AI transcript
0:00:05 The content here is for informational purposes only, should not be taken as legal business
0:00:10 tax or investment advice, or be used to evaluate any investment or security, and is not directed
0:00:16 at any investors or potential investors in any A16Z fund. For more details, please see
0:00:17 a16z.com/disclosures.
0:00:23 Hi, everyone. Welcome to the A6NZ podcast. I’m Sonal. So this week, to continue our
0:00:28 10-year anniversary series since the founding of A6NZ, we’re actually resurfacing some
0:00:32 of our previous episodes featuring founders Mark Andresen and Ben Horwitz. If you haven’t
0:00:36 heard our latest episode with Stuart Butterfield turning the tables as the entrepreneur interviewing
0:00:43 them, please do check that out and other episodes in this series on our website at a6nz.com/10.
0:00:49 But this episode was recorded at our annual Innovation Summit in 2017 and features writer
0:00:53 Stephen B. Johnson interviewing them about everything from their relationship to creative
0:00:54 inspirations.
0:01:01 All right. I’m delighted and honored to be here with you. And we’ve got a lot to cover.
0:01:04 And what the kind of architecture for this conversation is, in a sense, we’re going to
0:01:09 kind of zoom out. We’re going to start on a more personal level and broaden out to think
0:01:13 a little bit about tech cultures inside a given organization, and then start thinking a little
0:01:18 bit more about broader social trends coming out of technology and looking into the future
0:01:21 a little bit. But I wanted to start with something actually just listening to your conversation
0:01:25 with JJ, who I don’t know at all, but I’m going to call JJ. He was talking about that
0:01:31 first kind of literally magical moment going and seeing Universal Studios and then getting
0:01:35 into magic and how that was so transformative as an eight-year-old. And it occurred to me,
0:01:39 do you guys have a memory of something like that with tech at any point where you really
0:01:44 saw something? For me, it was late. It was hypercard, sophomore year in college, where
0:01:49 I was just like, oh, there is this whole possibility that I hadn’t imagined could happen on a screen.
0:01:53 Do you have similar stories? It’s funny. This is an embarrassing question because I’m sitting
0:01:59 next to Mark, but one of the ones I remember most vividly was seeing Mosaic because for
0:02:04 years in tech, there were all these ideas about like if you were in computer science
0:02:10 about what was possible from all the things that you ought to be able to do, but you could
0:02:15 never actually quite get them to work. And hypercard was like that in that way, but Mosaic
0:02:19 was really it. It was all there on it. And when you downloaded it, you were like, oh,
0:02:25 my God, the whole world is like right there. I can reach the world. That’s the most craziest
0:02:29 thing ever. But I hate to say that with him sitting here because I go right to his head.
0:02:33 Well, it was a really striking point because up until, certainly for me, and I think for
0:02:38 a lot of people, there was discussion about hypertext that had been circulating through
0:02:43 different subcultures. But I would say probably 80% of the preceded received at that point
0:02:48 was strangely enough about hypertext fiction. It was people who were writing these nonlinear
0:02:54 stories. And when you saw Mosaic for the first time, you’re like, oh, this isn’t some obscure
0:02:58 avant-garde postmodern literary device. This is the future of media.
0:03:01 I have a much better answer than that. So I actually just mentioned on stage, but like
0:03:05 the early PCs really were the mystery box, the magic box and that really, that just,
0:03:08 you know, the flesh and cursor had me from go. So that sense of potential was a really
0:03:14 big deal. The other, I swear to God that this is true. Knight Rider, who remembers Knight
0:03:16 Rider, Knight Rider. There we go. Knight Rider outstanding.
0:03:17 You’re talking about kids?
0:03:22 Kids. Holy shit. So I was, I forget, I was, it was 82. So yeah, I was 10, right? And so
0:03:25 this shows on, and I don’t know, it’s this guy in the leather jacket. And I don’t know,
0:03:29 he seems cool, whatever. But they did this very clever thing, the mystery box thing.
0:03:31 And then there was no internet, no, nothing couldn’t find anything. You just saw a few
0:03:36 commercials. They did not tell you that the car was like that special. And if you go back
0:03:41 and watch the pilot, it’s like 45 minutes in. And like, it’s the whole thing has happened.
0:03:44 He’s been shot in the face. He’s had reconstructive surgery. He’s got the new name. He’s got
0:03:49 the mission. He’s got the car. He’s driving along 45 minutes in the car talks. And like,
0:03:52 I think I fell out of the couch. Like, I think I just like literally, I was like, the car
0:03:56 is talking. Right. And then I started to remember what that felt like. And then I have to remember
0:04:00 the screens, like the dash on that thing, right? It was like being in the space shuttle.
