Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:00
I remember getting off the call and Josh was
0:02
also there and we looked at each other like, wow, like
0:04
the way companies are going to get changed
0:06
is going to be incredible. Software is going to
0:09
look so much different. That kind of transformation
0:11
is just so rare in our job, company after
0:13
company. I think OpenAI is quickly figuring out
0:15
the cost curve. They have done a lot in the open
0:18
source community on regulation,
0:20
security and safety. I think they're being pretty
0:22
proactive. This is 20VC, The
0:24
Memo with me, Harry Stebbings, and today we're focusing
0:26
on the most discussed company on earth right now,
0:29
OpenAI.
0:29
We're joined by the investing partner who led the
0:32
financing for Thrive Capital in OpenAI's
0:34
latest round, Vince Hanks. Vince
0:36
is a partner at Thrive where he's led the firm's
0:39
investments in OpenAI, Melio and
0:41
Airplane.dev. And he also sits on the
0:43
board of Airtable, Benchling, Lattice
0:45
and Melio. And prior to joining Thrive, Vince
0:47
was an investor at Tiger Global where he learned
0:49
the craft of venture from the legend that is Leafix,
0:52
Slud Edition now. But before we dive
0:54
into the show's day, we need to talk about SANA.
0:57
SANA is an AI-powered learning and knowledge sharing platform.
0:59
Think of it like chat GPT for all
1:02
of your company's knowledge. SANA integrates
1:04
with all your company's apps in under five
1:06
minutes and can search through every single file,
1:09
doc, pull request, video and more
1:11
in under 100 milliseconds. Assistant
1:13
is generative AI at its most useful.
1:16
Say you need to create a course on
1:18
OKR fundamentals for your employee
1:20
onboarding program and you're just really short
1:22
on time.
1:22
Assistant can generate the outline
1:25
and contents from scratch,
1:27
complete with relevant imagery. You can tweak
1:29
it, check it and then ask assistant to publish
1:31
it in five other languages and assign
1:34
it to all new hires in five markets.
1:36
And they've raised over $50 million to date
1:38
from some of the best. And you can request a free
1:41
trial at sanalabs.com
1:43
forward slash 20VC. And
1:45
if SANA really unlocks the knowledge in your team,
1:47
marketer hire brings that incredible knowledge
1:49
to your team. Have you been spending months looking
1:52
for the perfect hire?
1:53
Marketer hire is a talent platform
1:55
to hire expert marketers on demand.
1:58
The hiring process takes less than a week.
1:59
From initial consultation call to kicking
2:02
off work, it's free to use and you only
2:04
pay if you hire someone and they know quality
2:06
of supply is everything and so the application
2:09
process for talent is extremely rigorous.
2:11
Over 5000 marketers apply every
2:13
month and only 3% are accepted,
2:16
over 25,000 successful matches
2:18
and counting have been made and they are the number
2:20
one marketing only hire platform in
2:22
the world and one of the fastest growing B2B
2:25
tech companies of the last decade and Marketer
2:27
Hire is offering our listeners a
2:29
$1000 credit for first time customers. Go to
2:32
marketerhire.com, that's marketerhire.com
2:35
forward slash 20 VC and use
2:37
the code 20 VC to get your $1000 credit that's marketerhire.com
2:43
slash 20 VC with the code 20
2:46
VC. Three,
2:49
two, one,
2:49
zero. You
2:52
have now arrived at your destination.
2:55
Vince, I am so excited for this. I've heard
2:57
so many good things from Kareem, from Josh,
3:00
from Jack Altman, from Brad at OpenAI.
3:03
I've clearly got far too much free time, but
3:05
thank you so much for joining me today. Thank
3:07
you so much for having me Harry. I'm really excited to be here.
3:09
I'm a big fan of the show. That is very kind
3:11
of you, but I want to start with a little bit of context.
3:14
How did you make your way to Thrive and become
3:16
a partner at Thrive where you are today? Started
3:19
my career like many folks. I studied business
3:21
and accounting and undergrad. I went to
3:23
a big bank out of school, which was Goldman.
3:25
In
3:25
a lot of ways, going to these big
3:27
banks is like doing an MBA. It's a two-year
3:29
program for the most part. If you go there,
3:32
you want to work on these really complicated
3:34
big companies. Most of what I worked on was
3:36
that. It was things like AT&T
3:38
and Verizon or actually Dell
3:40
announced it was going to merge with EMC at the time.
3:42
They shipped us down to Austin. We
3:44
worked on carving out a bunch of software companies
3:46
to go finance that big transaction.
3:49
After doing that for about 18 months, I ended
3:51
up wanting to do something different. I got staffed
3:53
on Flipkart, which is an e-commerce business
3:56
in India. Lo and behold, the largest
3:58
shareholder was Tiger. I worked
4:00
pretty closely with the folks at Tiger on
4:02
Flipkart for about three or four months and
4:04
as I was going to leave Goldman, the stars
4:06
kind of aligned and I ended up joining Tiger
4:09
to go work really on private company investing
4:11
for the Skyly Fixall. As luck would have it,
4:13
the first company he handed to me was a software
4:15
company. And so I spent the first three years
4:17
of my career with him looking at lots
4:19
of software companies and trying to really find what were
4:22
the next generational big companies there.
4:24
Lee ended up leaving Tiger and I was
4:26
also chasing Airtable at the time
4:29
and the folks at Thrive,
4:29
Josh and Miles, had led a round of the
4:32
company and what I've learned from Thrive
4:34
is always be recruiting. One coffee chat
4:36
led to another coffee chat and about four years
4:38
ago I ended up joining the team at Thrive. There's
4:40
so many things for me to unpack that. I think Thrive's recruitment
4:43
machine is just incredible. I'm always trying to unpack
4:45
from Josh how he does it. He never quite tells
4:47
me. I do want to ask Lee is one of the most special
4:49
people in this business in my eyes. I love
4:52
him as a person. He's gifted in many
4:54
ways. What did you learn from working with Lee
4:56
and from your time at Tiger? First of all, I feel
4:58
really lucky to have started my career working
4:59
with Lee's. I credit a lot of where I got started
5:02
to working closely with him. Tiger also
5:04
is in the news a lot more recently but the firm's
5:06
been around for more than 20 years and if
5:09
you look at how it got off the ground,
5:11
it is very much in these kind of hedge fund
5:14
roots. Chase Coleman who started the firm, he was in
5:16
his mid-20s, he was really young. He had his hedge
5:18
fund mentality, mindset coming out of Tiger management,
5:21
post the dot-com bubble. The
5:23
way we thought about investing in companies
5:25
was very financial. Look at the P&L
5:28
and understand deeply how the numbers
5:30
tell the story of what the business does and
5:32
how does that ultimately make a good business, a great investment.
5:35
I learned a lot just from the financial rigor of
5:37
doing that. What's interesting is when you contrast
5:39
that to what we do at Thrive and how Josh has
5:41
built the firm, we started from the roots
5:44
of where Josh was, which was a founder.
5:46
He was the same age as Chase when he started Thrive
5:48
but ultimately he was trying to build a startup.
5:51
The mindset was very much how do you empathize
5:53
with the entrepreneur and we focus much
5:55
more on the product and the customer
5:58
and ultimately how does that manifest itself. into
6:00
a business. At the end of the day, we're looking for the
6:02
same thing. I was looking for a Tiger, which is iconic
6:04
market-leading companies that are going to generate great returns.
