Transcript: Marc Andreessen on Change, Constraints, and Curiosity

gistfile1.txt

Marc Andreessen on Change, Constraints, and Curiosity
@ Stanford Graduate School of Business

VIDEO: https://www.youtube.com/watch?v=P-T2VAcHRoE
RECORDED: Tuesday, November 8, 2016

DESC: If you want to be successful as a venture capitalist, you need to be ruthlessly open-minded, constantly re-examining your assumptions, shared Andreessen Horowitz Cofounder and Partner Marc Andreessen. Read more leadership lessons from his Stanford GSB View From The Top talk on Tuesday, November 8, 2016
____________________________________________


0:00  [MUSIC]
0:00  [APPLAUSE]
0:10  >> Well, Mark welcome back to Stanford and
0:10  from one Midwestern to another, we're honored to have you here.
0:20  >> Great, thank you, thanks everybody.
0:20  >> So, I'd like to begin by exploring your intense intellectual curiosity,
0:20  recent wired article profiled your library And instead of it being in your home,
0:30  it's actually in the lobby of A16Z.
0:30  So could you unpack for
0:30  us why you put it in a place open to visitors as opposed to your home?
0:30  >> Yeah, well, so two reasons, one is if you've been to other Venture Capital
0:40  offices, they look a little bit like insurance companies.
0:40  It's all tombstones of long dead companies that they sold and
0:40  took public once upon a time, so.
0:40  We just thought that was kind of depressing, so
0:50  we decided to do something different.
0:50  So the message that we're trying to send with the library, and
0:50  it's a very deliberate message, kind of goes as follows.
1:00  Which is, the great thing about the valley, especially in our time,
1:00  is the sense of newness and the sense of the future.
1:00  And the sense of audacity that there are radical new ideas and
1:10  that there are new people that come to the valley all the time and
1:10  can pursue things that have never been done before.
1:10  So, very future oriented and that is an enormous strength.
1:20  Elon Musk talks about this, kind of most vividly, he describes,
1:20  he says you always want to think for first principles.
1:20  You want to kind of not take any received wisdom from the past,
1:30  instead you want to think from scratch and he makes very good method assumptions.
1:30  And the past may no longer be true and so if you rethink things from scratch,
1:30  you can reach different conclusions and
1:40  do things differently than could've be done before.
1:40  And I think there's a huge strength to that,
1:40  I think there's also a big problem with that, or it's an incomplete theory.
1:40  Which is, there are also thousands of years of history in which lots and
1:50  lots of very smart people worked very hard and ran all kinds of experiments
1:50  on inventing new technologies, or creating new businesses,
1:50  or new ways to manage or new ways to lead or all kinds of things.
2:00  And they ran all these experiments, and
2:00  they ran these experiments throughout their entire lives.
2:00  At some point, somebody put them down in a book, and for very little money and for
2:10  a few hours of time,
2:10  you can literally learn from somebody's accumulated experience.
2:10  And I think there's just, there is so
2:10  much more to learn from the past than I think that we often realize.
2:20  And so I think you could productively spend literally all of your time,
2:20  all your life, just reading the experiences of great people that have come
2:20  before, and I think you learn every time.
2:30  >> So, speaking of history,
2:30  one of the significant elements in that library are the history of Hollywood.
2:30  And I'm curious why that fascinates you, and what parallels, if any,
2:30  you might see to the development in Silicon Valley.
2:40  >> Yeah, so in a lot of ways, Hollywood is very unlike the Valley.
2:40  Probably the biggest different is in the Valley,
2:40  if you say that somebody's startup is just all story.
2:50  And no substance, it's a very offensive thing to say.
2:50  In Hollywood, they take that as a huge compliment,
2:50  because the whole point is to tell a great story.
2:50  And so they view that as a clean win,
3:00  there's another big difference actually which I find very, very instructive
3:00  to think about which is people in the valley often think our lives are harder.
3:10  Or it's harder to start these companies or it's hard to compete for the business.
3:10  Hollywood is much more difficult, it's a much more actually hardcore competitive,
3:20  even vicious business environment.
3:20  And so it's always good, I think, to have perspective of people who have it harder
3:20  than we do and still manage to pull things off.
3:20  I think that's true of the big successes in Hollywood.
3:30  The thing that they have in common is we in tech and
3:30  they in entertainment are the two big original California industries.
3:40  Right, I think between the two industries we kind of embody the spirit of California
3:40  and the spirit of the West in the US.
3:40  Once upon a time there was a gold rush, then we pulled out all the gold and
3:50  we're like, okay now we've gotta make some new gold.
3:50  And we make it through technology and they make it through storytelling.
3:50  But very similar entrepreneur personalities, very, lots and
4:00  lots of people.
4:00  Up here everybody's got a start up idea, down there everybody's got a screen play.
4:00  Lots and lots of people come from all over the country, migrate in,
4:10  almost everybody's from out of town.
4:10  People have a real sense of what great is, one of the things about the valley that I
4:10  think is amazing is, it's impossible to be in the valley and
4:20  not have a sense of what a great technology is.
4:20  because you have your nose rubbed in it all the time, it's like yup,
4:20  there's Google, okay, all right, that's the bar.
4:20  In Hollywood, and for what they do, it's the same thing, they made everything from
4:30  Casa Blanca to Star Wars all the way up to all the great movies today.
4:30  And so it's just an incredibly high quality, high caliber,
4:40  high achieving culture.
4:40  Incredibly energetic people and between these two industries, California is
4:40  the I think the sixth largest economy in the world as a stand alone.
4:40  If it were a stand alone country it would be the sixth largest economy in the world.
4:50  And I think we are the two industries that power that.
4:50  >> So as you think about the adventure capitalists you take in a lot of
4:50  information from a lot of different companies, different trends.
5:00  This seems to be a parallel to how you approach learning
5:00  taking in a wide variety of history and other elements from books and stuff.
5:10  How do you think about learning?
5:10  How do you go about actually figuring out what you want to focus on and, develop?
5:10  >> Yeah, so there's part of what we do, so
5:10  the way modern adventure firms work is kind of strange and interesting.
5:20  So we get inbound pitches, so
5:20  we get inbound 2000 pitches a year, from qualified referred entrepreneurs.
5:20  So somebody who's kind of hitting the basic criteria, have been
5:30  able to network in the Valley, and been able to connect with someone we know.
5:30  A very qualified set of founders, so we see 2000 a year, and so a lot of it is
5:40  just meeting with all those founders, and trying to survive the tsunami.
5:40  I read, there's all these books, there's a great book, I disagree with the book, but
5:40  it's a very well-written book called The Rise and
5:50  Fall of American Growth by an economist named Robert Gordon.
5:50  And he says, basically, innovation is over, and all the great ideas have
5:50  been taken, and from here on out just everything's going to get boring.
5:50  There's no new fundamental technological breakthroughs on the horizon.
6:00  Some days it would be a relief if that were true,
6:00  because out here, we're just drowning in it, right.
6:00  It's just this constant tsunami of new ideas and
6:10  new thinking, and new people with new thinking.
6:10  And so a lot of it is just being in that fire hose, that's part of it and
6:20  then like I said, I try to compliment that trying to be proactive.
6:20  And so, one is trying to explore new areas of science in particular and there's,
6:20  I think, some very interesting new areas of science right now.
6:30  Many of which are happening right here at Stanford, and then the other is history,
6:30  I always go back to history.
6:30  And I always go back to try to understand, okay,
6:30  when you had a fundamental, I'll give you an example, self-driving cars, right.
6:40  So self-driving cars are a fundamental advance in computer technology.
6:40  They're a fundamental advance in the transportation industry.