0:04:03 And to this day, when I get in a car, that, you know, the modern cars are like that, right?
0:04:06 They’ve got up to and excluding the fact that they talk to you now. But you know, they got
0:04:08 all the screens in the distance of that and the dash and the tuzzling, the whole thing.
0:04:14 It’s still, I always still feel like I’m getting behind the dash of kit. That is the best answer.
0:04:20 So there’s a great thing about your, the relationship that you guys have, it’s a long enduring one,
0:04:24 incredibly productive one. There’s a line in the hard thing about hard things in your book,
0:04:28 not to embarrass you, Mark, but I just wanted to quote it here. This is, you’re talking
0:04:34 about the relationship. And what you said is, even after 18 years, he upsets me almost
0:04:41 every day by finding something wrong in my thinking. And I do the same for him. It works.
0:04:48 So first off, is that true? But more than that, are you guys, is there something predictably
0:04:52 wrong? Are you guys wrong? And are you finding yourselves correcting each other in ways that
0:04:57 are kind of, are there patterns to the way in which you disagree? Do you tend to err
0:05:01 on the side, this side, where Mark errors on another side?
0:05:06 You know, I think it’s, you know, we’re close enough in personality, but different enough
0:05:11 kind of in skills that we often see things from different angles. And then a lot of it
0:05:16 is Mark himself, which is like, Mark always likes to take the other side of the argument,
0:05:21 whatever side, like he just enjoys taking the other side. That’s his thing. And so, you
0:05:26 know, it just kind of goes that way. I think that the real key to it is that we somehow
0:05:33 got to a level of trust where we can really go at it in a way that would, for most people,
0:05:38 you just go like, if you like, you can’t talk to me that way. Like how, you know, like so
0:05:42 disrespectful, like you’re stepping on me, you’re asking me these questions that hurt
0:05:47 my feelings. But you know, for us, you know, it has still like, you know, sometimes like,
0:05:51 get close to that, but not, not all the way.
0:05:54 I think the big thing is the thing I decided at a certain point, because we get asked a
0:05:58 version of this question by the founding teams that we work with, or if we bring a CEO into
0:06:01 a company, help a founder, bring in a CEO, and they’re going to have a partnership that
0:06:03 hopefully works something like this, you know, get kind of asked kind of, how do you make
0:06:07 it work? Because it is so easy for the conflict, for the emotion to, to drive people apart.
0:06:12 And so the way I think about it is, it’s more important to me that we have the successful
0:06:17 partnership than it is that I’m right on any particular issue. And I’m proud to say that
0:06:21 Ben, of course, is the exact opposite. It’s far more important for him to be right than
0:06:27 absolutely. And so it meshes perfectly right hand and glove. I’m joking. That was a joke.
0:06:29 And so we both will argue it all the way up, but each of us will defer to the other. At
0:06:32 the end of it, if it’s an argument, it’s over which one, which of us is going to defer to
0:06:37 the other one, with each of us volunteering to do it, say most of the time.
0:06:42 And that’s really like, sometimes the argument will not resolve, but we’ll kind of know
0:06:48 who knows more about that thing. And we’ll yield in that way. And that’s been super productive.
0:06:53 And there are ongoing disputes about where the technology world is heading. Are there
0:06:57 kind of senses like, oh, no, you think this thing is going to be huge, but this is the
0:06:59 old argument we’ve been having for five years. It’s never going to happen.
0:07:02 Well, we both believe a lot and disagree and commit, right? And so it’s important. Like
0:07:05 as an example, one version of the question you asked is like, what if we’re arguing about
0:07:08 some startup we funded? And whether it’s, you know, we’re going to have some argument
0:07:11 about like that was a mistake or not or whatever. Like we basically, I don’t think ever have
0:07:15 those arguments. And the reason is because we may argue whether, and this is true of
0:07:18 our partnership or broadly, we may argue about whether to make the investment, but once we
0:07:21 make it, we’re in. And then at that point, it’s important that it’s the dynamic sort
0:07:25 of implicit promise in the team and including between the two of us as we’re all in this,
0:07:29 we’ve all committed. And I think that’s really critically important because that’s how you
0:07:32 maintain, that’s how you don’t have, I told you so.
0:07:35 And backbiting and talking about people when they’re not in the room and that kind of thing.
