Investing in 10+ Unicorns: Kevin Hartz on Finding and Crafting Winning Founding Teams

Kevin Hartz is definitely one of the best and most gifted investors in the Silicon Valley.

He invested in PayPal, in Uber, in Airbnb, in DoorDash, in SpaceX and many, many more

companies.

But he would definitely kill me if I'd say he's an investor.

He's on the other hand founder.

He is the founder of Zoom with an X in the beginning, not a Z, which was sold for around

a billion dollars to PayPal later in the journey.

He co-founded Eventbrite together with his wife Julia and they went public.

And now he's building Sauron, his new venture.

We will talk about this later while he is running his venture fund ASTAR.

So one of the topics that I wanted to talk about with Kevin was founding teams because

he did it from both sides.

He found the best teams that build ventures.

I think a look at his portfolio definitely shows that.

At the same time, he built successful companies, two companies that were unicorns.

And he's building the next one.

So all the learnings of his life about founding teams combined, I think are super, super interesting

for every founder starting to build a company and focusing on the early hires.

The key hires is so, so important for the trajectory of your venture that I've spent

a lot of time with Kevin on this topic.

And then later on, we looked into his thesis on the AI industry.

Where's the value creation?

Is the value creation and the foundational layer?

Is it in the application layer?

What is important when you build such companies?

How do you identify good ventures from an investor perspective?

And then also, what is the opportunity creation currently?

And why is he building a new company?

How does he do it?

What is the specialty that he's looking forward when we talk about building a new company?

Afterward, I asked him sadly off the record about his energy management and he was like,

yeah, I'm not the founder in my twenties anymore.

So I'm not out hustling everyone anymore.

But I think as a CEO, my duty is to inspire others by giving them a vision, giving them

the direction that we want to run with and try to inspire and then dive deep into the

things where I can be the most helpful to unblock the team.

So I really like those words.

So I wanted to include them even when I don't have them on the record.

At the same time, I think we are more than ready to jump into this interview.

You're listening to the unicorn bakery.

My name is Fabian Tausch and here's Kevin Hartz.

Thank you.

It's a pleasure to be here.

So I thought of a long time what to ask you first and I have one specific thing in mind.

You are investing in companies for more than 20 years now.

You're building companies for more than 20 years now.

You've seen it all in quotation marks.

Like you've seen teams that you believed in that did very well.

You've had a lot of incredible investments.

There were teams that you didn't invest in that did amazingly well and you are constantly

talking to new founders.

You're building a new company.

How do you take what you learned throughout the last 20, 30 plus years and apply it to

what you're currently focused on, but at the same time are staying open for all the changes

and new things happening?

Yeah, great question.

I think that I was always concerned with this kind of expression.

You can't teach an old dog new tricks and I always wanted to have a flexible mindset.

Always wanted to be thinking about always want to be open to change and just see a lot

of things go wrong when one loses that flexibility and so I've had the good fortune of investing

in starting companies since the mid late 90s.

That's a long time and while things have dramatically changed, they also really haven't dramatically

changed at all in terms of the type of people that I like to be in business with and love

to work with.

The same spirit is essential and present whether it's the original PayPal team of Peter and

that kind of historic group of people that came together to build PayPal or Airbnb or

Pinterest or Enduro or Ramp or others.

So I look for those immutable qualities and they really don't change at the same time.

Now I'm out in the market looking for that next set of leaders and builders.

One of the nice things I'm doing is building a company with a team the company is called

Soron.

It's reimagining home security and it's an area that's kind of falling on disrepair and

that really helps just keep in mind what it means to be building and have that spark.

So when you look at companies either building or investing, it all at first comes to the

founders and the founding team as a second layer.

You said there are characteristics that are not changing that much.

I ask myself, are there things that were relevant 20 years ago that are not as relevant today

in the founder team?

Well, there's the why now and that changes depending on the period, you know, when great

companies start, there's usually an opportunity in time and place, PayPal started towards the

tail end of the dot-com boom and e-commerce was certainly a segment to bet on.

Now it wasn't clear at the time because in 2000 you had a dramatic market crash of tech

stocks and the baby was thrown out with the bathwater, the Googles, the Amazon, Netflix,

PayPal were all kind of considered dot-bombs and not viable businesses.

But in fact, they were absolutely the stalwarts that we see today.

But you're looking for a reason why a company should exist and a lot of times there is a

platform that one is building on.

My simple thesis is to look and see, is it Airbnb and Uber taking advantage of iOS and

the pervasiveness of iOS and growing on mobile?

Is it this new slew of companies building on top of open AI, a platform?

It's often kind of bemoaned as, oh, there's all these companies that are light wrappers

on top of open AI, but I love people building on the application layer.

That's where an immense amount of value is created and that's where I focus a lot of

my time is a new platform emerges and developers build on top of it and build into something

rather substantial.

That's how my companies Zoom, XOOM and Eventbrite really started on top of building on top of

payment platforms, building on top of PayPal and later Stripe in terms of payments or pervasive.

What kind of applications can we build to take advantage of these new payment platforms?

One thought I had when we think about the best founding teams, would you say the best

teams will also pick the best opportunity?

