Latest / a16z AI Policy Brief / Inside Little Tech

Transcript

Inside Little Tech · a16z AI Policy Brief

0:00
SPEAKER_00: Welcome back to the AI Policy Brief.
0:02
SPEAKER_00: I'm Matt Peralt.
0:03
SPEAKER_00: Today I'm joined by Andrew Chen,
0:05
SPEAKER_00: General Partner at A16z Speedrun,
0:07
SPEAKER_00: our program for the earliest edge of little tech.
0:10
SPEAKER_00: Andrew works with founders who are turning an idea into a company,
0:13
SPEAKER_00: often before they've even incorporated,
0:14
SPEAKER_00: hired a team,
0:16
SPEAKER_00: or raised outside capital.
0:17
We talk about what life really looks like for two- and
0:19
three-person startups
0:20
SPEAKER_00: building from kitchen tables,
0:22
SPEAKER_00: why those founders are often missing from policy debates,
0:25
SPEAKER_00: and how regulation can shape how companies choose to build.
0:28
Andrew also shares his lessons from building Tech Week programs
0:31
and what makes
0:32
local startup ecosystems work and why supporting innovation is
0:35
ultimately a choice.
0:37
SPEAKER_00: Here's my conversation with Andrew Chen.
0:45
SPEAKER_00: Andrew, welcome to the a16z AI Policy Brief.
0:50
SPEAKER_00: Thank you for having me.
0:51
SPEAKER_00: So you lead our speedrun program.
0:53
SPEAKER_00: So can you tell us a little bit about what speedrun is?
0:55
SPEAKER_01: Yeah,
0:56
SPEAKER_01: so,
0:57
SPEAKER_01: you know,
0:57
many folks in the audience will know that startups have to come
1:01
from somewhere.
1:01
SPEAKER_01: And so,
1:02
SPEAKER_01: you know,
1:02
SPEAKER_01: our job is to try to find them on day one,
1:05
SPEAKER_01: when the founders are just starting to think about their companies.
1:08
SPEAKER_01: And so what's great is we, we
1:11
a program based in the San Francisco office where we will
1:16
announce to the broader
1:18
internet and through our marketing channels and podcasts and
1:21
Substack D's letters
1:22
and everything else that we'll be investing up to a million
1:25
dollars into brand new
1:26
SPEAKER_01: startups.
1:27
And the ones that we really focus on are the ones where folks
1:31
are just getting going.
1:33
SPEAKER_01: Folks that maybe have
1:34
Full Time Jobs and they're starting to come up with you know
1:38
with with with a new
1:39
SPEAKER_01: idea with their best friend and they want to go start something.
1:42
It could be folks that are have already left their you know
1:45
roles or they just
1:45
SPEAKER_01: graduated from school and they are are coming up with something
1:49
SPEAKER_01: A16z
2:16
with founders as kind of the beginning of a lifelong
2:19
relationship with them.
2:20
SPEAKER_01: That's that's the idea.
2:21
How do you decide what looks like a company that you want to
2:23
have as part of the
2:24
SPEAKER_00: speedrun program?
2:25
SPEAKER_01: At its core in the earliest days,
2:27
there are just so many famous examples of the ideas of ideas
2:33
changing all the time.
2:34
You know,
2:35
many of you guys know that a product we use every day,
2:38
Slack originally started out
2:40
SPEAKER_01: A16z
2:56
And so we're looking for folks that have had some really unique
3:00
experience in their past.
3:04
And that could be anything from enormous athletic
3:08
accomplishments on their way to a career.
3:11
in Tech or it could be that they started a GitHub repo that has
3:16
now thousands of
3:16
stars and it's growing incredibly fast or it could be some of
3:19
the traditional
3:19
markers where they have gone to elite universities and had sort
3:24
of learned elite
3:27
like incredible unique insights at some of our best AI
3:31
companies and then they're
3:32
SPEAKER_01: ready to use that in a new company as well.
3:35
I think we're very open minded and so we actually have mostly
3:39
folks out of the Bay Area
3:41
SPEAKER_01: A16z
3:57
SPEAKER_01: A16z
4:14
SPEAKER_00: So what is,
4:15
I'm kind of interested in the range of types of sizes of
4:19
companies that might
4:20
SPEAKER_00: participate in speedrun.
4:21
SPEAKER_00: It sounds like on the small side,
4:22
SPEAKER_00: it might be a couple of people who just decided to,
4:26
SPEAKER_00: who just have an idea,
4:27
SPEAKER_00: like really the smallest of the small,
4:29
SPEAKER_00: how large do the companies get?
4:31
SPEAKER_01: This is truly little tech.
4:33
SPEAKER_01: When we talk about little tech, this is it.
4:36
SPEAKER_01: I think the average team is two to three people.
4:41
They're running their companies not in their office,
4:45
not in a co-working space.
4:46
SPEAKER_01: They're running it at the kitchen table.
4:48
SPEAKER_01: It's really sort of the source of a lot of the
4:53
SPEAKER_01: Success that we've seen in the overall,
4:56
SPEAKER_01: you know,
4:57
SPEAKER_01: American industry has been,
4:58
SPEAKER_01: you know,
4:59
SPEAKER_01: really starts in these very small places.
5:01
And,
5:01
you know,
5:02
and then and then,
5:03
yeah,
5:03
we invest up to a million dollars.
5:04
SPEAKER_01: And then hopefully,
5:06
hopefully they raise more demo day and hopefully they keep
5:08
going and eventually
5:09
SPEAKER_01: they become incredible employers and,
5:14
SPEAKER_01: you know,
5:14
SPEAKER_01: pillars of the industry.
5:16
SPEAKER_01: And and we're we want to be with them from that journey from day one.
5:19
And what are organizations that would be too large to apply for
5:24
the speedrun?