0:07:36 That’s just bad.
0:07:40 Do you all have a, I’m actually in the middle of writing a book about long-term, complex
0:07:46 decision-making. So I have my own kind of bias in this question. But do you have, when
0:07:48 you’re confronting a decision to say, for instance, like should we fund this company
0:07:52 or should we follow in this round or other life decisions, do you find that you have
0:07:59 a process for that decision-making act that you go through and think about as a series
0:08:04 of stages? Or is it something that’s more fluid and conversational and intuitive?
0:08:09 Yeah. So it’s interesting. This business is different than our last. So running a company,
0:08:16 you try to be more structured in how you do this. In some ways, in that speed is really
0:08:22 important. So if you’re running a company, your output is decisions and you rate it
0:08:27 on quality and speed. And if you have to make the trade, which you always have to, you generally
0:08:32 go towards speed because you have a lot of decisions to make. And if you don’t make them
0:08:39 fast, then you freeze the entire organization. In our new business, basically quality is everything.
0:08:45 And so we’ll go around the horn 50,000 times if we have to to make sure that we’ve explored
0:08:51 every corner and every crevice of the discussion and we’ve not missed something. So I would
0:08:57 say in some ways, we have a lot of a framework in our minds about how we think of investments
0:09:03 and deals and so forth. But we’re willing to go in many loops where we would never do
0:09:04 that in a company.
0:09:08 One of the things that I love investigating and talking to people about is their kind
0:09:13 of creative workflow and where they find inspiration. There’s a lot of research out there that some
0:09:16 of which that I’ve done and other people have done about the importance of kind of diversity
0:09:22 of influences in your kind of worldview, leading to more creative thinking. So I’m just curious
0:09:27 about your kind of daily information diet in a sense, beyond the kind of the routine
0:09:32 of the meetings that you have with the founders and the pitch meetings and so on. Where do
0:09:36 you find that kind of outside influence in new ideas?
0:09:41 So we sort of cheat in a sense, which is we have, we see 2000 inbound startups a year.
0:09:44 These are by definition and 2000 are the smartest people in the world in all the domains that
0:09:48 they’re operating in. And so, I mean, honestly, after that, it’s just, it’s hard to pick up
0:09:52 like a magazine and open it with any level of enthusiasm because it’s like, you know,
0:09:54 you kind of have this, you know, you’re kind of seeing the stuff months or years before
0:09:58 it shows up in the press. And so that’s part of it. Personally, I’ve been running this
0:10:02 year a big experiment and I’ve always been a big reader and sort of information on the
0:10:05 board. And it just, you know, I’ve always tried to kind of balance short term, long term,
0:10:09 you know, different kinds of different time horizons of material, different kinds of material.
0:10:12 So I’ve been running a big experiment this year, which is I’ve been trying to do a bar
0:10:17 bell. I’ve been trying to polarize it. And so I’ve stopped completely reading newspapers,
0:10:22 magazines, basically anything that has a time horizon, basically greater than let’s say
0:10:26 five minutes to, you know, anything basically between five minutes and five years, which
0:10:30 is to say I basically only read social media on the one hand and then only books on the
0:10:34 other hand, right? And just polarize it and gap it way out. So what’s interesting about
0:10:38 that is of course, being on social media like that process, you know, necessarily you end
0:10:42 up consuming a lot of news and that a lot of what’s there notwithstanding the false
0:10:46 reports of the death of the web, a lot of what social media is, is links to things that
0:10:49 are interesting, right? People who you’re following are interested in. And so, you know, I do
0:10:52 end up reading basically everything. But one of the experiments was, does it matter? Like
0:10:55 if you don’t see the homepage of the newspaper, do you miss things? And it turns out if you
0:10:58 follow the right people, you really don’t, because they surface all the interesting
0:11:01 stuff anyway. And you get to see a lot of stuff that you wouldn’t necessarily see looking
0:11:04 at the homepage. But the other side, honestly, and you know, you’re accomplished book author,
0:11:08 the other side of it is just books, you know, books that probably become the great underestimated
0:11:12 source of information relevant to our daily lives that just gets, you know, as there is
0:11:17 just such a surplus of kind of near term information and consumption. And let’s just say, as the
0:11:21 real world is getting continuously more interesting in real time, you can spend all day long just
0:11:24 following the ins and outs of what’s happening in the political scene or what’s happening
0:11:26 in the sports scene or what’s happening in, you know, the business world or whatever.
