> 00:01:05,700
Ejaaz:
And so maybe you guys can spend a few minutes painting a picture of who you

18
00:01:05,700 --> 00:01:07,160
Ejaaz:
are. And now maybe you can kick us off.

19
00:01:07,460 --> 00:01:11,280
Anil:
Yeah, for sure. So yeah, we're all co-founders of Delphi. The Delphi that everyone

20
00:01:11,280 --> 00:01:13,480
Anil:
knows of today has like three main companies, right?

21
00:01:13,760 --> 00:01:16,060
Anil:
Delphi Research, Delphi Ventures, Delphi Labs.

22
00:01:16,460 --> 00:01:20,160
Anil:
Jan heads up the ventures, he's managing partner there. Jose heads up labs,

23
00:01:20,160 --> 00:01:21,940
Anil:
and I mostly focus on research and ventures.

24
00:01:22,780 --> 00:01:25,900
Anil:
Essentially, what Delphi does is we're very research focused,

25
00:01:26,120 --> 00:01:29,660
Anil:
right? We started back in 2018 with research embedded in our DNA.

26
00:01:30,320 --> 00:01:33,880
Anil:
Jan and I and a couple of our other co-founders all met at our first job out

27
00:01:33,880 --> 00:01:35,080
Anil:
of college at Bloomberg.

28
00:01:35,420 --> 00:01:40,360
Anil:
We did a lot of like TradFi there with research and did leverage finance at Deutsche Bank.

29
00:01:40,560 --> 00:01:44,000
Anil:
Essentially, it fell down the crypto rabbit hole when we realized that maybe

30
00:01:44,000 --> 00:01:47,320
Anil:
the future that TradFi promised wasn't all that great.

31
00:01:47,920 --> 00:01:51,360
Anil:
And yeah, just kind of like fell in love with kind of the promise that crypto provided.

32
00:01:51,940 --> 00:01:55,060
Anil:
We put our jobs in 2018, started Delphi as a research firm. And like,

33
00:01:55,500 --> 00:01:58,620
Anil:
you know, first few years, basically, no one really paid us for research because

34
00:01:58,620 --> 00:02:01,740
Anil:
like there weren't that many fundamental investors in the space.

35
00:02:02,360 --> 00:02:04,980
Anil:
You know, shout out like Multicorin and Hash. They were kind of like our first

36
00:02:04,980 --> 00:02:07,300
Anil:
two, you know, real paying customers.

37
00:02:07,580 --> 00:02:11,260
Anil:
But where we really bootshopped was actually helping design and kind of consulting

38
00:02:11,260 --> 00:02:13,060
Anil:
a lot of these protocols.

39
00:02:13,560 --> 00:02:17,760
Anil:
A lot of what you saw in DeFi and work with like protocols like Aave,

40
00:02:18,020 --> 00:02:22,260
Anil:
Lido very early on. And then especially with gaming as well,

41
00:02:22,460 --> 00:02:24,360
Anil:
with, you know, projects like Axie and Yield Guild.

42
00:02:24,740 --> 00:02:29,000
Anil:
Over the years, Delphi has kind of morphed into this like, you know,

43
00:02:29,240 --> 00:02:33,620
Anil:
three kind of like pronged layer where we build, we research and we invest.

44
00:02:33,780 --> 00:02:36,740
Anil:
Right. And I think these three different perspectives work really well together

45
00:02:36,740 --> 00:02:42,260
Anil:
because it lets us have, you know, as many like different hands on the elephant as possible.

46
00:02:42,320 --> 00:02:45,200
Anil:
So we can really feel what crypto is and where it's going and,

47
00:02:45,340 --> 00:02:46,520
Anil:
you know, have a really good pulse of it.

48
00:02:46,520 --> 00:02:50,520
Ejaaz:
It sounds like Delphi was extremely focused on the Web3 crypto world,

49
00:02:50,700 --> 00:02:53,740
Ejaaz:
right? And that's been your bread and butter since you guys have been incepted.

50
00:02:54,500 --> 00:02:58,500
Ejaaz:
And then over the last two years, you've been like dialing in very much on AI. I'm curious, like,

51
00:02:58,920 --> 00:03:01,740
Ejaaz:
like, what kind of like parallels run between the two technologies?

52
00:03:01,980 --> 00:03:04,700
Ejaaz:
Like, do you just see the AI stuff happening and thought like,

53
00:03:04,780 --> 00:03:06,560
Ejaaz:
huh, I just want to kind of like peek over the fence?

54
00:03:06,720 --> 00:03:08,900
Ejaaz:
And then did you get kind of like more involved in that? Like,

55
00:03:08,900 --> 00:03:10,940
Ejaaz:
what made you more interested?

56
00:03:11,380 --> 00:03:14,920
Anil:
Yeah, definitely. I'd say that like, you know, when we first fell down the crypto

57
00:03:14,920 --> 00:03:18,020
Anil:
rabbit hole, it was almost it wasn't even just obvious to us.

58
00:03:18,100 --> 00:03:22,100
Anil:
It was just like, you know, what else could we work on or spend our time doing

59
00:03:22,100 --> 00:03:24,620
Anil:
other than this, right? It felt like nothing else mattered.

60
00:03:24,860 --> 00:03:27,920
Anil:
And I think over the past two years, you know, shout out Tom,

61
00:03:28,120 --> 00:03:30,020
Anil:
one of our other venture partners and co-founders.

62
00:03:30,160 --> 00:03:34,680
Anil:
He really was early to the AI trend and everything like that.

63
00:03:34,820 --> 00:03:39,280
Anil:
And a lot of people within the hive mind that dealt by got nerds signed by AI

64
00:03:39,280 --> 00:03:41,120
Anil:
and it felt we had that same feeling

65
00:03:41,120 --> 00:03:46,180
Anil:
where it was like, how can we not be infatuated and obsessed with this?

66
00:03:46,660 --> 00:03:50,420
Anil:
And yeah, I think there are a bunch of parallels. I mean, obviously the speed

67
00:03:50,420 --> 00:03:55,740
Anil:
of innovation, just like when we entered crypto, just like today,

68
00:03:55,740 --> 00:03:58,460
Anil:
even with a team of almost 100, right?

69
00:03:58,500 --> 00:04:01,240
Anil:
We have around 88 people across three companies at Delphi.

70
00:04:01,360 --> 00:04:05,760
Anil:
There's just no way we can kind of, you know, keep up with every single thing happening in crypto.

71
00:04:05,980 --> 00:04:09,240
Anil:
I think that's the same thing that we see in AI and why we think,

72
00:04:09,460 --> 00:04:12,960
Anil:
you know, we'll get into this later, why we think it's really important to have

73
00:04:12,960 --> 00:04:17,220
Anil:
a team focused on it and kind of like separate the signal from the noise.

74
00:04:17,560 --> 00:04:22,000
Yan:
Yeah. No, in terms of parallels, I think just when you see something that,

75
00:04:22,460 --> 00:04:27,460
Yan:
looks so glaringly obvious in terms of, you know, its growth and application,

76
00:04:27,720 --> 00:04:32,240
Yan:
but at the same time, there isn't really any widespread adoption of it or it's

77
00:04:32,240 --> 00:04:36,140
Yan:
still, you know, orders of magnitude away from what it'll eventually be.

78
00:04:36,280 --> 00:04:39,260
Yan:
Your eyes tend to light up because you start to think about all the possibilities

79
00:04:39,260 --> 00:04:41,620
Yan:
on the growth building and investing side.

80
00:04:41,720 --> 00:04:45,980
Yan:
And so I think what we saw that with crypto, you tend to see here with AI.

81
00:04:46,560 --> 00:04:50,640
Yan:
And then there definitely overlaps the two in terms of implementation and where

82
00:04:50,640 --> 00:04:51,940
Yan:
they can be synergistic.

83
00:04:52,100 --> 00:04:56,140
Yan:
But I think, you know, holistically, you tend to, I think that positioning is

84
00:04:56,140 --> 00:04:58,840
Yan:
really what gets you excited at first, because, and you know,

85
00:04:58,940 --> 00:05:00,940
Yan:
for and all the reasons you're bullish on it are,

86
00:05:01,180 --> 00:05:05,120
Yan:
you know, fundamental reasons, and then kind of put that against a backdrop

87
00:05:05,120 --> 00:05:10,120
Yan:
of the fact that it's still barely permeated, and there's still very minimal adoption of it is.

88
00:05:10,240 --> 00:05:13,980
Yan:
And I think that position is what really excited us about it in the first place.

89
00:05:14,380 --> 00:05:18,800
Ejaaz:
Well, I think what something that's really interesting is your focus on kind

90
00:05:18,800 --> 00:05:20,840
Ejaaz:
of crypto and web three for the initial fund.

91
00:05:21,460 --> 00:05:24,180
Ejaaz:
Crypto is an incredibly fast changing technology, right?

92
00:05:24,520 --> 00:05:28,060
Ejaaz:
And the whole point around it is it's meant to rebuild a ton of different sectors,

93
00:05:28,500 --> 00:05:31,120
Ejaaz:
finance, media, you know, you name it.

94
00:05:31,500 --> 00:05:35,700
Ejaaz:
AI is exactly that as well. So I'm not, I can't say I'm exactly surprised that

95
00:05:35,700 --> 00:05:37,460
Ejaaz:
you guys are marrying both technologies together.

96
00:05:38,100 --> 00:05:41,660
Ejaaz:
You're doing a ton of stuff in this space. So you just mentioned a few arms,

97
00:05:41,840 --> 00:05:46,280
Ejaaz:
Anil, You're doing the research side, the investing side, and also the incubating side.

98
00:05:46,540 --> 00:05:49,280
Ejaaz:
I saw that you guys are incubating a bunch of AI companies.

99
00:05:49,680 --> 00:05:52,960
Ejaaz:
Maybe you guys can speak more to that. Jose, maybe I can pass this to you.

100
00:05:53,300 --> 00:05:58,080
Jose:
Yeah, very similar to these guys. I had my crypto-pilled moment in 2017 with

101
00:05:58,080 --> 00:06:01,880
Jose:
Ethereum and pretty much had the same experience last year, actually,

102
00:06:01,880 --> 00:06:05,760
Jose:
a bit later than I think some of the other people at Delphi when I read situational awareness.

103
00:06:06,020 --> 00:06:09,300
Jose:
I'd been playing with MidJourney, obviously, and ChatGPT, but I was just so

104
00:06:09,300 --> 00:06:12,640
Jose:
busy and kind out of deep into crypto that, that, uh,

105
00:06:13,430 --> 00:06:17,130
Jose:
I think I didn't realize just how momentous this thing was.

106
00:06:17,490 --> 00:06:20,370
Jose:
And then, yeah, last year, once I read Situational Awareness,

107
00:06:20,490 --> 00:06:21,530
Jose:
it really clicked into place.

108
00:06:22,150 --> 00:06:27,070
Jose:
And we pretty soon decided with labs that we had to start doing some stuff in AI.

109
00:06:27,430 --> 00:06:31,410
Jose:
So we put together our thesis on crypto AI, spent a lot of time on that,

110
00:06:31,530 --> 00:06:36,210
Jose:
just figuring out where the good places for overlap was, and then ended up partnering up with Nir.

111
00:06:36,670 --> 00:06:40,690
Jose:
Ilya is obviously an OG in AI. He's one of the original authors of Transformers paper.

112
00:06:40,690 --> 00:06:43,610
Jose:
And yeah we we partnered with them to run our

113
00:06:43,610 --> 00:06:46,650
Jose:
first accelerator in ai which was really really great

114
00:06:46,650 --> 00:06:51,210
Jose:
had some insanely uh strong founders that applied just through ilia's network

115
00:06:51,210 --> 00:06:58,210
Jose:
um and then did a second one uh about finished about two months ago with with

116
00:06:58,210 --> 00:07:02,650
Jose:
the cyber fund guys which was also awesome um yeah we always think that the

117
00:07:02,650 --> 00:07:06,430
Jose:
best way to i mean like anil said the the old

118
00:07:06,890 --> 00:07:09,190
Jose:
elephant groping metaphor that Anil likes.

119
00:07:09,350 --> 00:07:11,610
Jose:
We like having a lot of hands on the elephant.

120
00:07:11,830 --> 00:07:17,110
Jose:
And I think researching is awesome and we're all kind of researchers in, is it our core?

121
00:07:17,370 --> 00:07:21,770
Jose:
But building, you get a really unique perspective. And that happened for us in crypto too.

122
00:07:22,010 --> 00:07:24,930
Jose:
Like there was things we learned by building protocols and being really deeply

123
00:07:24,930 --> 00:07:26,750
Jose:
involved that we really couldn't have learned any other way.

124
00:07:26,950 --> 00:07:30,410
Jose:
And it's been the same exact thing with AI. So it's just been really interesting.

125
00:07:31,350 --> 00:07:34,310
Jose:
And also it's similar to crypto. It's an entirely new paradigm.

126
00:07:34,730 --> 00:07:37,550
Jose:
Like crypto building is like very different

127
00:07:37,550 --> 00:07:40,530
Jose:
from web2 building like you have this these smart contracts they're

128
00:07:40,530 --> 00:07:43,890
Jose:
immutable well uh they used to be anyway nowadays protocols

129
00:07:43,890 --> 00:07:47,290
Jose:
take a slightly different approach but still a lot of them are immutable and so

130
00:07:47,290 --> 00:07:50,330
Jose:
it ends up being more like hardware like you have to be really careful

131
00:07:50,330 --> 00:07:53,230
Jose:
you have to spend a lot of time researching um and then

132
00:07:53,230 --> 00:07:56,350
Jose:
writing like uh it's less of an iterative approach

133
00:07:56,350 --> 00:07:59,030
Jose:
and more of uh you know once this is out there it's it's out

134
00:07:59,030 --> 00:08:02,030
Jose:
there for anyone to exploit and ai is like a different paradigm still

135
00:08:02,030 --> 00:08:05,030
Jose:
where these things unlike like most

136
00:08:05,030 --> 00:08:08,090
Jose:
of software before it aren't deterministic they're

137
00:08:08,090 --> 00:08:11,510
Jose:
probabilistic and so it's really hard to ensure

138
00:08:11,510 --> 00:08:15,650
Jose:
like a uniform user experience and like they're not even standards for like

139
00:08:15,650 --> 00:08:19,950
Jose:
unit tests or anything like that um i really think the the metaphor of it being

140
00:08:19,950 --> 00:08:23,750
Jose:
a new kind of computer is great so it's just been really useful diving in and

141
00:08:23,750 --> 00:08:30,630
Jose:
and learning um like with our hands in the yeah with just just just getting stuck in and building so

142
00:08:30,630 --> 00:08:35,070
Ejaaz:
There's a lot going on in AI right now, new frontier models are being released

143
00:08:35,070 --> 00:08:36,630
Ejaaz:
like every week at this point.

144
00:08:37,110 --> 00:08:40,890
Ejaaz:
Billions of dollars are being spent to train these things. There are numerous

145
00:08:40,890 --> 00:08:42,770
Ejaaz:
consumer applications that are out there.

146
00:08:43,110 --> 00:08:46,790
Ejaaz:
And I can't help but think that this is like an incredibly expensive game to play.

147
00:08:47,230 --> 00:08:51,730
Ejaaz:
So I'm kind of curious, what's your unique edge when it comes to investing in AI?

148
00:08:51,850 --> 00:08:54,730
Ejaaz:
How do you view the market right now? And where do you think you guys can make

149
00:08:54,730 --> 00:08:56,410
Ejaaz:
the biggest impact with what you're doing?

150
00:08:56,610 --> 00:08:59,890
Ejaaz:
You obviously have the whole Web3 crypto background, and maybe it's something

151
00:08:59,890 --> 00:09:04,150
Ejaaz:
to do along with those kind of principles of investing that you had with that fund.

152
00:09:04,250 --> 00:09:07,510
Ejaaz:
But I'm curious whether there's anything new you guys are seeing in the market right now.

153
00:09:07,650 --> 00:09:10,850
Jose:
I don't think we have an edge right now. I think we're sort of

154
00:09:11,460 --> 00:09:15,760
Jose:
hoping to build our edge over time. We've definitely made a lot of investments

155
00:09:15,760 --> 00:09:17,880
Jose:
in crypto AI. I think we have edge there.