5:25
SPEAKER_00: Like what's the upper bound?
5:26
SPEAKER_01: Yeah, the upper bound.
5:27
SPEAKER_01: I mean,
5:28
SPEAKER_01: honestly,
5:28
SPEAKER_01: for us,
5:29
SPEAKER_01: we would say as soon as you've taken capital elsewhere,
5:34
SPEAKER_01: you know,
5:35
SPEAKER_01: we would consider that maybe not disqualifying,
5:37
SPEAKER_01: but it would be sort of like,
5:39
SPEAKER_01: you know,
5:40
SPEAKER_01: maybe.
5:40
SPEAKER_01: So you're kind of at a point where the question is,
5:43
SPEAKER_01: well,
5:43
SPEAKER_01: how many folks can you get to bootstrap the company together?
5:46
SPEAKER_01: Yeah.
5:47
So if you have a group of five friends that all went to college
5:51
and they trust each
5:52
other a lot and they want to spend six months on their own dime
5:56
working on
5:56
SPEAKER_01: something,
5:57
SPEAKER_01: of course,
5:57
SPEAKER_01: we would love to back a team like that,
5:59
SPEAKER_01: especially one with that much trust.
6:01
SPEAKER_01: But very rare.
6:02
Clearly,
6:02
would we see a team above,
6:05
you know,
6:05
three,
6:05
four or five people at that point?
6:07
SPEAKER_01: That's at some point,
6:08
SPEAKER_01: you know,
6:08
SPEAKER_01: you kind of don't want to,
6:10
SPEAKER_01: you don't want to divide the company's ownership to be too small,
6:13
SPEAKER_01: especially because people come and go.
6:15
SPEAKER_01: You probably need to,
6:16
SPEAKER_01: you know,
6:17
some people's life situations will be such that you'll need to
6:19
actually pay them.
6:20
And so yeah,
6:21
so we again,
6:23
really,
6:24
it's like a two or three person thing,
6:26
typically.
6:27
SPEAKER_00: So we invest at different phases of company life cycles.
6:31
SPEAKER_00: We have a growth team that's when companies tend to get larger.
6:33
SPEAKER_00: We have a venture team when companies are smaller.
6:35
Speedrun,
6:36
as you're describing it,
6:36
is like the littlest of the little tech.
6:38
Why did we as a firm and why did you decide that this was an
6:41
important part of a
6:42
SPEAKER_00: life cycle to support?
6:43
Yeah,
6:44
I think that the way I see the mission for A16Z Speedrun is
6:51
that we're in the
6:53
SPEAKER_01: You know,
6:54
we want to get as close as possible to the moment that founders
6:57
are actually
6:58
SPEAKER_01: created and new companies are created.
7:00
You know,
7:02
my first couple of years at the firm,
7:03
I spent more in series A,
7:05
series B.
7:06
SPEAKER_01: And so you're writing checks there where these are,
7:08
SPEAKER_01: you know,
7:09
SPEAKER_01: mature,
7:10
SPEAKER_01: mature businesses,
7:10
SPEAKER_01: or there's at least a glimmer of a business there.
7:13
SPEAKER_01: And so you're investing
7:15
SPEAKER_01: 10, 15, 20 million dollars.
7:16
SPEAKER_01: And so by that point,
7:18
SPEAKER_01: yeah,
7:18
SPEAKER_01: you have customers,
7:19
SPEAKER_01: you have metrics to look at,
7:23
SPEAKER_01: you have a team,
7:24
SPEAKER_01: etc.
7:25
SPEAKER_01: And that's a wonderful place to invest.
7:27
SPEAKER_01: And historically, it's been maybe the best place to invest.
7:30
SPEAKER_01: For us,
7:31
SPEAKER_01: one of the really magical things,
7:32
SPEAKER_01: something that I take a lot of just personal pride and excitement in,
7:36
is the idea that we are not just waiting for entrepreneurs to
7:40
show up at our
7:41
SPEAKER_01: doorstep,
7:43
but we're actually helping create the companies in the first
7:45
place so for example
7:46
in the last speedrun bash there were 70 companies and of the 70
7:50
there was probably
7:51
a dozen where you know these were folks that had full-time jobs
7:54
and they had to
7:55
turn in the laptops and their badges and we couldn't even wire
7:58
the money over
7:58
because you know they they didn't have a company to even
8:01
incorporate and so we sort
8:03
of you know helped incept these the idea of like hey like let's
8:08
let's create the
8:09
SPEAKER_01: most
8:10
SPEAKER_01: Powerful and best path to starting the company.
8:16
And I think a lot of those founders would have eventually
8:19
started businesses.
8:20
SPEAKER_01: When you're a founder, sometimes you just can't get that bug.
8:23
SPEAKER_00: You know, out of your ear.
8:25
But I would like to think that we maybe,
8:29
you know,
8:29
pulled some of that activity forward.
8:31
And so maybe folks that maybe would have waited another for
8:34
another five or 10
8:35
SPEAKER_01: years of work experience and financial stability,
8:38
SPEAKER_01: you know,
8:38
SPEAKER_01: maybe they would have taken that jump.
8:40
SPEAKER_01: And what an amazing time in the AI era,
8:43
SPEAKER_01: you know,
8:43
SPEAKER_01: to encourage them to,
8:45
SPEAKER_01: you know,
8:45
SPEAKER_01: jump in at the right moment.
8:48
Can you give a little texture to what a day in the life looks
8:50
like for a speedrun
8:51
SPEAKER_00: company,
8:52
SPEAKER_00: like when they're part of the program?
8:55
SPEAKER_00: AI Policy Brief a16z
9:19
SPEAKER_01: You know, pick your pick your city.
9:22
SPEAKER_01: We also have had teams teams from, you know, from Utah.