For example, now building on top of the platform of open AI or something as relevant because

it's definitely not the only thing that you could currently build, or would you say sometimes

the best teams are also working on stuff that is not as visible, the next big thing?

How do you think about a great team and then also the market opportunity that they're growing

after?

Well, great teams will find their way to the biggest market opportunity, hearkening again

back to PayPal.

It didn't start as PayPal, it started as FieldLink, which was security for mobile devices.

I think there were palm pilots at the time and it went through a number of quick iterations

before finding the right market, which was online payments and merchant processing.

It took off from there.

Similarly, the best teams, it makes me think of what Jeff Bezos would say, he wants to

hire people that are mostly right, meaning when they make decisions, they can write more

times than not, you can't always get things right and you've got to take some risk and

so on, but the best performers are mostly right and move into that right direction to

find those right markets.

Now the contradiction here is that it's a red ocean of competition around open AI, a

building on top of these platforms.

There's so many different generative AI tools and platforms and so on out there that how

do you, why would you run into this very competitive space and so there's somewhat

of a conundrum there, but there's still many opportunities.

When we devised SOAR on home security, we were looking more ahead at computer vision,

which is not even intensively focused around LLMs and that was really our thesis is that

computer vision side will see this revolution at a staggered pace, but it's absolutely coming.

One of the key elements of building companies but also investing is uncovering potential

where others might not see it, like assessing people in a way that you're like, okay, I

know this person can fulfill the job and grow within the opportunity that we have at our

company or the team of founders or just one founder that I love and the team that he puts

together have the ability to build something great.

Do you have a framework in mind on how to assess potential?

I do look a lot at kind of the slope of growth.

So I'm willing to give a lot of opportunity to finding raw talent and there's a number

of reasons to do this and it's maybe if you're in baseball, it's how do you build a magnificent

team you can't go out and recruit all day or pay the price of a Yankees pitcher and

so on. So it's finding very new, very raw talent that's going to build and you see in

all the great companies, whether it's all the way back to Microsoft in the 90s or Facebook

in the 2000, 2010s and beyond, Google is finding extraordinary talent to build this and it

tends to be very, very young talent. I just saw Fiji speak. She's the CEO of Instacart,

a French leader who kind of came out of eBay in her 20s to join Mark Zuckerberg and he

just continued to give her more and more responsibility as she grew with it and there's

no better feeling. I've seen that type of talent succeed and now she's running Instacart

as a public entity. So back at Zoom, for example, I made these wagers and hired Shyam Sankar

who headed a business development and country expansion for our money remittance business

and he went on to volunteer in a CTO, COO role and has been very instrumental there and

you want to find such good talent that you see them grow, that you see them progress

on into the future.

How do you find them as role diamonds?

Grad school I think is part of it. Yeah, we were recruiting folks out of grad school.

There is this kind of birds of a feather that you see these clusters of talent kind of come

together and certainly Peter was remarkably astute at it and in his case he recruited

a lot from the original Stanford review crew like David Sacks or Keith Rebois or others

and then Max Lepchin, his co-founder, brought talent from that UIUC group that a lot of

the crew originally involved in working on the original Mosaic browser and that was just

a kind of magical combination of the two.

It's so interesting because on the one hand I totally get it, because you said you took

some wagers as well and I think grad school in itself is a wager because it's people

without much experience, probably they had some projects with other students and so what

are you looking for when you're interviewing for a role, no matter how much ownership you

want them to have immediately from day one, but what are the questions you ask to understand

if they have the grit that it needs to then also excel through the opportunities that

you give?

Yeah, it's interesting because California labor laws are bleak and prevent a lot of

the examining questions that I think are most helpful, but when you're speaking with

founders that are building a company, you can ask any array and I think it's very important

to understand the founder's background, where they came from, you see a lot of overcoming

obstacles, you see competitiveness, you see academic and intellectual achievement, where

the parents, doctors or professors or what kind of climate did they grow up in and you

just see these same kind of stories repeated over and over with founders that become a

very clear path.

It's a little harder to do when you're hiring on the employee side.

In what way are you comparing founders, for example, to the Airbnb founders when you're

like, okay, I invested very early, now they're at this point, do I believe the founders that

I'm talking to right now could achieve the same growth and development to what extent

is this helpful or harmful?

Yeah, I mean, you always need to benchmark and there's always gives and takes and those

are the big questions, but it's like you look for kind of N of one, extraordinary founder

or founders that just kind of break the mold, that see the world in a different way, that

have our kind of app to move fast, they want to build, they want a sense of urgency that

is unbridled and a passion towards what they're building.

We always miss the mark on these, but I would say it's clearer over time and seen such a

wide pool of founders.

When I look at companies, especially building out of Europe, then often their investors

say, hey, hire as senior as possible, which would be the complete opposite of like giving

grad students the chance to move forward.

In what moments would you say it's experience over potential, like when do I hire experience

and when do I hire for potential and grit?

Probably that's the better framing.

Yeah, I mean, the vast majority, I would go with kind of speed and grit and in training

somebody up, there are certain areas like we're bringing on a perception engineer and

that's a very specific area that we just need somebody of that caliber, but for the most

part, I would always choose that slope or that high growth candidate over somebody that's

been in the market, somebody that's been so-called experienced is not going to necessarily move

at the same pace, has a certain way of doing things, is not as malleable, is not as forward

thinking and it sounds like ageism to a certain extent and maybe the valley, when I say the

valley, that includes all in tech, the metaphorical valley, but I think that's where you find

just the greatest variability in growth.