156
00:09:18,120 --> 00:09:21,080
Jose:
We've made a couple of investments in AI, but I think we all sort of recognize

157
00:09:21,080 --> 00:09:24,940
Jose:
that we're sort of paying tuition right now and getting to know the industry,

158
00:09:25,060 --> 00:09:29,060
Jose:
getting to know as many founders as possible and kind of building our edge over time.

159
00:09:29,520 --> 00:09:33,240
Jose:
That's the goal of intelligence really. Same as when we started in crypto,

160
00:09:33,540 --> 00:09:38,000
Jose:
you guys didn't want to start a fund straight away, wanted to kind of build

161
00:09:38,000 --> 00:09:40,240
Jose:
your edge, build your knowledge, and then go for that.

162
00:09:40,240 --> 00:09:43,640
Jose:
And I think it's similar here, except now we have some capital behind us.

163
00:09:43,720 --> 00:09:45,600
Jose:
So it makes sense to invest and start building that.

164
00:09:46,680 --> 00:09:50,000
Jose:
So yeah, the hope is that we build the brand with Delphi Intelligence,

165
00:09:50,280 --> 00:09:51,740
Jose:
get some really tough researchers on.

166
00:09:52,620 --> 00:09:55,400
Jose:
And then we're also doing a couple of other things. Like we've been investing

167
00:09:55,400 --> 00:10:00,540
Jose:
in young fund managers in AI, sort of looking to, like when we started Delphi

168
00:10:00,540 --> 00:10:02,880
Jose:
Ventures seven years ago, it was really hard to raise.

169
00:10:03,300 --> 00:10:06,980
Jose:
And we know firsthand both how hard it is to be a first-time fund manager raising

170
00:10:06,980 --> 00:10:09,300
Jose:
and also how much edge you can have as a first-time manager.

171
00:10:09,300 --> 00:10:12,520
Jose:
And so we're kind of looking to find those people that were in the same position

172
00:10:12,520 --> 00:10:16,900
Jose:
we were in seven years ago in AI and back them and then kind of benefit from

173
00:10:16,900 --> 00:10:18,020
Jose:
that deal flow and that learning.

174
00:10:18,700 --> 00:10:22,100
Jose:
And I think the way we're thinking about it internally is we would like to aim

175
00:10:22,100 --> 00:10:25,200
Jose:
to have edge and to really start accelerating our investment pace.

176
00:10:25,800 --> 00:10:31,000
Jose:
12 to 24 months from now, something like that. So yeah, this whole thing is

177
00:10:31,000 --> 00:10:33,120
Jose:
sort of us aiming to build that edge.

178
00:10:33,240 --> 00:10:36,360
Anil:
I think like even when we first got started and we were writing reports,

179
00:10:36,880 --> 00:10:40,340
Anil:
you know, if we put out a report on say, synthetics or something like that,

180
00:10:40,580 --> 00:10:43,560
Anil:
people would always message us afterwards and say, damn, you guys really knew

181
00:10:43,560 --> 00:10:44,480
Anil:
synthetics really well.

182
00:10:44,620 --> 00:10:47,300
Anil:
That's why the report came out. So, you know, great or anything like that.

183
00:10:47,600 --> 00:10:51,780
Anil:
It's quite the opposite, right? Like we learn about, you know,

184
00:10:51,900 --> 00:10:55,180
Anil:
whatever we're researching when we're putting together of this report that we

185
00:10:55,180 --> 00:10:59,200
Anil:
know is going to get like, you know, picked apart on places like crypto Twitter

186
00:10:59,200 --> 00:11:01,840
Anil:
or by the team or by competitors. Right.

187
00:11:02,320 --> 00:11:06,680
Anil:
So that's why we really do love having research embedded into our DNA,

188
00:11:06,680 --> 00:11:11,120
Anil:
because like it almost provides like this check and kind of this like,

189
00:11:11,240 --> 00:11:14,900
Anil:
you know, high bar that anything we publish, we know is going to be looked at

190
00:11:14,900 --> 00:11:19,160
Anil:
by, you know, people either building the space or other investors in the space, et cetera.

191
00:11:19,440 --> 00:11:23,180
Anil:
So we want to make sure that the research is not just really good for us to

192
00:11:23,180 --> 00:11:27,920
Anil:
use and build conviction, but also meets this bar where it won't get ripped apart.

193
00:11:28,340 --> 00:11:31,640
Anil:
And that kind of fear or intimidation, I think, is really powerful.

194
00:11:31,900 --> 00:11:32,220
Jose:
Yeah.

195
00:11:32,380 --> 00:11:37,140
Yan:
If I had to pick an edge, just to give you some answer to that question,

196
00:11:37,240 --> 00:11:39,220
Yan:
I'd say it comes from a few areas.

197
00:11:39,420 --> 00:11:44,840
Yan:
One, just from investing for however many years we've been doing it,

198
00:11:44,900 --> 00:11:48,820
Yan:
and granted, that's an edge that's kind of consistent across anyone who's been

199
00:11:48,820 --> 00:11:51,280
Yan:
doing it, So it's not necessarily a big one.

200
00:11:51,400 --> 00:11:57,360
Yan:
I think we do have a decent variety of backgrounds and ways of thinking as well.

201
00:11:57,440 --> 00:12:01,640
Yan:
And that's been an edge for us in crypto and should continue to be one here.

202
00:12:02,060 --> 00:12:06,440
Yan:
And I think just being able to operate as a group is a big edge where we're

203
00:12:06,440 --> 00:12:10,060
Yan:
able to take a variety of learnings that each of us are doing,

204
00:12:10,480 --> 00:12:16,000
Yan:
bring them to the table and get kind of immediate feedback and have just a variety

205
00:12:16,000 --> 00:12:18,840
Yan:
of points of view. I think that that's probably one of the bigger ones.

206
00:12:18,980 --> 00:12:24,560
Yan:
And then patience, I think is another one that we've kind of learned over time

207
00:12:24,560 --> 00:12:25,980
Yan:
in crypto in particular.

208
00:12:26,200 --> 00:12:29,780
Yan:
And so here we realize we don't really have an edge and we're trying to understand

209
00:12:29,780 --> 00:12:32,420
Yan:
is where the best opportunity is, right?

210
00:12:32,500 --> 00:12:39,140
Yan:
Is it early stage or does early stage really take too long to get a proper payback?

211
00:12:39,140 --> 00:12:45,560
Yan:
Is it makes sense to kind of invest in some of these growth-stage higher value or higher.

212
00:12:46,580 --> 00:12:51,440
Yan:
Valuation but lower risk type plays where you have a pretty kind of cemented

213
00:12:51,440 --> 00:12:56,160
Yan:
path to becoming a large company. And so that's still something we're exploring.

214
00:12:56,300 --> 00:13:01,100
Yan:
I don't think we have really have an answer there yet, but I think it's just the patience.

215
00:13:01,320 --> 00:13:06,700
Yan:
And I think what's helped with crypto is that you go through so many cycles so quickly.

216
00:13:06,960 --> 00:13:13,120
Yan:
And I think you can draw parallels to kind of other online experiences versus

217
00:13:13,120 --> 00:13:14,480
Yan:
physical ones. So if you think about like,

218
00:13:15,410 --> 00:13:19,930
Yan:
online poker guys have have seen an insane amount of hands right and so they

219
00:13:19,930 --> 00:13:23,070
Yan:
have a lot more experience than someone who plays live despite you know having

220
00:13:23,070 --> 00:13:27,490
Yan:
a long-term career so i think you know there is some benefit uh in terms of

221
00:13:27,490 --> 00:13:31,590
Yan:
taking that from crypto and understanding those cycles and trying to uh draw parallels there.

222
00:13:31,590 --> 00:13:34,730
Jose:
Yeah i think we all agree i definitely

223
00:13:34,730 --> 00:13:37,570
Jose:
agree with jan i think being a venture investor is like a skill that's sort

224
00:13:37,570 --> 00:13:40,790
Jose:
of generalizable across sectors like a lot of it dating founders

225
00:13:40,790 --> 00:13:43,510
Jose:
understand it but you you kind of need to understand the sector to be

226
00:13:43,510 --> 00:13:46,590
Jose:
able to properly do diligence the founder and not get bamboozled

227
00:13:46,590 --> 00:13:53,690
Jose:
by a high by a charismatic um you know sort of charlatan i guess um and so i

228
00:13:53,690 --> 00:13:57,510
Jose:
think what we all agree with is that uh we all agree that this is going to be

229
00:13:57,510 --> 00:14:01,690
Jose:
i think the biggest bubble that that like humanity has ever seen i think just

230
00:14:01,690 --> 00:14:04,490
Jose:
like all the ingredients are there isn't it like already

231
00:14:04,490 --> 00:14:08,610
Ejaaz:
A bubble this this was being said like last year and it's just been up only

232
00:14:08,610 --> 00:14:13,170
Ejaaz:
i think what nvidia crossed like four trillion in market cap this week.

233
00:14:13,710 --> 00:14:16,450
Ejaaz:
I feel like how big do you think this boat was going to go?

234
00:14:16,550 --> 00:14:19,770
Ejaaz:
Because I agree with you, charismatic founders are super important,

235
00:14:19,770 --> 00:14:25,890
Ejaaz:
but I see a bunch of these VC investors talk about theses for decades,

236
00:14:25,930 --> 00:14:27,690
Ejaaz:
right? The next 30 years is going to look like this.

237
00:14:27,830 --> 00:14:32,090
Ejaaz:
AGI, we're going to achieve it in whatever, 2027, or they're arguing about that.

238
00:14:32,470 --> 00:14:35,310
Ejaaz:
How important is the founder when it comes to all of these kinds of things?

239
00:14:35,450 --> 00:14:37,710
Ejaaz:
I'm guessing quite a lot. Yeah.

240
00:14:39,030 --> 00:14:43,030
Jose:
To me, we have different I think focuses even as investors.

241
00:14:43,170 --> 00:14:47,190
Jose:
To me, the founder is the most important thing, especially at the stage that

242
00:14:47,190 --> 00:14:49,030
Jose:
we invest in, which is normally seed or pre-seed.

243
00:14:49,270 --> 00:14:51,210
Jose:
The idea is going to change a lot.

244
00:14:51,830 --> 00:14:53,510
Jose:
And you're really betting on

245
00:14:53,510 --> 00:14:58,270
Jose:
a founder and you want someone that is just exceptional and has a history.

246
00:14:58,530 --> 00:15:01,510
Jose:
And exceptional people leave breadcrumbs. You can sort of look at their past

247
00:15:01,510 --> 00:15:07,270
Jose:
and be able to see some evidence of exceptional behavior before.

248
00:15:07,590 --> 00:15:11,470
Jose:
And ideally, you're looking for the things that are like, he was insane at a

249
00:15:11,470 --> 00:15:16,030
Jose:
video game or something in their youth, some sporting thing,

250
00:15:16,270 --> 00:15:18,990
Jose:
those things are generally better because they're not as priced in as someone

251
00:15:18,990 --> 00:15:21,550
Jose:
having done a successful startup and exited it or whatever.

252
00:15:21,790 --> 00:15:25,370
Jose:
And you're really looking for these kind of freaks, basically,

253
00:15:25,590 --> 00:15:28,570
Jose:
that are insanely motivated, that are able to...

254
00:15:29,960 --> 00:15:34,660
Jose:
Go through walls to achieve what they want. And so that pattern of like,

255
00:15:34,760 --> 00:15:37,780
Jose:
we've seen a few with ventures over the years, and those have been our big winners.

256
00:15:37,940 --> 00:15:40,240
Jose:
And we're just looking for more kind of an AI.

257
00:15:40,480 --> 00:15:43,280
Jose:
And then on the bubble comment, I don't think so.

258
00:15:43,380 --> 00:15:48,300
Jose:
I mean, I think when you look at where, I look at 2000 as my mainly,

259
00:15:48,620 --> 00:15:53,380
Jose:
like, maybe the biggest comp, like the price to earnings ratios of the Mag7

260
00:15:53,380 --> 00:15:57,860
Jose:
equivalent, we're still like, you know, two to three EX what they are now.

261
00:15:58,280 --> 00:16:01,560
Jose:
And then And I think in the private markets, there's definitely a few bubbly

262
00:16:01,560 --> 00:16:04,160
Jose:
things, but there's also like insane growth and fundamentals,

263
00:16:04,360 --> 00:16:09,180
Jose:
you know, like CatcherPT is the fastest company ever to a hundred billion in

264
00:16:09,180 --> 00:16:11,140
Jose:
revenue, to a billion in revenue, to 10 billion in revenue.

265
00:16:11,580 --> 00:16:15,220
Jose:
Cursor, I think was the fastest actually company ever to half a billion in revenue.

266
00:16:16,140 --> 00:16:20,440
Jose:
And you're seeing multiples of these, right? With DAUs, like actual revenue.

267
00:16:21,100 --> 00:16:26,200
Jose:
I do think there's some bubbly behavior and some stuff that's kind of reminiscent

268
00:16:26,200 --> 00:16:30,660
Jose:
of 2000 with these valuations, but I do think there's just a long way to go

269
00:16:30,660 --> 00:16:33,520
Jose:
just because, first of all,

270
00:16:33,780 --> 00:16:36,680
Jose:
you have the most profitable companies in the history of the world that are

271
00:16:36,680 --> 00:16:40,840
Jose:
stuck in this game-theoretic arms race where they're incentivized to spend every

272
00:16:40,840 --> 00:16:45,660
Jose:
single dollar of free cash flow into training better AI models because otherwise

273
00:16:45,660 --> 00:16:48,860
Jose:
they might miss AGI and have their company destroyed.

274
00:16:49,200 --> 00:16:53,680
Jose:
And that's a dynamic that's just going to be a constant tailwind to making these models better.

275
00:16:53,840 --> 00:16:59,560
Jose:
And every startup in the ecosystem benefits from better models. So there's that.

276
00:17:00,160 --> 00:17:04,300
Jose:
And then I think there's just the fact that this stuff, like the internet was

277
00:17:04,300 --> 00:17:08,940
Jose:
kind of like, people got really excited in 2000, but there was all this infrastructure

278
00:17:08,940 --> 00:17:13,220
Jose:
that still needed to be built for the killer apps that people imagined in 2000 to work, right?

279
00:17:13,400 --> 00:17:17,240
Jose:
You needed people to have mobile phones to build Uber. You needed payment rails.

280
00:17:17,500 --> 00:17:22,060
Jose:
You needed like GPS working. You needed all these different enabling technologies.

281
00:17:22,380 --> 00:17:25,700
Jose:
And with AI, it really feels like you don't. like everyone

282
00:17:25,700 --> 00:17:28,840
Jose:
has a smartphone everyone has a has a computer fast

283
00:17:28,840 --> 00:17:31,820
Jose:
internet like um it there's nothing

284
00:17:31,820 --> 00:17:34,740
Jose:
in the way of this thing just scaling like

285
00:17:34,740 --> 00:17:38,780
Jose:
it's really limited just by the quality of applications uh for people to use

286
00:17:38,780 --> 00:17:41,820
Jose:
and there's so much talent going into it there's so much compute going in there's

287
00:17:41,820 --> 00:17:46,980
Jose:
so much like spending uh happening that i just think it's it's gonna stay extremely

288
00:17:46,980 --> 00:17:53,060
Jose:
uh it's gonna keep moving extremely fast uh yeah so i don't think this is the bubble the bubble yet.

289
00:17:53,460 --> 00:17:58,080
Yan:
Yeah. And on the bubble point, I think you can kind of think of it in multiple phases, right?

290
00:17:58,200 --> 00:18:03,100
Yan:
So right now you have this kind of scenario where the markets are.

291
00:18:04,190 --> 00:18:08,050
Yan:
Really giving credit for just capex. So, so margins are coming down on some

292
00:18:08,050 --> 00:18:11,150
Yan:
of these bigger players and, and it doesn't matter because they need to spend

293
00:18:11,150 --> 00:18:15,730
Yan:
and, and spend and spend and just get to this point where, um.