9:26
SPEAKER_01: We've had teams from from from all of all over the US,
9:29
SPEAKER_01: quite a few from New York,
9:30
SPEAKER_01: quite a few from from Pacific Northwest as well.
9:34
So kind of imagine you have you have two people that get
9:36
accepted to this program.
9:38
SPEAKER_01: They're very excited.
9:38
SPEAKER_01: They they they have to incorporate the company.
9:41
SPEAKER_01: AI Policy Brief a16z
9:56
SPEAKER_01: a16z
10:19
you know waking up they're often you know for the duration of
10:22
program they're
10:23
roommates with their with their co-founder so you're you're you
10:26
are spending 24 7
10:27
together like truly you know you're not you're not living
10:30
somewhere fancy you're
10:31
SPEAKER_01: you're you're somewhere you know
10:34
SPEAKER_01: typically close to the A16z office.
10:37
A lot of these folks are either working together at their homes
10:42
and we also offer
10:43
SPEAKER_01: access to the A16z offices.
10:44
SPEAKER_01: So sometimes they'll end up walking in or taking an Uber in,
10:51
SPEAKER_01: taking a Waymo into the office and then they end up in,
10:55
SPEAKER_01: we have a couple of co-working spaces where you'll spend time.
10:58
And the only thing that they should be spending time on really
11:01
is just to figure out
11:03
is this business going to work like how do we they got to build
11:06
the product and
11:07
they got to sell the product and they got to do it in a short
11:11
period of time and so
11:12
typically you'll have maybe two co-founders one will be more of
11:15
a business
11:16
co-founder one will be more of a technology and product
11:19
co-founder the technology
11:20
and product co-founder is typically using a lot of AI right now
11:24
a lot of AI coding
11:26
SPEAKER_01: so rather than going out and needing to
11:30
Ad a lot of cost and and you know maybe outsourcing you know
11:33
parts of your your
11:34
coding or hiring a bunch of you know young people to to to do
11:38
coding you can just
11:39
focus on just you know hey how effective can one or two people
11:42
be you know right
11:43
now and like let's just make this business stable and survive
11:45
let's then let's go
11:46
hire a bunch of folks then the business co-founders often out
11:49
there interviewing
11:50
SPEAKER_01: customers talking to customers trying to make deals happen
11:54
And so there's often this dichotomy that we see of kind of like
11:57
the outwards facing
11:58
business co-founder and then the inwards facing product and
12:01
technology founder as
12:02
SPEAKER_01: well.
12:02
What we do with Speedrun is once per week,
12:05
we end up having some formal programming.
12:08
SPEAKER_01: We will typically have a speaker.
12:09
SPEAKER_01: We've had some amazing founders, the founder of Zynga or Vercel.
12:27
SPEAKER_01: C-level folks from OpenAI.
12:29
We've had,
12:32
of course,
12:33
Mark Andreessen and Ben Horowitz with us many times as well.
12:36
SPEAKER_01: That's always fantastic.
12:38
SPEAKER_01: And then we do these office hours where they will sit in a group of 20,
12:42
30 companies and they will talk to each other about problems
12:48
they're trying to
12:48
SPEAKER_01: solve.
12:49
SPEAKER_01: So we run these office hours.
12:50
SPEAKER_01: We talk about everything from hiring your first
12:53
you know your first employees to signing your first customers
12:57
how do you launch on
12:57
social media how do you what's the difference between selling
13:01
into a large
13:02
enterprise versus selling into SMB and all the things that you
13:04
do and of course how
13:06
is everyone using AI you know what are all the new tools that
13:09
everybody's using
13:11
SPEAKER_01: We spend a lot of time on that as well.
13:13
And then hopefully the rest of the week they're out and about
13:17
working and they're
13:17
SPEAKER_01: being immersed in kind of the milieu of the SF tech ecosystem.
13:23
So they're meeting a ton of people,
13:24
meeting a ton of investors,
13:26
learning all of that.
13:27
And so yeah,
13:29
we're very excited about just everything that can happen in
13:32
just a couple of weeks.
13:34
SPEAKER_00: So I know like obviously
13:36
the most exciting thing to talk about is success the success
13:39
part of the ecosystem
13:40
but in venture like there's a lot of not success right it's
13:42
like sort of the nature
13:43
of the business and I assume the majority of the business and
13:47
the earlier you go in
13:48
the life cycle the more true that is more companies don't
13:51
succeed so what does that
13:52
look like for speedrun participants like when companies come
13:56
through the program
13:57
and then they're not able to launch a successful business at
13:59
least the one that we
14:01
SPEAKER_00: funded as part of the program what does that pathway look like
14:04
SPEAKER_01: Yeah, that's right.
14:05
SPEAKER_01: Yeah,
14:05
SPEAKER_01: I think historically in the venture capital business overall,
14:10
SPEAKER_01: you often,
14:11
SPEAKER_01: you know,
14:11
the colloquial description is something like half the companies
14:17
don't work out at
14:18
SPEAKER_01: all,
14:18
SPEAKER_01: right?
14:18
SPEAKER_01: They just kind of outright, you know, fold.
14:22
Then you'll have another,
14:23
you know,
14:24
if that's like five and 10 companies,
14:26
you'll have another
14:28
2 or 3 where you make a little money and then all the money is
14:32
made in that kind of
14:34
top decile the 1 out of 10 where you get the grand slam and
14:39
that just seems to be
14:40
SPEAKER_01: true mathematically which is just
14:43
SPEAKER_01: Amazing.
14:43
SPEAKER_01: It's been very consistent over many decades.
14:45
So,
14:46
you know,
14:46
to your question that on on,
14:48
you know,
14:48
what happens to the to the ones that fold?