0:11:29 And so you can really get, you know, let’s talk about myself, I can get really trapped
0:11:33 in the present. And so the ability to at least have some time to be able to go back and be
0:11:37 able to read things that were written five or 10 or 50 or 100 years ago, that have stood
0:11:40 the test of time in the form of books has been I think is very valuable.
0:11:43 It has been very interesting. I mean, the book business is actually quite healthy and
0:11:47 people are reading, you know, reading print books, there’s a kind of return to print books.
0:11:51 And it does feel as if I think one of the things you don’t realize until you write them,
0:11:54 particularly with nonfiction books, but it’s true fiction as well that when you meet someone
0:12:00 who’s read one of your books, they have been living inside your mind for 12 hours, 20 hours,
0:12:05 depending how long the book is. And so it is still an unrivaled way to get complicated
0:12:09 ideas into other people’s minds. And so it’s been, I think a sign of health in the culture
0:12:13 of that books are actually thriving in the midst of all this kind of minute by minute
0:12:14 social media.
0:12:16 And also, by the way, as you well know, like audio books, right, I think there’s a renaissance
0:12:21 in audio books, which is just having the smartphone and now the wireless, you know, ear pods makes
0:12:26 it so much more convenient for your content, long form audio content. And podcasts, obviously
0:12:29 are a big part of that. But audio books in the course is drive time and wait time and
0:12:32 this time and, you know, morning time and so forth completely fit into my life in a
0:12:34 way that books didn’t use to.
0:12:38 I also wanted to ask you, Ben, about music, can you talk a little bit about that in terms
0:12:40 of your own kind of creative view of the world?
0:12:44 Yes. Well, it’s interesting and it’s very specific to hip hop for me and hip hop is an
0:12:49 unusual music form in that it’s a very kind of capitalistic form of music, which is completely
0:12:55 kind of unheard of in popular music. And that the main theme of hip hop, if you go through
0:12:59 all the great rappers is like, how do you build something out of nothing? You know, how do
0:13:03 you compete these kinds of things as opposed to R&B, which was maybe love songs and like
0:13:09 rock and roll, which is more communist. But it’s perfect. It’s a perfect analog to entrepreneurship.
0:13:13 It’s kind of the exact kind of motivational soundtrack for entrepreneurs. And that’s really
0:13:19 how I started with it, because any theme I wanted to write about, like it was a great
0:13:26 way to find inspiration. But it led to, if you say I made a contribution to the management
0:13:33 literature, it actually came out of rap music in that the big thing that was different in
0:13:38 my book was that the logic of management is not very complicated. Then you can understand
0:13:44 all the management theory. It’s just not that hard. But the emotional, psychological complexity
0:13:49 of doing it is incredibly difficult. And you know, we see tremendous fallout from brilliant,
0:13:55 brilliant people who can never get over that. And so the big challenge for me was like,
0:14:01 how do you communicate the emotional part of the lesson? And hip hop is great for that
0:14:06 because it carries the emotion. And it’s all about kind of the capitalism. So I wrote
0:14:12 a post, how do you handle politics in a company? And I went through all the things that cause
0:14:16 politics and the subtle things, like how somebody asking for a raise can do it and how you deal
0:14:22 with that technique and so forth. But a lot of it is the attitude of the manager. And
0:14:27 so the rap quote that I used was Rick Ross, who do you think you’re fucking with? I’m
0:14:33 the fucking boss. And like, once you get that, then you know how to do it. That’s great.
0:14:39 Okay, so let’s zoom out a little bit now. You were asking JJ Abrams about LA as the
0:14:43 kind of epicenter of the movie business. So with all the changes that we’ve seen in the
0:14:50 tech sector and all the volatility, the one constant really for half a century has been
0:14:55 that the Bay Area and Silicon Valley have been the epicenter of the technology world
0:15:00 really without any near arrival, probably for 50 years, I think it’d probably be fair
0:15:03 to say, despite the fact that it has gone through all these different revolutions and
0:15:08 you had big computers and then personal computers and then the web and then social media. So
0:15:13 really two questions I think, why, why they are, like what was it about that particular
0:15:18 configuration that rooted tech in that world? And do you think we’re going to look back
0:15:21 in 30 or 40 years and it’s going to have the same concentration?
0:15:25 Yeah. So the why, so the why is I think it’s history, right? And so just the fact that
0:15:28 it’s been a network effect, right? It’s been a snowball rolling down the hill, picking
0:15:32 up momentum now for 56, actually turns out 50, 60, 70, 80 years. A lot of ways it goes
0:15:37 back to the 1920s, 1930s, the early defense contractors. Steve Blank has a whole series
0:15:41 of videos called the secret history of Silicon Valley. He traces it all the way back almost
0:15:42 a hundred years.