294
00:18:16,570 --> 00:18:21,170
Yan:
The, like the next kind of wave is proving out that the spend is actually valuable.

295
00:18:21,470 --> 00:18:23,730
Yan:
And I think you're, you're starting to see elements of that,

296
00:18:23,770 --> 00:18:26,510
Yan:
but the, the market is kind of very forgiving right now.

297
00:18:26,650 --> 00:18:30,350
Yan:
And, and so, um, you, you know, for the first time in a while you have this

298
00:18:30,350 --> 00:18:33,690
Yan:
technology that can improve efficiency by an order of magnitude.

299
00:18:33,850 --> 00:18:35,730
Yan:
And it just gets captured in so many ways, right?

300
00:18:36,030 --> 00:18:39,930
Yan:
You'll have the big guys who leverage their distribution to just improve margins

301
00:18:39,930 --> 00:18:42,930
Yan:
because they need to reduce headcount or just become more efficient.

302
00:18:43,230 --> 00:18:47,850
Yan:
On the startup side, you have these smaller teams that can get to unicorn status

303
00:18:47,850 --> 00:18:52,210
Yan:
without really needing these longer term cash raises.

304
00:18:52,490 --> 00:18:57,250
Yan:
And so I think the fact that it's kind of happening across multiple areas is

305
00:18:57,250 --> 00:19:00,490
Yan:
what'll give it legs for quite some time.

306
00:19:00,610 --> 00:19:05,190
Yan:
But yeah, in the interim, you have basically this massive spend phase and that

307
00:19:05,190 --> 00:19:08,330
Yan:
doesn't seem like it's going to be slowing down anytime soon once we're starting

308
00:19:08,330 --> 00:19:11,010
Yan:
to see that there are actual improvements to be made to the base models, right?

309
00:19:11,070 --> 00:19:14,370
Yan:
There was that concern up front where, okay, it was actually kind of solved.

310
00:19:14,650 --> 00:19:17,310
Yan:
And then when there were these big breakthroughs, then everyone,

311
00:19:17,590 --> 00:19:19,610
Yan:
you know, the CapEx got turned back on again.

312
00:19:19,670 --> 00:19:22,510
Yan:
And so it doesn't seem like that's really going to slow down anytime soon.

313
00:19:22,510 --> 00:19:27,550
Yan:
But at the same time, you're having real efficiency gains at the early stage.

314
00:19:27,690 --> 00:19:31,690
Yan:
And so, yeah, I think that the trickiest part is probably the very late stage

315
00:19:31,690 --> 00:19:37,190
Yan:
investing side in the world where they don't necessarily need to bring on that capital.

316
00:19:37,970 --> 00:19:41,670
Anil:
Yeah. The one thing I'd add here too is like, Bubble has this very like negative

317
00:19:41,670 --> 00:19:42,730
Anil:
connotation to it, right?

318
00:19:42,830 --> 00:19:47,150
Anil:
I think like one reason we're really excited is because we actually do exactly what Jan said.

319
00:19:47,330 --> 00:19:50,030
Anil:
We think they're going to be insane efficiency gains. We think there's going

320
00:19:50,030 --> 00:19:53,430
Anil:
to be this huge period of abundance, right?

321
00:19:53,530 --> 00:19:56,470
Anil:
Obviously with this new innovation. And I think I think like,

322
00:19:56,890 --> 00:19:59,230
Anil:
you know, one thing that we think about and we were talking about just this

323
00:19:59,230 --> 00:20:03,070
Anil:
past week at our founders retreat is like, you know, there's this like the churn

324
00:20:03,070 --> 00:20:09,210
Anil:
rate of Forbes 500, the Fortune 500 company every decade has just been going up and up and up. Right.

325
00:20:09,430 --> 00:20:11,690
Anil:
So even if you use the churn rate from like the last decade,

326
00:20:11,890 --> 00:20:15,970
Anil:
I think, you know, probably half of the companies would be kind of churned in

327
00:20:15,970 --> 00:20:17,430
Anil:
the next like 10 years. Right.

328
00:20:17,940 --> 00:20:23,200
Anil:
We actually think, or this is at least my stance, I think that churn rate is

329
00:20:23,200 --> 00:20:25,540
Anil:
going to increase exponentially because of AI.

330
00:20:25,880 --> 00:20:32,640
Anil:
And I think you may even see 350 to 400 of the top 500 companies get churned

331
00:20:32,640 --> 00:20:34,760
Anil:
out in the next decade, which what does that mean?

332
00:20:34,860 --> 00:20:38,320
Anil:
That just means there's immense value creation happening in other areas of the

333
00:20:38,320 --> 00:20:41,820
Anil:
market and capturing even a little bit of that upside. I think it's just going

334
00:20:41,820 --> 00:20:45,280
Anil:
to be the craziest thing that you could have ever hoped for as an investor, right?

335
00:20:45,760 --> 00:20:50,000
Anil:
So yeah, I think we are excited for some of these big companies that already do exist.

336
00:20:50,180 --> 00:20:54,680
Anil:
Obviously, like the Max 7, Fang, they're obviously fighting very hard to hold

337
00:20:54,680 --> 00:20:57,380
Anil:
on to their spots, and there will be a lot of efficiency gains there.

338
00:20:57,500 --> 00:21:01,440
Anil:
But I think more excitingly and obviously going to be much harder to figure

339
00:21:01,440 --> 00:21:07,020
Anil:
out are the companies that will go from zero to some of these top 500 companies

340
00:21:07,020 --> 00:21:10,620
Anil:
in areas all across the map.

341
00:21:10,780 --> 00:21:15,080
Anil:
So yeah, honestly, we're just super excited. But yeah, I think it's going to

342
00:21:15,080 --> 00:21:17,720
Anil:
be challenging, but that's why we're kind of pumped.

343
00:21:18,780 --> 00:21:22,520
Ejaaz:
Yeah. So one of these words that I keep hearing all three of you mention is the word edge.

344
00:21:23,080 --> 00:21:25,980
Ejaaz:
And it's like looking to find the edge. And what I want to ask,

345
00:21:26,040 --> 00:21:28,800
Ejaaz:
because I think this is what I'm personally interested in, a lot of people who

346
00:21:28,800 --> 00:21:32,580
Ejaaz:
are listening, is what the process looks like in finding an edge and what type

347
00:21:32,580 --> 00:21:35,120
Ejaaz:
of topics you guys are interested in pursuing where you can find that.

348
00:21:35,200 --> 00:21:37,600
Ejaaz:
Because a lot of times there are episodes, we're interested in just exploring

349
00:21:37,600 --> 00:21:40,200
Ejaaz:
different frontiers, but there's a lot of different pillars in the world of

350
00:21:40,200 --> 00:21:42,040
Ejaaz:
AI. There's so many different industries and categories.

351
00:21:42,300 --> 00:21:44,620
Ejaaz:
Is there a particular spot you're excited about?

352
00:21:44,920 --> 00:21:48,780
Ejaaz:
And within that spot, how do you go about finding an edge and getting an advantage?

353
00:21:49,060 --> 00:21:54,740
Jose:
It's honestly like a lot of trial and error and being very honest with yourself about where you sit.

354
00:21:54,900 --> 00:21:57,700
Jose:
I think that's something crypto really gives you like to survive and thrive

355
00:21:57,700 --> 00:22:02,000
Jose:
in crypto. You need to be very honest about whether you have edge or not and where you have edge.

356
00:22:02,940 --> 00:22:06,760
Jose:
And in AI, I think for us, it's just been a process of, I think,

357
00:22:06,920 --> 00:22:10,800
Jose:
first of all, we started looking at, obviously, we did crypto AI where we thought,

358
00:22:10,880 --> 00:22:13,860
Jose:
you know, there's an overlap here with crypto, we have an existing brand.

359
00:22:14,460 --> 00:22:18,040
Jose:
The sector is exciting here. I think it's pretty clear that we can have edge,

360
00:22:18,460 --> 00:22:19,960
Jose:
like we're very early to it.

361
00:22:20,400 --> 00:22:23,660
Jose:
And then we started trying to do more AI direct investments. And I

362
00:22:25,170 --> 00:22:29,090
Jose:
Uh, the, the bigger challenge. We were, we were like, some of the stuff was

363
00:22:29,090 --> 00:22:31,390
Jose:
hard for us to get our head around.

364
00:22:31,650 --> 00:22:37,490
Jose:
Um, but also it was unclear to us, um, like whether we had edge and that's always

365
00:22:37,490 --> 00:22:39,610
Jose:
like a bad sign. Like you should, you should kind of know.

366
00:22:39,810 --> 00:22:43,190
Jose:
Um, I guess we know the feeling of, of having edge to some extent.

367
00:22:43,390 --> 00:22:47,470
Jose:
And I think it's a mixture of, um, there's like some reason,

368
00:22:47,890 --> 00:22:51,010
Jose:
something that other people aren't seeing here, which I definitely think we're,

369
00:22:51,090 --> 00:22:54,730
Jose:
we're like more bullish on AI than the average person, but probably not than the average VC. Right.

370
00:22:55,230 --> 00:22:58,930
Jose:
So then we thought, okay, I think this direct investment, there's some negative

371
00:22:58,930 --> 00:23:03,950
Jose:
selection happening here, like the deals that we're seeing are potentially not the best ones.

372
00:23:04,330 --> 00:23:07,870
Jose:
And so we started to look at, I mean, first of all, we started to look at fund

373
00:23:07,870 --> 00:23:10,990
Jose:
managers, which I think was an interesting one where we saw,

374
00:23:11,210 --> 00:23:13,950
Jose:
okay, there's these fund managers raising small funds,

375
00:23:14,350 --> 00:23:18,030
Jose:
first-time fund managers, they're really struggling too, because no one wants

376
00:23:18,030 --> 00:23:22,410
Jose:
to back a first-time fund manager generally, and the fund of funds are very risk averse.

377
00:23:22,670 --> 00:23:27,790
Jose:
And so, and we started seeing, wow, there's some guys here who are super plugged

378
00:23:27,790 --> 00:23:29,130
Jose:
in, insanely well-networked and

379
00:23:29,130 --> 00:23:32,330
Jose:
hungry, and really remind us of kind of ourselves seven years ago in AI.

380
00:23:32,550 --> 00:23:35,530
Jose:
And this could be a way that we can have some edge, like not only will these

381
00:23:35,530 --> 00:23:40,190
Jose:
guys perform, but also the deal flow that we get through them is going to be

382
00:23:40,190 --> 00:23:44,890
Jose:
like pre-vetted and give us some access that kind of overcomes that negative selection problem.

383
00:23:45,150 --> 00:23:49,630
Jose:
So we've been kind of digging into that now, and we think that there's edge there for us.

384
00:23:50,370 --> 00:23:53,150
Jose:
We're also looking at China, like we've been looking at China for a while,

385
00:23:53,250 --> 00:23:56,850
Jose:
one of our, actually, both our members of the investment team spend a lot of

386
00:23:56,850 --> 00:24:00,450
Jose:
their time, of the intelligence team spend a lot of their time in China.

387
00:24:00,790 --> 00:24:03,910
Jose:
I believe China is producing like over half of AI engineers.

388
00:24:04,210 --> 00:24:08,710
Jose:
And also the, it's much, the rounds are much cheaper there because there's just

389
00:24:08,710 --> 00:24:10,590
Jose:
less capital, like the US investors are

390
00:24:11,490 --> 00:24:16,110
Jose:
really able to invest in China, like institutionals. And there's obviously concerns

391
00:24:16,110 --> 00:24:17,890
Jose:
like geopolitical concerns and stuff like that.

392
00:24:18,070 --> 00:24:21,130
Jose:
So you've kind of been looking there and figuring out whether there's a way

393
00:24:21,130 --> 00:24:26,210
Jose:
for us to have edge there and to add some value in helping kind of these founders go global.

394
00:24:26,370 --> 00:24:28,950
Jose:
So I think for me, I'm curious what the other guys think actually.

395
00:24:29,150 --> 00:24:33,410
Jose:
And then we're also looking at kind of these secondaries of the big names,

396
00:24:33,510 --> 00:24:36,990
Jose:
the Anthropics, the the groks um the

397
00:24:36,990 --> 00:24:39,830
Jose:
the open ais and kind of figuring out you know

398
00:24:39,830 --> 00:24:43,650
Jose:
whether we have edge there because i think there we're more just trying to capture

399
00:24:43,650 --> 00:24:50,150
Jose:
the the beta versus have a lot of edge but um yeah for me it's a trial it's

400
00:24:50,150 --> 00:24:54,290
Jose:
a trial and error process of like thinking through things going in doing some

401
00:24:54,290 --> 00:24:58,630
Jose:
research and then figuring out being very honest with ourselves if we if we think we have edge or not

402
00:24:58,630 --> 00:25:04,370
Yan:
Yeah no i think the the honesty is is the important one um edge comes in many forms, right?

403
00:25:05,010 --> 00:25:09,710
Yan:
It's selection edge, it's timing edge, it's some informational edge.

404
00:25:09,910 --> 00:25:14,710
Yan:
And then there's some that comes with experience in terms of bet sizing and everything else.

405
00:25:14,810 --> 00:25:20,490
Yan:
And so for us, what we're in the process of doing now is basically trying to

406
00:25:20,490 --> 00:25:22,650
Yan:
understand where we can have an edge.

407
00:25:22,810 --> 00:25:28,910
Yan:
And I think even that on its own is very valuable or it could even be considered

408
00:25:28,910 --> 00:25:33,030
Yan:
an edge and now we're like using this in a very nebulous way but so you know

409
00:25:33,030 --> 00:25:38,330
Yan:
timing wise it's it's it's on the early side for sure right so i i think that's

410
00:25:38,330 --> 00:25:41,430
Yan:
certainly one having the luxury to commit a.

411
00:25:44,710 --> 00:25:48,850
Yan:
To look at this without necessarily needing to generate a return immediately,

412
00:25:48,850 --> 00:25:50,710
Yan:
I think is a huge benefit, right?

413
00:25:51,070 --> 00:25:57,270
Yan:
Where to some extent, other managers as part of their job, they're forced to deploy, right?

414
00:25:57,390 --> 00:26:02,750
Yan:
And so that I think comes with a disadvantage where you might be deploying in

415
00:26:02,750 --> 00:26:03,910
Yan:
areas you don't necessarily want to.

416
00:26:04,030 --> 00:26:08,750
Yan:
So I think the patience itself is a huge benefit and should give us the opportunity

417
00:26:08,750 --> 00:26:10,990
Yan:
to find those unique plays.

418
00:26:11,190 --> 00:26:14,310
Yan:
I think one of the biggest things, and and this is another learning in crypto,

419
00:26:14,530 --> 00:26:17,810
Yan:
is so much of it comes down to bet sizing, right?

420
00:26:17,890 --> 00:26:22,610
Yan:
And it's like, it's really knowing what the opportunity is and whether you're

421
00:26:22,610 --> 00:26:29,190
Yan:
allocating one, five, 10, 50% to a position is really what makes or breaks a

422
00:26:29,190 --> 00:26:31,710
Yan:
lot of these or what really drives, I think, the outperformance.

423
00:26:32,230 --> 00:26:36,610
Ejaaz:
How do you personally figure that out though, Jan? I know you say that and that's

424
00:26:36,610 --> 00:26:39,890
Ejaaz:
what all the investors say, but I want to get inside your head.

425
00:26:40,670 --> 00:26:43,550
Ejaaz:
What's the difference between you being like, you know what,

426
00:26:43,690 --> 00:26:45,450
Ejaaz:
I'm going to give you around $1 to $5 million.

427
00:26:45,770 --> 00:26:48,970
Ejaaz:
And then you're going, you know what, I'm going to pump in $20 million into

428
00:26:48,970 --> 00:26:52,330
Ejaaz:
your thing, which is not something you guys are unknown to, right?

429
00:26:52,550 --> 00:26:53,970
Ejaaz:
So what is that difference?

430
00:26:54,710 --> 00:26:58,530
Anil:
Jan is a great person to ask this question to be honest.

431
00:26:59,290 --> 00:27:02,830
Yan:
The big one is just risk. And so it's understanding, you know,

432
00:27:02,970 --> 00:27:06,290
Yan:
how can this go wrong? And realistically, what is my downside?