14:50
SPEAKER_01: Well, a bunch of great things happen, right?
14:52
SPEAKER_01: I mean,
14:53
SPEAKER_01: first,
14:53
SPEAKER_01: like we,
14:55
SPEAKER_01: we think that even though we as a firm make the money on the,
15:00
SPEAKER_01: you know,
15:00
SPEAKER_01: on on on the power law kind of exceptional companies that break through,
15:05
SPEAKER_01: you know,
15:06
SPEAKER_01: we make our reputations with how we work with all the founders that are
15:11
um you know working hard working a hundred hours a week they're
15:15
um you know they're
15:17
they're they're completely plugged in they're working weekends
15:20
like the whole thing
15:21
and and it just doesn't work out you know and sometimes it's to
15:23
no fault of their
15:24
own sometimes there's just a market downturn you know they're
15:27
starting a pre-ai
15:28
SPEAKER_01: company
15:29
SPEAKER_01: when everybody's going AI.
15:31
SPEAKER_01: It's because their co-founder got sick.
15:35
It's because of all these many,
15:38
many different potential things that could happen.
15:40
SPEAKER_01: And so as a result,
15:42
SPEAKER_01: the best thing we can do for those companies is number one,
15:45
SPEAKER_01: should we back the founders again?
15:47
That's one I think very special thing about that Silicon Valley
15:51
mentality is like,
15:52
okay,
15:52
great.
15:53
SPEAKER_01: You've just spent
15:54
SPEAKER_01: a million bucks educating these founders about how to start a company.
15:59
SPEAKER_01: Hopefully, they can start another company.
16:02
SPEAKER_01: And by the way,
16:03
whether you're talking about Uber or Slack or Microsoft or
16:09
OpenAI or any of these
16:11
SPEAKER_01: companies,
16:11
SPEAKER_01: when you look at the CEO and the founders involved,
16:16
SPEAKER_01: very frequently,
16:17
SPEAKER_01: more often than not,
16:18
SPEAKER_01: they have been involved
16:19
SPEAKER_01: A16z
16:44
SPEAKER_01: And what that means is they went at it completely, just 150%.
16:49
And what they need is they need a little bit of time to rebuild
16:54
their finances.
16:55
SPEAKER_01: They probably weren't paying themselves that well.
16:58
SPEAKER_01: They probably need some time to focus on their health.
17:02
We find that a lot of these founders become very hot
17:05
commodities because that means
17:07
SPEAKER_01: that you can get somebody that works hard
17:10
SPEAKER_01: AI,
17:10
SPEAKER_01: works on unstructured problems,
17:12
SPEAKER_01: is completely current on all of the newest technology,
17:16
SPEAKER_01: and you can get them at your company.
17:18
And so what we've been seeing is a lot of teams will end up
17:22
hiring each other.
17:24
SPEAKER_01: If one doesn't work out, you respect the founder, great, you hire them.
17:28
SPEAKER_01: And then maybe they spend two or three years with you.
17:31
SPEAKER_01: Maybe they lead some new initiatives and new products in your business.
17:34
SPEAKER_01: And then a few years later,
17:36
they spin out and they're ready to start a new company and then
17:38
hopefully we as
17:39
a16z get another shot at them to work together so yeah that's
17:44
sort of all by design
17:46
and I think it's perfectly expected as outcomes and we just
17:50
want to partner with
17:50
SPEAKER_01: these founders for their whole professional career
17:53
so I'd love to turn to policy now since we're this policy
17:58
Substack this is the
17:59
policy yeah it's policy brief let's do it and so I'd love to
18:02
like understand more
18:04
about how the companies you work with think about policy issues
18:08
this is something
18:08
that we we obviously do to some extent but you're like living
18:12
and breathing the
18:14
founder side of it on a daily basis and we've tried to kind of
18:17
channel the concept
18:18
in different ways like we recently wrote this piece called
18:20
Greetings from a Garage
18:21
SPEAKER_00: about
18:23
what it looks like from a policy perspective if a couple of
18:25
people go into a garage
18:26
to build the next great tech tool in California what is the
18:31
regulatory landscape
18:32
that they have to deal with and it was a really helpful
18:35
exercise because often
18:38
policy teams legal teams will think about just incremental
18:41
burden like if we pass
18:43
there's a new law that's been proposed what happens if we pass
18:45
this new law how in
18:47
SPEAKER_00: the risk or not is it going to be how helpful is it going to be
18:50
SPEAKER_00: And I think the insight that the head of our state team,
18:53
SPEAKER_00: Kevin,
18:53
SPEAKER_00: had when we were thinking about this piece is like,
18:56
SPEAKER_00: it's not just the incremental burden.
18:59
It's all the things that a founder has to face from the moment
19:02
they're building
19:03
SPEAKER_00: with the technology.
19:04
So like in California,
19:05
that doesn't just mean SB53,
19:07
which was passed last year.
19:08
SPEAKER_00: It means all these data provenance requirements from the session before.
19:12
SPEAKER_00: It means privacy law in California that was passed several years before.
19:16
SPEAKER_00: It's all accumulative.
19:17
And the moment that you start building the tool,
19:19
you have to comply with all these things.
19:21
SPEAKER_00: And so really the policy picture is the composite of all those things.
19:25
SPEAKER_00: It's not an individual one of them.
19:27
SPEAKER_00: So I'm curious what you,
19:29
SPEAKER_00: like,
19:29
SPEAKER_00: you know,
19:29
SPEAKER_00: when you're at the stage,
19:30
SPEAKER_00: like you're literally at the,
19:31
SPEAKER_00: like,
19:32
SPEAKER_00: we just met in the coffee shop where,
19:33
SPEAKER_00: you know,
19:33
SPEAKER_00: we're in the garage,
19:34
SPEAKER_00: we're roommates,
19:36
SPEAKER_00: we're,
19:37
SPEAKER_00: you know,
19:37
SPEAKER_00: we're in the shared Waymo on the way to the A16z office.