6:07
I think every investor wants that stuff, but to
6:09
make it tangible, at Tiger, I would have asked a
6:11
question if you were the founder and I'm trying to figure
6:13
out who your customer is. I would say, who's your
6:15
ICP? Or like, who's your core customer?
6:17
That's a very investor lens. At Thrive,
6:19
the way we'd ask that same question is, if I'm a SDR
6:22
on your sales team and I'm trying
6:24
to qualify a prospect, what am I looking
6:26
for? At the end of the day, it's really the same question. You
6:29
could even argue that the way
6:29
we do it at Thrive is less efficient, but I think
6:32
it shows a much deeper level of empathy with
6:34
the founder and it's a very different
6:36
mindset. I think the words we use are very important.
6:38
I often kind of change my tone and say, we,
6:41
what do we need to do to achieve the
6:43
next core milestone? And these little things like
6:46
changing from you to we and how
6:48
you address that question, which I think actually make
6:50
a big difference in how tones go
6:52
in terms of founder relations. So I totally
6:54
get you there. Vince, I heard you had a fascinating
6:56
background and so before we dive into many
6:59
more technical things,
6:59
I have to ask, I always believe that we're
7:02
all a function of our past, which means
7:04
we're all running from something. What are you running
7:06
from, Vince? Everyone's motivators
7:08
are to actually unpack and understand. We
7:11
talk about this a lot in founder assessment and
7:13
I just like you're saying, I believe that we're all a byproduct
7:15
of these accumulated experiences and
7:18
I'm sure have had a lot. I grew up in Michigan, I've
7:20
got three brothers, I've got two parents, they're
7:22
divorced and so there's lots of things in that I'm
7:24
sure manifest into who I am today,
7:26
but I'm sure like a lot of people that are listening
7:28
or listening to your show, I'm trying to figure out
7:30
who I am and I'm trying to figure out how that
7:33
impacts the decisions I make and how I react
7:36
to what I'm seeing as I go through the world. And
7:38
I think for me, personally, the motivator
7:40
is just maintaining this mindset. How
7:42
do you maintain the steepest slope possible
7:44
that you're learning on? One of the things I've
7:47
grown on a lot over the last eight years
7:49
of my career is I think when you start,
7:51
you take a very solo mentality to
7:53
doing things because the ship on your shoulder, you want
7:55
to accomplish things and so you do that really
7:58
on your own, but over time, I've realized
8:01
your friends, your spouse, your coworkers,
8:03
the people you're around and so forth, all
8:05
of those folks, you can compound your learning
8:07
curves together. And so for me, I think
8:09
a big part of where I've come from and where
8:11
I'm going is trying to make sure I utilize
8:14
all of these people are ecosystem to really compound
8:17
how I learn in that slope. And ultimately
8:19
that's been, I think the most powerful thing for me over the last
8:21
10 years of my short career and really my entire
8:24
life. How do you approach trust, Vince?
8:26
Trust is a tough one. It's very difficult.
8:29
It's like hard to gain, easy to lose. How
8:31
do you approach that? We talk about this as a team
8:33
from our culture. A lot of when you join the Thrive
8:35
team, the focus is how do you build trust
8:38
with the organization? And trust
8:40
is one of these things where in greater organizations,
8:42
it can be given by default in many organizations.
8:45
You have to earn trust. I think we have a culture
8:47
in Thrive that definitely people get a lot
8:49
of trust by default because of how small our team
8:52
is and how autonomous our model is. For me personally,
8:54
I actually had really great investors that you guys would know told
8:56
me something that was interesting, which is trust with
8:58
founders is actually just being very predictable.
9:01
People want to not feel like they're
9:03
getting surprised by how you're thinking. They
9:05
want to understand how you think, how
9:07
you're going to react and feel like they
9:09
understand you. And that ultimately kind of breaks
9:11
down these boundaries between people and
9:13
allow you to have some mutual trust and empathy
9:16
with each other. And so I think with founders in
9:18
particular, when we talk about building trust, it's
9:20
like any partnership. You've got to increase
9:22
the reps, get in the water in the trenches with them.
9:24
And they have to understand how you think. And ultimately,
9:27
I think you've got to telegraph how you're going to
9:29
make decisions
9:29
with them and make sure that they get there
9:32
alongside of you. They don't feel like you're
9:34
super imposing things top down on them. And
9:36
they don't feel like they're getting surprised at a left field because
9:39
if they feel that mutual level of partnership
9:41
with you, I think it's really actually pretty easy
9:43
to have trust with founders. I spoke to many of
9:45
the founders who you have that trusting relationship
9:47
with. And I promise we're going to bring it back to schedule, but
9:49
I'm enjoying this too much. And they said bluntly,
9:51
your ability to keep a level head
9:53
is actually one of your strongest points as
9:56
an investor. How do you think about
9:58
maintaining an even keel?
9:59
in terms of mindset. I think part
10:02
of it is you're a byproduct of your environment. I think
10:04
for me, I've gone through a bunch of different waves in
10:06
my career and that's helped me understand
10:08
what volatility looks like, feels like. And so I don't
10:10
think you just become level-headed as a person. I think you
10:12
kind of build into that psyche over time.
10:15
Someone said to me once, which is the line that I've been saying
10:17
to the team a bunch internally is things
10:20
are never as good as they seemed and they're never
10:22
as bad as they appeared. And I think just
10:24
keeping in mind that the rest of the
10:26
environment around you does react to these
10:28
peaks and troughs of your emotion in the market
10:31
and the volatility. And ultimately, in
10:33
good times, people over extrapolate and in bad times,
10:35
people under extrapolate. If you maintain
10:38
this kind of more balanced approach, I
10:40
do think it helps you hold more clearly
10:42
what are we ultimately looking for and solving
10:44
around for a given investment or
10:47
person or situation. I've
10:49
just found that it's not productive to necessarily get
10:51
caught up in the emotions. You got to try to think clearly.
10:54
And if you can remove that noise from the volatility,
10:56
the emotion, it allows you to focus on the
10:58
core a lot more easily.
10:59
But that said, I think you do need to trust your gut
11:02
and you do need to be emotional and react that
11:04
way. And so I wouldn't say it's all about being just
11:06
this kind of robotic level-headed person. You
11:08
need to figure out one of the times to do that and one
11:10
of the times to lean into your instinct. Speaking
11:12
of kind of leaning into instinct, I think everyone's
11:15
leaning into the instinct around AI
11:17
being the fundamental next platform
11:19
that changes all of human history, which it
11:21
very much could be. But I just want to ask on the hype
11:23
cycle there and the ups and downs that we mentioned.
11:26
I think for AI has never been greater. Is
11:28
AI the next big thing
11:29
or is it the new hype cycle
11:32
that will fade? This is the question I think every
11:34
investor is probably asking themselves
11:37
right now or certainly the ones I talked to. And
11:39
it's the quintessential question of like, do you sit
11:41
on the sideline and be patient and stay disciplined?
11:44
Or do you jump in on the gold rush? Because
11:46
if we don't get in now, we're going to miss all the seed in
11:48
Series A companies that create tens of billions of
11:51
dollars of value, which understandable. I
11:53
think there's a reason why we're in the business of being optimist
11:55
in venture capital. And I think for us and
11:58
really thrive in general, we try to get. into
12:00
the psyche of why are people so
12:02
excited and less about chasing
12:04
the next deal and more about what's
12:07
the core value to the customer? What's the product?