433
00:27:06,810 --> 00:27:10,690
Yan:
And then I think sometimes, you know, when things are going well,

434
00:27:10,810 --> 00:27:18,890
Yan:
it's also knowing when to, like on paper, you should be taking position down.

435
00:27:19,070 --> 00:27:26,070
Yan:
But I think there's an edge in understanding the position outside of it relative

436
00:27:26,070 --> 00:27:27,810
Yan:
to the rest of your portfolio, right?

437
00:27:27,910 --> 00:27:31,470
Yan:
And saying, sure, by the book, I should probably be downsizing,

438
00:27:31,630 --> 00:27:35,210
Yan:
but it's more about how is this position relative to the rest of the market?

439
00:27:35,610 --> 00:27:37,530
Yan:
Is everyone else underexposed?

440
00:27:38,350 --> 00:27:42,510
Yan:
Will there be a lot of money coming in. And so I think that ends up really,

441
00:27:42,690 --> 00:27:47,410
Yan:
it's like, it's understanding that your winners are winners and they should

442
00:27:47,410 --> 00:27:50,590
Yan:
remain that way. And so you're either doubling down or leaving them as is.

443
00:27:50,890 --> 00:27:56,270
Yan:
And so it's not often that you get really convicted and it's kind of in those

444
00:27:56,270 --> 00:27:59,130
Yan:
scenarios where a lot of those edges line up, right?

445
00:27:59,230 --> 00:28:05,050
Yan:
I happen to be down this rabbit hole and I found this, it's going to be a lot

446
00:28:05,050 --> 00:28:06,790
Yan:
harder to get access to this in the future.

447
00:28:06,850 --> 00:28:10,550
Yan:
I think it's de-risked more than people actually think.

448
00:28:10,750 --> 00:28:16,690
Yan:
And so it's when the stars align in those scenarios that you really need to just kind of have...

449
00:28:16,690 --> 00:28:18,090
Jose:
You're talking about optronic here?

450
00:28:19,490 --> 00:28:22,790
Yan:
That's one of them. Yeah. And where you just have a lot of...

451
00:28:22,790 --> 00:28:27,370
Yan:
And I think the risk tolerance is a big one too, where thankfully from crypto,

452
00:28:27,690 --> 00:28:29,730
Yan:
you kind of get numb to the volatility.

453
00:28:29,970 --> 00:28:34,250
Yan:
And I think that ends up being a huge edge as well, where you're just able to

454
00:28:34,250 --> 00:28:36,450
Yan:
tolerate swings where if it goes wrong, it goes wrong.

455
00:28:36,610 --> 00:28:40,870
Yan:
But ultimately, more often than night, it will go right. And you really want

456
00:28:40,870 --> 00:28:42,710
Yan:
to be able to capitalize on those opportunities.

457
00:28:43,430 --> 00:28:47,350
Jose:
Yeah, I think that Jan's really good at this. It's probably one of his biggest strengths.

458
00:28:47,510 --> 00:28:50,110
Jose:
And we definitely have a lot of experience just from, in fund one,

459
00:28:50,190 --> 00:28:53,510
Jose:
we started with one position in the fund just by virtue of our size.

460
00:28:53,670 --> 00:28:56,470
Jose:
And the rest of the cycle was us just selling that position to buy others.

461
00:28:56,890 --> 00:29:01,550
Jose:
And so we just, you really, from that, like understand deeply,

462
00:29:01,590 --> 00:29:03,530
Jose:
like the importance of bet sizing.

463
00:29:03,650 --> 00:29:06,370
Jose:
And you also naturally have this like hurdle rate, right? Like,

464
00:29:06,470 --> 00:29:08,390
Jose:
is this thing going to outperform DoorChain?

465
00:29:09,010 --> 00:29:12,710
Jose:
Which was our position at the time. But I think the sizing, that's,

466
00:29:12,810 --> 00:29:15,070
Jose:
yeah, Yeah, one of the biggest things is also one of the biggest things I look

467
00:29:15,070 --> 00:29:20,590
Jose:
for in fund managers, like people who are going to be concentrated and not afraid to take big swings.

468
00:29:20,770 --> 00:29:23,490
Jose:
And it's also one of the biggest mistakes early fund managers make.

469
00:29:24,190 --> 00:29:29,010
Jose:
They want to kind of, and like, concentration just drives all the right behaviors.

470
00:29:29,270 --> 00:29:33,270
Jose:
Like it forces you to think about whether this founder is going to be able to

471
00:29:33,270 --> 00:29:36,530
Jose:
return the fund for you, whether this is someone you want to spend a lot of time with.

472
00:29:37,090 --> 00:29:41,330
Jose:
It forces you to actually add value to the founder. It forces you away from

473
00:29:41,330 --> 00:29:45,830
Jose:
like indexing and just following in to around because Sequoia is in or whatever.

474
00:29:47,470 --> 00:29:54,650
Jose:
So, and then the other thing is just like conviction is, it's like a feeling, right?

475
00:29:54,750 --> 00:29:58,970
Jose:
That you build through research and speaking to someone and thinking about it.

476
00:29:59,110 --> 00:30:02,710
Jose:
But when you have it, it's really important to recognize it because conviction,

477
00:30:02,810 --> 00:30:07,850
Jose:
at least for me, it's not like you can have sort of 10x more conviction in something

478
00:30:07,850 --> 00:30:09,310
Jose:
than you have on anything else.

479
00:30:09,310 --> 00:30:13,190
Jose:
And a lot of people will feel that and size them equally anyway,

480
00:30:13,510 --> 00:30:15,450
Jose:
right? Or like I have to have 10 positions or whatever.

481
00:30:15,770 --> 00:30:19,130
Jose:
But actually if you're conviction, if you have 10X more conviction in something

482
00:30:19,130 --> 00:30:23,910
Jose:
else, you should size it appropriately because those things don't come along that often.

483
00:30:24,150 --> 00:30:27,410
Jose:
You know, there's only probably three to five, if you're lucky,

484
00:30:27,590 --> 00:30:31,070
Jose:
spots a year where you really find that kind of conviction where the stars line up.

485
00:30:31,490 --> 00:30:34,690
Jose:
And when you find it, it's really important to size things correctly.

486
00:30:34,790 --> 00:30:38,770
Jose:
And it's kind of the biggest difference, I think, in performance for people.

487
00:30:39,310 --> 00:30:43,030
Anil:
That's why we wanted to build this research team build this conviction right is like

488
00:30:43,510 --> 00:30:48,530
Anil:
we think we feel confident in our ability to see these opportunities.

489
00:30:48,950 --> 00:30:52,530
Anil:
But if you don't have the conviction, you may not take the swing at the right size.

490
00:30:52,690 --> 00:30:56,150
Anil:
Right. And I think that's going to be really important for us.

491
00:30:56,210 --> 00:30:59,850
Anil:
And then, you know, going back to Josh's question about, you know,

492
00:30:59,950 --> 00:31:01,530
Anil:
obviously, we've been using the word edge a lot.

493
00:31:01,610 --> 00:31:03,930
Anil:
I'll say that, like, you know, EJ started it. So

494
00:31:06,850 --> 00:31:10,510
Anil:
not totally our fault. But the only thing I'd add to what these guys said is

495
00:31:10,510 --> 00:31:15,310
Anil:
like, For me, I think one of the biggest edges that we founded with Delphi is

496
00:31:15,310 --> 00:31:16,230
Anil:
just different perspectives.

497
00:31:16,350 --> 00:31:21,050
Anil:
And I think that's what we're going to seek out with Delphi Intelligence as well.

498
00:31:21,190 --> 00:31:26,530
Anil:
And I think that's not even just within our team, which we really do like building

499
00:31:26,530 --> 00:31:28,890
Anil:
those perspectives and insights within the team.

500
00:31:29,010 --> 00:31:32,110
Anil:
But I think more so just within our trusted network.

501
00:31:33,090 --> 00:31:37,610
Anil:
Within crypto, we lean on our network all the time. And that really helps scale

502
00:31:37,610 --> 00:31:40,770
Anil:
the amount and, you know, the speed at which we learn.

503
00:31:41,070 --> 00:31:43,630
Anil:
That's definitely going to be something we lean on, you know,

504
00:31:43,730 --> 00:31:45,790
Anil:
within other areas that we're trying to explore and learn about.

505
00:31:46,050 --> 00:31:49,710
Ejaaz:
Yeah. So as you guys move into the world of AI, I'm curious if Delphi,

506
00:31:49,830 --> 00:31:53,970
Ejaaz:
as a company, if you individually, you have a framework or a structure of how

507
00:31:53,970 --> 00:31:55,050
Ejaaz:
you think about these opportunities.

508
00:31:55,330 --> 00:31:58,050
Ejaaz:
Because AI is divided into a lot of big categories.

509
00:31:58,370 --> 00:32:01,350
Ejaaz:
I mean, on the show, we like to talk about it as a layer cake almost where you

510
00:32:01,350 --> 00:32:04,130
Ejaaz:
have the chips layer, then you have foundation models and you have dev tools

511
00:32:04,130 --> 00:32:06,710
Ejaaz:
and like infrastructure. and then the top's the application layer.

512
00:32:06,950 --> 00:32:09,630
Ejaaz:
And there's all these different worlds that you could explore,

513
00:32:09,630 --> 00:32:11,050
Ejaaz:
I guess, to get that edge.

514
00:32:11,150 --> 00:32:14,930
Ejaaz:
And I'm curious if any of you or if there's a company-wide kind of tooling or

515
00:32:14,930 --> 00:32:18,570
Ejaaz:
a way that you explore these opportunities and find order in the chaos when

516
00:32:18,570 --> 00:32:19,470
Ejaaz:
you're evaluating everything.

517
00:32:19,730 --> 00:32:23,490
Jose:
I definitely think we have, different people have different perspectives on this.

518
00:32:24,570 --> 00:32:29,090
Jose:
We've looked at things across the layer cake.

519
00:32:29,330 --> 00:32:33,270
Jose:
I think personally, I'm most interested in the top and the bottom.

520
00:32:33,270 --> 00:32:36,510
Jose:
Um i just think that's like

521
00:32:36,510 --> 00:32:39,390
Jose:
um those are the places that tend to be

522
00:32:39,390 --> 00:32:43,850
Jose:
the most defensible so we've looked at a couple of we haven't actually pulled

523
00:32:43,850 --> 00:32:46,150
Jose:
the trigger on any although i actually made a mistake on one of them but we've

524
00:32:46,150 --> 00:32:50,430
Jose:
looked at a bunch of chip startups and and and people doing new new architectures

525
00:32:50,430 --> 00:32:54,010
Jose:
and stuff which have been really interesting um and then for me i'm really bullish

526
00:32:54,010 --> 00:32:57,570
Jose:
on the application layer like i think i think chat gpt rapper is

527
00:32:58,250 --> 00:33:01,490
Jose:
uh people use it as a as sort of

528
00:33:01,490 --> 00:33:04,590
Jose:
you know to throw shade but i think chachapati wrappers

529
00:33:04,590 --> 00:33:06,930
Jose:
are going to be insanely valuable and you're kind of already seeing it with

530
00:33:06,930 --> 00:33:12,770
Jose:
with cursor you know and and others like it and to me ai the capabilities that

531
00:33:12,770 --> 00:33:17,530
Jose:
it has already it could do um probably like 100x more than what people are using

532
00:33:17,530 --> 00:33:22,070
Jose:
it for right now and that gap to me is the product opportunity

533
00:33:22,630 --> 00:33:25,650
Jose:
of creating like verticalized applications

534
00:33:26,230 --> 00:33:28,030
Jose:
with really clean products,

535
00:33:28,670 --> 00:33:34,010
Jose:
with really smooth like context engineering and to solve like particular pain points.

536
00:33:34,030 --> 00:33:38,810
Jose:
And I think you're going to have those across every single vertical and they're

537
00:33:38,810 --> 00:33:40,750
Jose:
going to be, yeah, really, really, really big opportunities.

538
00:33:40,990 --> 00:33:43,410
Jose:
So that's one I'm really excited about.

539
00:33:44,730 --> 00:33:47,990
Jose:
But yeah, we look at stuff all across the stack, I think just to,

540
00:33:48,130 --> 00:33:51,230
Jose:
at this point, just to kind of build knowledge. I mean, actually,

541
00:33:51,370 --> 00:33:53,950
Jose:
in the crypto AI area, we did look at a lot of data stuff, too.

542
00:33:54,210 --> 00:33:58,130
Jose:
We kind of had an intuition that that would be somewhere that crypto would have

543
00:33:58,130 --> 00:34:02,770
Jose:
a particular advantage, like being able to, it's always been kind of a crypto

544
00:34:02,770 --> 00:34:06,690
Jose:
thesis, right? Initially, it was this idea of Web3 Social where everyone would

545
00:34:06,690 --> 00:34:08,750
Jose:
own their own data and you'd get paid for it.

546
00:34:09,330 --> 00:34:13,570
Jose:
But I think the idea of coordinating a bunch of humans to provide valuable data

547
00:34:13,570 --> 00:34:18,350
Jose:
to train AIs always was like an obvious or seemed like an obvious crypto AI idea.

548
00:34:18,530 --> 00:34:20,730
Jose:
So we did make a lot of bets there too.

549
00:34:21,210 --> 00:34:24,870
Jose:
I think we're a little bit more cautious now just given where things are going

550
00:34:24,870 --> 00:34:30,010
Jose:
with synthetic data and just RL and we're being a bit more cautious there.

551
00:34:30,370 --> 00:34:33,390
Jose:
But yeah, those are two that kind of came to mind.

552
00:34:34,190 --> 00:34:37,610
Ejaaz:
So you mentioned that you're bullish ChatGPT wrappers.

553
00:34:38,410 --> 00:34:41,330
Ejaaz:
Can you just give us the bull case for them?

554
00:34:41,470 --> 00:34:45,590
Ejaaz:
Because I, like you, have seen so many people shit on them, basically.

555
00:34:45,930 --> 00:34:47,330
Ejaaz:
Yeah. Why are you so bullish?

556
00:34:48,010 --> 00:34:53,570
Jose:
The sort of precondition for me being bullish on a ChatGPT wrapper is the founders,

557
00:34:53,990 --> 00:34:57,010
Jose:
or the app gets better as the models improve, right?

558
00:34:57,190 --> 00:35:00,390
Jose:
So it actually becomes more useful as the models get better.

559
00:35:00,650 --> 00:35:03,290
Jose:
And there's a lot of examples where that's the case. There was a lot of examples

560
00:35:03,290 --> 00:35:06,930
Jose:
initially to be where you're just building some scaffolding on ChatGPT to do

561
00:35:06,930 --> 00:35:08,930
Jose:
code or therapy or something.

562
00:35:09,110 --> 00:35:12,970
Jose:
And that's not interesting. All that stuff will get picked off by the models.

563
00:35:13,650 --> 00:35:20,210
Jose:
What is interesting is just verticalized applications, which improve as the models get better.

564
00:35:20,410 --> 00:35:24,210
Jose:
And some of them, I think even the more interesting ones or the most interesting

565
00:35:24,210 --> 00:35:26,090
Jose:
ones are the ones which actually don't work right now.

566
00:35:26,250 --> 00:35:30,330
Jose:
They're actually just betting on the models improving enough that one day they'll work well.

567
00:35:30,730 --> 00:35:33,550
Jose:
And there was a bunch of examples of that initially, but I think there's There's

568
00:35:33,550 --> 00:35:35,550
Jose:
some interesting ones now too.

569
00:35:35,970 --> 00:35:39,890
Jose:
But to me, the bull case is just, yeah, kind of what I said earlier,

570
00:35:40,550 --> 00:35:43,390
Jose:
what i said before that to get the most out of

571
00:35:43,390 --> 00:35:46,650
Jose:
out of models is actually hard work like you need pretty

572
00:35:46,650 --> 00:35:49,630
Jose:
good system prompts for whatever uh vertical you're

573
00:35:49,630 --> 00:35:53,150
Jose:
using it for right like if you're using a model for for therapy it needs to

574
00:35:53,150 --> 00:35:56,570
Jose:
not be so agreeable uh it needs to actually like tell you hard truths and stuff

575
00:35:56,570 --> 00:36:00,270
Jose:
like this whereas if you're using the model to write you uh i don't know a twitter

576
00:36:00,270 --> 00:36:03,310
Jose:
or shill post or something then maybe you want it to be persuasive and and stuff

577
00:36:03,310 --> 00:36:06,930
Jose:
like this um if you're using a model for investment due diligence you needed

578
00:36:06,930 --> 00:36:08,510
Jose:
to have access to all your investment notes.