19:41
At that phase,
19:43
how do the people that you work with think about policy issues?
19:46
For these founders,
19:48
they are so mission-focused to just make their company,
19:54
make their product.
19:54
SPEAKER_01: I mean, it's not even a company at that point.
19:57
SPEAKER_01: They might even think of it as a project or as a product.
19:59
There's no company because it's just the two guys or gals
20:03
trying to make things work.
20:05
SPEAKER_01: And so for them from a policy perspective,
20:07
SPEAKER_01: look,
20:07
I think every obstacle that is put in front of them along the
20:13
way for them to just
20:15
SPEAKER_01: prove that their initial product can even work just adds friction,
20:19
SPEAKER_01: you know,
20:19
SPEAKER_01: for the most part.
20:20
SPEAKER_01: So I think most of the founders,
20:22
SPEAKER_01: when you talk to them,
20:23
SPEAKER_01: they really see all that stuff as like,
20:25
SPEAKER_01: oh,
20:25
SPEAKER_01: wow,
20:26
SPEAKER_01: I've incorporated,
20:27
SPEAKER_01: I'm doing this thing,
20:27
SPEAKER_01: I'm hiring these people.
20:28
SPEAKER_01: Okay, you get the stack of paperwork this big.
20:32
and you know a lot of these laws a lot of the paperwork is
20:35
designed for companies
20:36
that are much much larger than them right and have the ability
20:39
to comply because
20:40
they have teams of lawyers and teams of you know people experts
20:44
you know that that
20:45
are that are in there and of course we we as a16z try to try to
20:48
help them um but
20:49
for for for for most of these folks all they're trying to do is
20:53
they're just trying
20:54
SPEAKER_01: to to to get going at all
20:56
SPEAKER_01: and to survive as a business.
20:59
And so generally their interaction on kind of the regulatory
21:06
side of the world is
21:08
usually negative because these are things that actually create
21:12
protections often
21:13
for either big tech or for other parts of the industry that
21:16
they end up trying to
21:20
SPEAKER_01: disrupt.
21:21
SPEAKER_01: So I'd say that's kind of one major part.
21:22
SPEAKER_01: The other part is
21:23
is that for many of these founders,
21:25
they just don't have time to participate.
21:29
SPEAKER_01: These guys don't have lobbyists.
21:35
SPEAKER_01: They're not involved in the political process.
21:38
They're not really represented in all these ways because,
21:42
frankly,
21:42
they just don't have time.
21:44
They may not even have time to shower or to have a nice dinner
21:50
for themselves.
21:52
SPEAKER_01: They're working that hard.
21:54
much less to do things that are kind of sit outside their kind
21:58
of just day-to-day
21:59
SPEAKER_01: of just survival at that point.
22:01
SPEAKER_00: And that's a compounding thing.
22:02
SPEAKER_00: I mean,
22:03
I think that's a dynamic that we encounter all the time because
22:06
we show up,
22:07
SPEAKER_00: we represent little tech,
22:08
SPEAKER_00: but we're not,
22:09
SPEAKER_00: or I'm not,
22:10
SPEAKER_00: a founder.
22:11
And so when we show up,
22:14
people are like,
22:14
well,
22:14
we'd rather talk to the startup.
22:16
SPEAKER_00: We don't want to talk to the investor in the startup.
22:19
SPEAKER_00: But the challenge is, as you're saying, the startups are
22:23
not showering working 100 hour weeks like doing their they're
22:27
not going to go and
22:27
like make the drive to Sacramento or the flight to Washington
22:30
or flight to Brussels
22:32
and represent what it's like for them to try to build under the
22:37
in the current
22:38
regulatory environment that's right yeah well and and and their
22:42
timeline a startup
22:43
a two-person startups timeline not only is it not a year like
22:47
you may not know if
22:48
SPEAKER_01: you have a company in a year
22:51
SPEAKER_01: It may not even be six months.
22:53
SPEAKER_01: It may not even be three months.
22:54
SPEAKER_01: You might be working on really small time schedules.
22:59
And so to make the investment to fly to one of these places and
23:02
to be involved,
23:05
SPEAKER_01: I do think it's really hard to carve out time and do it,
23:08
which is why I think both it's really important to have some
23:12
mechanism to represent
23:13
SPEAKER_01: these people because these are the future job creators,
23:17
these are the companies that are going to be employing folks
23:19
and continuing
23:20
SPEAKER_01: Continuing America's success on the technology field.
23:24
SPEAKER_01: But also on top of that,
23:27
the folks that maybe do show up in all these places and do
23:34
advocate for tech,
23:36
SPEAKER_01: you get a very skewed view because you don't hear from little tech,
23:41
SPEAKER_01: but you might hear a lot from big tech.
23:43
You might hear a lot from the industries that are getting
23:47
disrupted and very little
23:50
from the 30-year-olds that are spending night and day trying to
23:55
make that happen.
23:56
So as long as we actually get hostile reactions from
23:58
policymakers when we raise
24:00
SPEAKER_00: that point,
24:01
SPEAKER_00: they'll say,
24:02
Well all of industry agrees on this and when they say all of
24:04
industry what they
24:05
means is big tech companies or sometimes it might be big
24:08
financial services
24:09
companies like large organizations and they say you have a
24:12
different view and we
24:13
need you to get on board and we'll say in the policy
24:15
conversation like it's sort of
24:17
by design at least some percent of the time like there's lots
24:20
of time where we
24:21
agree with big tech companies but we'll like when we disagree
24:24
we'll say like it's
24:24
by design we're representing a different part of the ecosystem
24:27
so of course we
24:27
SPEAKER_01: don't agree on everything that's right that's right absolutely
24:31
So one of the things about the time horizon that you're talking
24:33
about is often
24:34
there's a gap between the lobbying phase and the effective date
24:37
of a piece of
24:37
SPEAKER_00: legislation.