6:40  They're obviously going to be a very big impact on the automotive industry,
6:50  which is a very big global production industry, consumer industry.
6:50  It's like okay, well what happened,
6:50  the thing about the impact of self driving cars.
6:50  Well let's start by understanding what were the impact of cars, right.
7:00  And there were very, very interesting lessons from how cars rolled out that I
7:00  think can be applied 100 years later to figuring out some of the,
7:00  both the issues and the opportunities of self driving cars.
7:10  And so I try to go back and look for
7:10  the patterns that can kind of be more helpful in predicting the future.
7:10  >> Speaking of science, backstage you mentioned that at this point in our
7:10  history, 90% of the scientists that have ever existed and engineers live right now.
7:20  >> Yeah, living or working today.
7:20  It's 90% of all scientists, 80 or 90%, I think it's 90%, of all scientists and
7:30  engineers, who have ever lived are living and working today.
7:30  >> So what does that imply for humanity, going forward the next two decades, and
7:30  even the next century?
7:30  >> Well, this is why I'm such a gigantic bull on future innovation.
7:40  So there's two basic theories, the Gordan theory is basically,
7:40  it's a question of low-hanging fruit.
7:40  And so electrical power was low-hanging fruit, steam engine was low hanging fruit.
7:40  Electromagnetic waves were low-hanging fruit, and from here on out,
7:50  everything gets much more complex.
7:50  And so it's harder, you're putting more, diminishing returns.
7:50  You're climbing higher and higher, and putting more and
7:50  more effort in for less results.
8:00  And actually, by the way, what Gordon points to that actually justifies that
8:00  point of view is Measured productivity growth across the economy is quite low.
8:00  Right? And so, right,
8:00  because basically there's two arguments,
8:10  critics of technology basically make two arguments.
8:10  One is there's no innovation, there's no productivity growth.
8:10  And the other is, there's so
8:10  much innovation that it's going to destroy all the jobs.
8:10  Right, and what's important to realize is they're diametrically opposed
8:20  arguments, right?
8:20  They literally, factually, logically can't both be true.
8:20  Either you have super low productivity growth and nothing changes, right,
8:20  on one side of the argument, or you have super high productivity growth through
8:30  technological change and everything changes.
8:30  But they can't be reconciled, they can't both be true.
8:30  However, there is one way they actually can both be true,
8:40  which is where things get really interesting.
8:40  Which is they vary by industry, right?
8:40  And I actually think this is the part of his argument that I agree with,
8:40  which is that there are industries like technology, consumer electronics,
8:50  media, retail as examples,
8:50  where there's extremely rapid productivity growth in technological adoption.
8:50  And then there are other industries like healthcare,
9:00  education, law, government where there's a little technological adoption,
9:00  a very little productivity improvement.
9:00  And so one of the things we try to do is try kind of navigate through
9:00  the differences industry by industry and try to kind of understand how to apply
9:10  the technology particularly the areas that where it hasn't been applied before.
9:10  >> To the point of jobs in our economy,
9:10  one of the main criticisms of self-driving cars in the future is there's 5 million
9:20  truckers all around America that the role of the economy may diminish.
9:20  We see companies like Facebook and Twitter compared to their equivalent
9:20  market capitalization companies in the past decades having far fewer employees.
9:30  So are you of the mindset that the employment opportunities will actually
9:30  increase as technology proliferates, and why do you think that?
9:40  >> Also there's an argument from logic which is sort of the doctrine
9:40  of free trade, specialization of labor,
9:40  it's kind of well established in economic theory why that would be true.
9:50  Then there's observation by history,
9:50  which as we're sitting here today after 300 years of technological change and
9:50  there are more jobs in the world than ever before, right?
9:50  There are more jobs in the US today than ever before,
9:50  there are more jobs around the world than ever before.
10:00  There are more people who are employed.
10:00  By the way, income levels have never been higher, right?
10:00  For all of our issues of stagnation of income, overall income levels across
10:00  the entire population of the world are at an all-time high.
10:10  And so, if technological change were going to cause elimination of jobs,
10:10  one presumes we would have seen it by now.
10:10  >> Yeah. >> Right, again,
10:10  the counter argument to that is but you're not taking an account like some dramatic
10:20  breakthrough in artificial intelligence.
10:20  At which point I start accusing my opponents in this argument of just
10:20  hand-waving.
10:20  If it's so easy to make a super genius AI robot, why don't you give it a shot?
10:30  We are looking to try to fund it,
10:30  in fact come in and pitch us, we will happily invest.
10:30  It may be the last investment we ever make, but
10:30  it will be a very successful one.
10:40  >> [LAUGH] >> So, we're totally on board with that.
10:40  >> Yeah.
10:40  >> So, there's sort of the long run historical view,
10:40  which I think is pretty well established.
10:50  Then the questions get very specific.
10:50  And these are some of the questions that have come up in this election cycle,
10:50  which is okay, the 5 million people involved in driving, one form or
11:00  another are transportation jobs in the US, transportation delivery services,
11:00  trucking, things like that.
11:00  And you're like, wow, 5 million jobs seems like a lot of jobs,
11:00  and it is a lot of jobs.
11:10  This is where I think people underestimate.
11:10  Even with our low rate of productivity growth today,
11:10  people underestimate the rate of change in the economy that happens anyway.
11:10  And the thing that happens there is that what gets reported are always headline
11:20  numbers and the headline numbers are always the net numbers.
11:20  And so any given month, they'll report.
11:20  It just came out, I think, two days ago.
11:20  It was last month or I forget, 170,000 new jobs or something like that in the US.
11:30  And so you're like, jeez, you're not creating very many jobs.
11:30  How can you possibly create 5 million?
11:30  So that's the net number, it's not the gross number, right?
11:40  So the gross number is much higher.
11:40  Both in terms of jobs destroyed, and jobs created.
11:40  And the top line number there, is every year in the US, on average,
11:40  about 21 million jobs are destroyed, and about 24.5 million jobs are created, for
11:50  a net add over time of about 3.5 million jobs a year.
11:50  And so the real answer to how do you replace 5 million jobs is we already
12:00  replaced that in less than a quarter.
12:00  Today, just this quarter today, just this calendar quarter with the technology we
12:00  already have, we're already going to create more than 5 million jobs.
12:00  5 million gross new jobs.
12:00  And so change is happening in the economy all the time, right?
12:10  And by the way, it's happening whether or not we elect Hillary or
12:10  Trump, it's happening whether or not we have steel here, somewhere else,
12:10  it's happening whether or not Silicon Valley rises or falls.
12:20  The change is just going to keep coming.
12:20  >> Yeah.
12:20  >> And then to me that gets back to our day job and what I think is so
12:20  important to focus on which is okay.
12:20  So, then it's saying okay, the world is going to change,
12:30  the jobs are going to change.
12:30  How do we set people up to be able to take advantage of the change?
12:30  Right? How do we have change work for people?
12:30  How do we expand opportunities?
12:40  Why education is so important.
12:40  That's why financial services actually is so important.
12:40  That's why it's so important to be able to raise money for new businesses.
12:40  That's why it's so important for
12:40  people to be able to move from company to company to be able to get new jobs.
12:50  And then the conversation I think we ought to be having is the next hour's worth of
12:50  conversation on all the things that we could do to have more people have access
12:50  to all the opportunity that the new technology in many cases is creating.
13:00  >> Yeah, you mentioned Robert Gordon.
13:00  Who else is in your intellectual diet that you turn to when you want to think
13:00  about deep issues like change and funding in our economy?
13:10  >> So I have a little mental model of Peter Thiel.
13:10  I have a little simulation of Peter Thiel.
13:10  He lives on my shoulder right here.
13:10  >> [LAUGH] >> And I argue with him all day long.