579
00:36:08,510 --> 00:36:11,370
Jose:
You needed to know what the thesis is behind your firm.

580
00:36:11,890 --> 00:36:16,070
Jose:
So there's all this, people call it prompt engineering. I like context engineering,

581
00:36:16,330 --> 00:36:19,830
Jose:
which is a combination of meta-prompt and context.

582
00:36:20,090 --> 00:36:22,390
Jose:
And that stuff is actually really hard.

583
00:36:22,670 --> 00:36:26,310
Jose:
It's hard to get the most out of a model. And there's going to be applications

584
00:36:26,310 --> 00:36:30,070
Jose:
that optimize that process for a specific vertical and just give users really

585
00:36:30,070 --> 00:36:33,610
Jose:
refined experiences for it. Cursor is a great example, I think.

586
00:36:33,950 --> 00:36:38,250
Ejaaz:
But also Anthropic just released Claude Code recently, right?

587
00:36:38,710 --> 00:36:43,570
Ejaaz:
And so I'm curious about your thought around how much of the application layer

588
00:36:43,570 --> 00:36:47,790
Ejaaz:
you think the model makers can actually kind of take, right?

589
00:36:47,970 --> 00:36:49,990
Ejaaz:
So I'll give you another example.

590
00:36:51,190 --> 00:36:56,730
Ejaaz:
XAI just launched GROC 4 and they have this huge distribution network, right? Which is X.

591
00:36:56,990 --> 00:37:01,670
Ejaaz:
And granted, Elon is a very unique case because he's just buying everything.

592
00:37:02,150 --> 00:37:05,370
Ejaaz:
He's probably going to be influencing the chip sector at some point as well.

593
00:37:05,510 --> 00:37:07,830
Ejaaz:
He's putting chips into our brain, blah, blah, blah.

594
00:37:08,350 --> 00:37:12,730
Ejaaz:
And he's building up a massive competitor in terms of data centers.

595
00:37:13,330 --> 00:37:18,570
Ejaaz:
What edge do you think application builders that either you're investing in

596
00:37:18,570 --> 00:37:24,710
Ejaaz:
right now or that you're looking for right now have over what model producers

597
00:37:24,710 --> 00:37:26,650
Ejaaz:
can just kind of replicate themselves?

598
00:37:26,870 --> 00:37:30,030
Ejaaz:
Is it in the context engineering that you're talking about, Jose?

599
00:37:30,490 --> 00:37:36,650
Ejaaz:
Is it the fact that these founders can basically and intuitively describe how

600
00:37:36,650 --> 00:37:39,790
Ejaaz:
an app should behave? Because a lot of this is just around social behavior.

601
00:37:41,090 --> 00:37:46,310
Ejaaz:
The thing that makes an app successful is if you go on it and a bunch of people

602
00:37:46,310 --> 00:37:49,470
Ejaaz:
like it and really vibe with it. That's it.

603
00:37:49,630 --> 00:37:52,030
Ejaaz:
OpenAI just launched their agent yesterday.

604
00:37:52,390 --> 00:37:55,670
Ejaaz:
And the number one bit of feedback I've seen was, this is cool,

605
00:37:55,790 --> 00:37:57,110
Ejaaz:
but what am I going to use it for?

606
00:37:57,210 --> 00:38:01,270
Ejaaz:
And if you have your potential target market saying, what am I going to use it for?

607
00:38:01,490 --> 00:38:04,610
Ejaaz:
You haven't nailed the application there. So I'm wondering whether like there

608
00:38:04,610 --> 00:38:09,510
Ejaaz:
is like, you know, maybe just a list of items that you think separates kind

609
00:38:09,510 --> 00:38:13,850
Ejaaz:
of like founders that are building applications in AI versus like model producers

610
00:38:13,850 --> 00:38:15,930
Ejaaz:
that are just going to like steal their stuff eventually.

611
00:38:16,250 --> 00:38:19,450
Jose:
I think it's a great question. It's kind of the golden question if you're investing

612
00:38:19,450 --> 00:38:23,750
Jose:
in AI applications, like is this something that the models can do?

613
00:38:24,230 --> 00:38:30,710
Jose:
I think coding is an interesting one where, like, if, I think if Claude turns

614
00:38:30,710 --> 00:38:35,310
Jose:
out to be the best coding model for everything, it's going to be hard for,

615
00:38:35,310 --> 00:38:38,170
Jose:
for Cursor to, to win, right? Right.

616
00:38:38,350 --> 00:38:41,850
Jose:
If it's just literally a Claude wrapper, although there's still like cool stuff

617
00:38:41,850 --> 00:38:44,850
Jose:
that Cursor's built, like the, the rules, you know, which are,

618
00:38:45,010 --> 00:38:46,590
Jose:
which I think is a really interesting primitive.

619
00:38:46,590 --> 00:38:52,230
Jose:
I don't know if you guys have used cursor much, but it's a very interesting

620
00:38:52,230 --> 00:38:54,910
Jose:
UX framework that they've built and there's other stuff like that.

621
00:38:54,990 --> 00:39:02,810
Jose:
And I think there's definitely advantages to being laser focused on just pretty

622
00:39:02,810 --> 00:39:04,730
Jose:
much user experience and not having to build your own models.

623
00:39:05,330 --> 00:39:09,210
Jose:
It's hard to answer in the abstract and in the general. I think you have to

624
00:39:09,210 --> 00:39:11,510
Jose:
go kind of like application by application.

625
00:39:11,770 --> 00:39:18,350
Yan:
Yeah, user experience is a big one in the sense. I think one parallel is looking at Gemini, right?

626
00:39:18,450 --> 00:39:24,030
Yan:
And how underutilized it is because it's just the UX is tough, right?

627
00:39:24,230 --> 00:39:29,750
Yan:
And so it's kind of clunky. It doesn't really, it's not as widely used as you'd

628
00:39:29,750 --> 00:39:33,650
Yan:
expect it to be considering how many people are using Gmail and all of that.

629
00:39:33,770 --> 00:39:38,110
Yan:
And so I do think, you know, the UX is a big component.

630
00:39:38,110 --> 00:39:44,090
Yan:
And so it depends on how much of the value is just in the raw processing ability

631
00:39:44,090 --> 00:39:49,110
Yan:
of the model versus how much of the value in the product is in building out

632
00:39:49,110 --> 00:39:52,910
Yan:
everything else around it and making the experience fluid.

633
00:39:53,110 --> 00:39:57,570
Jose:
There's a lot, for instance, Harvey's an interesting one where they've just

634
00:39:57,570 --> 00:40:01,830
Jose:
built a lot of scaffolding, as I understand it, a lot of scaffolding to make

635
00:40:01,830 --> 00:40:05,450
Jose:
the document creation for lawyers extremely fast and seamless.

636
00:40:06,190 --> 00:40:11,410
Ejaaz:
So Harvey AI, just for context for the listeners is like ChatGPT for lawyers. Is that right, Jose?

637
00:40:12,190 --> 00:40:18,490
Jose:
Yeah, basically, for creating memos and stuff like this. And you want to be able to have your firm's

638
00:40:19,790 --> 00:40:22,690
Jose:
standard boilerplate stuff and like whatever the style

639
00:40:22,690 --> 00:40:25,670
Jose:
is that your firm writes in the key

640
00:40:25,670 --> 00:40:28,510
Jose:
documents and you want to go document by document because this isn't

641
00:40:28,510 --> 00:40:31,330
Jose:
this is like very high uh fake stuff that

642
00:40:31,330 --> 00:40:34,670
Jose:
you don't want to get wrong and and i think that's going to be the case for

643
00:40:34,670 --> 00:40:39,590
Jose:
like almost every vertical is going to have this uh and because like reliability

644
00:40:39,590 --> 00:40:43,070
Jose:
is also a huge thing kind of talked about that before but these these models

645
00:40:43,070 --> 00:40:47,310
Jose:
are not uh they're getting more and more but they still have hallucinations

646
00:40:47,310 --> 00:40:50,670
Jose:
and not super consistent um that that's another thing that the

647
00:40:51,010 --> 00:40:54,670
Jose:
kind of verticalized applications can can help fix with really good scaffolding

648
00:40:54,670 --> 00:40:58,310
Jose:
and system prompts and stuff um but yeah i think harvey and cursor probably

649
00:40:58,310 --> 00:41:04,490
Jose:
the two biggest examples of ones so far that i think have have built cool stuff

650
00:41:04,490 --> 00:41:08,570
Jose:
on top of um like a basic wrapper nice

651
00:41:08,570 --> 00:41:11,850
Anil:
Yeah i do also think customization is going to be a big key and i'm you know

652
00:41:11,850 --> 00:41:15,150
Anil:
i wanted to jump in after you because I think like, this is something I go back

653
00:41:15,150 --> 00:41:18,830
Anil:
and forth on a lot is a lot of these model creators obviously have a lot of

654
00:41:18,830 --> 00:41:20,930
Anil:
data on, you know, who is paying for compute,

655
00:41:21,050 --> 00:41:24,770
Anil:
how much they're paying and, you know, can very quickly figure out why,

656
00:41:24,970 --> 00:41:27,210
Anil:
you know, if this person is paying, they're obviously building something that

657
00:41:27,210 --> 00:41:28,750
Anil:
is valuable. Let's go copy and paste that.

658
00:41:28,970 --> 00:41:32,430
Anil:
And yeah, to Ejaz's point, obviously a lot of, you know, these guys are all

659
00:41:32,430 --> 00:41:36,170
Anil:
going towards this agent space, towards like creating something that is scalable

660
00:41:36,170 --> 00:41:37,350
Anil:
to, you know, the masses.

661
00:41:37,590 --> 00:41:40,850
Anil:
I think, you know, the last decade was very much about, you know,

662
00:41:40,910 --> 00:41:43,250
Anil:
there's an app for that. and I think

663
00:41:43,790 --> 00:41:46,590
Anil:
upcoming decade will be very much like there's an app for you,

664
00:41:46,870 --> 00:41:48,570
Anil:
right? So very like custom app.

665
00:41:49,110 --> 00:41:53,410
Anil:
Maybe Jose, like, I don't know if you want to leak or share some of the conversations

666
00:41:53,410 --> 00:41:57,490
Anil:
we were having this week about like, something labs is building for Delphi itself.

667
00:41:57,670 --> 00:42:00,590
Anil:
I don't know if you want to go into that. But like, I think that's a great example

668
00:42:00,590 --> 00:42:04,050
Anil:
of like something that, you know, yes, we know a lot of these model creators

669
00:42:04,050 --> 00:42:07,750
Anil:
will have something that will probably accomplish 70 to 80%,

670
00:42:07,750 --> 00:42:10,470
Anil:
if not, maybe even more, you know, in the future for us.

671
00:42:10,590 --> 00:42:13,350
Anil:
But it's something that, you know, So I think Louds wanted to roll up their

672
00:42:13,350 --> 00:42:16,810
Anil:
sleeves, get their hands dirty and build something custom fit for us that would

673
00:42:16,810 --> 00:42:19,010
Anil:
be, you know, fulfill basically 100% of our needs.

674
00:42:19,170 --> 00:42:21,890
Jose:
There's like Delphi, we operate, we like to call it like the hive mind,

675
00:42:22,030 --> 00:42:24,150
Jose:
right? It's also the name of our pod.

676
00:42:24,310 --> 00:42:27,890
Jose:
And it really operates that way where there's a bunch of people in different

677
00:42:27,890 --> 00:42:30,950
Jose:
divisions, some doing research, some building stuff, some investing that are

678
00:42:30,950 --> 00:42:32,310
Jose:
having a bunch of interesting calls.

679
00:42:33,450 --> 00:42:39,450
Jose:
And right now it's the sort of bandwidth between surfacing the interesting conversations

680
00:42:39,450 --> 00:42:42,530
Jose:
for the whole firm to benefit from is really slow.

681
00:42:42,810 --> 00:42:46,130
Jose:
Like we have to schedule these like bi-weekly calls. And then by the time that's

682
00:42:46,130 --> 00:42:47,750
Jose:
happened, people have forgotten about it.

683
00:42:47,910 --> 00:42:54,030
Jose:
And so I think the initial sort of vision is for it to be sort of an organizational

684
00:42:54,030 --> 00:42:58,510
Jose:
knowledge base, or like we call it, you know, Delphi OS or Hivemind OS,

685
00:42:58,750 --> 00:43:02,650
Jose:
which can just, first of all, like have all the conversations that people are

686
00:43:02,650 --> 00:43:08,270
Jose:
having across the firms in a retrievable and like queryable format and then

687
00:43:08,270 --> 00:43:10,190
Jose:
building like intelligence on top of that.

688
00:43:10,370 --> 00:43:13,330
Jose:
So this thing can, for instance, generate IC memos really easily.

689
00:43:13,470 --> 00:43:17,850
Jose:
Like I have a bunch of calls of the project and then it has our IC memo format.

690
00:43:18,030 --> 00:43:23,190
Jose:
Maybe I can put in podcasts that the founder's done and then I can answer some

691
00:43:23,190 --> 00:43:26,330
Jose:
questions to the AI and then it can just generate an IC memo format,

692
00:43:26,430 --> 00:43:28,390
Jose:
you know, something that takes me kind of hours to do.

693
00:43:28,770 --> 00:43:34,090
Jose:
You might have the same with research. Or for instance, if we want to have a

694
00:43:34,090 --> 00:43:37,850
Jose:
kind of CRM of all the companies that we've ever spoken to, we can see all the

695
00:43:37,850 --> 00:43:41,010
Jose:
conversations people have had with people at this company and also all the conversations

696
00:43:41,010 --> 00:43:42,930
Jose:
people have had about this company, right?

697
00:43:43,050 --> 00:43:47,350
Jose:
We can sort of search this and see, oh, this founder actually leaked to Malifaux.

698
00:43:47,570 --> 00:43:52,030
Jose:
Like these guys are not performing well. They ended up using a different service provider or whatever.

699
00:43:52,510 --> 00:43:57,570
Jose:
Like, and we want to have, and I think every company will basically have this in the future.

700
00:43:57,570 --> 00:44:02,150
Jose:
Like it'll all the knowledge of the company will feed into this to this uh central

701
00:44:02,150 --> 00:44:06,070
Jose:
like memory knowledge base whatever you want to call it and then there'll be

702
00:44:06,070 --> 00:44:09,350
Jose:
various kinds of agents you can run on it that both help the company operate

703
00:44:09,350 --> 00:44:13,690
Jose:
better and just automate and augment its people to be able to to be able to do more you

704
00:44:13,690 --> 00:44:17,190
Anil:
Know you could kind of see this getting kind of crazier as time goes on right

705
00:44:17,190 --> 00:44:20,830
Anil:
like recently we just had this big founders retreat and we always like to like

706
00:44:20,830 --> 00:44:23,410
Anil:
kind of like share a book that we all read and stuff like that and this book

707
00:44:23,410 --> 00:44:26,930
Anil:
for this last week was Essentialism by Greg McCohen, right?

708
00:44:27,110 --> 00:44:33,610
Anil:
And you could see us using all this data that this knowledge base like fills

709
00:44:33,610 --> 00:44:37,530
Anil:
and then in our chat add an agent that is based off Greg McCohen who like kind

710
00:44:37,530 --> 00:44:41,610
Anil:
of follows our calls and then kind of shits on us whenever we're drifting away

711
00:44:41,610 --> 00:44:43,790
Anil:
from, you know, what the thesis of his book is.

712
00:44:44,010 --> 00:44:46,950
Anil:
So it's not like us holding each other accountable, but this agent almost holding

713
00:44:46,950 --> 00:44:49,490
Anil:
us accountable to the decisions we're making at an org level.

714
00:44:49,790 --> 00:44:52,630
Anil:
So yeah, I think we're super excited to play around with it.