Is there a point in time of the life cycle of a company, like when you come above the

typical startup phase and you're scaling up and it becomes, you're still pushing, but

it's also managing like what's there, does it change then do you then look for different

candidates and different kind of talent or would you still like, where is the, is there

a flipping point of talent?

Yeah, I mean, there's, you know, a company is like a, you know, a book, it has many different

chapters and each chapter is kind of different and the story progresses over time and you

do need to hire differently at different times, like the challenges is that, you know, as

a company matures, like there's just a, the momentum is it's harder to keep things like

in forward momentum and to keep things from becoming too sedentary.

And so, you know, you want to create, keep that sense of urgency driving, you know, people

forward, like at a time where companies just naturally, you know, want to slow down.

And so like it's incredibly admirable of companies that have kind of just continued to grow

through, you know, these sort of periods.

But, but yeah, I think that hiring methodology, you know, lasts for most of the lifetime of

a company. And yeah, I would also kind of posit that there is, you know, other little

traps and things, you know, we've seen, you know, maybe on the AI side, if you're, there's

been a lot of PhDs and so on entering the market, there you have to kind of gauge, like

is this, this person has like a lot of specific knowledge of, you know, some area of that

AI, but is there a kind of commercial sense to it? So always kind of gauging in that core

team, is there a commercial sense?

I think the one, one side we talked about is finding talent. I think the other side is

retaining talent when we're talking about company building. And you mentioned that you're happy

to trust talent with opportunities and give them more and more ownership and like what

projects they can go forward with. What would you say is the most important part about like

helping talent to succeed in inside an organization?

Yeah, there's, there's this, you know, kind of proverbial lifeboat exercise where you

find that, you know, maybe it's a relatively small amount of engineers are having an enormous

impact on, on the organization. And, you know, you want to recognize that and, and work to

retain and challenge those engineers. They're usually the ones that you can steer on, you

know, kind of blue sky, totally new projects. And, and they'll go after this with, with

bigger, you know, so, so that kind of recognition is, is important. What would I say, outside

of that? Yeah, and, and I've always found it, you know, you can never give too much stock

to the most talented, you know, just loading them up with, with options or stock is, you

know, worth its weight of gold for your, you know, kind of top tops. But I think that in, in

general, when you're at a startup, you all the way through maturity and beyond, it's, you

know, it's, you go there because of the people you're working with, and the things that you're

learning, and then, you know, compensation and these sort of ideas as well. And so just from a

leadership, a founder mentality is like, are you providing that to the team? Or do you have

people just slaving away in obscurity and in some de minimis part of the project, you know, and so I

think kind of keeping that kind of empathy in mind of, of, and in understanding the battle that the

team is going through is, is critically important. I also just love to give, you know, those that

are performing really well, just, you know, like, put them in front of investors or put them in front

of directly in front of customers or other areas of exposure and just see how they do. Because

often those like stars just will shine on all dimensions.

How would you describe your own leadership style?

I would describe it as like giving it a tremendous amount of, of kind of, of yielding to kind of my

direct reports, yielding to them to make magic happen. And, you know, that's especially important

now because, you know, I'm building Sauron, but I've also got, we just launched our second $300

million A star fund. And, you know, and so I've got a fair amount on my plate. And so it's a

forcing function to actually delegate. And I'm probably an overly big fanboy of Peter Teal. But

Peter did this incredibly well, where, you know, he didn't seem to ever want to manage people, but

did a great job at hiring and letting the talent, you know, operate on our own. And then you see,

you know, in his first go around at it, you saw, you know, Max Levchin, the CTO and roll up both of

the CFO now running Sequoia and Keith, now at COSLA and Reid Hoffman created LinkedIn. I mean,

just incredible, incredible, incredible, dense talent. And, and so, you know, aspire to be even in

the zip code of that type of talent.

So talking about Sauron and building it, how do you, how would you describe your own role? I think

you did it in a form function right now, in the last answer. But what are the parts that you with

all the things on your plate can do? And what are the parts that you can't do with building a new

company? Yeah, I hope I hope it's more inspiring. You know, I'm as much as I don't know, I was up late

last night. We were searching for 3D designers for like game designers for the way that we're building

our interface, depends on that. But yeah, I'm, you know, part steward of capital, part setting this

initial vision for what the product is. And then, you know, pulling as much of the team along as, you

know, but, but to be honest, like the team's remarkable self starters, you know, they're just going 24

seven, they're building there. So, so in a lot of ways, I just have to be patient. We have a, we're going

to give some demos on July 9. And that's, that's, you know, all that needs to be said is a lot has to

happen between now and July 9 to be able to, to be at the point where we present.