13:20  >> Who wins?
13:20  >> I always win.
13:20  >> [LAUGH] >> Face to face,
13:20  it's slightly more challenging.
13:20  So, I'm a big believer, goes back to history.
13:30  Charlie Munger talks a lot about being able to learn from the past.
13:30  He always says, the way he describes it, basically he calls it a build of mental
13:30  model of a person and you want to kind of try to understand.
13:40  And it's hard because the people we're doing this for are very smart people.
13:40  And so you have to try to simulate people who [LAUGH] in a lot of cases are smarter
13:40  than you are.
13:40  We want to build a mental model, and that could be a mental model of,
13:50  in modern times, Steve Jobs or Jeff Bezos or somebody, Larry Page, Elon Musk.
13:50  We have mental model of somebody from the past, Thomas Edison, J.P. Morgan.
13:50  It can be a model of just somebody you admire,
14:00  a great philosopher or religious figure.
14:00  And you want to kind of construct a model of how they think.
14:00  And be able to be very objective and
14:00  fair where you can kind of think things through from their standpoint.
14:10  And then you kind of you have your own view on things and
14:10  then you try to run through kind of in your head what you know of them and
14:10  so okay here are the conclusions that they would reach.
14:10  And if you put enough time into that,
14:20  you start to be able to have these conversations I find sort of
14:20  with yourself.
14:20  People might look at you funny while it's happening but
14:20  you get to kind of engage in this dialogue and so I'm a big believer in that.
14:30  The three people in the valley who I do that with most often, my shoulder,
14:30  Peter, Elon Musk and Larry Page.
14:30  The three of them I find pushed the boundaries of what technology can do and
14:40  what Silicon Valley can do further than anybody else.
14:40  I think they're the three most audacious people, I think,
14:40  probably who have ever worked in the Valley.
14:40  >> Yeah.
14:50  >> And then you've got your models of the great founders.
14:50  Steve Jobs, obvious, and many other folks who built amazing companies
14:50  as entrepreneurs and then the new one for us is the great CEO's, right?
15:00  So Andy Grove being kind of our base case for that.
15:00  So, and whenever we talk about management in our firm, we're always trying to index
15:10  back on okay, what would Andy Grove say if he were sitting here today.
15:10  Good news that Andy is that he was very forceful in expressing his views.
15:10  And so he's on the record with on most of the important topics.
15:20  And as a very good kind of reference point for thinking about those issues.
15:20  >> What I'm struck by is you love to engage with people who may
15:20  disagree with you?
15:20  >> Yeah.
15:20  >> What's the broader lesson here,
15:30  for those of us here in business school who are thinking about these things?
15:30  >> Yeah, so I just think, and again, this is the kind of thing where, this way.
15:30  Adventure Capital, [LAUGH] venture capital.
15:40  Venture capital is a funny business.
15:40  Let me say why I feel so strongly on this.
15:40  So in venture capital, there's two kinds of mistakes you can make.
15:40  There's a mistake of omission.
15:40  Or there's a mistake of commission.
15:40  Which is mistake most people make because I make a mistake.
15:50  I make a decision, I invest in a company, I lose all my money.
15:50  It's the mistake everybody kind of thinks about.
15:50  And then there's the mistake of omission.
15:50  Which is Mark Zuckerberg walks in the door at venture capital firm XYZ in 2004.
16:00  And, you know, like what is this little kid doing?
16:00  And this idea's crazy.
16:10  And Fredster proved that this could never work.
16:10  And this is ridiculous, which is what he got told by a lot of people.
16:10  In the venture capital business,
16:10  every highly successful VC has made mistakes of omission, really big ones.
16:20  Of companies that they had the chance to invest in, they should've invested in,
16:20  they didn't invest in.
16:20  It turns out the mistakes of Co-Mission, they matter, but
16:30  they don't scar you for life.
16:30  >> [LAUGH] >> They just fade into history.
16:30  The mistakes of omission, right?
16:40  It's a asymmetric payoff.
16:40  When these companies work, you know, they get a thousand extra.
16:40  If you lose, you lose 1 x, if you win, we win 1000 extra, 10,000x.
16:40  The other is, the companies succeed, right?
16:50  The companies you passed on succeed, and they torture you, right?
16:50  >> [LAUGH] >> Because, you know.
16:50  >> So, who's tortured you recently?
16:50  >> No comment >> [LAUGH]
16:50  >> I can't even bring
17:00  myself to talk about it.
17:00  You pull up Tech Meme in the morning or whatever your choice of thing is, and
17:00  you're just like, shit, that one again.
17:00  >> [LAUGH] >> And so
17:00  they just torment the hell out of you.
17:10  And so, it's a humbling job.
17:10  I think the same basic principle applies to, but much more broadly,
17:10  I think the same principle, especially for people kind of in this ecosystem,
17:20  I think the same principle applies to even people, that mentality applies
17:20  to people with kind of what you consider to be more normal careers, right?
17:20  So, as you think about having a career over 50 years,
17:30  you're going to make basically a sequence of bets, right?
17:30  And you're going to make a sequence of bets of the places you choose to go and
17:30  the people you choose to work with.
17:30  And you're going to screw some of those up,
17:30  and you're going to make a series of bets and decisions for
17:40  places you don't want to go and people you don't want to work with.
17:40  And you're going to make those mistakes on both sides, and
17:40  you're going to have the exact same feeling on the other side.
17:50  When you had the chance to be employee number four at Google and
17:50  you didn't take it, right?
17:50  That sticks with you for a while.
17:50  >> [LAUGH] >> That's a big one,
17:50  rhat one tends to wear on you.
17:50  And so, I think that mentality, we call it the slugging percentage mentality.
18:00  Which is basically, take the bat, lose 1x, don't take the bat,
18:00  possibly miss on 1,000x.
18:00  That mentality is a very powerful mentality.
18:10  If you're going to do that, you have to be kind of ruthlessly open-minded, right?
18:10  And then this is sort of my view, when we make mistakes of omission,
18:10  why do we do it?
18:20  And I think it's almost always because we have some theory for why something's not
18:20  going to work, and you try to investigate that theory, but you know,
18:20  the problem with human nature, you develop a theory and you tend to want to prove it.
18:30  A confirmation bias, they call it.
18:30  And so, you develop an idea, and then you look for
18:30  all the evidence that supports it, and you ignore all the evidence that disproves it,
18:40  and so you get locked into your ideas, right?
18:40  And then by the way, pattern matching works against you, right?
18:40  Because things that didn't work in the past, might work now.
18:40  And this is the problem with history, this is why Elon focuses so much for
18:50  first principles.
18:50  Just because like MySpace, for example, didn't reach Facebook levels,
18:50  the scale didn't mean that Facebook wouldn't be able to.
19:00  So, you have to be ruthlessly open-minded,
19:00  you have to be constantly willing to re-examine your assumptions.
19:00  And to do that you have to try to figure out a way to not get emotionally tied to
19:10  your beliefs.
19:10  So as I always like to say, there's this software term called sandbox, and
19:10  there's the idea that you can basically run code on bare metal,
19:20  you can run code on the chip, or you can run code on a sandbox.
19:20  The idea of a sandbox is, it's a contained environment, where if the code goes bad,
19:20  if it's Malware, or it's an AI that wakes up or whatever,
19:30  you can nuke the sandbox and nothing bad happens.
19:30  I think we need to run ideas in sandboxes.
19:30  Beliefs should be run in sandboxes.
19:30  As much as possible, you shouldn't self identify with beliefs.
19:40  We should treat them all as sort of more abstract objects.
19:40  And be willing to pull them in, think about them, and then put them back on
19:40  the shelf, as opposed to saying, this is part of my identity.