24:39
SPEAKER_00: So I was just looking this up.
24:41
SPEAKER_00: Colorado,
24:42
SPEAKER_00: which has the most comprehensive AI law at the state level,
24:46
was actually passed in 2024 and is set to go into effect in
24:49
June of this year.
24:51
SPEAKER_00: That's a little bit unusual.
24:52
SPEAKER_00: It's been delayed a couple of times for various different things.
24:54
But I can imagine if you're a startup in Colorado trying to
24:57
make the calculus about
24:58
SPEAKER_00: whether you engage in 2024,
25:01
The law is going to go into effect a couple years down the road
25:04
and to your point
25:05
you're not even operating in one year time horizons let alone
25:08
multiple year time
25:09
horizons the likelihood that you're going to prioritize
25:12
negotiate like the time
25:14
SPEAKER_00: consuming process of negotiation and making your voice heard is low
25:17
SPEAKER_01: Absolutely, absolutely.
25:19
SPEAKER_00: Just from an ROI, time ROI standpoint, it just doesn't make any sense.
25:23
I think there's also a really unique thing because these teams
25:27
are so small and
25:28
SPEAKER_01: they're kind of at their inception,
25:30
SPEAKER_01: which is they get to pick where they start the company from.
25:33
SPEAKER_01: I moved to the Bay Area in 2007.
25:39
SPEAKER_01: And in 2007,
25:41
the world revolved around Palo Alto and Mountain View and kind
25:46
of the peninsula,
25:47
SPEAKER_01: what we think of as Silicon Valley.
25:49
SPEAKER_01: But folks,
25:51
SPEAKER_01: I think,
25:51
are familiar with this now that we saw a whole-scale migration
25:57
of the center of the
25:58
SPEAKER_01: startup ecosystem being from
26:00
SPEAKER_01: The Peninsula to San Francisco proper.
26:04
SPEAKER_01: And then,
26:04
SPEAKER_01: subsequently,
26:05
we've seen it now spread where New York is now one of the huge
26:11
startup hubs that's
26:12
SPEAKER_01: emerged.
26:14
SPEAKER_01: And all the
26:15
SPEAKER_01: over the last couple of years,
26:16
all the major venture capital firms had opened offices in
26:22
various places in London
26:24
SPEAKER_01: and Europe.
26:25
SPEAKER_01: There's been other major hubs that have created.
26:29
And so I think there is also an interesting thing where that
26:31
almost becomes part of
26:33
SPEAKER_01: one of the most important choices that you can make as a startup
26:36
which is where are you going to plant roots where are you going
26:38
to hire people from
26:39
where are you going to do that and if you end up working on an
26:43
AI company and you
26:44
know from day one that there's going to be a bunch of extra
26:48
rules about how you can
26:49
SPEAKER_01: use AI you know
26:52
SPEAKER_01: The two guys, two gals startup, they're very mobile.
26:59
SPEAKER_01: They can pick and choose where they want to go, and they do.
27:03
SPEAKER_01: And historically, the Bay Area has benefited from that mobility.
27:07
I think there's some interesting stats about how nearly 50% of
27:13
venture-backed
27:13
SPEAKER_01: startups are first-generation immigrants,
27:15
SPEAKER_01: for example,
27:16
SPEAKER_01: people who relocate in order to do it.
27:19
SPEAKER_01: A16z
27:36
It's not certain that California will always be the best place
27:41
to start companies.
27:42
And so I think it's important for us to preserve this ecosystem
27:45
because it is very,
27:47
SPEAKER_01: very special.
27:47
SPEAKER_01: It is something that is unique,
27:49
SPEAKER_01: not just in the world,
27:50
but even within the US in order to have this fertile of an
27:56
ecosystem for startups
27:57
SPEAKER_01: and new technology.
27:58
Are you hearing founders talk about regulatory issues as one
28:02
factor that they
28:03
SPEAKER_00: consider and where they end up citing the business?
28:05
SPEAKER_01: Yeah, I think that it is sometimes indirect.
28:10
SPEAKER_01: I think, for example, just cost of living is a really simple one.
28:17
SPEAKER_01: Or if you're building a deep tech company,
28:18
SPEAKER_01: if you're building the next generation of manufacturing,
28:21
you have to go somewhere where people are building that there's
28:25
large-scale
28:27
SPEAKER_01: Lab and Warehouse Space at reasonable prices.
28:31
SPEAKER_01: And so I think you're seeing these pockets like El Segundo in LA.
28:35
SPEAKER_01: You're seeing folks building things in Texas for many of those reasons.
28:42
SPEAKER_01: I think the Bay Area has hugely benefited from
28:47
SPEAKER_01: Just the AI wave, you know, really centered here.
28:50
So I think that's really extended the life of this ecosystem in
28:53
a big way.
28:53
A lot of people have been coming back to the Bay Area since the
28:56
sort of COVID dispersal.
28:59
SPEAKER_01: And then I do think that...
29:02
You know the ability for again another indirect kind of you
29:05
know force is that the
29:07
startups have to really move where there's investors it turns
29:10
out it's a lot easier
29:11
to code and it's a lot easier to get your first dollar revenue
29:14
and it's a lot
29:14
SPEAKER_01: easier to hire your first employee you can kind of do that anywhere
29:18
But actually getting capital to really scale your business
29:21
requires the ability for
29:23
SPEAKER_01: investors to be there.