What were the exact positions that you were, or founding team members that you were looking for, for

Sauron, just to get an understanding for the specific case, what were the first people you brought

on board? Well, Vasu is our co-founder. So it's myself, Jack Abraham, who I've known for years. He's,

he's the creator and founder of Atomic, which is a startup studio. And I, ASTAR did this in conjunction

with, with Jack. So it's the two of us. And our initial search was all around finding Vasu. And

she's extraordinary. So she's a roboticist. She is Caltech post grad. She was at Neuro. She was at

Zooks. And then most recently, in four and a half years at Zipline. So she's, she's, you know, right out

of central casting for what we're looking for on the robotic side, and perception side for this. And,

you know, and, and so she checks all the boxes for like the skills, but it's also kind of the

attitude. And it's the kind of growth mindset, the mentality of wanting to learn, you know, wanting to

build something really caring, and you can kind of see, you know, you just want that in a founder. So

we're pretty delighted we have Vasu on board.

I listened to a talk that you gave at some point, and you were quite strongly opinionated on incubations.

And you said 99.9% of incubations suck besides snowflake. And it sounds a bit like an incubation,

what you're doing. So why?

Yes. Well, well, hopefully I can prove myself wrong here. But incubations in general just tend to be very

kind of stiff. They lack a soul. They're kind of assembled together. You know, where instead of

pulling together founders, you're pulling together, you know, hired guns. And at the end of the day,

do people care enough to end the mission? Do they care enough? And so, yeah, we have a lot to prove

here. But I'm, I'm seeing some very non incubation like patterns for our incubation, which is good,

because it's, it's, it's, it's somewhat terrifying if you look at the just absolute track record of

things of incubations out in the market. But yeah, I wonder at the end of the day, if it is much

different, maybe we don't have as much of large sample size, but, you know, most companies, most

startups fail. And, you know, maybe incubations aren't, aren't, you know, much different.

What do you say are the things other incubations do that you tried not to fall into?

Well, I think, I think there's a, like this, you know, the, it's developing an ownership

mentality. And there's an expression, nobody ever washed a rental car. And Americans know that,

you know, like, get that immediately, because there's this like, when you, when you rent a car,

you know, you'll run it into a post or you'll back over a fire hydrant or something, and you'll

say, oh, no worries, it's a rental, you know, it's because you don't care. And, you know, there's

like an Airbnb example where somebody during the holidays, like nailed a Christmas tree into a

hardwood floor, you know, of somebody else's house. And again, it's like, when it's a rental,

people don't care. And so how do you create that ownership mentality that when push comes to shove,

like that, that somebody really cares that this is their work that this is meaningful and impactful,

and they're going to go to any lengths to make this succeed. And that's the usually what, what

misses out in incubations is that you get kind of bored of the idea or your mind drifts,

or you go chase the next big AI or crypto thing and it's done.

I think all the Germans listening right now are like, yeah, but we had Rocket Internet and they

incubated or copied and incubated a lot of things and and then like Zalando, Hello Fresh and others

were still, still like did well. And also the founders are dummy, please don't be sad, like the

CEO and co founder of Hello Fresh, but like not having that much equity and still being the public

company CEO. I have no clue if there were new incentive packages and how they reiterated on

that. But at the beginning, there was not much of equity packages and still somehow they made it work,

not for every company, like a few of them, but Rocket Internet is an interesting example.

Well, I think the somewhere's kidnapped and threatened their children. No, I'm just kidding.

I know Ali and the team, they actually invested way back when when we were starting Eventbrite,

they invested in Eventbrite and that was like a good way to ensure that they didn't copy Eventbrite.

But you know, that is a model like they're like I was always impressed with how fast and was such

urgency, how quickly the somewhere's could could move and could drive and they knew how to find a

certain archetype of, you know, ambition, you know, up and coming ambitious, you know, founder,

maybe out of consulting, maybe out of McKinsey, maybe out of banking, high IQ and run with it.

And so you can say a lot of things about the somewhere's like I remember when they were trying

to compete with Airbnb and it was a lot of bitterness, but they had a style of building

companies that worked enough as Zolando and Hello Fresh exhibit.

On the other hand, those were, I think, a bit more execution heavy companies than, for example,

what you're building with Sauron. So there's a difference in the kind of incubation because

they saw a proven business model and rebuilt it from the German market or then in different

markets as well, like when they copied it and build something in the South American market

or Asian market or so those were execution heavy businesses, I think not much R&D.

So we can also with everything, it's not black and white.

So we can layer here and say, OK, for those kind of businesses, it might work for those

it might whatever.

So it's super hard to say it's either or, but it's an interesting example from that point.

And when we look at Sauron, I mean, knowing that you invested in a few good companies,

so therefore also you have the opportunity to finance and fund your own startups if you

want to.

How do you think about capital when you're building a new company?

Yeah, I think that like having a fund raising formal rounds is important.

And having tried it both ways where I've just kind of slopped my own capital into things,

it's helpful to have like the checks and balances of doing a C to series A and so on.

It's helpful in investing to have, I find to have LPs and it's a way of keeping score

and keeping honest your real returns because what investors tend to do when they're not

obsessively calculating actual results is that you tend to fool yourself into everything

is an Airbnb and a Pinterest and you forget all the ones that went sideways or didn't

make it or so on.

Like capital and our LPs are a great way for us to ensure we methodically drive returns.

Now, we don't want to overthink about that.

Everything shouldn't be turned into like an IRR, MO, IC or so on because but at the end

of the day, like the end result is those numbers will look very good.