715
00:44:52,630 --> 00:44:54,930
Anil:
And I think it will be super useful for other companies.

716
00:44:55,090 --> 00:44:58,350
Anil:
And at the same time, to answer your question, do I think this is something

717
00:44:58,350 --> 00:45:02,350
Anil:
that like the big models like OpenAI, Anthropic, et cetera, you know,

718
00:45:02,450 --> 00:45:04,110
Anil:
Crop are not going to build in?

719
00:45:04,210 --> 00:45:08,250
Anil:
No, of course, they're obviously building it right now, as we've seen with all

720
00:45:08,250 --> 00:45:09,110
Anil:
these recent announcements.

721
00:45:09,290 --> 00:45:11,690
Anil:
But I think the customization is something that's really special.

722
00:45:11,810 --> 00:45:15,510
Anil:
I think will be like, you know, again, what I said earlier, an app for everyone

723
00:45:15,510 --> 00:45:18,050
Anil:
rather than here's an app for, you know, you.

724
00:45:18,960 --> 00:45:21,640
Ejaaz:
Yeah, this is a super interesting point, right? Because you were able to build

725
00:45:21,640 --> 00:45:26,160
Ejaaz:
Delphi OS using AI, and that would previously have been something that you'd

726
00:45:26,160 --> 00:45:29,960
Ejaaz:
have to go to a larger company or use a lot of resources in-house to develop.

727
00:45:30,100 --> 00:45:31,120
Ejaaz:
It's become much easier.

728
00:45:31,380 --> 00:45:34,260
Ejaaz:
And then you mentioned that, well, Grok is probably going to integrate this.

729
00:45:34,640 --> 00:45:37,960
Ejaaz:
ChatGPT will probably see these types of tools in. I'm curious where you see

730
00:45:37,960 --> 00:45:43,020
Ejaaz:
the most forming, because a lot of the new innovations tend to become commoditized fairly quickly.

731
00:45:43,280 --> 00:45:45,880
Ejaaz:
And I think one of the most that we've seen perform the best,

732
00:45:46,000 --> 00:45:48,940
Ejaaz:
at least in the consumer world, which is what a lot of the people who are listening

733
00:45:48,940 --> 00:45:51,420
Ejaaz:
to. are involved in, is ChatGPT's memory function.

734
00:45:51,740 --> 00:45:54,740
Ejaaz:
And memory is amazing because it includes all the context of previous conversations

735
00:45:54,740 --> 00:45:57,760
Ejaaz:
you've had, and it really locks people into that platform.

736
00:45:57,980 --> 00:46:03,820
Ejaaz:
But outside of memory, I haven't really seen many other moats that make me want to use a model.

737
00:46:03,920 --> 00:46:07,320
Ejaaz:
So I'm curious what your takes on moats are, if they're possible to capture

738
00:46:07,320 --> 00:46:12,320
Ejaaz:
a large amount of a user base, or is it just going to be commoditized software

739
00:46:12,320 --> 00:46:15,440
Ejaaz:
all the way up, all the models get better, they all kind of copy everyone's features.

740
00:46:15,880 --> 00:46:18,080
Ejaaz:
Is there any moats that you guys are excited about?

741
00:46:18,080 --> 00:46:25,140
Yan:
One funny one is there's a big moat to the brand and what kind of gets normalized, right?

742
00:46:25,280 --> 00:46:30,260
Yan:
So as we kind of all agreed on earlier, we are using a very,

743
00:46:30,360 --> 00:46:32,460
Yan:
very small fraction of the potential of these, right?

744
00:46:32,620 --> 00:46:36,020
Yan:
And so if you think about the earliest adopters of this tech,

745
00:46:36,220 --> 00:46:42,900
Yan:
which, you know, ChatGPT has an insane amount of users, but the penetration is still pretty low.

746
00:46:43,060 --> 00:46:48,140
Yan:
And that's why it's so valuable. And so the first cohort is going to be kind

747
00:46:48,140 --> 00:46:52,740
Yan:
of the most diligent about figuring out, okay, this one is better for this.

748
00:46:52,840 --> 00:46:53,840
Yan:
This one is better for this.

749
00:46:55,380 --> 00:46:59,780
Yan:
But each incremental onboarder is going to be less particular.

750
00:47:00,120 --> 00:47:02,960
Yan:
And at the same time, all of the models will keep getting better.

751
00:47:03,160 --> 00:47:08,320
Yan:
So what that basically means is each one will continue to use less and less incremental.

752
00:47:08,580 --> 00:47:11,320
Yan:
Of the potential of this thing and and and they're all going

753
00:47:11,320 --> 00:47:14,460
Yan:
to be relatively commoditized for their use case and so what

754
00:47:14,460 --> 00:47:17,340
Yan:
it'll boil down to is what gets normalized you

755
00:47:17,340 --> 00:47:20,400
Yan:
know uh going back to you know use xerox

756
00:47:20,400 --> 00:47:26,180
Yan:
uh like for copying then google everywhere you google it and then now like chad

757
00:47:26,180 --> 00:47:30,280
Yan:
gpt has has won that so far right that's just kind of the one that comes into

758
00:47:30,280 --> 00:47:34,300
Yan:
mind for anyone who's looking to start dabbling in this and i think you know

759
00:47:34,300 --> 00:47:38,380
Yan:
that as an onboarding tool and as a customer acquisition tool can't really be slept on?

760
00:47:38,540 --> 00:47:43,160
Jose:
In general, AI stuff has less of a network effect than the web two giants did, right?

761
00:47:43,280 --> 00:47:47,900
Jose:
Like social media and ad based stuff has way bigger network effect where it's

762
00:47:47,900 --> 00:47:49,080
Jose:
just much harder to disrupt.

763
00:47:50,160 --> 00:47:54,860
Jose:
But I think, I mean, the moat in AI, there's some things that have a data moat, right?

764
00:47:54,980 --> 00:47:59,720
Jose:
Someone like Tesla that has like so many hours of driving data and there's other

765
00:47:59,720 --> 00:48:02,660
Jose:
like robotics companies that we've looked at where that's a moat.

766
00:48:02,780 --> 00:48:07,120
Jose:
I think OpenAI in itself, the amount of chats that they have and the ability

767
00:48:07,120 --> 00:48:11,300
Jose:
to use that for training and things like this is also somewhat of a moat.

768
00:48:11,940 --> 00:48:17,820
Jose:
But I do think in AI that the main moat is just going to be UX and speed,

769
00:48:18,260 --> 00:48:24,240
Jose:
the team that is the best at constantly shifting to where the meta is and building the next thing.

770
00:48:24,660 --> 00:48:27,600
Jose:
Ideally, you don't want your memory to sit with ChatGPT or whoever.

771
00:48:27,800 --> 00:48:30,420
Jose:
And this is, I think, pretty visceral for people when they're sharing.

772
00:48:30,540 --> 00:48:33,720
Jose:
I've shared some pretty personal stuff with ChatGPT.

773
00:48:35,000 --> 00:48:38,880
Jose:
I think we all have. Yeah, like more personal than I ever thought I would have.

774
00:48:41,340 --> 00:48:45,660
Jose:
So I think ideally, remember, we would actually sit and we have a project that

775
00:48:45,660 --> 00:48:47,360
Jose:
we're incubating that's actually building this.

776
00:48:47,560 --> 00:48:53,760
Jose:
Ideally, you would have private memory built on TE or ideally FHE once that works.

777
00:48:53,920 --> 00:48:58,580
Jose:
And then you would give in sort of a cursor like UX, you'd be able to choose

778
00:48:58,580 --> 00:49:02,540
Jose:
which model you want to give permission to access certain parts of that context

779
00:49:02,540 --> 00:49:04,000
Jose:
to answer a query, right?

780
00:49:04,280 --> 00:49:06,980
Jose:
I mean, the ideal ideal would just be you have a model that runs locally,

781
00:49:07,000 --> 00:49:08,080
Jose:
but I think that's going to be

782
00:49:08,750 --> 00:49:11,390
Jose:
super tough um so i think

783
00:49:11,390 --> 00:49:14,230
Jose:
that's like one interesting area but i agree in

784
00:49:14,230 --> 00:49:16,970
Jose:
general like the moats and that's why we've also been looking at

785
00:49:16,970 --> 00:49:20,970
Jose:
deep tech stuff i do think the moats sort of end up also moving to like hardware

786
00:49:20,970 --> 00:49:27,090
Jose:
to ip um to just things that in the past were seen as not sexy you know like

787
00:49:27,090 --> 00:49:30,790
Jose:
uh it's not software it's it's too hard but i think those things will actually

788
00:49:30,790 --> 00:49:36,770
Jose:
have like some of the most persistent moats in an era of of uh of ai and just insanely

789
00:49:37,510 --> 00:49:38,710
Jose:
deflated cost of software.

790
00:49:39,070 --> 00:49:43,750
Anil:
Yeah. I'd say that like on the memory front, I really hope that's not a moat, right?

791
00:49:43,890 --> 00:49:47,810
Anil:
Like I, if memory is a moat, that just means that you're kind of like stuck

792
00:49:47,810 --> 00:49:51,050
Anil:
into the, one of these ecosystems and you're really relying on that one builder

793
00:49:51,050 --> 00:49:53,930
Anil:
to build every, you know, the best app of everything.

794
00:49:54,190 --> 00:49:58,090
Anil:
Whereas like, you know, um, yeah. So, you know, to Jose's point,

795
00:49:58,190 --> 00:50:01,410
Anil:
yeah, we are incubating a project that is, you know, based off this thesis that

796
00:50:01,410 --> 00:50:03,710
Anil:
memory won't be locked in in one place and won't be disemoted.

797
00:50:03,930 --> 00:50:10,210
Ejaaz:
So I feel like this whole memory term is just like another term to describe data, right?

798
00:50:10,350 --> 00:50:14,990
Ejaaz:
And that's what all the top social media technology platforms have nailed so

799
00:50:14,990 --> 00:50:18,310
Ejaaz:
far, right? They just aggregate the most amount of data.

800
00:50:18,510 --> 00:50:22,170
Ejaaz:
I mean, Jose, you just mentioned that you use so much personal stuff or you

801
00:50:22,170 --> 00:50:23,610
Ejaaz:
say so much personal stuff to ChatGPT.

802
00:50:23,770 --> 00:50:27,050
Ejaaz:
I am talking to this thing for hours on end, right?

803
00:50:27,210 --> 00:50:31,190
Ejaaz:
So at this point, I'm just like naturally inclined to use ChatGPT,

804
00:50:31,350 --> 00:50:35,170
Ejaaz:
even though there's like another model that comes out. I really hope the portability

805
00:50:35,170 --> 00:50:38,930
Ejaaz:
gets figured out Anil, to your point I just don't know what the incentives would

806
00:50:38,930 --> 00:50:41,410
Ejaaz:
be for some of the bigger model producers

807
00:50:41,990 --> 00:50:45,130
Jose:
It's sort of different from social media though because in social media it's

808
00:50:45,130 --> 00:50:48,230
Jose:
not just the data it's the fact that all your friends are on there so you don't

809
00:50:48,230 --> 00:50:51,430
Jose:
just have to port over your data you have to get all your friends to sign up

810
00:50:51,430 --> 00:50:55,190
Jose:
to whatever new thing whatever web through social media thing you're using whereas here,

811
00:50:55,370 --> 00:51:00,770
Jose:
portability you actually have access to all your chatchipity chats and it's

812
00:51:00,770 --> 00:51:04,590
Jose:
not that heavy no matter how much you've talked to it It's text data,

813
00:51:04,710 --> 00:51:06,610
Jose:
you know, it fits on any computer.

814
00:51:07,070 --> 00:51:12,090
Jose:
So I do think if someone builds a great user experience here,

815
00:51:12,310 --> 00:51:17,230
Jose:
it's something where it can actually win because it's a better product,

816
00:51:17,310 --> 00:51:19,610
Jose:
like fundamentally. It just has to work really well.

817
00:51:20,180 --> 00:51:23,420
Anil:
I also think, Ejaz, you put out this tweet, I don't even know if it was today

818
00:51:23,420 --> 00:51:25,180
Anil:
or yesterday, it's been a long week,

819
00:51:25,280 --> 00:51:28,900
Anil:
but you talk about the different personalities of these models, right?

820
00:51:29,160 --> 00:51:32,540
Anil:
I think that's an interesting way to think about Emote as well, right?

821
00:51:32,940 --> 00:51:36,300
Anil:
The conversations I have with Rock are way different than the conversations

822
00:51:36,300 --> 00:51:38,020
Anil:
I have with like O3, right?

823
00:51:39,000 --> 00:51:41,860
Anil:
So yeah, I think that's an interesting way to think about it as well.

824
00:51:42,120 --> 00:51:45,300
Ejaaz:
No, that's a good point. For those of you who are wondering what this tweet

825
00:51:45,300 --> 00:51:49,940
Ejaaz:
said, I basically described all the top models as having different personalities.

826
00:51:50,220 --> 00:51:55,740
Ejaaz:
So I said, Grok was kind of like, whatever, naughty and rude and extremely horny, just to be frank.

827
00:51:56,000 --> 00:51:59,980
Ejaaz:
And then ChadGBT was like kind of this incredibly agreeable personality.

828
00:52:00,640 --> 00:52:04,520
Ejaaz:
Claude was kind of like, yeah, there you go. There you go. It's much more human,

829
00:52:04,640 --> 00:52:05,380
Ejaaz:
right? It's much more intuitive.

830
00:52:05,560 --> 00:52:08,480
Ejaaz:
Anyway, they have a bunch of different personalities and it kind of like attracts

831
00:52:08,480 --> 00:52:12,420
Ejaaz:
a certain type of audience or it kind of like secretly molds you into being

832
00:52:12,420 --> 00:52:14,000
Ejaaz:
some one type of a user, right?

833
00:52:14,020 --> 00:52:17,220
Ejaaz:
You end up saying information to one model that you went to another.

834
00:52:17,400 --> 00:52:21,520
Ejaaz:
And it just kind of like Chris's weird kind of sociodynamics that I think are interesting.

835
00:52:21,700 --> 00:52:26,380
Ejaaz:
But kind of moving on, guys, I remember when you first started your fund in

836
00:52:26,380 --> 00:52:31,640
Ejaaz:
2019, the stuff that you guys were investing in, I thought you guys were insane.

837
00:52:31,820 --> 00:52:34,920
Ejaaz:
And this is coming from someone that like worked in the space, right?

838
00:52:35,280 --> 00:52:38,960
Ejaaz:
And then of course, years passed, and it turns out that you guys nailed it.

839
00:52:39,140 --> 00:52:42,880
Ejaaz:
So my natural question now that you're focusing on AI and investing so much

840
00:52:42,880 --> 00:52:46,680
Ejaaz:
in AI is, and I'm going to put each of you in the hot seat, so prep your answer,

841
00:52:46,680 --> 00:52:53,960
Ejaaz:
is what is one emerging contrarian trend in AI right now that you think everyone is missing,

842
00:52:54,180 --> 00:52:58,180
Ejaaz:
but they should 100% focus on because it's going to become a big thing over

843
00:52:58,180 --> 00:52:59,020
Ejaaz:
the next couple of years?

844
00:52:59,180 --> 00:53:06,640
Jose:
I guess one thing I'd say is I don't think you necessarily have to be contrarian in venture, actually.

845
00:53:07,120 --> 00:53:11,720
Jose:
I think you have to be right, but not necessarily contrarian.

846
00:53:12,760 --> 00:53:18,260
Jose:
Although it helps for sure. Like it definitely is helpful when you're looking

847
00:53:18,260 --> 00:53:20,760
Jose:
at something and you're really bullish on it and no one else happens to be.

848
00:53:21,820 --> 00:53:24,640
Jose:
But I do think just, yeah.

849
00:53:25,940 --> 00:53:29,960
Jose:
I mean, the area I'd say is the one I already spoke about, which is like GPT wrappers.

850
00:53:30,100 --> 00:53:33,080
Jose:
I think a lot of people are sleeping on them and I think they're going to be

851
00:53:33,080 --> 00:53:36,600
Jose:
absolutely like giant kind of businesses.