0:15:43 Fantastic.
0:15:46 Fantastic shares. And the point of it is, it’s just, it’s this kind of network effect
0:15:49 that’s just kept rolling, right? And so it’s been this place where it’s just like, it’s
0:15:53 the place where the next really smart engineer programmer or, you know, equivalently salesperson,
0:15:58 marketing person, west door contact, whoever they are, finance person on the margin, right,
0:16:01 is more tempted to move to the valley than many other places, which isn’t to say that
0:16:04 there aren’t many capable people all over the world. It’s just on the margin. Many of
0:16:07 the ones who are super ambitious end up at the valley. And of course, I’m an example
0:16:10 of that. And as a consequence, right, it’s a story of imports, right? And so another
0:16:14 thing just to read, I’m sure if people are interested, Tom Wolf, the great novelist,
0:16:18 journalist wrote a piece in the 80s in Esquire about literally Bob Noyce, who was the original
0:16:22 CEO of Intel, one of the fathers of Silicon Valley and literally grew up in Iowa, grew
0:16:26 up in the Midwest and was the Silicon Valley import. And actually Wolf ascribes a lot of
0:16:31 modern Valley culture to literally Bob Noyce importing, interestingly, Midwestern culture,
0:16:35 right, including, by the way, egalitarianism, right. So the whole open floor plan thing,
0:16:39 stock option ownership, everybody owns a share in the company. He traces that actually back
0:16:43 to Midwestern culture. And so it just got established and it developed this ethic and
0:16:46 it’s probably not an accident that it’s the frontier, right? It’s probably not an accident
0:16:49 that this sort of gold rush happened, right? It’s just kind of this frontier out the mentality
0:16:53 has continued. So that’s the good news, right? The bad news is, as I discussed with JJ, like
0:16:56 it’s just number one, we’re just bursting at the seams, like it’s just become a hard
0:16:58 place to do business. And the number two is there’s great people all over the world and
0:17:02 like why on earth? So the joke in the valley is, you know, help wanted, right? Software
0:17:06 company puts up Silicon Valley, software company puts up a help wanted out on the internet
0:17:09 or whatever and says, you know, help wanted, you know, software engineer to work on new
0:17:14 collaboration software tool, online collaboration software tool that will enable people to work
0:17:19 together independent of geography all over the world. So in real time, PS must relocate
0:17:22 to San Francisco to apply. And so it’s this weird incongruity, which is we’re building
0:17:26 the technologies that in theory should let this stuff spread. And yet for some reason
0:17:30 in the last 10, 20 years, it’s actually been concentrating more and more. And so I’ve come
0:17:33 to believe it’s a maybe this is obvious to some people, but I would come to believe it’s
0:17:38 a human dynamics question. It’s a psychology, sociology question, not a technology question
0:17:42 in a lot of ways, which is just like how do people best work together, right? And it just
0:17:46 so happens that at least for the form of traditional companies, which you just see over and over
0:17:50 again is just when you can get everybody in the same room physically in the same room,
0:17:53 right, with the level of, say, fidelity of communication interaction where we’re sitting,
0:17:57 you know, it’s why, by the way, it’s why we’re all physically here. And there are a few successful
0:18:00 distributed companies, but there really aren’t very many as a consequence of that. And so
0:18:05 my hope is that we’re going to get there in the next, you know, let’s say 10 or 20 years,
0:18:09 my hope is that we’re going to get telepresence, right, in the form of video conferencing and
0:18:14 telepresence robots and VR and AR and all these things to collaboration software and
0:18:17 work group software and Slack and GitHub and all these amazing technologies are building
0:18:20 for collaboration. My hope is we’re going to get it to the point where it’s just going
0:18:23 to be obvious that we don’t all have to be in the same place. If that happens, you could
0:18:27 say it’s quote bad for the Valley in the sense of like maybe Silicon Valley is not central
0:18:31 anymore, but it would be so good for the world for that to be the case and we would all benefit
0:18:34 so much from that. I think it’s a very worthwhile thing to pursue and something I’m very fired
0:18:35 up about.