It's interesting because in the German market, for example, and I'm comparing here because

it's interesting to discuss a bit of the differences.

A lot of the founders that then have money, try to build on their own to own more equity,

to don't like have to grow as fast, to have a bit more time.

And then there are examples, for example, like Scotcha Kohn, a good friend of mine,

who's one of the co-founders of GitHub, who then at some point self-financed his venture

back and then realized at some point, okay, I self-financed it, but now that I would need

to take a leap, I'm not at the point where others want to finance it and fund it.

So I self-financed it to a point where I cannot continue it.

So he discontinued or let all the people go now.

It's like running with one person and then it's profitable.

But having the issue of not running for the metrics for a few months or years could push

you off track that if you want to go back, it's hard to.

So it's interesting to see that, and in the German market, there is a trend towards,

hey, let's build bootstrap businesses, our cash incentives are better than,

because if I do 10 million in ARR and it's profitable, it's sometimes a better exit

than a mediocre venture that I built.

It's interesting how the incentives of building companies are shifting around.

Yeah, it tends to fluctuate with a feel in the market, but on the one hand,

you have David at mid-journey who had a challenging time.

He created a great product in lead motion, early haptics.

He was a hardware engineer and didn't have a great experience with investors in venture

capitalists, and so in his next go-around with mid-journey, he's bootstrapped that.

He owns 100% of that business.

And likely doing a few million a day in that revenue, like it's extraordinary,

but the downside of that is that he's got to drive himself.

As the sole shareholder, I'm not sure whether or not he grants stock to employees.

I'm not sure if he does, but as the sole shareholder, he doesn't have as much of the accountability,

which is great on the one hand, but can you build a world-class organization on your own?

Like therein lies kind of the trade-off, but I guess I'm not going to feel sorry for him

to be in the position he is.

It's incredible what he's built.

And then you have founders that have built their companies and been deluded down to almost nothing

when they could have built a business and did 2 million in ARR and just dividend it out to themselves

in a much more effective manner.

There's always that trade-off, and you just have to ensure that you're not kind of, you know,

that you don't constrain things too much if you take a more bootstrappy mode.

Like bootstrapping is always the best way to build a company.

You're building in a very capital-efficient manner.

And if you can do that, you absolutely should do that.

And you should have a thought and a path and a direction to do that.

In some cases, you don't have that option.

Certain businesses, you know, hardware, software, so on, would be harder to bootstrap.

But, you know, like in today's world, like the InfaCost, you know, I think it was Max Lefkin

a number of years back published a blog on when will we have our first one-person,

billion-dollar-valued company?

And it must have happened somewhere.

Maybe that's mid-journey, but I guess that's not one employee.

But just the fact that you can outsource so much infrastructure, software services

and build such a kind of, like I kind of think of it as a perpetual motion machine

that can just run on its own and rack up higher and higher revenues and returns

is kind of a great thing.

But in a great aspiration, you just have to ensure it doesn't come at too much cost

or that then competitors just run you over when they see the opportunity

and know you're undergone in many ways.

You mentioned the opportunity to outsource many things and you're building a new company.

And the question is for everybody, like how much do I, like what are the things

that I want to build internally and have to have as a mode?

And what are the things that I can work on with freelancers

or software that solves issues or infrastructure layers that solve issues?

How do you decide what's crucial for building internally versus, okay,

it's faster and better to, probably it's more expensive at first or over the long term,

but it's not as crucial to build our mode and we therefore go for freelancers

or even software and infrastructure layers.

Yeah, I don't really think another Peter Thielism in the late 90s,

he just sold, early 2000s, he had sold PayPal to eBay

and he bought a place in the Four Seasons residence in San Francisco,

never did his laundry and he would say, Kevin, you outsourced your non-core competencies.

And so in Peter's case, he wanted to think and build.

He wanted to play chess and spend time with friends and family and so on.

And so he outsourced everything else, laundry, dishes, the whole thing.

And you can kind of look at that same way where the company is,

like what can anyone do, what's kind of a commodity.

In the case of Soron, like we want to kind of outsource,

we want to not reinvent the wheel unless we absolutely have to

and where we absolutely have to reinvent the wheel,

that would be a proprietary advantage and so we'd spend time on that.

But we're building hardware, software, we have insulation,

we'll have insulation teams and so on.

But in our first incarnations, it's to get out in the market

and so we're going to use off-the-shelf hardware,

we're going to use off-the-shelf sensors and LiDAR and radar

and all these different systems.

But over time, we'll learn from that and boil them down

and in time with success, we'll build our own hardware,

contract manufacture towards building our own hardware.

But if you go out the gates and try to do everything and boil the ocean,

you just won't be successful.

There's too many points of uncertainty, too many variables loose.

So thinking about the problem in that source is like,

where can you apply the most unique advantage in your expertise?

That inevitably falls around that software AI side.

I think a question that arises more often...

Oh, good. We can cut that out easily.

I think one of the questions that arises more often also on the AI side,

but it fits perfectly towards your case as well,

is the question of, is the go-to market capacity

and then the customer base that I'm building

and everything from that point onward

would be enough defensibility for the start to build on top of?

Because I think in your case with OK,

we're not also building our own hardware,

we're going with off-the-shelf at first.