29:24
SPEAKER_01: And so I think that's why when you look around the world,
29:30
SPEAKER_01: a lot of the biggest startup hubs happen to be kind of where you have
29:33
SPEAKER_01: Great Universities,
29:35
SPEAKER_01: co-located with customers,
29:36
SPEAKER_01: co-located with investors,
29:38
SPEAKER_00: kind of all in one spot.
29:40
And I wonder and I worry about things like the wealth tax that
29:46
are currently under
29:46
SPEAKER_01: discussion in its advantage.
29:48
SPEAKER_01: AI Policy Brief a16z
30:07
So we've been talking a lot about the one-to-one relationships
30:11
that you've
30:12
SPEAKER_00: developed as part of the Speedrun program,
30:13
but you've also been really invested in trying to create
30:16
ecosystems through the
30:17
SPEAKER_00: Tech Week program.
30:19
SPEAKER_00: What was the idea behind that program?
30:20
So we started Tech Week and it's a program that sits in our
30:25
team here within the
30:27
SPEAKER_01: Speedrun umbrella.
30:29
And what Tech Week really does is it's really a celebration of
30:35
startups and
30:35
SPEAKER_01: and technology and progress in all the major cities.
30:39
And so we hosted in San Francisco,
30:41
we hosted in LA,
30:42
we hosted in New York.
30:43
And then for the first time this year,
30:44
we're actually hosting it in Boston as well.
30:47
We have over 100,000 people going to these events through the
30:56
year,
30:56
which is just staggering.
30:58
SPEAKER_01: It sounds extraordinary.
30:59
SPEAKER_00: Yeah, it sounds like a fake number.
31:00
SPEAKER_00: 100,000 people.
31:00
SPEAKER_00: Yeah, 100,000 attendees.
31:01
SPEAKER_00: Yeah.
31:03
SPEAKER_00: Yeah, wow.
31:04
SPEAKER_01: That's right.
31:04
SPEAKER_01: That's right.
31:05
SPEAKER_01: And, you know, almost almost 20,000 companies.
31:09
SPEAKER_01: There's, you know, 3000 plus events.
31:12
The way that we do it and the way the way that it's been very
31:16
powerful is we actually as a firm
31:18
SPEAKER_01: will kind of just designate a particular week as tech week.
31:23
And then what we do is then we start talking to a bunch of our
31:28
partners and whether
31:31
SPEAKER_01: that's other investors,
31:32
SPEAKER_01: whether that's larger companies,
31:34
SPEAKER_01: whether that's portfolio companies,
31:35
SPEAKER_01: and we ask them,
31:36
SPEAKER_01: hey,
31:36
SPEAKER_01: would you like to
31:37
you know do events during that week in a way where then we can
31:40
we can all you know
31:41
we can all do stuff together and we'll host a couple big things
31:45
and then we end up
31:46
also we end up also you know just partnering and people just
31:51
spring up and start
31:52
doing things so this year's Boston date is going to be May 26
31:57
to 31 and then New
31:58
SPEAKER_01: York is the next week after so June 1st to 7th
32:01
And then we just bring together just a massive number of events
32:04
and people.
32:06
SPEAKER_01: And,
32:06
SPEAKER_01: you know,
32:07
SPEAKER_01: we've partnered with a bunch of folks in the ecosystem,
32:10
SPEAKER_01: whether that's,
32:11
SPEAKER_01: you know,
32:12
SPEAKER_01: Fenwick and HSBC and IBM and a lot of the big companies out there,
32:16
SPEAKER_01: whether it's Adobe and AWS and Deal.
32:19
Google and many others to sort of anchor the overall program in
32:25
a way where then
32:26
SPEAKER_01: we're able to create a very wide breadth of different types of events.
32:33
SPEAKER_01: So we have things like hackathons.
32:35
SPEAKER_01: We have things that are just like mixers.
32:36
SPEAKER_01: We have speaking events.
32:38
SPEAKER_01: People go out and do a bunch of fun stuff.
32:39
SPEAKER_01: Like in LA, there was a surfing event
32:43
you know events or folks doing yoga or you know so it's sort of
32:47
part social part
32:48
serious and then you know these days what's been amazing and
32:51
kind of the fit with
32:52
speedrun has been it's sort of become our way to discover brand
32:57
new founders before
32:58
maybe they even know that they're founders and and it's also a
33:01
celebration of tech
33:02
kind of within all these communities and so so what what does
33:05
that mean how do you
33:06
SPEAKER_00: discover a founder before they know if they're a founder
33:09
SPEAKER_01: Well,
33:10
we just feel like if you're a type of person with certain kinds
33:15
of work experience,
33:16
SPEAKER_01: certain resume,
33:17
SPEAKER_01: and you're finding yourself attending a bunch of startup events,
33:21
SPEAKER_01: you know,
33:22
SPEAKER_01: like the probability of you starting something,
33:25
SPEAKER_01: like you're probably startup curious,
33:28
but you're probably going to get there like sometime in the
33:31
next few years.
33:32
SPEAKER_01: So that's what we found, which has been fantastic.
33:36
SPEAKER_00: Is policy showing up at Tech Weeks?
33:40
Are you seeing it in the various different local ecosystems
33:42
that you're tapping into?
33:44
SPEAKER_01: Yeah, so we would love to do more.
33:46
It's definitely exciting to see more and more of an ecosystem
33:53
around policy.
33:54
SPEAKER_01: So for example,
33:55
SPEAKER_01: I think across all of the tech weeks last year,
33:58
SPEAKER_01: we had 4000 attendees attending various policy related events.
34:03
And that includes over 1000 folks from government,
34:06
over 1000 policy professionals.
34:08
SPEAKER_01: And then oftentimes we've had events where
34:11
SPEAKER_01: You know, local government will come together.