19:50  And it's a very hard thing.
19:50  It's almost a zen-like, you have to take the ego out of ideas,
19:50  which is a very hard thing to do.
20:00  >> Absolutely.
20:00  You never went to business school.
20:00  You're in an audience full of folks who, obvious, are in business school.
20:00  What do you hope that we leave here knowing?
20:10  Both about the Valley and technology that you learned from actual experience?
20:10  >> Yeah, so the big thing, it all sounds so easy, from a distance.
20:20  >> [LAUGH] >> It all sounds so straight forward,
20:20  like the case makes it all so clear.
20:20  So, strategy is important.
20:30  Strategy is thinking about like,
20:30  okay what are the decisions that should be made over time, is very important.
20:30  I like to come at it, particularly, in terms of running something,
20:30  like running a company.
20:30  Making strategic decisions that affect your company.
20:40  The strategic decisions sound more obvious than they are, and
20:40  one is there's just a lot of context at the time that's missing.
20:40  Even in the histories it's usually missing.
20:40  The other thing I think, the biggest thing I find newly minted MBAs have a hard
20:50  time wrapping their heads around is, the abstract idea of what the right
20:50  thing to do is in a business situation is all well and good,
20:50  but as an actual manager in an actual business, you have other constraints.
20:50  In particular, you have organizational constraints.
21:00  You have constraints as to what your actual organization can do.
21:00  And those constraints have to do with your employee base,
21:10  how well your company's organized and run, the existing commitments you've made to
21:10  those employees, the expectations of your investors, the expectations of your board.
21:20  Andy Grove had this sort of way the he described it, which is basically he said,
21:20  as the CEO's in place,
21:20  basically on day one a CEO has complete latitude to do whatever they want.
21:20  On day two that's no longer true,
21:30  because you've started to make commitments to people.
21:30  You've started to tell the board, okay here's my plan, I'm going to do x.
21:30  You told investors, here's my plan, I'm going to do x.
21:30  You've told employees, here's what we're going to do, here's the plan.
21:40  So then as a CEO, you get 6 months in, 12 months in, 18 months in, 24 months in, and
21:40  you change your view on strategy.
21:40  And you go back to all the people and you tell them we're going to do something
21:50  different and they all say, but you just got done telling us the opposite last
21:50  week, or last month, or last quarter, like what's wrong with you?
21:50  And so, basically, what happens is the people who take the idealized approach
22:00  of shifting strategies when it seems logical, typically end up getting fired.
22:00  They literally end up getting either fired explicitly, their board will fire them or
22:00  their investors will fire them, because they appear that they're lurching all over
22:10  the place and whip-lashing the organization.
22:10  Or they get fired, they just get abandoned, right?
22:10  Like the employees and the investors all get a vote, and they get to leave and
22:20  they quit.
22:20  I'll have this conversation sometimes with founders where I'm like,
22:20  that's all well and good, but
22:20  like look around you, you're going to be sitting here by yourself, right?
22:20  How do you feel about that?
22:20  With this big new decision that you think you're going to make.
22:30  And so, the real world constraints are very intense, and
22:30  they're very context specific to that company, at that moment in time,
22:30  and kind of how it got to where it got to.
22:40  Now, Andy's response to that was very crisp and solid,
22:40  which was basically, he literally said this.
22:40  He's like, well okay, if the next guy is going to come in here and
22:40  he's going to do all these decisions, why don't we just go outside and
22:50  come back in through the door and we'll just go ahead and
22:50  we'll just breach all the old commitments and we'll do the clearly right thing.
22:50  And so, you're the leader, you're responsible for it, the answer has to be,
23:00  you have to be able to work your way through it.
23:00  You have to be able to break the glass to be able to do that.
23:00  It's just that's much harder and riskier than it seems from the outside.
23:10  >> Speaking of context in leadership, you ran a very successful company with
23:10  technical employees, and now you run a very successful VC with a lot of partners
23:10  who were successful entrepreneurs in their own right.
23:20  What are the differences in leadership you have to display in both those
23:20  environments that you've learned to bounce, I guess?
23:20  >> Yeah, so the big thing, Siri is weighting in.
23:30  Siri has opinions on management here in the front row.
23:30  So, the big thing as an operator, what I always found as an operator, it's like
23:30  a 90, 10 thing, where you're spending 90% of your time making decisions.
23:40  Like 90% of the time you're sitting there and customers, and issues, and
23:40  you're getting employee product reviews, and employee feedback.
23:50  And you're making hiring decisions, firing decisions, market expansion decisions,
23:50  financial decisions, and so it's just an unending series of decisions all day long.
23:50  And then you get maybe 10% of your time to sit and think.
24:00  If you're lucky and if you're disciplined you carve out 10%.
24:00  You kind of think and read, and get caught up on kind of your own.
24:00  Your own theories on things.
24:00  And you kind of have to do that,
24:00  because these markets tend to be brutally competitive, and
24:10  if you just sit around thinking, your company's not going to go anywhere.
24:10  As an investor, and anybody who's been an investor will tell you,
24:10  if you take that approach to investing, you'll blow yourself to bits, right?
24:20  If your approach to investing is 90% action,
24:20  10% thought, right, then you're like every other schmo in the stock market.
24:20  You're just churning through stocks and you're going to, ultimately,
24:30  as a stock market investor,
24:30  you're going to do terribly because you're going to churn your portfolio.
24:30  As a VC, you're just going to blow yourself to bits.
24:30  You'll do the first ten deals and then that's it and then they'll all fail and
24:30  you're done, your career's over.
24:40  And so a biased transaction as an investor is a very dangerous thing,
24:40  generally a very bad thing.
24:40  And so, the great VCs and the great, by the way, great VCs, the great public
24:50  market managers, the great hedge fund managers, anybody with long term horizon.
24:50  They seem to spend 90% of their time thinking, thinking and arguing.
25:00  There's thinking and then there's the actual process of all the people around
25:00  you having to try to convince them of your crazy idea so
25:00  the arguing part ends up dominating.
25:00  So 90% of your time arguing and 10% of your time, or maybe even less,
25:10  making decisions.
25:10  One of the things we ask ourselves is,
25:10  how many decisions a year do we make that actually matter?
25:10  And it's probably 20?
25:20  It's probably the new investments we make, which is about 20 a year and
25:20  that's probably it.
25:20  And there's lots of things we advise on and help with, but the decisions we make,
25:20  it's probably two a month, at max.
25:30  And so, it's a radically different balance of time.
25:30  By the way, if you've in an operating job, it sounds very appealing because you don't
25:30  have to sit there all day long and be involved in all this activity and
25:40  you get all this time to think and read and talk to people and it sounds great.
25:40  Except a lot of operators are highly action-oriented and
25:40  their entire success in their career has come from taking action right?
25:50  Crisply and aggressively.
25:50  And so a lot of operators can't make the jump to ever being an investor.
25:50  because it's just too frustrating, right?
25:50  Six months in, they're just like crawling the walls, right?
26:00  They'll go home at the end of the day and they're literally,
26:00  I didn't do anything today, right?
26:00  This day was a total write off, I made no decisions, right?
26:00  And that's true of about 28 out of 30 days kind of by definition.
26:10  And so I'd say most, I'm sorry, most investors can't make the jump operator,
26:10  because they would turn into a little puddle.
26:10  >> [LAUGH] >> Of plasma under their desk.
26:20  Most operators can't become investors,
26:20  because they can't deal with the frustration of not taking the action.
26:20  And so we're trying to straddle it somewhere in the middle.
26:30  This is the speech I give when we only have former founders or
26:30  CEOs as GPs at our firm.
26:30  So we have all been through this transition, and this is the little speech
26:30  that I give every time, and they all tell me, yep, you got it, I understand.