852
00:53:36,820 --> 00:53:41,440
Ejaaz:
What's a GPT wrapper that isn't like a coding wrapper that you think people

853
00:53:41,440 --> 00:53:43,280
Ejaaz:
should focus on or pay attention to?

854
00:53:44,140 --> 00:53:48,160
Jose:
Um, I mean, this, this application that, that we're building internally,

855
00:53:48,180 --> 00:53:51,220
Jose:
and there's a couple of teams that we've spoken to that are, that are building it.

856
00:53:51,560 --> 00:53:54,740
Jose:
Uh, one of them is, is Den. It's like a YC company.

857
00:53:55,000 --> 00:53:57,720
Jose:
So it's kind of like, think about it as cursor for, for work,

858
00:53:57,960 --> 00:54:01,620
Jose:
right? It ingests like all your work data, your emails, your memos, your calls.

859
00:54:01,800 --> 00:54:05,840
Jose:
Um, and, and, and then you're able to use any model to like run on that data.

860
00:54:06,180 --> 00:54:10,220
Jose:
They also built a Slack clone, which I think is really interesting because the

861
00:54:10,220 --> 00:54:13,660
Jose:
idea being that you're chatting with these models anyway, and actually in the future.

862
00:54:14,280 --> 00:54:18,000
Jose:
And so you can open these chat groups with a model and your team in them.

863
00:54:18,140 --> 00:54:20,920
Jose:
And you can all chat to the model together in these groups and have different

864
00:54:20,920 --> 00:54:22,940
Jose:
models in the different chats, which I think is really interesting.

865
00:54:23,080 --> 00:54:26,960
Jose:
The idea that you're chatting already, why not have a chat app where you can

866
00:54:26,960 --> 00:54:30,220
Jose:
have group chats with the models and they can be on calls and stuff like this.

867
00:54:31,260 --> 00:54:35,240
Jose:
I think various versions of those, I think you'll have a new Slack.

868
00:54:36,160 --> 00:54:41,780
Jose:
I think all the company CRM stuff that Salesforce does right now is going to be rebuilt around AI.

869
00:54:41,960 --> 00:54:46,820
Jose:
I have to think of more some more examples of good rap. Those are the ones I've mainly been focused on.

870
00:54:47,080 --> 00:54:50,680
Jose:
But I think in hiring, for instance, you're definitely going to have something

871
00:54:50,680 --> 00:54:55,680
Jose:
like that that's just going to know exactly what kind of person you're looking for.

872
00:54:55,820 --> 00:54:58,260
Jose:
It can do the interviews for you, sort candidates for you.

873
00:54:59,720 --> 00:55:05,660
Jose:
In every vertical you can think of, AI is going to have, you're going to want

874
00:55:05,660 --> 00:55:08,520
Jose:
AI to do a huge percentage of the work

875
00:55:08,520 --> 00:55:12,000
Jose:
and there's going to be an app that facilitates that workflow, I think.

876
00:55:12,900 --> 00:55:16,140
Anil:
You're giving more high-level ideas, though. I feel like Ejaz wanted specific,

877
00:55:16,440 --> 00:55:20,100
Ejaaz:
Right? I want specific. A specific company. Yeah, exactly. Yeah, yeah, yeah.

878
00:55:21,100 --> 00:55:24,800
Ejaaz:
And the crazier, the better, honestly. Yeah, the crazier, the better. Just lean in.

879
00:55:25,200 --> 00:55:28,100
Anil:
Mine aren't going to be crazy, and I hope Jan and Jose will make up for that.

880
00:55:28,200 --> 00:55:32,000
Anil:
But going off of what Jose said, which is the contrarian part,

881
00:55:32,000 --> 00:55:34,240
Anil:
I think, is over-indexed.

882
00:55:34,540 --> 00:55:38,040
Anil:
And I think in crypto venture, it definitely worked out really well for us.

883
00:55:38,120 --> 00:55:41,580
Anil:
But also, I think nowadays in crypto, there's not much stuff that is contrarian.

884
00:55:42,000 --> 00:55:46,820
Anil:
Every conversation you have, people are bullish, hype or pump or something like that.

885
00:55:47,700 --> 00:55:50,720
Anil:
But I think, you know, when it comes to, you know, just generally AI,

886
00:55:50,960 --> 00:55:54,700
Anil:
I think for us, we thought the contrarian thing was, we thought even the most

887
00:55:54,700 --> 00:55:57,380
Anil:
bullish people were going to be underexposed, right? So for us,

888
00:55:57,440 --> 00:55:58,540
Anil:
we just want to be underexposed.

889
00:55:59,280 --> 00:56:02,660
Anil:
And, you know, the thing that I go back and forth on, you know,

890
00:56:02,720 --> 00:56:07,540
Anil:
to Jan's point is like, I think finding alpha here is going to be extremely difficult.

891
00:56:08,240 --> 00:56:11,560
Anil:
Obviously, we're up for the challenge, but I think it's going to be extremely difficult. difficult.

892
00:56:11,820 --> 00:56:15,580
Anil:
So for me, what I've been kind of pushing internally, and I think,

893
00:56:15,640 --> 00:56:19,000
Anil:
you know, this is open to kind of like any anyone inside or outside of Delphi

894
00:56:19,000 --> 00:56:21,680
Anil:
is, you know, capture a lot of this beta exposure.

895
00:56:21,920 --> 00:56:26,200
Anil:
I think sometimes like investors and people just like to work very hard to,

896
00:56:26,320 --> 00:56:28,200
Anil:
you know, to feel like they're smart.

897
00:56:28,680 --> 00:56:31,680
Anil:
But I think almost like, you know, you can capture, you know,

898
00:56:31,760 --> 00:56:35,900
Anil:
a nice index of, you know, open AI, Anthropic, like Andurl, Neuralink,

899
00:56:36,000 --> 00:56:40,760
Anil:
all this stuff, and capture a lot of this beta upside in a lot of these like

900
00:56:40,760 --> 00:56:43,120
Anil:
sectors that you think are going to be massive, right?

901
00:56:43,740 --> 00:56:46,560
Anil:
Even in the public equities, I think like companies like Google,

902
00:56:47,040 --> 00:56:51,000
Anil:
you know, maybe Tesla and stuff like that, I think are worth like looking at.

903
00:56:51,380 --> 00:56:55,060
Anil:
You know, I'm super bullish Google, for example, even though people maybe,

904
00:56:55,320 --> 00:56:57,980
Anil:
you know, are dancing on their graves because they're thinking that,

905
00:56:58,140 --> 00:57:01,100
Anil:
you know, their big search is going to be like cannibalized by AI, right?

906
00:57:01,260 --> 00:57:04,720
Anil:
Or open AI is launching this browser, which is going to like kill Chrome or

907
00:57:04,720 --> 00:57:09,240
Anil:
something like that. Um, so yeah, I think, you know, again, definitely not contrarian, right?

908
00:57:09,300 --> 00:57:12,200
Anil:
I'm literally fucking talking about Google and, uh, you know,

909
00:57:12,600 --> 00:57:16,620
Anil:
open AI and stuff, but I do think that people will mid curve it and say,

910
00:57:16,740 --> 00:57:19,940
Anil:
you know, that's too easy or, oh, these things have run away.

911
00:57:20,060 --> 00:57:23,480
Anil:
Like maybe the 10 X is behind me or a hundred X is behind me or something like

912
00:57:23,480 --> 00:57:24,580
Anil:
that. So let me try and find that

913
00:57:24,580 --> 00:57:27,200
Anil:
a hundred X and then probably invest in things that go to zero instead.

914
00:57:27,540 --> 00:57:31,400
Anil:
Right. Um, so that's my kind of answer, but, um, and then, you know,

915
00:57:31,500 --> 00:57:36,700
Anil:
more in Jose's vein of like giving, broad ideas and not specific names.

916
00:57:36,860 --> 00:57:41,220
Anil:
I think one idea that I think will be massive in the next, I don't know,

917
00:57:41,300 --> 00:57:43,180
Anil:
12 to 18 months is I think...

918
00:57:43,910 --> 00:57:47,010
Anil:
If you're using Twitter nowadays, you kind of get really annoyed at all these

919
00:57:47,010 --> 00:57:51,250
Anil:
bots, right? And these agents that are like, in your replies, they're really bad.

920
00:57:51,370 --> 00:57:54,230
Anil:
And so a lot of people are kind of like looking for a social network that is

921
00:57:54,230 --> 00:57:58,230
Anil:
like, you know, people only, right? Maybe you do this world corner, whatever the fuck.

922
00:57:58,410 --> 00:58:02,050
Anil:
I think actually the opposite is even more interesting where it's like a one

923
00:58:02,050 --> 00:58:08,450
Anil:
on, you know, one where it's like you entering a social network where it's all agents, right?

924
00:58:08,610 --> 00:58:13,310
Anil:
And you basically can kind of like get these agents to have a conversation about

925
00:58:13,310 --> 00:58:17,190
Anil:
whatever you want based on personalities that you actually do follow, right?

926
00:58:17,450 --> 00:58:20,190
Anil:
Instead of, you know, people listen to all in podcasts, and you're waiting for,

927
00:58:20,330 --> 00:58:23,230
Anil:
you know, the topics that they're talking about, hoping to talk about a topic

928
00:58:23,230 --> 00:58:27,130
Anil:
that is maybe relevant to you, you can kind of create your own podcast of those

929
00:58:27,130 --> 00:58:28,730
Anil:
personalities, personalities you

930
00:58:28,730 --> 00:58:31,610
Anil:
do want to follow talking about the exact topic you want to talk about.

931
00:58:31,710 --> 00:58:34,610
Anil:
So I think something like that will be really cool. And I think will kind of

932
00:58:34,610 --> 00:58:36,150
Anil:
exist in the next like 12 to 18 months.

933
00:58:36,890 --> 00:58:40,190
Anil:
I don't know if the company exists yeah but um that's something that

934
00:58:40,190 --> 00:58:43,010
Ejaaz:
I i think like meta is that's that part

935
00:58:43,010 --> 00:58:46,250
Ejaaz:
of their strategy is just to kind of create a bunch of ai companions grok

936
00:58:46,250 --> 00:58:48,890
Ejaaz:
is launching them as well and i wonder i wish i

937
00:58:48,890 --> 00:58:54,230
Ejaaz:
could somehow track how much time each human user spends with some of these

938
00:58:54,230 --> 00:58:58,630
Ejaaz:
ai agents and companions as they go live i bet you like it's going to be incredibly

939
00:58:58,630 --> 00:59:03,590
Ejaaz:
sticky and what's really interesting about that anil um is that it's basically

940
00:59:03,590 --> 00:59:06,870
Ejaaz:
going to be a reflection of the person to an extent, right?

941
00:59:07,190 --> 00:59:11,490
Ejaaz:
And it depends on how much you dial up the sycophancy trait or if you dial it

942
00:59:11,490 --> 00:59:15,410
Ejaaz:
down and it becomes kind of like your mentor that kind of like abuses you every

943
00:59:15,410 --> 00:59:17,930
Ejaaz:
now and then and says like, no, you need to work harder or whatever that might

944
00:59:17,930 --> 00:59:20,270
Ejaaz:
be. All right, Jan, you're up next.

945
00:59:20,630 --> 00:59:25,190
Yan:
So one area I've spent a decent amount of time looking into and I'm super excited

946
00:59:25,190 --> 00:59:27,050
Yan:
about is the humanoid space.

947
00:59:27,290 --> 00:59:34,210
Yan:
So I think, you know, us speaking to a bunch of emerging managers and early

948
00:59:34,210 --> 00:59:41,130
Yan:
stage investors, It seems as if most of them are kind of fading it to some degree,

949
00:59:41,170 --> 00:59:48,370
Yan:
or they think it'll be more of a application-specific form factor that makes

950
00:59:48,370 --> 00:59:53,610
Yan:
more sense from a cost perspective, from a utility perspective.

951
00:59:53,610 --> 00:59:59,710
Yan:
Part of it is them talking their book, naturally, because building out the humanoid

952
00:59:59,710 --> 01:00:02,930
Yan:
component is very difficult and expensive.

953
01:00:03,230 --> 01:00:07,270
Yan:
And if you're doing early stage investing, it makes more sense to do these targeted

954
01:00:07,270 --> 01:00:11,390
Yan:
use cases that can get to market a lot more quickly and start to generate revenue.

955
01:00:11,390 --> 01:00:16,470
Yan:
And so I think there's a massive world where those make a lot of sense, right?

956
01:00:16,630 --> 01:00:21,670
Yan:
The unit economics can be very predictable because most of the tech already exists.

957
01:00:21,730 --> 01:00:24,870
Yan:
And I agree, there's a huge market for those.

958
01:00:25,030 --> 01:00:30,070
Yan:
But I think fading the humanoid side doesn't make much sense.

959
01:00:31,900 --> 01:00:38,740
Yan:
And the way to think about it is the market for the humanoid form factor is insanely huge.

960
01:00:39,160 --> 01:00:44,720
Yan:
I'm very aligned with the idea that there will be billions of these in probably

961
01:00:44,720 --> 01:00:47,500
Yan:
two decades just because of the amount of time it takes to build them.

962
01:00:47,700 --> 01:00:53,480
Yan:
But I think there will be a massive just supply crunch for them within the next

963
01:00:53,480 --> 01:00:55,920
Yan:
three to five years, realistically.

964
01:00:55,920 --> 01:00:59,180
Yan:
Um the the the the human

965
01:00:59,180 --> 01:01:02,900
Yan:
form factor makes a lot of sense because it can easily slot into everyday life

966
01:01:02,900 --> 01:01:05,920
Yan:
now i think uh the cost component is

967
01:01:05,920 --> 01:01:12,900
Yan:
starting to really get close to achievable so the the human form factor uh business

968
01:01:12,900 --> 01:01:18,800
Yan:
model usually fell off in the transition from uh prototype to scalable model

969
01:01:18,800 --> 01:01:23,340
Yan:
and and that makes a lot of sense right you have these insanely expensive robots that can breakdance,

970
01:01:23,460 --> 01:01:27,400
Yan:
but that's not really valuable from a business perspective.

971
01:01:27,640 --> 01:01:33,200
Yan:
Ultimately, what you want is reliability. So you're paying for hours worked, right?

972
01:01:33,300 --> 01:01:36,180
Yan:
That's kind of what really drives the value prop here.

973
01:01:36,260 --> 01:01:42,240
Yan:
And so I don't think there's a winner take all in this market because the demand,

974
01:01:42,380 --> 01:01:44,680
Yan:
I think, is nearly infinite, right?

975
01:01:44,760 --> 01:01:50,100
Yan:
And as they get better, the surface area for deployment and implementation only grows.

976
01:01:50,680 --> 01:01:54,760
Yan:
They all kind of gather within, you know, they all learn together,

977
01:01:55,080 --> 01:02:00,060
Yan:
which is, I think, something that isn't really appreciated enough where whatever

978
01:02:00,060 --> 01:02:03,240
Yan:
it's learning in one factory, it gets to apply everywhere else.

979
01:02:03,360 --> 01:02:08,360
Yan:
And so, and then, and you also, I think one of the things that gets faded on

980
01:02:08,360 --> 01:02:12,840
Yan:
the humanoid side is the fact that people think there will be kind of a societal

981
01:02:12,840 --> 01:02:15,240
Yan:
uprising, right? They're taking our jobs.

982
01:02:16,000 --> 01:02:21,720
Yan:
But for the foreseeable future, it just kind of amplifies productivity, right?

983
01:02:21,800 --> 01:02:28,040
Yan:
If you zoom out and think about demographics in terms of the population that

984
01:02:28,040 --> 01:02:30,840
Yan:
wants to do some of these roles, that's only going to decrease.

985
01:02:31,020 --> 01:02:35,100
Yan:
So cost of labor will increase. On the other hand, you have electricity costs

986
01:02:35,100 --> 01:02:37,840
Yan:
will come down, production costs will come down, reliability,

987
01:02:37,840 --> 01:02:38,920
Yan:
these things will come down.