12:09
What's happening? Crypto obviously had an
12:11
amazing run and I think has kind of pulled back
12:14
a decent amount. We've been comparing and contrasting
12:16
how much is AI potentially the thing people are latching on
12:18
to like crypto maybe for the marginal investor
12:21
a couple of years ago versus how much is
12:23
it real. I think just use crypto as an analogy.
12:26
It was very audially driven. Bitcoin
12:28
came out after a financial crisis and
12:30
it was all about take the centralized financial system
12:33
and make it decentralized. A lot of people take
12:35
their money anywhere, don't let the government insert
12:37
themselves in the financial ecosystem. And really
12:39
like you had to believe in that as
12:41
a concept to take all of the trade offs of
12:44
using crypto as a consumer is a pretty bad
12:46
user experience. And so when we think about
12:49
other hype cycles, I mean, if you go back to even
12:51
the dot com bubble when that ran
12:54
up, Microsoft, Intel, Cisco
12:56
were the top three technology companies in the peak
12:58
of the dot com bubble. They were all infrastructure
13:01
related in some way and they were 50 percent
13:04
of three trillion a market cap in the hype cycle. And
13:06
so people really latch on to these companies
13:08
and they can run for a long time when people believe
13:10
and speculation feels more speculation. And so
13:13
at the end of the day, you know, if you invested in Microsoft 2000,
13:15
it would have taken 15 or 20 years to
13:17
break even on your investment. And so it's hard to
13:19
really time and predict hype cycles.
13:22
And so we ask ourselves, what's the core premise of AI?
13:24
We don't know them all today, but we can try to figure
13:26
out what some of them are and we can really try to
13:28
understand products. And there's obvious benefits
13:31
to those things today. Obviously all the search that's
13:33
coming with chat, LGBT and the LLMs,
13:36
but also just like companies that allow you to trigger
13:38
actions without having to do 20 clicks or
13:40
I mean, for your show, I'm sure you're using AI
13:42
in some way to edit or create content or
13:45
something like that. The marginal cost of content
13:47
production has come down a lot with these tools.
13:49
I think when these are the foundational questions you're asking
13:51
because of technology shift, it forces
13:54
every company to think about what could
13:56
happen to their business over the next five years. That
13:58
would be really disruptive or.
13:59
how should we be thinking about re-disrupting ourselves
14:02
to ultimately take advantage of the platform? And
14:04
all the boardrooms I'm in and many of the founders I work with
14:07
are thinking about that question. The question which
14:09
really came to me was someone posted on Facebook,
14:11
this picture, put a picture of themselves below
14:13
and said, is it AI or is it real?
14:16
And I genuinely was like, I don't
14:18
know, that's not sure. And that's
14:20
a real realization moment of like where
14:22
we are today. I do want to focus specifically
14:25
on open AI, obviously thrive very prominent
14:27
in the list round. Tell me how did the
14:29
deal go down? How did it come to
14:31
be, Vince? Yeah, we really
14:34
first started focusing on the company maybe 18 months
14:36
ago, because there were a number of
14:38
startups that were really rising with
14:40
software products with these kind of tools that
14:43
you could use to do marketing copy, or you could create
14:45
blog posts with them. I think many of these companies,
14:47
we spent a bunch
14:48
of time with them. And we kept coming back
14:50
to the fact that they were thin user
14:52
experiences on top of this cool thing, which
14:54
was this model, which really no one talked about,
14:57
at least not mainstream 18 months ago.
14:59
And so that triggered us to go spend time with the company.
15:02
And ultimately, maybe in the classic Sam
15:04
kind of way, the way we kicked off the round
15:07
was he did almost like a closed demo with lots
15:09
of investors on a couple of calls of
15:11
the technology they're working on and ultimately GPT
15:13
for and we were on that call
15:15
like many other investors. And I think I remember
15:18
getting off the call and Josh was also there. And
15:20
we looked at each other like, wow, like the way
15:23
companies are going to get change is going to be incredible.
15:25
Software is going to look so much different. And
15:28
we were reflecting on it. And that
15:30
kind of transformation is just so rare in
15:32
our job company of our company. And so many
15:34
things seem marginal. But when you see these things that
15:37
are discontinuous, or seem so different, we
15:39
trigger this inner instinct to pause focus
15:41
reflect. And ultimately, we have to spend more time
15:43
learning about the financials and the products they were releasing
15:45
in the customers. But that's what encouraged us
15:47
to really lean
15:48
in. You said you're one of several firms
15:50
who saw this. Why do you think they chose
15:52
you? Because this was one of the most hotly competitive
15:55
rounds to finance. Why do you think they chose
15:57
you? I think there's a lot of people that said no, as much as
15:59
I
15:59
I like to think we want a really competitive deal. It
16:02
was not obvious to everybody. And even now I don't
16:04
think it's still obvious to everybody, even folks
16:06
that use the product. Why do you think it's not obvious?
16:08
Cause you have to assume that essentially we'll move
16:11
from a search interface to a chat
16:13
interface as the primary UI of engagement.
16:15
Is that why, or are there other reasons why it's not obvious?
16:18
I think a lot of investors get tripped up on
16:20
trying to be so precise on TAM
16:22
and market
16:23
and defensibility and the moats
16:25
around businesses and trying to map
16:27
that all to price. And those are so
16:30
important in the investment decision-making process. But
16:32
when it's so early in a technology
16:34
cycle like this, it almost is a little bit more
16:36
of a venture mindset where there are going to be 50 reasons
16:39
you can say no, and we're not going to have answers to every
16:42
question, but we need to really think about
16:44
the things that can go right. We're not talking
16:46
about building the next unicorn or decacorn.
16:49
We're talking about disrupting search
16:51
or Google. I mean, that's a trillion dollar opportunity.
16:53
Sure. I can't put a TAM around
16:55
chat GBT, but I can tell you like we're
16:57
not talking about a small price at the end of the day. And
17:00
so I think when we say it's not obvious to people are thinking
17:02
about it in this lens, they're thinking about it in the box
17:05
of what does this chat interface do? And oh,
17:07
those use cases aren't that valuable. And I think
17:09
it does take a higher level of creativity
17:11
or imagination to ultimately think about the world
17:14
that way. I totally agree with you in terms of applying
17:16
that kind of different lens and mindset. But
17:18
then there are also core that you have to do,
17:21
which we all do when we make an investment. When
17:23
you thought
17:23
top down on market analysis, how
17:25
did you approach top down market analysis
17:28
when you were sitting with Josh on this one? Yeah, as
17:31
these new technologies scale, it's so
17:33
hard to be precise about a TAM. And
17:35
so more of what we got into the psyche
17:37
of thinking about is if you looked
17:40
at other big technology movements,
17:42
how did they scale? The iPhone went from
17:44
a million phones a year kind of post-launch
17:47
to a hundred million in five years. AWS
17:49
took six or seven years to get to $100 million of
17:51
run rate, but that went from 100 million to 10 billion
17:54
in six or seven years. Google went from
17:56
nothing to 10 billion in the first six
17:58
years of monetizing. Obviously this is
18:01
rarefied error we're talking about. These are three of the most
18:03
transformational companies on the planet, but the
18:05
technology, if you really think about
18:07
it, and the zeitgeist it's captured, it is of
18:09
that elk of the transformation it can have
18:11
of the world. And so for us, it's less
18:14
about thinking about the exact TAM, and
18:16
it's more about if we think about what
18:18
the world's gonna look like in five or 10 years from
18:20
now, how are we gonna look back and say, wow,
18:22
I can't imagine the world without this kind of thing. When
18:25
you have experiences like that, that's what gets us
18:27
out of bed. These are the kinds of things that create new categories,
18:29
that was what a lot of the discussion was,
18:32
and maybe that sounds less precise to people, and
18:34
they think it's a little bit too finger in the air, but I
18:37
think for us, a lot of this is instinct mapped
18:39
with why it's a really great business. I think
18:41
something that dictates whether
18:43
it achieves that scale and
18:46
enterprise value is whether it is actually kind of
18:48
the one defining model to rule them
18:50
all, or whether there's an alternative mechanism
18:53
that will kind of bifurcate the market. Now,
18:55
when we look at open AI, it fundamentally
18:57
says that there will be one model that rules them all,
18:59
versus the world of many models
19:02
with hugging face and the like being others.