26:40  And then six months later, they're like, I see.
26:40  >> So you mentioned crazy ideas.
26:40  What are some of those crazy ideas that you guys are banting around in the A16Z
26:50  offices that you think will define the future of the Silicon Valley and
26:50  tech here?
26:50  >> Yeah, the way we sort of go is we call ourselves extremists thinking about ideas.
27:00  Generally, smart people in the industry are going to have figured out everything
27:00  that's going to happen the next five or ten years.
27:00  It's one of the things like, [LAUGH] Peter,
27:10  my little argument with my shoulder Peter, Peter talks a lot about secrets.
27:10  And he talks about secrets being something that you know or
27:10  believe that other people either don't know or don't believe.
27:20  And I actually think like over a five or ten year period,
27:20  there actually aren't very many secrets.
27:20  Most of the good ideas are already kind of obvious.
27:20  And by the way, they've probably been tried before and they probably failed, but
27:20  the time is going to come where they're going to work.
27:30  And so we in the industry, the iPad is a spectacular new idea,
27:30  except it wasn't a new idea.
27:30  Apple had something called the Newton 20 years earlier,
27:30  but it was basically the same thing.
27:40  It failed then, it worked in 2009.
27:40  So tablet computing was not a new idea,
27:40  it just happened to take 20 years to get it to work.
27:40  But for that entire 20 year period, it did look like a good new idea.
27:50  So most of the good ideas are kind of out there.
27:50  They're circulating or,
27:50  by the way, they're running in labs, over here in the engineering department.
27:50  They're already up and
27:50  running as prototypes and grad students are working on them.
28:00  So that part of it is not, and we swim in that world and trying to understand it and
28:00  all that, but
28:00  it's not that there are huge undiscovered ideas with a short term timeframe.
28:00  So I think we spend a lot more of our time on the idea side, trying to think, okay,
28:10  ten plus years out, right?
28:10  Sort of take the extreme view, right, of okay, it's sort of a [INAUDIBLE] of
28:10  the venture philosophy, which is, forget whether or not it will work.
28:20  Ask the question about what if it did work, and
28:20  then kind of push that question out as far as you can, right?
28:20  So I mentioned self-driving cars.
28:20  So let's just assume for a moment that self-driving cars actually work.
28:30  Okay, what are the consequences of that?
28:30  Well, for example, one consequence is,
28:30  potentially, cars changed our idea of geography, right?
28:30  Before cars, everybody used to have to live in the city.
28:40  Cars created the idea of a suburb, right?
28:40  because you could actually commute.
28:40  We all sit here 80 years later,
28:40  wishing that nobody had thought of that because commutes are horrible.
28:50  The number one correlator to job satisfaction is commute time, right?
28:50  And so, you see this every day.
28:50  You're sitting there in traffic, sitting there next to,
29:00  and so everybody has the same epiphany.
29:00  It's like, it's eight in the morning, I'm sitting there, and I'm like, okay,
29:00  I'm sitting, there's car over there with one person in it,
29:00  there's a car over here with one person in it.
29:00  And I'm sitting here along.
29:00  Like somebody should really come up with a car pooling app.
29:10  Every morning, right, every driver has that idea and we all still commute, right,
29:10  in our cars.
29:10  And said we all hate it, and so, and it's all wasted time.
29:10  And we have people at our companies now who commute two or three hours a day, and
29:20  it's just completely wasted time.
29:20  I mean, now they can play Pokemon Go, right,
29:20  while they drive, but that has other issues.
29:30  So then he said, okay, self-driving cars.
29:30  Well, self-driving cars, maybe you can reclaim all that time.
29:30  Maybe you can reclaim all the commute time, right?
29:30  And so maybe all of a sudden you can have the idea that maybe an hour long commute
29:40  is actually a big perk, right?
29:40  Because instead of driving, instead of having to sit and focus and
29:40  merge through traffic, what if your car is a rolling living room, right?
29:40  And what if you get to spend that hour playing with your kid or
29:50  reading the news or watching TV or actual working, right?
29:50  Because you don't have to worry about driving.
29:50  Or what if you had different kinds of cars?
30:00  What if some cars are rolling offices?
30:00  What if I can take a nap, I could sleep.
30:00  I can sleep for six hours at home, get in the car, sleep for another hour and
30:00  a half on my way to work, and be all set, right?
30:10  And so in that version of the world, all of a sudden, maybe suburbs,
30:10  like maybe now we go to exurbs.
30:10  Maybe now geography becomes actually much more protractible, and
30:20  we can have these urban environments get much, much larger.
30:20  And then we're trying to think, okay, then back from that,
30:20  what would be the consequences of that?
30:20  What would be the consequences of that in terms of how these company get built?
30:20  What would be the infrastructure that has to get built to support that
30:30  kind of thing?
30:30  What are the kind of early signals that kind of show that that kind of thing is
30:30  either starting to happen or not.
30:30  And then try to chart at least some view of how the future will unfold.
30:40  And so, yeah.
30:40  >> So, you mentioned a lot of us have great ideas we want to execute on.
30:40  You've also noted that capital is abundant and opportunities are scarce.
30:50  And I think venture capitals are only a $50 billion industry amongst
30:50  a multi-trillion dollar economy.
30:50  What do you make of the different investment choices being made by larger
30:50  companies vice what's happening here?
31:00  >> Yeah, so this is the big thing, I think, about tech.
31:00  [LAUGH] l So, once upon a time, we had a tech bubble.
31:00  Some of you may have noticed or read about it.
31:10  It was followed by a catastrophic crash in 2000.
31:10  It was an actual bubble, the numbers are very clear on that.
31:20  There was about four years of nuclear winter, where very little happened, and
31:20  then starting in 2004, tech started to work again.
31:20  And between 2004 and today, it's just been constant commentary of bubble, bubble,
31:30  new bubble.
31:30  Tech bubble 2.0, all the way up and to the right,
31:30  as tech has worked constant screaming of bubble.
31:30  Up to and including the chairman of the Federal Reserve.
31:40  Janet Yellen has developed a theory that tech's in a bubble,
31:40  which we really appreciate out here.
31:40  >> [LAUGH] >> I think she might have slightly larger
31:40  issues to deal with.
31:50  And by the way, with imminent predictions of a crash, right?
31:50  Which hasn't happened, right?
31:50  And it's every six months, imminent predictions of a crash, no crash, and
31:50  it might actually be starting to give up on this.
32:00  Which ironically would probably be the sign of an actual bubble.
32:00  >> [LAUGH] >> In any event,
32:00  the result of all that is exactly what you said.
32:10  So all of new tech in the US, all of new tech venture capital, growth,
32:10  investing, all of what we would consider to be what Silicon Valley does.
32:20  Total capital in is about $50 billion a year.
32:20  And again, sounds like a big number, is a big number on an absolute basis.
32:20  It's a drop in the bucket from a national or global standpoint.
32:30  So the US economy, the US economy is, I don't know, like $6 trillion or
32:30  something like that and then just look at financial assets.
32:40  I think the number is now it's like $14 trillion of negative yielding
32:40  government debt in the world.
32:40  So bonds that people have to pay for the privilege of holding and
32:50  then just even looking in corporate America.
32:50  This year alone corporate America will discourage more than a trillion dollars of
32:50  cash back to shareholders.
33:00  Let's just think about it this way, like a trillion dollars in cash was coming out of
33:00  the Fortune 500 that literally can't figure out anything to do with the cash
33:00  other than give it back, anything productive to do with it.
33:00  So they give it back, dividends and buyback.
33:10  So a trillion dollars comes out and then 50 billion,
33:10  basically of that goes back into tech.
33:10  Where does the other 950 billion go, right?