It's a bit like OK, an AI wrapper or an application layer,

where you're like, OK, this might be something

that some other LLM or AI infrastructure company

might solve in the next months,

but we have our customer base until then built to a certain point

and we can now also iterate towards the needs faster

with the new technology capacities.

So the question stays,

is the go-to market motion and capacity defensibility in itself?

Well, in the case of Sauron,

it's a very complex business.

We see a lot of players in the market that have stumbled

just in terms of poor execution of installations.

If you have to send an installer out more than once,

like already you're losing money,

and so there's a multi-points that we can differentiate,

but those are also multi-points which make us vulnerable to lose,

and we have to really show that where we see Sauron really shining

and our strong belief is that we're going to have this kind of...

post-World War II film, The Wizard of Oz, was released

and it was in black and white towards the beginning

and then Dorothy lands on Oz and the door opens and the world is in color,

and so the first time the world saw a film that was in color

and it was this incredible moment,

and we see that with Sauron of how we're representing data,

we're doing it in a way that hasn't been represented before,

and I'll leave it a little bit at that,

but we're very excited to unveil that to the world

and that's what our target is on July 9th.

Getting back to the AI topics and moving a bit forward in our conversation,

before I come back to how you think about AI,

I think we talked about staying open to new founders

and understanding why they might be the ones that change the world

and investing in them.

The same happens with topics and industries that are created

and staying open-minded about, okay, that's happening right now,

building your own opinions and making your own opinions

and figuring out, okay, what to go with and what to bet on.

How do you try or how are you assessing,

on the one hand trends, but also technologies and potential industries

to understand if you think they're worth a bet, they're not worth a bet

and staying open-minded to that and discovering those new platforms

that you talked about earlier?

Yeah, great question.

It's so fun in tech to really see a new phenomenon start to arise

and it can manifest itself in so many different ways,

like the sharing economy in the case of Airbnb.

We looked at it like when we had invested,

we had put a term sheet into Airbnb after demo day and so did Sequoia.

They went with Sequoia, but they liked us enough

and rule off and the team cut us in.

But our thesis around that was very much a business

that we saw as like a distributed storage system.

So you can have these big EMC storage systems of the past

which were on-prem and you'd have to cool them and maintain them

and have people out to fix them and update parts and so on.

Or you can have distributed storage throughout the internet,

in which case you have this perfect system of not having to maintain all of that.

Airbnb was almost like that conception of it.

As we thought about it, we were very excited about that possibility

and it has absolutely become that and in spades.

If I think about it, it's a lot like I have a great mentor, Michael Ovitz,

who was the founder of CAA in the mid-70s.

During that golden era of Hollywood, he was the sort of kingmaker.

He would put together these films, Jurassic Park and others,

and actors, Tom Cruise and so on.

And I see the tech businesses like that.

You're trying to find the great stars, the great talent of tomorrow.

You're trying to find the Beatles in a sense and help this all come into fruition.

And you do that around these different trends that are forming,

these platforms are forming, and that's why you hear this

why now expression of like, okay, it's 2007 or 2008 and Steve Jobs releases an iPhone.

That's an opportunity, rush to it and build.

But sometimes that always doesn't happen.

I remember it was so exciting at the Facebook F8 conference,

the first time they opened their API and then they said,

well, let's not be too generous with developers and they effectively shut it down.

Maybe Zynga and a couple of companies snuck through there.

But if finding what is going to be that next Airbnb or that next great company

is what we're always on the lookout for.

Without throwing in any notions from my side, it's June 4th, 2024.

So the question that I'll ask might your answer might not be the same in 6 to 12 or 18 months.

But what would you say?

Where does the value creation in the whole topic and industry of AI happen

throughout the next months and years?

Well, I mean, it is at the application layers.

And, you know, I'll call that like being a thin wrapper is like a great thing to be

on top of open AI because you start out there, but you build something of substance over time.

And, you know, most of these things start out incredibly simplistic.

If you look at, you know, the first incarnation of PageRank and Google Search

or the first incarnation of Zuckerberg's Facebook directory of people,

its sophistication and defensibility grows tremendously over time.

And, you know, we backed a company just less than a year and a half ago in Madrid.

It's called Lucia, and it's a Latam-focused chatbot.

And, you know, we put a quarter million dollars in at like a two million pre

as they were just getting Alvaro and the team were just getting started.

And, you know, today it's doing a few million in DAUs and dominant in Latam.

And, you know, we see that as just the beginning for a lot of these opportunities,

whether it's building specific vertical chatbot like Lucia

or whether it's, you know, kind of enterprise sales aids

or whether it's customer service in terms of a company we're working with called Decagon,

which has automated tens of thousands of customer service inquiries a month to Eventbrite

and seen that work extraordinarily well and them doing it for Substack and many other great services.

Why are there so many people thinking that the application layer or building on top

and building in an eye wrapper or application layer is not enough?

I think it's a bit of engineering arrogance like that everyone wants to build a foundational model

and that you have to build a foundational model to kind of build something successfully

and you don't, like they're out there, like you can build on top of them.

And so building on top of them seems like too light and cheap and airy,

but that's, you know, again what we did when PayPal, we convinced David Sacks to open.