19:04
How do the two views differ,
19:07
and why do you think there will be one
19:09
model to rule them all? I mean, obviously, these are my opinions,
19:11
not the company's, just to make that clear
19:14
to everybody. I don't think I would characterize
19:16
open AI as one model to rule them all, and
19:18
obviously, there's all kinds of talk right now about it being
19:21
closed source, and the model should be open
19:23
source for the value of the community and all of those factors.
19:26
But I'd say, actually, if you give the company credit
19:28
for this stuff, which they've released, they have done
19:30
a lot in the open source and back to the
19:32
community. Clip was a model that they
19:34
put out there that helped a lot of the image generation
19:36
open source models. Whisper was for
19:38
audio. They open sourced the inference
19:41
framework called Triton, on top of
19:43
GPUs and CUDA, to ultimately drive more efficacy
19:46
and scalability of inference. And so
19:48
they have done a bunch of that side. They obviously have
19:50
kept the core GPT model behind
19:52
an API and paywall. I think thinking
19:55
about it as this entirely closed ecosystem, in
19:57
my mind, doesn't really give credit to what the company
19:59
has done in the open source. side, founders we
20:01
at least talked to are choosing their model has
20:03
quickly changed from who's got the biggest
20:05
model and who has the lowest cost
20:07
model to these new dimensions, which
20:10
is if you're going to build a company today, Harry,
20:12
you don't want to have to think about the scalability
20:15
of it or the reliability of the infrastructure
20:17
or supporting all of that stuff. If
20:19
you use an open source model, you've got to go figure out all of these infrastructure
20:22
problems. And so now if you talk to OpenAI and
20:24
you think about what they're spending time on, it's
20:26
how do you support this onslaught
20:28
of scale that has come into their
20:31
system? And that's infrastructure engineering
20:33
problems. It's the same thing we talked about in cloud 10 years
20:36
ago. I'm sure there were banks out there saying,
20:38
you know, we're not going to move to public cloud because
20:40
we can run all our infrastructure ourselves and do it really
20:42
well. You would never have told a startup 10
20:44
years ago to go run your own infrastructure if Amazon
20:47
was there doing it for you. I think we're kind of
20:49
getting into this world now where sure, there are lots
20:51
of open source models and they can use for great things.
20:54
And some companies will choose that path. But
20:56
by and large, for the vast majority of companies, particularly
20:58
ones that have scarce resources, which are many of
21:00
the ones I work with, I don't want you spending time
21:02
building infrastructure and scaling open source models.
21:05
I want you just getting your product out there with lots of value
21:07
to your customers. And I think OpenAI is
21:09
quickly figuring out the cost curve, which is probably the
21:11
biggest competing interest against that. They
21:14
put out 3.5 turbo, which is I think at least one,
21:16
if not two orders of magnitude less expensive than
21:19
before. And in my opinion, I'd bet on these ecosystems
21:21
to scale the infrastructure and ultimately that's
21:23
becoming more important, the decision factors. You
21:25
mentioned that kind of the model still being kind of behind
21:28
lock and key. Can I ask, people
21:29
often suggest like the commoditization of the model
21:32
as a challenge. When you thought about that
21:34
when making the investment, how did you get
21:36
comfortable in terms of model commoditization
21:39
and long term edge for OpenAI?
21:41
This is a question we talked about a lot. I
21:43
think part of it, we're already seeing the evidence of,
21:45
in my opinion, where commoditization of
21:48
the core model output is not really
21:50
what people will make decisions out over time. I'm
21:52
sure there are lots of companies that give similar search results to
21:54
Google, but the reason these companies
21:57
get built up to scale is because the ecosystem
21:59
around.
21:59
them grows. And you're even seeing this, they launched
22:02
plugins about a month ago and chat GBT
22:04
has taken off and taken the world by storm. I think it's the
22:06
fastest company in 100 million users ever. And
22:09
so even if the raw model
22:11
does get competed against, there are maybe
22:13
companies like Google or Facebook or
22:16
Microsoft or some of these startups that can
22:18
compete on that. I do think, again,
22:20
the dimension by which people are going to think about this
22:22
in five years from now looking back isn't going to be
22:24
about the commoditization of the model
22:26
and the raw output. It's going to be, oh, the ecosystem
22:29
around the model has become much more
22:31
robust such that you can do a lot more with
22:34
the model than just get text outputs or
22:37
the convenience of putting these things together
22:39
such that it's not just text. It's now text in
22:41
image or video and audio all through
22:43
one interface. That's complicated.
22:46
And again, it's not to be dismissive that the model
22:48
doesn't matter in terms of accuracy or size,
22:51
but it's to say that people are going to optimize
22:53
for companies, builders are going to optimize
22:55
for the inputs that allow it to
22:57
be much easier from the build and much
22:59
easier
22:59
for them to capture this new channel of
23:02
hundreds of millions of users that are flooding here. And
23:04
that is not about commoditization model. That's
23:06
about ecosystem and the kind of classic things that
23:08
make great businesses. In terms of the ecosystem
23:11
around it, you always highlight competitors or
23:13
competitive threats when making until
23:15
you've literally going down investment number here. It's
23:18
great training for any one of the investors. But like
23:20
going down the competitive threats, who did you
23:22
identify as the competitive threats that
23:24
you saw as having potential to
23:26
be noticeable? Elephant in the room
23:29
is big tech companies.
23:29
This is not something where they're
23:32
sitting flat footed Facebook, Google,
23:34
even Microsoft. I know they're great partners. I
23:37
have Tom Tungus on the show recently, and he said that
23:39
Google have been the most disappointing of all. And
23:42
they were his former employer. And he said, terribly
23:44
disappointing. We've seen AWS partner
23:47
with Hugging Face. Who do you think
23:49
is doing well? I think it's hard to be
23:51
dismissive of these companies. If you think about
23:53
where the best talent in AI is right now, I
23:55
think it's open AI. And then I think everyone
23:58
would tell you it is Google and
23:59
Facebook and Microsoft and Amazon. And
24:02
maybe there are folks that have kind of dripped their way into
24:04
the startup ecosystem, but by and large,
24:06
the talent is so clustered in these
24:08
big tech companies. And so I get it that they're
24:10
tripping over themselves, trying to figure out how to navigate
24:13
these giant organizations they've created.
24:15
But let's not also kid ourselves. We've talked
24:17
about this, maybe outside of the context of opening
24:20
up level it to AI, we work with lots
24:22
of startups that compete on things like
24:24
presentations or content creation.