33:10  Into negative yielding bonds.
33:10  And so which is what, to me, it's like okay, why are interest rates so low?
33:20  I think interest rates are so low for a fundamental reason,
33:20  which is the world is awash in capital.
33:20  Basically, the 20th century worked.
33:30  The 20th century created enormous amounts of wealth all around the world and
33:30  that process continues.
33:30  And now that money is all out there, sitting in these giant pools.
33:40  Primarily, people's retirement savings.
33:40  These huge pensions or these big sovereign wealth funds that basically represent
33:40  people's future retirement.
33:40  And that money, all needs to earn a return of 6 or 7 or 8% per year to pay for
33:50  everybody's old age.
33:50  And that money is seeking out investment opportunities and
33:50  it can put 50 billion into tech, and then where else is it going to go?
34:00  And so I not don't think there's a tech bubble,
34:00  I think there's a massive tech bust.
34:00  I think we're running a critical shortage of new technology in the world and I think
34:00  you can view through a technological lens in productivity growth, but I think you
34:10  can also view that on a financial lens, which is there literally are not.
34:10  We as a society cannot come up with enough places to productively deploy capital and
34:20  this is the thing,
34:20  like this just seems completely bizarre like do we have any problems?
34:20  Yes, we do.
34:20  We've got massive education problems, massive healthcare problems.
34:20  There are big problems all over the way still to this day.
34:30  [LAUGH] Funny problems are funny.
34:30  We use to have a massive global hunger problem.
34:30  Now of course, we have a massive global obesity problem.
34:30  Good news, people aren't starving to death as much anymore.
34:40  Bad news, now everybody's going to die of diabetes and heart attacks, but
34:40  like it's still a problem.
34:40  So now, we have a big problem on the other side.
34:50  And so, we've got the money.
34:50  Through most of recorded history, you'd have these problems and
34:50  you wouldn't be able to find an answer to them.
34:50  So we've got the money, we can fund the answers.
34:50  We've got the problems, we know what we have to go do.
35:00  It's the process in the middle that's kind of all screwed up.
35:00  And so, that's where I just think like a tax on Silicon Valley like being bad at
35:00  innovation or whatever is kind of beside the point.
35:00  Even if you're right, even if the valley, even if the critics are all right and
35:10  a tax on Silicon Valley is a joke and it's all just a bunch of photo sharing apps and
35:10  self-driving cars will never happen and we're also it's like okay, so then what?
35:20  If like if we suck, okay, fine.
35:20  Like who else is going to do it?
35:20  By the way, the Chinese are going for it.
35:20  But again there, the dollar there's a big tech investment happening in China.
35:30  But again, the absolute dollar is just not that relative to even the Chinese economy,
35:30  much less our economy.
35:30  And so, I think it's a tech depression more than it's a tech bubble.
35:40  >> Is that risk aversion on the parts of investors or is it rational behavior?
35:40  >> Well, it's a combination.
35:40  It's a bunch of factors.
35:40  I think like for sure,
35:40  we've been on a risk reverse equity market since the 2000 crash.
35:50  If you look at the history of crashes and
35:50  what happens after people fight the last war.
35:50  And so, what we know from 2000 is that stocks crash.
36:00  So, everyone's constantly fearful.
36:00  The headlines are always about the market's about to crash,
36:00  like all the way up.
36:00  And then since 2008, on top of that, we now have a massive aversion to debt.
36:10  We're incredibly scared about debt.
36:10  We're scared about real estate.
36:10  It's really interesting.
36:10  You talk to like our investors or some of these big,
36:20  like the Stanford endowment here manages $25 billion.
36:20  And if you talk to folks like the guys who run the Stanford endowment not
36:20  them specifically, but that class of people.
36:20  It's just like, they literally, they go through.
36:30  Here are all the things we can invest in like public equities,
36:30  terrible performance, 2008 crash horrible bonds, we just had a financial crisis.
36:30  Bonds are largely unsafe and don't yield anything, anyway.
36:40  Real estate, blew up in 2008.
36:40  That can blow up at any moment.
36:40  Commodities, it looked like oil was going to go to the Moon.
36:40  Now, looks like the world's awash in an oil glut.
36:50  The Saudis are dumping on the market at any price.
36:50  So, you can't invest in commodities.
36:50  Hedge funds, activist hedge funds were working.
36:50  And now, they don't work any more and they're blowing up.
37:00  Venture capitals, we hear good things, but everybody says, it's a joke.
37:00  So maybe we can only put a small amount of money in it, anyway.
37:00  And so, they just literally like a process of elimination.
37:10  It's like, okay, where is the money going to go?
37:10  And so it's this weirdly perverse like it's like I get really pessimistic about
37:10  this, because I'm like the transmission mechanism is not working properly.
37:20  Like somehow, there's something between the money, the problem and
37:20  then the transmission mechanism to connect the money to the problem
37:20  that involves ideas, people, companies, technology, management.
37:30  Something in there is broken, but I also get very optimistic.
37:30  because it's like boy, if we could figure out how to get that transmission mechanism
37:30  working better, we could go do much bigger things.
37:30  And that, by the way is a very big reason for
37:40  optimism in the valley as you do have a lot of founders in the valley now who
37:40  are going after much bigger problems that people used to go after.
37:40  Like it did not use to be the case that the valley would go after ideas like just
37:50  with the huge effort, it's not going on to buffer different areas of healthcare or
37:50  transportation as an example or even space.
37:50  With SpaceX.
38:00  It was just inconceivable that those ideas would have been tried 20 years ago.
38:00  And so, there are a group of people who do have this view that we can now go do
38:00  these things.
38:00  >> I could sit here for hours.
38:10  Unfortunately, I have time for one last question before Q&A from the audience and
38:10  this is deeply personal to me.
38:10  I know you have a 20-month old son, I have a 22-month old son.
38:10  How should I raise him in the next five,
38:20  ten years to be relevant in the 21st century?
38:20  >> [LAUGH] >> Can I get back to you in [LAUGH] 20 or
38:20  25 years once we find out how it went on our end?
38:30  Look, I think curiosity.
38:30  I mean, I just think curiosity.
38:30  I think it's a cliche, but I think it's true,
38:40  which is just whatever the assumptions are about how the world's working today.
38:40  So many kids just get brought up, I think Douglass Adams had this old line.
38:40  He said, whatever existed before you were born is just like ancient and irrelevant.
38:50  Whatever happens when you are, whatever happens, new technologies,
38:50  new ideas happen when you're an adult.
39:00  When you're a young adult are exciting and fresh and whenever ideas that
39:00  appear past the age of 50 are against the holy order of things and it's this
39:10  natural kind of human kind of view of like it's weird, history is very long.
39:10  We show up a bunch of stuff has happened.
39:10  We draw a bunch of conclusions from that, a bunch of other stuff happens.
39:20  We have a relatively narrow window to have an impact on that.
39:20  And so I think that curiosity just is really, really critical curiosity and
39:20  an ability to learn.
39:20  Frankly, everyone has an opinion on education.
39:30  My opinion is our education system is still dominated by industrial age theory
39:30  of we're creating cogs for a machine.
39:30  I mean, luckily at places like Stanford, there are exceptions to that, but
39:40  a lot of K through 12 a lot of that is fundamentally to get people to work on
39:40  farms or in factories.
39:40  And if you trace the ideas behind education,
39:50  people sitting in rows in a classroom and somebody that's a senior authority
39:50  figure sitting in front lecturing and the whole thing.
39:50  That's very kind of 1890 kind of era philosophy.
40:00  Frederic, what's his name?
40:00  It's mass production.
40:00  >> Yeah.
40:00  >> Applied to education.