He was the head of product at the time of PayPal and Alan Braverman and I convinced him to open up an API.

And so we were the very first developers on PayPal and if you looked at our initial apps that we built,

they were incredibly crude, but, you know, they gained in complexity and elegance and defensibility over time.

And so I think when people see this initial starting point, they just think it's something flimsy

that open AI or Microsoft or anyone else will just run over.

As we talked about Peter Thiel before and Peter, I recently saw a snippet where he talked about how he thinks

AI might decrease the importance of math skills and increase the importance of word skills,

like being able to, yeah, be great with words.

So he compared it to, in the, I think in the 90s, he said he wanted everybody to,

or he thought it would be enough instead of testing for math skills to test for chess skills.

And then at some point, because he was in his chess era and at some point he realized,

yeah, maybe that's not a good idea later on, but his statement was very clearly on,

he thinks word skills will become more important than math skills.

Have you seen the snippet? Not that I'm.

I haven't. I'm trying to understand why word skills would be more important.

I can quickly switch it up.

Maybe it still seems like eloquence can be parroted from NLLM very quickly, though.

I've interviewed people before where I know they were reading off of LLMs for answers and technical answers.

Yeah, let me find it and we can talk about it once you saw it. I think that's easier.

Where do I have it?

My intuition would be it's going to be quite the opposite, where it seems much worse for the math people

than the word people.

And what people have told me is that they think within three to five years,

the AI models will be able to solve all the US math Olympiad problems.

And that would shift things quite a bit.

Silicon Valley in the early 21st century, it's way too biased towards the math people.

I don't know if it's a French Revolution thing or a Russian sort of Straussian secret cabal control thing,

where you have to prioritize it.

But that's the thing that seems deeply unstable.

And that's what I would bet on getting reversed where, you know, it's like, isn't it like the place where math ability, like, you know,

it's sort of, it's the thing that's the test for everything.

It's like, if you want to go to medical school, okay, we weed people out through physics and calculus.

And like, I'm not sure that's really correlated with your, I don't know, your dexterity as a neurosurgeon.

I don't really want someone operating on my brain to be, you know, doing prime number of factorizations in their head while they're operating on my brain or something like that.

And so, you know, in the late 80s, early 90s, I had a sort of a chest bias, because I was a pretty good chess player.

And my chest bias was you should just test everyone on chess ability.

And that should be the gating factor. Why even do math? Why not just chess?

And that got undermined by the computers in 1997.

And isn't that what's going to happen to math?

And isn't that a long overdue rebalancing of our society?

I mean, is it investor in, is it standard cognition in Devon?

We're implementing Devon right now, which is, you know, the code writer at Eventbrite.

By the way, it's an irony that founders fund back that because they were so anti the conventional AI path of there's so much noise in the market and then to go into one of the most crowded markets.

But they did so I would imagine based on the team that's been put together there, which are all IOI math Olympiads.

And so, at least Peter, yeah, at least Peter and Napoleon still very much value like the mass skill in building.

But I guess, you know, maybe I would suppose that, you know, at some point Devon will surpass the whole team itself in terms of writing code and and all of a sudden the cult of like the killer IOI engineer kind of goes away.

And I have one last topic that I thought about.

And also, I think it changes probably, potentially, also a bit with all the potential around AI.

How do you think about the life cycle of companies?

Life cycle in terms of speed or.

Sort of how long does it take to get to a certain point like, let's say, getting IPO ready, of course, knowing that it depends on windows, like when is the IPO window, like what kind of company am I building and whatever.

At the same time, also, the life cycle and the chances to build a generational company versus a company that might be a great exit at some point, but might be not a generational company.

Well, Amazon started in maybe 93 and went public 95 or 96.

And John Doher used to say that if a company, you know, this is 80s and 90s rubric if a company had six quarter, either three or six quarters of increasing revenues, then it should go public.

And so there just used to be one, you know, you didn't have all the regulations, our reins oxley and all the challenges of getting public.

But it was considered the natural progression to get there.

And now you just have a lot of founders that don't want to go public like they could very easily.

But they just don't want to.

And, you know, there's an argument that going public really elevates the game.

It puts you in front of, you know, the public markets to be on display and a chance for you to really compete and have a liquid stock currency to recruit and hire.

But there's very little interest by very few founders.

You know, Steve took Reddit public recently good for him.

It's holding up.

It's doing doing well.

But he seems to be more of an anomaly in that we're in this world where we just continue to create more private companies that swell up.

And that's where you see all these crossover funds and different public funds that are now spending much more time on the private side.

This is, you know, my I would say my miscalculation of SPACs was like I had thought SPACs would be the way to kind of fix this illiquidity problem is to make a simple way to do that.

What happened is that it became it became an attraction of a negative attraction.

You got the wrong companies public or public too early.

And that's a whole different ballgame.

Although we just sold we sold our second SPAC a while back rather than going through the pain anguish of de-spacking and that company is now trading it over $300 a share, which is bizarre.

This is crazy.

Do you think SPAC will come back as a vehicle like just an other option of getting and going public?

No, I soundly believe it won't.

It's just for a lot of different reasons.

I think most distinctly a good friend, Kevin Warsh, who was on the Fed Reserve Board, you know, before I was going to before launching nice back, I wanted to talk to him and see if it was the right thing to do.