24:27
What's scary to me right now is if you're
24:29
a startup, large companies are
24:31
shipping product. The kind of canonical
24:34
examples were that, oh, the big incumbent
24:36
can't move and they're slow footed and
24:38
you got years to execute before they do something.
24:40
Microsoft's 200,000 person company, they've
24:43
shipped AI and Bing, AI and PowerPoint,
24:45
they're rolling out of the other products. Adobe's got
24:48
the content creation in their product already.
24:50
Even the big startups like Notion, those
24:52
folks I've been so impressed with how much they've shipped
24:55
in their product so quickly. And
24:57
so if you're a startup, I get that you
24:59
have the advantage
24:59
of speed, but you need a multi-year execution
25:02
window to be able to take advantage of all that product
25:04
you can ship quickly. If these big companies are
25:06
able to ship this quickly, they've got so much
25:09
distribution, so much talent, I just
25:11
think it's gonna be very hard to compete with them. So
25:13
the reason I'm also so bullish on open AI is
25:15
like Alice Rampel says very wisely, the
25:17
big question when investing in startups is
25:20
will the incumbent acquire innovation
25:22
before the startup acquires distribution?
25:25
And I think the challenge for everyone with open AI
25:27
is, it's still really a startup in terms
25:29
of processes
25:29
and a lot of its structure,
25:32
but it's called the distribution of a large incumbent.
25:35
So it can move as fast with the
25:37
distribution. That is a very
25:39
real and competitive threat that I would
25:41
not want to come up against. My question to
25:43
your point there though is, actually in
25:45
this next wave of AI, is the
25:48
value captured predominantly by
25:50
incumbents or is it captured by startups?
25:53
I think it's too early to call, maybe that's
25:55
a cop out answer, but to put it in context,
25:58
if you go look back in time,
25:59
at these big technology cycles, they
26:02
are great moments to create
26:04
new categories. And so maybe
26:06
that's a better lens to look at it through. Where
26:08
will new categories get created where being an incumbent
26:11
or being a startup won't matter? And if you look back
26:13
at the dot com, Google, PayPal came
26:15
out of it. Google about finding information, PayPal
26:17
about it, Bank Online. Social came
26:20
after. Facebook was connecting all these people
26:22
online and Twitter was the town square for
26:24
broadcasting information. Mobile,
26:26
the answer wasn't Salesforce was gonna build mobile
26:28
CRM and that's where all the value was. The answer
26:31
was you put a computer in everyone's pocket. DoorDash
26:33
and Uber came out of that because there's geolocation,
26:35
the ability to have connectivity. In my view, what we're
26:37
looking for, if I always answer that question is
26:40
yes. If you're going toe to toe right now with
26:42
an incumbent on their home turf and
26:44
their shipping, I think it's a hard bet to
26:46
take the opposite side of, you know, in our business.
26:48
But if you're creating something that's totally a different user
26:51
experience that no incumbent has
26:53
today, I'd bet on a startup 10 times
26:55
out of 10 in that case. The reason the show
26:57
is successful, because I just get to interview smarter people
26:59
than me and
26:59
just ask questions that I'm thinking. Brilliant
27:02
model I've built. But my question is, if the value
27:04
accrues like incumbent versus startup,
27:06
the other question is value accrual at
27:09
the infrastructure layer versus the application
27:11
layer. How do you think about value accrual
27:14
there? Because you quite rightly mentioned earlier, three
27:16
companies, two trillion, you know, and I think
27:18
in the prior application layer, there's about
27:20
a similar market cap, two trillion, but 50
27:22
companies, so much more distributed kind of
27:25
enterprise value. Where does the value accrue
27:27
infrastructure or application layer? I have
27:29
to believe
27:29
it's gonna accrue mostly to the application
27:32
layer. I think if you think about infrastructure
27:34
companies as platforms, you know, if you think about
27:36
the software market as the relatives, the
27:38
value of AWS and Microsoft
27:41
and Google Cloud versus the software
27:43
companies built on top or all of the internal
27:45
software tools built on top, I think it's probably
27:47
an order of magnitude to one, you know, so I just
27:49
have to believe where value gets created to end
27:52
customers is where it will get captured. And
27:54
the nice part about the infrastructure business is there are
27:56
effectively toll roads on that
27:58
whole ecosystem.
27:59
applications might compete more aggressively with each
28:02
other and open AI or Amazon
28:04
or the infrastructure provider might be able to clip a coupon
28:07
on it That's really valuable We might capitalize it
28:09
at a high multiple But I have to believe that
28:11
there's gonna be a lot more value in the new experiences
28:13
and applications than the infrastructure How
28:15
do you think about investing in the space
28:18
when it moves so fast for your
28:20
investor and benchling air table? Atlantis
28:23
these markets are not changing on a daily
28:25
basis So how do you get
28:27
comfortable with the rapidly changing
28:29
market
28:29
like we haven't seen before it's hard
28:32
I think at the early stage, you know,
28:34
we think about investing in founders first
28:36
and foremost you want extraordinary
28:39
people in really great
28:41
sandboxes and those folks will
28:43
be able to figure out how to pivot and adjust
28:45
and iterate to Take advantage opportunity in
28:47
this environment If you told me would I rather be in
28:49
the sandbox with the great founder? Experimenting
28:52
and iterating or sitting on the sidelines I'd
28:54
rather be iterating even if things changed a bunch
28:57
or underneath you at the early stage, but that's
28:59
how I think you can't make every Investment without
29:02
knowing what's gonna happen. So some things there
29:04
is an obvious Evolution that makes
29:06
your business less interesting than it was six months
29:08
ago If you think about where we were on infrastructure companies
29:10
nine months ago There were all kinds of companies
29:13
branding themselves as ML ops or these
29:15
end-to-end solutions for AI
29:17
now in this environment We're in now there's specialized
29:20
companies that are doing individual components
29:22
of the infrastructure stack Whether it's laying chain
29:24
on connecting all these tools and stitching them together
29:27
or these vector databases that are allowing you to
29:29
store information And so things specialize,
29:31
you know our reaction It's not that hard
29:34
to say I have a friend and mentor
29:36
of mine that I spent a lot of time with who says
29:38
there's only two questions that matter for a company who
29:41
is the customer and What is the product
29:43
and you'd be shocked at how many folks can't
29:46
really answer that question with clarity? And if
29:48
you can answer that question with clarity Everything
29:50
else about the business stems from there And so if you're
29:52
really answering that question you're struggling because
29:54
you think in two weeks from now There's a chance
29:56
that the
29:57
customer doesn't care or the product won't
29:59
have any
29:59
any value, that's hard to build a company around.
30:02
And so in any of those environments, I think he'd rather
30:04
just wait for that. There'd be more clarity than
30:06
try to keep firing investments into
30:08
that market. I made an investment and I lost money
30:10
in it. And I lost money because of an externality
30:13
that was outside of our control, which
30:15
is a very frustrating reason
30:17
to lose. How do you think about
30:19
the impact in coming year
30:21
or so of regulation? Elon has been
30:23
very clear in saying we can't wait until
30:26
it's in the hands of everyone because then it is
30:28
too late. How do
30:29
you think about regulation in the next six
30:32
to 18 months? I don't want to speak for the company
30:34
on anything, because I know this is a topic that's out
30:36
there in the mainstream. I think regulation
30:39
will come to AI and it will be necessary.