988
01:02:39,380 --> 01:02:43,140
Yan:
And these businesses become pretty profitable pretty quickly,

989
01:02:43,260 --> 01:02:46,540
Yan:
especially when you think about their creative kind of forms of financing so i

990
01:02:46,540 --> 01:02:51,040
Yan:
think that space isn't really um as

991
01:02:51,040 --> 01:02:53,720
Yan:
as as appreciated and so realistically in the

992
01:02:53,720 --> 01:02:56,560
Yan:
u.s there are basically three major players for it

993
01:02:56,560 --> 01:03:00,000
Yan:
right you have tesla as the leader with optimus uh figure

994
01:03:00,000 --> 01:03:04,580
Yan:
is second in line they just did a val they did a raise at 40 billion that's

995
01:03:04,580 --> 01:03:08,540
Yan:
kind of getting wrapped up and then i think eptronic is the clear third and

996
01:03:08,540 --> 01:03:12,840
Yan:
um that that's they're trying to do another race soon and that's the one uh

997
01:03:12,840 --> 01:03:17,700
Yan:
we're really excited about internally because we see a lot of value there we um.

998
01:03:18,660 --> 01:03:23,600
Yan:
We think what they excel in is the actuator side, which is basically the joint of the robot.

999
01:03:23,860 --> 01:03:25,980
Yan:
And that's something they've been building for quite some time.

1000
01:03:26,080 --> 01:03:32,140
Yan:
And I think there is a moat in that because of how that contributes to the dollar

1001
01:03:32,140 --> 01:03:36,100
Yan:
spend per hour's worked formula and in terms of what it does for reliability.

1002
01:03:36,720 --> 01:03:41,540
Yan:
And then on the other hand, they're partnering with Google and plugging in Gemini, right?

1003
01:03:41,700 --> 01:03:46,820
Yan:
So you have the physical humanoid and then the model and the two needs to work

1004
01:03:46,820 --> 01:03:49,400
Yan:
in tandem. And so you can try and build the model from scratch,

1005
01:03:49,560 --> 01:03:53,340
Yan:
which is what Figur is doing after their kind of separation from open AI.

1006
01:03:53,760 --> 01:03:58,340
Yan:
But I think partnering with someone and focusing on your strength makes a lot of sense.

1007
01:03:59,440 --> 01:04:01,300
Yan:
And so, yeah, it turned into an electronic shell.

1008
01:04:01,300 --> 01:04:05,920
Ejaaz:
That point around the actuator, Jan, is such a crazy thing to think about.

1009
01:04:06,060 --> 01:04:09,740
Ejaaz:
Can you imagine in the Industrial Revolution when humans were just working at

1010
01:04:09,740 --> 01:04:15,420
Ejaaz:
factories, that they were each graded by their ability to move their elbow or

1011
01:04:15,420 --> 01:04:17,620
Ejaaz:
whatever at a 90-degree angle? That's just insane.

1012
01:04:17,720 --> 01:04:22,040
Ejaaz:
The fact that you can program economics into these things is crazy.

1013
01:04:22,180 --> 01:04:23,000
Ejaaz:
And I think you're right.

1014
01:04:23,320 --> 01:04:29,620
Ejaaz:
Being able to picture and visualize these robots as actual, not some otherworldly

1015
01:04:29,620 --> 01:04:33,680
Ejaaz:
creature, but just functioning humans and then monetizing that is just,

1016
01:04:33,820 --> 01:04:37,060
Ejaaz:
it's just a new model to kind of like wrap yourself around.

1017
01:04:38,480 --> 01:04:39,380
Ejaaz:
It's just insane.

1018
01:04:40,220 --> 01:04:43,560
Jose:
I think humanoid is a really good one because you can kind of like,

1019
01:04:43,680 --> 01:04:48,180
Jose:
I think being in crypto so long, you can kind of identify what things cause a bubble.

1020
01:04:48,920 --> 01:04:53,900
Jose:
And I think obviously the thing has to have very strong narrative potential, right?

1021
01:04:54,120 --> 01:04:57,480
Jose:
Like humanoid robots replacing all physical labor has that. And then you also

1022
01:04:57,480 --> 01:04:59,080
Jose:
have to have a lot of hate.

1023
01:04:59,380 --> 01:05:04,840
Jose:
Like you kind of need, because it both forces people to talk about it and also

1024
01:05:04,840 --> 01:05:07,080
Jose:
creates like these really hated rallies.

1025
01:05:07,280 --> 01:05:10,860
Jose:
And I think humanoid robots actually has a decent amount of hate from like smart

1026
01:05:10,860 --> 01:05:14,120
Jose:
people who just think that specialized robots are gonna win out.

1027
01:05:14,820 --> 01:05:19,240
Jose:
So it's a very, I think, good contestant for that. I'd give you two names that

1028
01:05:19,240 --> 01:05:21,860
Jose:
I think are interesting, maybe contrarian.

1029
01:05:22,000 --> 01:05:24,340
Jose:
I think Anthropic is really valuable.

1030
01:05:24,880 --> 01:05:27,980
Jose:
It's like the least valuable of the model companies. I think you could get it

1031
01:05:27,980 --> 01:05:34,100
Jose:
at like 60 bill when I last looked a month or two ago versus three to 400 billion

1032
01:05:34,100 --> 01:05:39,380
Jose:
for OpenAI and 150 billion or so for Grok or for XAI now.

1033
01:05:40,300 --> 01:05:44,800
Jose:
And they're clearly the winners in coding. Like they have been over and over again.

1034
01:05:45,100 --> 01:05:49,300
Jose:
I think they have a lot of market share in coding, like every dev and any dev

1035
01:05:49,300 --> 01:05:50,560
Jose:
you speak to is using CodeCode.

1036
01:05:51,440 --> 01:05:55,960
Jose:
I think that's insanely valuable. If you think software has eaten the world,

1037
01:05:56,080 --> 01:05:59,300
Jose:
is going to continue to eat the world, and you are literally the world's software

1038
01:05:59,300 --> 01:06:04,400
Jose:
factory, where everyone is going to produce software, I think it's insanely valuable.

1039
01:06:04,620 --> 01:06:08,540
Jose:
It's also one of the things that's easiest to train on because you have these

1040
01:06:08,540 --> 01:06:12,600
Jose:
easy kind of RL loops that you can do. It's formally verifiable and stuff.

1041
01:06:13,800 --> 01:06:16,680
Jose:
So I think they're actually in a really strong position. And

1042
01:06:16,680 --> 01:06:19,320
Jose:
it's tough because they don't have their own users i think

1043
01:06:19,320 --> 01:06:22,100
Jose:
a lot of people use it via api and that's generally

1044
01:06:22,100 --> 01:06:25,280
Jose:
not a not a great place to be but i think if they win coding that's

1045
01:06:25,280 --> 01:06:28,260
Jose:
like i think tens of trillions of of

1046
01:06:28,260 --> 01:06:31,080
Jose:
dollars like use case like i think it's only going

1047
01:06:31,080 --> 01:06:35,660
Jose:
to get get bigger um and then the other one the one we're speaking about at

1048
01:06:35,660 --> 01:06:40,240
Jose:
a dinner is just it's in a hated sector it's not to do with ai but it's it's

1049
01:06:40,240 --> 01:06:46,920
Jose:
epic games um so those guys they're doing like six billion in revenue and um

1050
01:06:46,920 --> 01:06:48,660
Jose:
i haven't found supply for it yet,

1051
01:06:48,800 --> 01:06:51,080
Jose:
but it trades at something like 15 billion,

1052
01:06:51,700 --> 01:06:57,060
Jose:
which, you know, it's a very depressed multiple and just because gaming is not hard at all right now.

1053
01:06:57,420 --> 01:07:00,080
Jose:
Gaming is in kind of a secular decline for the last two years.

1054
01:07:00,360 --> 01:07:03,220
Jose:
Sort of the time people have spent, not just crypto gaming, but time people

1055
01:07:03,220 --> 01:07:07,280
Jose:
have spent gaming has gone down for two years straight, which no one really thought was possible.

1056
01:07:08,440 --> 01:07:11,240
Jose:
No one knows the reason either. A lot of people speculate it's literally just

1057
01:07:11,240 --> 01:07:17,100
Jose:
TikTok eating your leisure time that people used to be spending gaming.

1058
01:07:17,720 --> 01:07:20,800
Jose:
And people talked a lot about the metaverse in crypto.

1059
01:07:21,420 --> 01:07:26,840
Jose:
Fortnite has actually built the metaverse. It's not VR like most people expected,

1060
01:07:26,980 --> 01:07:30,100
Jose:
but they have the closest thing to a metaverse in terms of

1061
01:07:32,900 --> 01:07:38,280
Jose:
Just different worlds that are player created, all the different maps that are

1062
01:07:38,280 --> 01:07:40,980
Jose:
player created, like 500 million users.

1063
01:07:40,980 --> 01:07:45,840
Jose:
They're having concurrent players maps with like thousands of players and just

1064
01:07:45,840 --> 01:07:52,680
Jose:
a really thoughtful CO and I think like everything is going to be leveraged

1065
01:07:52,680 --> 01:07:55,380
Jose:
by AI and I think they will be too just in the speed of what they can do.

1066
01:07:55,740 --> 01:07:59,100
Jose:
I think it's an interesting one that like it's always interesting to look at

1067
01:07:59,100 --> 01:08:03,460
Jose:
sectors that people aren't excited about at all and I think gaming is one of them right now.

1068
01:08:04,000 --> 01:08:09,400
Ejaaz:
Awesome. Before we round up guys you made a big announcement this week around

1069
01:08:09,400 --> 01:08:11,440
Ejaaz:
something called Delphi Intelligence.

1070
01:08:11,900 --> 01:08:15,840
Ejaaz:
And you gave Josh and I access to the platform beforehand. And we have to say,

1071
01:08:15,940 --> 01:08:17,160
Ejaaz:
like, we were super impressed.

1072
01:08:17,740 --> 01:08:21,560
Ejaaz:
Maybe you could tell us a little more about what this is and why it's important

1073
01:08:21,560 --> 01:08:22,600
Ejaaz:
towards what you guys are doing.

1074
01:08:23,180 --> 01:08:26,480
Anil:
Yeah, definitely. Yeah, so obviously, we've talked about this a lot on the pod

1075
01:08:26,480 --> 01:08:29,860
Anil:
already. But like, research is just at the heart of everything we do.

1076
01:08:30,300 --> 01:08:33,820
Anil:
And to be honest, like, any decision we make, we kind of want to go in with

1077
01:08:33,820 --> 01:08:36,900
Anil:
conviction and as much like insight and knowledge as possible.

1078
01:08:36,900 --> 01:08:39,920
Anil:
So we know we're not only making the right decision, but when we are making

1079
01:08:39,920 --> 01:08:43,940
Anil:
that decision, can size it properly, right? And I think for us,

1080
01:08:44,320 --> 01:08:47,300
Anil:
you know right basically you know jose right after he

1081
01:08:47,300 --> 01:08:50,480
Anil:
he kind of like passed around the situational witness paper um

1082
01:08:50,480 --> 01:08:53,260
Anil:
which he actually read on a you know week off which is like

1083
01:08:53,260 --> 01:08:56,360
Anil:
probably when we get the most work done it's like our weeks off um

1084
01:08:56,360 --> 01:08:59,320
Anil:
to actually like read and think about you know the future of delphi and everything

1085
01:08:59,320 --> 01:09:02,460
Anil:
like that i think that's when we really you know probably nine ten months ago

1086
01:09:02,460 --> 01:09:07,040
Anil:
at this point realized that um you know this was like a no not an option for

1087
01:09:07,040 --> 01:09:12,780
Anil:
us right we think to be the best uh investors builders researchers in crypto

1088
01:09:12,780 --> 01:09:16,240
Anil:
and honestly any area, you kind of need to start building expertise in AI.

1089
01:09:16,880 --> 01:09:20,240
Anil:
So that's when we really started, you know, rolling up recipes and doing the

1090
01:09:20,240 --> 01:09:24,500
Anil:
hard work of building out a team and building out kind of like an MO,

1091
01:09:24,660 --> 01:09:28,980
Anil:
which is just publish a lot of like great work on in areas that we're interested about.

1092
01:09:29,000 --> 01:09:32,320
Anil:
So we can kind of build conviction and build expertise in this area to help

1093
01:09:32,320 --> 01:09:35,100
Anil:
us make these decisions. So that's what Delphi Intelligence is.

1094
01:09:35,280 --> 01:09:38,140
Anil:
It's a research platform, free to access for all. So you can,

1095
01:09:38,260 --> 01:09:42,920
Anil:
you know, go on delphiintelligence.io right now, put your email in and you'll

1096
01:09:42,920 --> 01:09:46,380
Anil:
get all of our research, you know, basically bi-weekly free.

1097
01:09:46,620 --> 01:09:51,220
Anil:
We already have two reports out, you know, one on just like AI in the era of

1098
01:09:51,220 --> 01:09:53,980
Anil:
entertainment, and then one on video generation models.

1099
01:09:54,680 --> 01:09:58,460
Anil:
Both are like great. We have another one coming out next week on AI powered

1100
01:09:58,460 --> 01:10:01,180
Anil:
browsers, which I think is going to be like really top of mind for a lot of people.

1101
01:10:02,240 --> 01:10:06,880
Anil:
And essentially like, you know, it's us open sourcing our learning to the world.

1102
01:10:07,460 --> 01:10:10,480
Anil:
And what's cool about it too, is it's not just going to be our team.

1103
01:10:10,480 --> 01:10:13,320
Anil:
We're going to be curating a lot of great reads from within our network and

1104
01:10:13,320 --> 01:10:17,500
Anil:
people we respect, including some of the fund managers that Jose brought up.

1105
01:10:17,660 --> 01:10:21,500
Anil:
So yeah, I mean, if you're interested, please subscribe, follow us on Twitter

1106
01:10:21,500 --> 01:10:24,620
Anil:
and everything like that. But we're really excited about it.

1107
01:10:25,100 --> 01:10:30,300
Ejaaz:
Awesome. Well, thank you all for spending time with Josh and I and kind of going

1108
01:10:30,300 --> 01:10:32,240
Ejaaz:
through your thoughts on the AI market.

1109
01:10:32,380 --> 01:10:38,400
Ejaaz:
As you can imagine, there's just so much going on and our Twitter feeds or rather

1110
01:10:38,400 --> 01:10:43,480
Ejaaz:
our X feeds are off the hook. We are talking to like five different AI models

1111
01:10:43,480 --> 01:10:44,820
Ejaaz:
for various different things a day.

1112
01:10:44,960 --> 01:10:50,520
Ejaaz:
And it's just not easy to think strategically and long term and have conviction

1113
01:10:50,520 --> 01:10:53,720
Ejaaz:
around investments, right? Investments are such a hard thing to kind of nail.

1114
01:10:54,200 --> 01:10:57,540
Ejaaz:
So, you know, hearing your perspectives has been hugely informative for us and

1115
01:10:57,540 --> 01:10:58,860
Ejaaz:
I'm sure for our audience as well.

1116
01:10:59,220 --> 01:11:03,340
Ejaaz:
For the Limitless listeners, thank you so much for joining us for another episode.

1117
01:11:03,340 --> 01:11:08,020
Ejaaz:
As you know, Josh and I are trying out something new, which is just put out

1118
01:11:08,020 --> 01:11:12,860
Ejaaz:
loads of content as and when it comes live, as and when the topic is trending.

1119
01:11:13,060 --> 01:11:14,860
Ejaaz:
So we appreciate you and your feedback.

1120
01:11:15,220 --> 01:11:20,160
Ejaaz:
The main bit of feedback that we've got so far is that you love the guest episodes

1121
01:11:20,160 --> 01:11:22,020
Ejaaz:
and we want to get more interesting guests on.

1122
01:11:22,180 --> 01:11:25,300
Ejaaz:
We hope you see this as one of those pushes towards that.

1123
01:11:25,860 --> 01:11:29,560
Ejaaz:
And again, if you have any friends or colleagues or whatever that might be interested

1124
01:11:29,560 --> 01:11:33,160
Ejaaz:
in this thing, we appreciate you sharing, liking and subscribing.

1125
01:11:33,680 --> 01:11:37,100
Ejaaz:
Thanks, folks, and we'll see you on the next one. See you guys. Thanks.