34:13
SPEAKER_01: So I know,
34:14
SPEAKER_01: you know,
34:15
SPEAKER_01: whether that's,
34:16
SPEAKER_01: you know,
34:16
SPEAKER_01: in the mayor's office,
34:19
SPEAKER_01: or,
34:19
SPEAKER_01: you know,
34:20
SPEAKER_01: folks that work on the national sphere,
34:22
SPEAKER_01: we also have had a bunch of the consulates,
34:25
SPEAKER_01: you know,
34:25
SPEAKER_01: involved as
34:26
as well because they're all very interested in building that
34:29
bridge between their
34:29
SPEAKER_01: countries and particularly the San Francisco system.
34:35
So the British consulate's been super involved on tons of stuff
34:40
and then many other
34:42
SPEAKER_01: countries as well.
34:42
So yeah,
34:44
it's been great to have that as a burgeoning part of the Tech
34:49
Week ecosystem.
34:50
So if you see a local policymaker or a representative from a
34:53
consulate or member of
34:55
SPEAKER_00: Congress at a Tech Week event,
34:57
SPEAKER_00: and you're chatting with them on the sidelines,
34:59
what would you relay to them about the realities of building a
35:03
competitive company,
35:04
SPEAKER_00: given what you know about what the early,
35:06
SPEAKER_00: early stage really looks like?
35:08
SPEAKER_01: Yeah, yeah.
35:09
SPEAKER_01: Well, I think I'd say a couple things.
35:13
I think the first is to just have them understand that it's a
35:20
choice whether or not
35:22
SPEAKER_01: each state or each city wants to have startups or not.
35:27
If you want to have startups,
35:28
you have to do things to make it conducive to having startups.
35:32
SPEAKER_01: And there's a long list of things that might be involved in there.
35:38
SPEAKER_01: And it's not just for the founders.
35:39
SPEAKER_01: It's for the angel investors.
35:42
SPEAKER_01: It's for the employees.
35:43
SPEAKER_01: It's for all these things that surround that whole group.
35:47
The second thing is,
35:49
as we were just talking about a couple of minutes ago...
35:52
I think a lot of policymakers are not having real conversations
35:56
with little tech.
35:58
You know it's it takes a little if there's a little bit more
36:01
friction you have to
36:01
kind of put in a little bit more time into it but again you
36:04
know if you if we want
36:05
to have if we decide you know hey the we want to have more
36:09
startups in America we
36:10
want more innovation America or we or or we're we're talking
36:13
about that kind of at
36:14
the city level you know we we have to figure out like what are
36:18
the things that we
36:19
can unblock for for a lot of these companies and those things
36:22
change you know
36:22
SPEAKER_01: software company will tell you something different
36:24
SPEAKER_01: than a AI native company,
36:26
which will tell you something different than a hardware company
36:29
and a robotics
36:29
SPEAKER_01: company.
36:30
You know,
36:30
it's very dynamic,
36:31
you know,
36:31
one year,
36:31
you know,
36:32
for example,
36:33
this year we are seeing
36:36
way more robotics companies than in any year past and they have
36:40
very different
36:40
needs uh compared to you know they have supply chain needs they
36:43
have you know real
36:44
estate needs they have a different set of investors they have
36:47
potentially different
36:47
set of partners um and and and so you know we have to we have
36:51
to actually you know
36:52
SPEAKER_01: dig dig into that in in in a real way and then
36:55
and maybe you know thirdly to just maybe talk our book I think
36:59
I think working with
37:01
folks like a16z and other venture capital firms and other
37:07
ecosystem kind of players
37:09
you know what we can do is we can certainly help open the door
37:12
to to creating those
37:16
conversations and so you know like open offer to the policy
37:20
folks that are that are
37:21
SPEAKER_01: other you know listening to this anybody that wants to come and
37:24
you know spend time with five or ten really curated you know
37:30
little tech startups
37:31
that are you know two three people and just kind of hear like
37:33
okay what do you need
37:34
to be successful like we're very happy to arrange that also a
37:38
lot of these
37:38
companies are very interested to you know like certain sectors
37:41
are particularly
37:42
interested in partnering you know with government so I think
37:46
that's all great
37:46
I'm a policy person who's interested in doing that I'm actually
37:49
coming out to talk
37:50
to a speedrun talk to a speedrun cohort next week so I'm
37:53
looking forward to it it's
37:54
been fun doing that in the past because like you said you
37:57
really see in those
37:57
conversations like when you talk through various different
38:00
policy issues sometimes
38:01
SPEAKER_00: they're not
38:02
I'm building into a use case that might face regulatory
38:03
headwinds but I think the
38:04
use case is really compelling and I don't think policymakers
38:06
understand what we're
38:06
trying to do and so it's like interesting to see that delta and
38:09
we've engaged with
38:09
the speedrun team in the past to make sure policymakers also
38:11
appreciate that delta
38:32
SPEAKER_01: Yes, I think that's right.
38:33
SPEAKER_01: And I think it's,
38:34
SPEAKER_01: um,
38:35
SPEAKER_01: you know,
38:35
SPEAKER_01: it's such,
38:36
SPEAKER_01: it's,
38:36
it is a wonderful thing to spend time with these teams because
38:39
they're often,
38:40
SPEAKER_01: you know,
38:40
SPEAKER_01: very,
38:41
SPEAKER_01: very much young people.
38:42
SPEAKER_01: You know, we, we have a lot of folks that are in the early twenties.
38:45
SPEAKER_01: I think we even have a team that's all 18 and 19 year olds,
38:48
SPEAKER_01: for example,
38:48
SPEAKER_01: and like in the current batch,
38:49
SPEAKER_01: you decided to
38:50
SPEAKER_01: a16z
39:20
Andrew,
39:21
thanks so much for joining the AI Policy Brief Awesome,
39:23
thank you for having me