30:41
No doubt that at all. In my mind, I think it
30:44
has to be done in partnership with the companies that are building
30:46
because your regulation for the sake of regulation
30:49
is not going to be what solves the problems people are
30:51
concerned about. You've got to go understand the nuances.
30:53
You have to understand the technology. And so I
30:55
would hope that other companies
30:58
building in AI are also working with regulators
31:00
to understand what is the technology and help educate
31:03
them to get to the right decision. I
31:05
do think there's kind of some negative stigma
31:07
going around open AI on regulation,
31:10
security and safety. And in my view, I
31:12
think they're being pretty proactive. As an example,
31:14
on GPT-4, we saw the demo
31:16
in the fall. They didn't just release the product then. They
31:18
took, I think, four or five months to test
31:21
and learn about safety and the edges of the model
31:23
and then ultimately released it to the world.
31:25
And so they're not going to get it perfect every
31:28
time. But I do think trying to let
31:29
people build with and understand it is important. And
31:32
poorly designed regulation on a technology
31:35
this early is not going to be effective
31:37
for anybody. And we've got to find the middle ground. My
31:40
concern is the chasm of knowledge
31:42
between private and regulator side
31:44
has never been greater. How
31:46
do we set effective regulation
31:49
when the regulators do not understand
31:52
so much of the infrastructure and
31:54
the opportunity? This is a solvable
31:56
problem. There's information out there. There's experts
31:58
out there regulators.
31:59
with how up to speed they've gotten on crypto
32:02
and they've really taken time to learn about it. And sure,
32:04
they're gonna make mistakes, but ultimately that
32:07
gap is solvable and it's gotta start
32:09
the dialogue between the people
32:11
working closest on AI and these regulators.
32:13
And it might take years to get to a solution so people gotta
32:15
be patient. But this isn't an unknown,
32:17
this is something that you guys can work together on. Final
32:20
one on open AI and then we are gonna move to kind of your investing
32:22
style before a quick fire. I have to ask,
32:25
I would get in trouble if I didn't and some
32:27
very charming British guy otherwise. The
32:29
price
32:29
was high, I reported $29 billion.
32:33
What was the discussion internally around
32:35
price? Cause you still have to see real upside.
32:38
What was the discussion and how did you
32:40
project upside scenario planning? We
32:42
certainly had a discussion, a heated discussion around
32:44
it. We have a very team oriented
32:47
firm. And so we disagree and commit
32:49
once we make decisions and we make decisions to this team,
32:51
we live by that. No one person makes a decision.
32:54
And the price here was high. It's on an
32:56
absolute basis. It doesn't matter what company you're investing
32:58
in at the prices of this round
32:59
was done. You're talking about very
33:02
upper echelon type outcomes to justify
33:05
good returns. But I kind of come back to
33:07
the intangible. It's really hard
33:09
to understand the pace of adoption
33:12
of these major technologies. And any
33:14
numbers I would put on paper for you would look
33:16
insane. You would look at them in a spreadsheet and tell
33:19
me there's just no evidence to support
33:21
this. But then you look at these iconic technologies
33:23
and the great ones all follow that insane
33:26
curve. And so the balance we try to hold,
33:29
I think what makes Thrive a really special firm is
33:31
we're able to kind of separate ourselves
33:33
from the kind of quantitative
33:35
rigor that we rely on
33:38
for a lot of investments we do and hold the tension
33:40
of what could go right, think creatively,
33:42
understand how this could look like the most transformative
33:45
things. And if that happens, again,
33:47
we're talking about search potentially getting
33:49
disrupted or something that enables
33:52
workflow automation for hundreds of billions
33:54
of dollars spent on different jobs
33:57
and categories out there. And even now, we
33:59
didn't know this at the time of the... investment, but with chat
34:01
GPT pro and the traction that seeing maybe
34:03
that won't be the durable revenue engine for the
34:05
company over time, but the velocity
34:08
with which that's ramping. You know, Kareem, my partner
34:10
Kareem has always saying it's the scalability
34:12
properties of what they have built are
34:15
so good that we'd rather bet
34:17
on the compounding upside of that scalability
34:20
than sit on the sidelines because once it's clear,
34:22
it'll get priced up very quickly. And
34:24
so we ultimately decided to lean into that. How
34:27
do you determine when to throw the financial
34:29
rigor that you do have in the team
34:32
out of the window versus when it's needed
34:34
to make a sound and wise investment?
34:37
It's less about throwing it out the window and
34:39
it's more about putting it in context. A
34:41
financial model is a tool to help
34:43
you make a decision. You need many tools
34:46
to make a decision. It's less about
34:48
is that the only thing we use and we
34:50
throw it out the window and it's more about what's
34:52
the weight by which we put on the financial
34:54
rigor and model as the tool to make the decision.
34:57
I think for us, there aren't that many
34:59
iconic companies that could create it. And so
35:01
no iconic company, at least that
35:04
we've been a part of, was it clear
35:06
and obvious in the early days of that technology,
35:08
they always get priced well ahead and
35:10
they look cheap in hindsight because
35:12
they defy the gravity of the model. And so it's
35:14
kind of the quintessential humans think linear and
35:17
the best things happen exponential. We have
35:19
to understand it going back to being level
35:21
headed. When do our psyche or
35:23
heuristics break down in an investment
35:25
decision making process? And when they do, we
35:27
need to compensate them with a different
35:29
tool
35:29
in our toolkit. Can I ask you a bit of a weird
35:32
personal one, Vince, but this is a big check
35:34
for you. It's a really big check and
35:37
it's kind of you leading. Were you nervous
35:39
about that? I think we had the same
35:42
level of excitement that we
35:44
have in most investments. We don't
35:46
have a culture where there's this
35:48
kind of sharp elbow mentality
35:50
or people feel on edge for making investments
35:53
because prices are high or checks
35:55
are really large. We have a culture
35:57
where we support each other and we have a growth mindset.
35:59
learn. And so if this ends up not working out,
36:02
I don't think we're going to look back on this and put all the onus
36:04
on one person or two people that made a decision.
36:06
I think we're going to look back on this and say we got there as a team.
36:09
It was the right thing at the time. But here's
36:11
the learnings from it. And here's how we're going to adapt our lens
36:13
and course correct in the future. It is intimidating.
36:16
It's nerve wracking sometimes to write large
36:18
checks and investments. But in the right team
36:21
and culture, this is how we enable ourselves
36:23
to make these kinds of transformational investments. And
36:25
frankly, I think there are a lot of firms that could
36:27
not have done this because their organizations
36:29
are not set up to make these
36:32
kinds of big decisions. Your ability
36:34
to make these big decisions is the one
36:36
single fund. It's the one single team.
36:38
So help me understand specifically with Thrive
36:41
about the structure that enables you to make a decision
36:43
that others maybe couldn't. I think the thing
36:46
that enables us to make this decision
36:48
is we have this single small
36:50
team that thinks extremely creative
36:53
about how the world works. And we
36:55
are ultimately trying to concentrate our
36:57
investments in the best products
36:59
that are out there in the world. And so when you
37:02
simplify it down to that and
37:04
we allow this autonomy and
37:06
focus to happen on a single team
37:08
across the entire investment cycle from
37:10
early stage all the way to growth, we are
37:13
unable to go find these kind of iconic companies.
37:15
And we're young in our career, obviously, but
37:17
we also we can appreciate when these kind
37:20
of transformational technologies come around. And
37:22
certainly this was the most transformational thing I've seen in
37:24
my career and being able to lean into that
37:26
and not be afraid that if it goes
37:28
wrong, we're going to
37:29
get fired or if it goes wrong,
37:32
we're going to be pushed down in the firm.