40:00  And so I think that the thing with kids is to try to get them
40:00  To the point where curiosity and learning is good and exciting and fun.
40:10  Right, and it sounds obvious, and yet, for most people, that never happens.
40:10  >> Yeah, well, let me turn it over to the audience for some questions.
40:20  >> Hi, Mark.
40:20  This is Sam, 2015 alumni.
40:30  I've seen the ecosystem for the startup VC.
40:30  Have changed, evolved significantly over the past ten years,
40:40  in terms more entrepreneurship, educational resources,
40:40  cheaper resources for start-ups, and the more incubator startups.
40:40  So, that probably make the competition, amount of VC's more serious nowadays.
40:50  So, I'm curious about what's your view about this ecosystem over the next five,
40:50  ten years?
41:00  And what matters most to to in this world.
41:00  Is it your vision, your PR marketing, or
41:00  your disciplines on the research and due diligence?
41:10  >> Yeah, so the big thing I think, from what you said,
41:10  that I agree with you that the startup ecosystem has changed a lot.
41:10  And generally, in a very positive way.
41:10  The big thing that's happened, as you said, the big thing that's happened is
41:20  the education is so much more widely available.
41:20  There are now so many more people, both here in the valley and around the world,
41:20  who now know how to start a company.
41:30  And that, again, maybe sounds obvious, but when I arrived in the valley in 1994,
41:30  I went to the book store, and
41:30  I tried to find a book on how to start a tech company.
41:30  And that was a long and lonely trip through the bookshelves.
41:40  >> [LAUGH] >> There was very little.
41:40  Luckily, Andy's book was there, but
41:40  Andy's book was about how to run Intel, which, it was a monopoly.
41:40  I couldn't really get a lot of benefit out of that for my startup at the time.
41:50  And so, the resources just simply didn't exist, right?
41:50  And today you've got, it's everything.
41:50  A lot of it comes out of Standford.
42:00  Just the CS-183 course at Stanford that various people have taught,
42:00  Peter taught, Simon has taught.
42:00  And then the fact that those classes are not just accessible to people at Stanford,
42:10  they're accessible to people all around the world.
42:10  And there are people all around the world watching, right?
42:10  Those videos all the time, learning how to start companies.
42:10  And so, the information's out, and
42:20  you have these new ecosystem kind of participants, like Y combinator,
42:20  in particular, which has had this big impact in the valley.
42:20  Y Combinator is now producing 240 startups a year in the Bay Area out
42:30  of their current program, and they're planning to ramp that up.
42:30  Ironically, for venture capital, this has all been a tremendous blessing,
42:30  because this is steel flow.
42:40  And so, the more startups there are,
42:40  the pond that we fish out of is very well stocked.
42:40  So, we are super, super thrilled that all these things are happening,
42:50  because we get to evaluate a far larger number of startups per year than used to
42:50  be the case.
42:50  I think the competition is actually much more intense between the startups.
43:00  And you actually see that, how many you'll see that, where there will
43:00  be a company that will be super hot in a YC class that we'll be doing an idea,
43:00  and then they'll be the belle of the ball, and they'll raise money, and it's all bad.
43:10  And then six months later, in the next class,
43:10  there'll be another company doing the exact same thing.
43:10  [LAUGH] And now they're the hot company, and the previous company's, what the hell?
43:10  And then they compete.
43:10  And then they both come to raise venture capital a year later.
43:20  And then we, as VCs, get to pick between the two.
43:20  And so that's, at least so far, it's been really great for VC.
43:20  If you're thinking about starting a company, therefore,
43:30  really stresses an idea,
43:30  and again, I'll use my Peter here on the shoulder, thing channel this through.
43:30  Something Peter says I think is really relevant, which is, a lot of founders
43:40  think about what it takes to get the second person into the company,
43:40  the second founder or the second engineer, or the third person or the fourth person.
43:40  He says that's no longer the challenge.
43:40  Everyone wants to start a company, and so, everybody can get together two or three or
43:50  four people to start a company.
43:50  He said the challenge is how do you get engineer number 20, right?
43:50  Because engineer number 20, right?
43:50  If they're in this ecosystem, right?
43:50  They could go start their own company.
44:00  And so, how can you convince somebody to join your company's,
44:00  essentially, you're number 20 with a tiny percentage of equity, right?
44:00  And not being in charge, as compared to starting their own company, and
44:00  getting the equity and being in charge.
44:10  Now, one of the things to do is you tell them they're going to have
44:10  the same problem.
44:10  But then they respond to you that you also have that problem, and
44:10  then you sit there and stare at each other awkwardly.
44:20  >> [LAUGH] >> So, that's the thing.
44:20  So, what you have right now is you just have a very large number of seed funded
44:20  companies that are experiments.
44:30  And they're experiments on their ideas, but more importantly,
44:30  they're experiments on who has the will and the drive and the horsepower, right?
44:30  And the salesmanship and everything else that goes into recruiting, right?
44:30  And I think actually Sam Altman, again, I talked to the other day.
44:40  He said, if you could wave a magic wand,
44:40  you would combine a lot of these companies together to get critical mass, but
44:40  mergers between startups almost never happen for a variety of reasons.
44:50  And so, it makes this sort of acute,
44:50  this recruiting problem is a very acute problem.
44:50  By the way, the temptation is to think about it as a recruiting problem,
44:50  which is we have to get really good at recruiting.
45:00  But it's not really that as much as it's like, okay, what about my company is going
45:00  to be so spectacular, and so special, and so unusual, and so distinct and
45:10  differentiated that it's going to be able to easily hire engineer number 20?
45:10  Away from starting his or her own company.
45:10  And so, to me, it basically translates into basically,
45:10  you just have to get better.
45:20  You have to be able to compete, and
45:20  you have to win in this really brutal initial battle.
45:20  The good news is the companies coming out of that sort of churning kind of, sort of
45:30  snakepit competition, the companies that come out of that are really strong,
45:30  because they had to be to come through it.
45:30  But it is tougher, and
45:30  it is the other side of how easy it's become to start the companies.
45:40  >> [SOUND] Thanks, Mark,
45:40  for your time.
45:50  I have a question about the international investing.
45:50  So, I see some on top VCs like Saquioya Perkins,
46:00  they went to China market many years ago.
46:00  But in the contrast, I think, and just pretty much focusing on the US market.
46:10  I want to your point of view of expending investing in
46:10  those eMarkets like China India,
46:20  South East Asia, just like to hear about your view.
46:20  >> Yeah, so, it's very, it's extremely tempting.
46:30  Because they're entrepreneurial opportunities and
46:30  great companies all over the world.
46:30  Increasingly distributed, right, all over the world.
46:30  Literally, and so it's very tempting.
46:30  But I'll tell you the typical problem, and
46:40  then I'll tell you the theoretical question that comes out the other side.
46:40  So, the problem is,
46:40  to the extent that venture capital investing is like we understand it today,
46:50  which is it's a very hands on process of really deeply understanding.
46:50  It's a people business.
46:50  It's a process of very deeply understanding the people
46:50  you are working with.
46:50  Deeply understanding the founders,
46:50  deeply understanding the executives in the company, the engineers,
47:00  the management team, the culture.
47:00  You both have to evaluate the company and also to help work with the company.
47:00  It has always been a hands on business like that, right?
47:00  When it's worked.
47:00  So, if it continues to be a hands on business like that,
47:10  then there's a problem with geographical remoteness, right?
47:10  Which is, if I'm not present in another geography,
47:10  do I really know the people well enough to make those decisions?
47:20  So, then, what a bunch of firms are trying to do is they then staff
47:20  local teams, right?