And he said, Kevin, you know, the best way to separate investors from their hard earned money is to launch a SPAC.

And he turned out to be exactly right on that.

He's mostly right.

I should have listened to you because you were very convinced, right?

I didn't want to bring this topic up, but now we're here and and you brought it up.

But I listened to like a few of the interviews you gave in that time and your plans.

And I think, as you mentioned, you did two, like A-star one and A-star two, I think were the names.

And those were the SPAC vehicles.

And it's interesting to see how the SPAC vehicle, like not yours specifically, but in general, like the opportunity or the option to bring up a SPAC got shattered.

In short amount of time, short amounts of time.

Yeah, it was a it was a SPAC bubble and it and it popped.

And, you know, I think we have this like interesting situation where we have like tech and information, you know, just move so quickly, even more quickly these days.

And there's so many more platforms being built that, you know, just so we had a SPAC bubble and now we're entering an AI bubble.

And hopefully re-entering a crypto bubble or I mean, I shouldn't hope for a bubble, but we can never just we capitalism never gets it just quite right.

It's either boom or bust.

It's the Kremlin described at once as a pendulum, you know, that never stays right in the middle.

It's it swings really hard one way or the other.

And, you know, and for that reason, in for all these other platforms, it's going to be an exciting five years ahead.

I mean, a tremendously prosperous five years, but it's also going to be wrought with, you know, there'll be a reckoning at some point as we set up and and people say it's different this time.

Coming back to the company building part.

I think you mentioned Max Lafton's article on the one person billion dollar company.

At the same time, we have all the upcoming solutions and things that are already there with all the AI stuff.

Devin and other companies.

And do you think like, do you think it's either the companies like will people and founders build more complex companies and tackle more complex problems because and take the same amount of time or longer to build them because there are new opportunities?

Or will they build ease in quotation marks easier to build companies that are like problems that are already a bit more familiar and known, but in less amount of time because of the capabilities of AI?

Yeah, I mean, theoretically, like what should happen is that we should be able to we should be able to build much faster.

You know, maybe to Peter's like outsourcing non core competencies, like I remember in the early 2000s building zoom and we had to build every aspect of, you know, fraud detection deterrence.

We had to build, build in KYC.

We had to build in filing suspicious activity reports SARS.

And now all that is obfuscated, you know, there's like, you know, there's like a library or there's like a, you know, a component or a platform to do any and all these sort of things.

And so, like being able to assemble together a lot of the components of a company, you still have like that key differentiation piece that you have to build.

Yeah, theoretically, we should, we should be able to move faster and build more.

But that's theoretically and often, you know, in, in, I like the Devon or other code creators, you should just start building up bigger and bigger platforms, you know, so event bright, I hope it's not under NDA, but we're implementing Devon.

And, you know, when we, we, so we have one instance or I'm not exactly sure how it's implemented, but, you know, when we're at 100 or 1000 Devons, like, what happens then, like how fast, you know, wouldn't we be moving at like nearly the speed of light and doing so much.

And so that's a very exciting future.

And, and yeah, but these things always like take much longer than we suppose kind of like AGI is always five years away.

One last question or topic.

I think there are not many people that are more deeply rooted in the Silicon Valley than you are.

What's your take on other ecosystems could be the US ecosystems like New York, but could also be Europe or even South America, Asia, whatever.

Yeah, ecosystems are really important.

I was talking to a friend had dinner friend, Karthik, who started Polychain, the crypto fund, and he was commenting on how like these sort of crypto havens had diminished after COVID.

And we were kind of just noticing that there was like a resettling in the Bay Area and even New York or Puerto Rico, you know, some people stayed in Puerto Rico for for different reasons for tax reasons.

But yeah, I mean, I'm really interested in kind of communities that come together around things.

I mean, but I still am not finding, I mean, outside of maybe Hollywood and that ecosystem, you know, which now the streaming leaders have all infiltrated like Netflix.

But you can see I'm very biased towards tech.

But, but yeah, I mean, I think the most interesting communities are the ones, you know, like what happened in Hayes Valley over the last couple years where, you know, it all really started in one shared hacker house.

And that hacker house kind of coming together around a community coffee shop and other community area and more people moving in.

And as I grew out, cerebral valley was really born.

And now you have HF zero and you have like all these startups sort of streaming in from in populating Hayes Valley in a way that like this whole new kind of community has has begun.

And and so in terms of movements, I don't know, it's like the left bank in Paris or something akin to that is is what I want to say.

And I'm and and I outside of maybe Bernie man, I can't think of any other like mass phenomenon like it.

For everyone not so familiar Hayes Valley is a spot and location in San Francisco where a lot of AI people met and gathered as you described and then everything else became or came a bit out of it.

And the last question for this podcast promised is if somebody's listening and says OK, what Kevin talks about seems to be very profound.

I would love to work with you and a star and who should reach out?

Who are you looking for at the moment?

What are you investing in?

Always looking for great talent.

Always looking for founders that that are out raising seed.

We primarily focus on on seed rounds.

We love to be first money in Kevin at a star dot co Kevin.

Thank you so much.

It's been such a pleasure.

Thank you.

A lot of fun.