37:34
I think having that psychological safety
37:36
that enables a firm to do it. And I think we have
37:39
that at Thrive. I think we change
37:41
so much as investors over time in terms
37:43
of what we value in the companies we invest
37:46
in and the founders we invest in. When it comes
37:48
to what's changed in what you appreciate
37:51
in an investment,
37:52
what has changed about that mindset? It's
37:54
actually very clear. The thing I've developed the
37:56
most on in my investing mindset is
37:59
this deep. empathy for the customer.
38:01
Trying to really think deeply about not
38:04
just what's the product and trying
38:06
to write that down on paper, but really to understand
38:08
how the business is going to get built and mapping
38:11
those nuances to who the
38:13
person is, what the product does, how
38:15
does that manifest itself in the business?
38:18
Do you have a sales intensive product
38:20
that's going to require people to be kind of constantly
38:23
out there and on edge and with their customers?
38:25
Do you have a product that's more middleware and so you need
38:27
somebody who's going to be willing to grind
38:29
it out and not be in the limelight?
38:31
Do you have a product that requires a lot of creativity?
38:34
That means you need to set up your org to be creative.
38:36
There are certain org structures that promote that more. And
38:38
when you think about where you can develop as an investor,
38:41
I think where I have developed the most is
38:43
continuing to understand the connectivity
38:45
between this rigorous financial
38:47
lens, very much of where I started at Tiger
38:50
to how do you build a company? And
38:52
at that intersection, it's really hard to get
38:54
to clarity on, but when you do, it is clear.
38:56
And I think the best companies have very
38:58
simple explanations. It trickles down
39:01
the company. The CEO can articulate in one
39:03
sentence, but the manager for
39:05
rung is down also can articulate it. And
39:07
that means that manager knows how to
39:09
go left when they should go right or
39:11
they know how to make that decision and communicate it through
39:14
a team. And so having appreciation for that, I think is where
39:16
I've grown the most on these investment decisions. And again,
39:18
it's hard to quantify, but I think it results
39:21
in just like a deeper empathy for
39:23
what a great company looks like.
39:24
Final two questions. I think we learn a lot
39:27
from wins and losses, and we
39:29
don't often analyze the wins as well
39:31
in the same detail. If we start on the loss, then we'll
39:34
move to the win. What's been the biggest investing
39:36
mistake for you and how
39:38
has your mindset changed as a result? I
39:40
made lots of mistakes. The one that I think stands
39:42
out the most to me, or at least comes back
39:45
to me a bunch in my psyche is when
39:47
I was at Tiger, I flew out to Sydney
39:49
and spent a bunch of time with the Canva team in
39:51
person. And obviously now
39:54
people know
39:54
about it. It's a remarkable company. At the time,
39:57
it was much smaller than it is now. And
39:59
at the same same time we were so focused
40:01
on investing in enterprise software as
40:03
an emerging category at Tiger. And
40:05
so, as we were spending time with Canva, I very
40:08
much, and this is early in my career, but I very much
40:11
let the pattern matching and the
40:13
DNA of what I was thinking about of a great enterprise
40:15
software company creep into us
40:18
looking at Canva. And the learning,
40:20
I think, just to distill it down to something, is every
40:22
company is unique, even within the bounds of enterprise
40:24
software companies, they're unique. And we
40:26
took this lens of what a great enterprise software company
40:29
was, we retrofitted to Canva, and
40:31
we said, okay, well, the churn
40:33
looks a lot higher than what grade looks
40:36
like. And the engagement looks very
40:38
whimsical relative to deep workflows
40:40
and integrations. And so, we shouldn't
40:43
value this highly recurring software
40:45
business, we should value a consumer subscription business.
40:48
And there's merit to that. But ultimately, because
40:50
we were so much in this one-dimensional
40:52
mindset of enterprise software, we missed
40:55
what was so special about the company. The
40:57
learning for me is just you can't walk in
40:59
biased about what you're looking for out of the
41:01
metrics of a company or what the
41:03
product should result in metrics. You
41:06
should walk in and try to have a very open
41:08
mind and say, do these metrics explain
41:11
the qualitative of the business that I'm
41:13
so excited about and reinforce why
41:15
it's special. And if you think those things are true,
41:18
I think then you should lean into it a lot more
41:20
and try to really get to the guts of it. It could be a great investment.
41:23
And I see this a lot with investors. When you get too pattern-matchy
41:26
on why companies should be the way they are, I
41:28
think it leads to a lot of mistakes. Flip
41:29
side, because that could lead into another
41:32
fucking hour. You can see why I do
41:34
this for a living. But
41:36
on the flip side, when you think about biggest
41:38
investing win or success, I
41:41
get it say earlier and DPI
41:43
takes time. When you think about biggest investing
41:45
win or success, what was that? And
41:47
what did you learn from that process? You're
41:49
only as good as your next investment, Harry. And
41:52
I've still got a lot to prove. So it's hard to
41:54
say we won. But I do think an important
41:56
part of winning and just understanding
41:59
how to embrace.
41:59
kind of this culture
42:02
of making great investments is you've
42:04
got to figure out how to be authentic to yourself
42:07
and map that to how you go win. You
42:09
know, what a founder might be attracted to for
42:12
me might be different than you, Harry,
42:14
it might be different than the next investor that's
42:16
listening to this podcast. And so you
42:18
can't just apply someone else's style
42:21
and say it's going to work for you and you're going to win. You
42:23
got to try to figure out how to map what's
42:25
really authentic to you and evoke that
42:27
emotion in the entrepreneur and get
42:29
them bought into why that could be a really
42:32
fruitful partnership for them. And it's hard
42:34
to do that. You can't do a lot of those things at once. You
42:36
can't treat it transactionally. You've got to invest the time.
42:39
There's no substitute for the time. The things
42:41
I look back on that were the most exciting
42:43
and rewarding for us to be a part of, we've
42:45
been able to establish that kind of ground
42:48
with whoever is most important in that process or
42:50
the team. And when you do that, I think those things
42:52
become obvious. And the decision is less about
42:54
all these other variables in the process. And it's much more emotional
42:57
because they want to work with you. Vince, that was bizarre.
42:59
Like, one, most people
43:02
are like, oh, I'm not going to give
43:04
an answer to the mistake, but I'll give
43:06
you the answer to the success. You did
43:08
the opposite. And then two, when you were
43:10
like being political at the beginning, I was like,
43:12
oh, come on, man. But that was a really
43:15
good answer, which was fantastic.
43:18
I really like that. I want
43:20
to move into a quick fire. So I'm going to say
43:22
a short statement and you give me your immediate thoughts. Does
43:24
that sound OK? Sounds great. Karim
43:26
told me that you are basically the encyclopedia
43:29
of business.
43:29
What does your content consumption
43:32
look like? What are the favorites? The
43:34
short answer is just reading anything. I think
43:37
for me, I like to try to subscribe to lots
43:39
of different angles of reading. So sometimes
43:42
it's podcasts on topics. Sometimes
43:44
it's newsletters. Sometimes you go find the
43:46
blog focused on the developer to go learn
43:48
about the thing. Most importantly, like learning
43:50
about the history of things is really important. All this
43:53
stuff is happening right now. And so where I've
43:55
been reading is trying to trace the key figures
43:57
from all the way back to like 1980s in a.