47:20  But then you have this fundamental problem where you have this selection problem,
47:20  where if that local team is really good, they can easily leave and
47:30  go run their own firm, which is actually the exactly
47:30  same problem I talked about with startups, it also applies to venture capital firms.
47:40  Say we had a really good team on the ground in China,
47:40  how can I convince them to stay with us as opposed to just?
47:40  Because they could very easily go down the street open up their own shop,
47:40  raise money.
47:40  There's plenty of investors in China who want to invest in venture capital.
47:50  And so, if they're good, they leave and start their own firm.
47:50  If they're bad, they stay working for me.
47:50  >> [LAUGH] >> Which has its own issues.
47:50  And so, most of those experiments have not worked.
48:00  It is very striking, Sequoia in particular, has worked really well.
48:00  And they have built a very good team, and
48:00  they figured out a very good dynamic on that.
48:00  But it is a very tough question, because you have to really dig into, okay,
48:10  why did it work for them, and why has it not worked for so many others,
48:10  if you want to think about doing that yourself.
48:10  And so, that's a live topic, or question, in the venture ecosystem.
48:20  Now, there's the assumption, go back to first principles,
48:20  the assumption underneath that,is that venture capital is a hands on business.
48:30  And should be a hands on business, the way I've described it, right?
48:30  And again, if you look at history, if you look at investment banks 100 years ago,
48:30  they were very hands on like this.
48:30  If you look at private equity firms in the 70s, when they were a boutique business,
48:40  they were very hands on like this.
48:40  At one point stock investing was probably a lot like this.
48:40  These days, in a lot of fields it's not hands on anymore,
48:50  now it's fully become computerized, systematized and computerized.
48:50  And so, the counterargument to my whole theory on this is I'm describing venture
48:50  capital the old-fashioned way.
49:00  It shouldn't be so based on people, it should be done through algorithms, right?
49:00  You should be able to do quantitative venture capital, right?
49:00  And so you should be able to either like open up a crowd funding marketplace,
49:00  and let the world compete with ideas,
49:10  and to be able to raise capital, which is what AngieList is trying to do.
49:10  Or you should, there's various firms now trying to do quantitative venture capital.
49:10  You should basically go gather all the data, you can look at everybody's
49:20  college transcripts, you can look at everybody's past career records,
49:20  their LinkedIn connections, or whatever data you want to look at.
49:20  And you can assemble a quantitative view to predict success or failure,
49:30  and invest that way.
49:30  People are trying, I would say we haven't yet
49:30  gotten to the point where we think we've figured out how to do that.
49:30  Again, I fantasize that somebody is going to come in and
49:40  pinch us on building that as a company and we can fund it, and then we can retire.
49:40  I think, that question of is it really going to continue to
49:50  be that hands on is the question underneath the question that you asked.
49:50  I think it's an open question.
50:00  >> Hi, my name's Sam Jackson.
50:00  A question about the question of capital wash.
50:00  So, what advice would you have for
50:10  people who have big ideas that might be very capital intensive.
50:10  And a couple of years ago I asked Elon Musk,
50:10  what was his advice if someone started a new SpaceX or Tesla, and
50:10  he suggested that I become an internet billionaire first.
50:20  >> [LAUGH] >> So,
50:20  I was wondering if you had any other, maybe, easier ideas.
50:20  >> [LAUGH] >> That is how he did it.
50:30  So, it is actually striking, it is striking.
50:30  The guys who are most prone to be like, yeah, this software stuff is useless and
50:30  stupid, and we should be going to space travel instead,
50:40  they all started doing the useless and stupid software stuff, so that is true.
50:40  I would like to tell you, again, I'll word this as a question.
50:50  I'll pose an open question.
50:50  Which we're trying to think through, and
50:50  I think a lot of people are trying to think through.
50:50  Here's how I think about this which is, okay.
50:50  There's a well known corporate investing model which is GE or Honeywell.
51:00  You go to one of those places or Lockheed, and they've got their big programs, and
51:00  you either work the bureaucracy to do your program or not, and
51:00  that's where most of these big capital intensive things happen.
51:10  And then there's the venture capital model, and
51:10  the venture capital model is sort of staging the capital over time.
51:10  It's milestone based investing, but the milestones are kind of undefined, and
51:20  then you're going to a new investor in every round.
51:20  And so you've kind of got this constant financing risk, which says well,
51:20  what if I can't raise the next round of money?
51:20  And then you've got the problem with these capital intensive projects where it's like
51:30  well, I know it's going to take a half a billion dollars to get to market.
51:30  If I go raise $20 million from VCs today,
51:30  I have this massive funding risk of $480 million in the future, right?
51:30  And then from your standpoint,
51:40  can I even recruit the team to be able to build the product?
51:40  Because I can't even tell them that the money will be there.
51:40  And then Andrisa stood on stage and
51:40  told me about the $14 billion of negative yielding government debt, and
51:40  why can't I get somebody to give me the $500 billion to do the thing.
51:50  And so, the idea that's floating around is sort of an idea to borrow
51:50  an idea from actually a large public infrastructure.
51:50  And private infrastructure gets built, which is project finance,
52:00  which is basically like the way a dam gets built.
52:00  Which is basically the way a lot of airplane projects get run, new airplanes,
52:10  is basically, like the Silicon Valley model,
52:10  we're going to have a new idea, we're also going to have milestones.
52:10  Unlike the Silicon Valley model, we're going to have a very,
52:20  very crisp understanding of what those milestones are, right?
52:20  We're going to have a program management office that's at a level of sophistication
52:20  that a typically valley start-up doesn't have.
52:20  And we're going to chart out, in detail, how that money is going to get spent,
52:30  what the milestones are along the way,
52:30  how the organization's going to get built to do that.
52:30  And it's going to be precise, we're going to know week by week what
52:30  all the work is, and all the parts, and how they fit together.
52:40  And we're going to try to get to a much higher level of predictability
52:40  to build the thing than a typical Valley start-up would have.
52:40  And then in theory, you could then raise money, and this is sort
52:50  of where there could be a new asset class, and call it like, tech project finance.
52:50  Where you could basically say, okay you could go and you could raise $500 million,
52:50  and the $500 million wouldn't show up in your bank account.
53:00  Instead enough money to get to the first milestone would show up, and
53:00  then there'd be a checkpoint, and then if you hit the checkpoint you get the money,
53:00  right, and so the money unlocks over time.
53:00  And so, then you could think about having a sort of tech project finance
53:10  venture fund, where you could have like 20 projects like that start.
53:10  And then you could assume that like 10 of them fail along the way and
53:10  get liquidated, right?
53:20  And just get shut down before most of the money is gone in.
53:20  And maybe 10 of them work and
53:20  then the fund is big enough to fund 10 to completion, but not 20.
53:20  And you kind of play it like a venture portfolio.
53:30  So, that idea is kind of floating out there.
53:30  And by the way, this is what I'm describing.
53:30  This is kind of how Google X works a little bit, and this is how
53:30  some of the advanced RNB projects, and some of big value companies, operate.
53:40  And so, that idea's out there, but
53:40  it doesn't fully exist in the form that I just described.
53:40  And I think one of the really interesting questions for the future of tech is,
53:40  are we collectively going to figure out a way to do that, right?
53:50  because if we can't we can't figure out a way to do that,
53:50  then the project's you're describing either have to be these flukes, right?
53:50  You have a once in a generation Elon Musk, Internet Billionaire,
53:50  who decides to build a space company and a car company at the same time.
54:00  because who would do that?
54:00  >> [LAUGH] >> Or it's just going to be big companies
54:00  doing all this stuff, right, which would be a depressing answer.
54:00  Mark, thank you so much for your time.
54:00  We appreciate you being here.
54:10  >> Good, good, thanks everybody, thank you.
54:10  >> [APPLAUSE]