43:59
AI to today. And so for me, it's
44:02
really just about variety. Simulate
44:04
the personas, simulate different people
44:06
that are in the ecosystem and go learn about
44:08
them. And wherever you can find content on that, I'm willing
44:11
to read it. Vince, you can invest in one
44:13
multi-stage firm other than Thrive.
44:16
Which ones you invest in and why then? Harry,
44:18
I'm only investing in Thrive. There's no other in that.
44:21
Like, you could do founders fund, and now it's upside.
44:23
Or you could do sequoia. Are we going for a sequoia?
44:26
Give me one. We're concentrated. We only
44:28
invest in ourselves. I can't give you
44:29
one, Harry. If you were to invest in one seed firm,
44:32
can you give me a seed firm? I can. I don't
44:35
know him personally, but Nat Friedman, who's
44:37
now focused a lot on AI and I think has
44:39
his own fund, I really respect some of the stuff that
44:41
I've seen him go do. And I would be certainly
44:44
interested in putting money in his fund. What have you
44:46
changed your mind on in the last 12 months,
44:48
Vince? This is not for the rapid fire, but one
44:50
thing that's become clear as we've kind
44:52
of gone into this more difficult
44:55
environment to operate in is in the
44:57
good times, we probably
44:59
over-attribute to teams and products
45:02
how good they are. And in the bad times,
45:04
you can't just blame everything on macro, but
45:06
I think it makes you reflect on the over-attribution
45:09
you probably did in the good times. And
45:11
one thing I've changed my mind on is you got to be
45:13
more balanced about how much credit you give
45:15
to the momentum of a company
45:18
from the market environment versus the
45:20
actual execution they're doing and know that
45:22
there's the balance there. And good execution
45:25
doesn't always mean that it leads to great momentum.
45:28
Sometimes great momentum is also influenced
45:30
by these market environments or
45:32
variables that are harder to quantify.
45:35
What do you think was the craziest thing we saw
45:37
happen in the low interest rate
45:39
environment of 2020, 2022? Crypto
45:43
stands out. We talked about peakmarketcap and
45:45
.com. I think crypto's peak market cap for
45:48
tokens was like 3 trillion, somewhat similar.
45:50
The most iconic company that people talked
45:52
about was a fraud. I think we're going
45:54
to look back on that, particularly from an investor lens and
45:57
say there's these hype cycles that lead to massive
45:59
speculation.
45:59
And sometimes even the things people think are
46:02
most real are just not. Okay, you're on incredible
46:04
boards. You can choose one board member
46:07
for your company. Who would you choose? The
46:09
person that stands out that I've learned
46:11
so much from is Eric Fichria. He's on
46:13
the benching board. He's on the board of this company I work with called
46:16
Airplane. Eric is, I think,
46:18
just an incredible blend of has the
46:20
operational instinct rigger, but also
46:23
is fun to be around. And he's able to land
46:25
his messages in a really effective way with entrepreneurs.
46:27
And he also just has a really great balance perspective
46:30
on being commercial and understanding how
46:33
all of that works and interplays with the strategy of the
46:35
company. And I found that perspective
46:37
to be something that I'm continually learning from
46:39
as I'm listening to him. And when you find those kinds of
46:41
people, I think you just want to surround yourself with them.
46:44
What was your biggest lesson from working with Josh,
46:46
then? There's so many lessons this hard. Josh is an amazing
46:48
person. This is not about investing,
46:51
but I think with Josh, biggest lesson is
46:53
really that, like, to be successful,
46:55
you don't have to compromise on all
46:57
of the things that are important that are not your career,
47:00
your family, your friends. Josh,
47:02
one of the most amazing parts of being around
47:04
him is his warmth and empathy
47:07
and his priority of his family
47:09
over everything else is so
47:12
obvious when you spend time with him. I've respected
47:14
that so much about him. And I think it's
47:17
even changed the way that I prioritize how I run
47:19
my life. Tell me, Vince, final one, what
47:21
are the next five years hold for you? When we sit down in 2028,
47:23
where do you want Vince to be then? Hopefully
47:27
we'll be talking about some amazing AI companies
47:29
we invested in that created lots of value
47:32
for us both. But I think we have a lot of
47:34
ambition at Thrive as a firm, and
47:36
I hope to be a big part of us building it
47:38
and ultimately going and backing some of the
47:40
next transformational companies, but also building
47:43
our team and maintaining this amazing
47:45
culture that I think we have and attracting some of
47:47
the most talented people that want to go
47:49
invest in these kinds of companies to come work with us.
47:51
And so I hope if we talk again in five years, we're
47:53
talking about those companies, we're talking about the people on our team.
47:56
And ultimately, we're really excited about all that stuff.
47:58
Vince, thank you so much for putting that on.
47:59
putting up with my prying questions
48:02
and kind of not letting you get off on some of them.
48:04
I really appreciate it, but this has been fantastic,
48:06
man. I've wanted to do it for a while because I've heard so many good things.
48:08
So thank you so much, man. Thank you,
48:10
Harry. It's really fun to chat.
48:13
I just love that discussion with Vince. And if you want
48:16
to see more from us on video, of course you can on YouTube
48:18
by searching for 20 VC. But before we
48:20
leave you today, we need to talk about Son.
48:23
Son is an AI powered learning and knowledge
48:25
sharing platform. Think of it like chat GPT
48:28
for all of your company's knowledge. Son
48:30
integrates with all
48:30
your company's apps in under five minutes
48:33
and can search through every single file,
48:35
doc,
48:35
pull request, video and more in
48:37
under 100 milliseconds. Assistant
48:39
is generative AI at its most useful.
48:42
Say you need to create a course on
48:44
OKR fundamentals for your employee
48:46
onboarding program and you're just really short
48:48
on time.
48:49
Assistant can generate the outline and
48:51
contents from scratch, complete
48:54
with relevant imagery. You can tweak it, check
48:56
it and then ask
48:56
Assistant to publish it in five other
48:59
languages and assign it to all new hires
49:01
in five markets. And they've raised over $50 million
49:04
to date from some of the best. And
49:06
you can request a free trial at saanalabs.com
49:09
forward slash 20 VC. And
49:11
if saanalab really unlocks the knowledge in your team,
49:13
marketer hire brings that incredible knowledge
49:16
to your team. Have you been spending months
49:18
looking for the perfect hire? Marketer
49:20
hire is a talent platform to hire expert
49:22
marketers on demand. The hiring process
49:25
takes less than a week from initial consultation
49:27
call to kicking off work. It's free
49:29
to use and you only pay if you hire someone
49:32
and they know quality of supply is everything.
49:34
And so the application process for talent is extremely
49:37
rigorous. Over 5000 marketers apply
49:39
every month and only 3% are accepted.
49:42
Over 25000 successful matches
49:44
and counting have been made and they are the number
49:47
one marketing only hire platform in
49:49
the world. And one of the fastest growing B2B
49:51
tech companies of the last decade. And marketer
49:53
hire is offering our listeners
49:56
a $1000 credit for first
49:56
time customers. Go to marketerhire.com.
49:59
That's marketerhire.com forward
50:02
slash 20 VC and use the code 20
50:04
VC to get your $1,000 credit. That's
50:07
marketerhire.com slash 20
50:10
VC with the code 20 VC. As
50:13
always, I so appreciate all your support and we have an incredible
50:15
episode coming out with AROD on Friday.
50:17
That's such a special
50:19
one.
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More