0:18:39 How much do you think, just to go back to the point about noise in the early days of
0:18:43 Silicon Valley and the history of it, to have written about this a little bit as well, how
0:18:51 much do you think that the participatory option granting culture, which is very different,
0:18:54 there were very few kind of East Coast firms that were doing that. So you had much more
0:19:00 traditional kind of top down equity systems in those corporate entities. How much do you
0:19:03 think that is part of the success of Silicon Valley? This is something I think that would
0:19:06 be interesting to go back and look at just economically.
0:19:10 So I think it ends up being very important because of the nature of technology companies.
0:19:18 So if you look at, there are other kinds of companies where the people are much more interchangeable
0:19:23 and this kind of gets into why the network effect is so important and so forth. And in
0:19:29 like a tech company, there’s lots of people who are extremely valuable and that innovation
0:19:36 as a way to get them their kind of proper compensation for their contribution, the great
0:19:39 conversation with Mark and Charles Koch, where he talked about like, you have to be
0:19:43 rewarded for what you contribute to others. And that really is key to any business and
0:19:47 any incentive system. And particularly in technology, because there are so many people
0:19:53 in the company who are so valuable and so fundamentally critical to the company’s success,
0:19:58 it really is one of a very few kinds of compensation systems that would work. And certainly, you
0:20:02 know, a lot of the systems on the East Coast would never work for tech companies to be
0:20:03 kind of world-class competitive.
0:20:11 So it’s been six years since Mark, you wrote the software eats the world essay. I went
0:20:15 back and looked at it and reread it. It was a great piece. It reminded me of, I’m sure
0:20:18 a lot of people have seen this. There was a great thing that was circulating on social
0:20:24 media a couple of years ago. It was an old kind of single page flyer for Radio Shack from
0:20:29 like 1988 or something like that. It was a list of like 30 products that Radio Shack sold.
0:20:36 And the answer machine was, you know, a VCR, an alarm clock, like a TRS 80 kind of descendant,
0:20:39 you know, a game console, something like that. And literally without exception, every single
0:20:43 one of them is now an app on your phone, right? The whole thing had gotten swallowed up by
0:20:48 software, which is of course a measuring productivity problem because all those things
0:20:55 in aggregate cost $30,000 in 1988. And now they’re free on a phone that costs $600, which
0:20:57 is actually progress, but doesn’t sometimes look like it.
0:21:03 So obviously I think that that was a very prescient forecast to make. Has anything kind
0:21:08 of surprised you six years later looking back on it? I mean, in it, you say the next big
0:21:12 stages are health and education. And I’m wondering, you know, particularly on those fronts, has
0:21:14 it lived up to the kind of promise you saw back then?
0:21:16 Yeah, they’re sort of the overall concept of software eats the world. But then there
0:21:20 was a specific framework that I proposed in the piece, which is sort of a weak form of
0:21:23 semi strong form and a strong form of this hypothesis, right? And so the weak form was
0:21:28 every product that can write every physical product will become a software product, right?
0:21:31 And that’s that’s exactly your radar check example. Things go from being physical products
0:21:37 to being apps. The second sort of semi strong version of that was therefore any company
0:21:41 that makes a product that can be turned into software will itself therefore have to become
0:21:44 a software company. Right. And in fact, I was thinking you could you could see this thing
0:21:47 for example, playing out right now in the car industry, right, where all the car companies
0:21:50 are spinning up software efforts, they’re buying software companies are spinning up
0:21:53 software and as fast as they possibly can because they see what’s coming with autonomy
0:21:58 and all these other software advances. And then the strong and sort of audacious slash
0:22:03 ambitious slash arrogant hubristic version of the thesis is in any industry as a result
0:22:07 of this dynamic in the long run, the winning company in the industry will be the best software
0:22:12 company, right, which is a provocative statement, right? Because in a lot of these industries,
0:22:16 and again, cars are a great example. You have incumbents who are really good at making cars
0:22:18 trying to become great software companies. And then you have great software companies
0:22:22 that have no idea how to build a car, right, who are going to start who are going to start
0:22:25 making cars, right? And then you’re going to have basically, right, this giant collision
0:22:29 between companies coming from two totally different backgrounds. And so I think that
0:22:34 you’re seeing lots of that first stage that week stage, lots of products transitioning,
0:22:37 you’re seeing lots of companies becoming software companies. I think we’re just entering in a
0:22:41 lot of industries were entering that Thursday’s where there’s this very interesting structural
0:22:44 battle that’s forming up. The other thing I says, yeah, I think you exactly nailed it
0:22:48 with healthcare and education, right, which is there are these giant sectors of the economy
0:22:53 in which not only is there no productivity growth, like overall in both healthcare and
0:22:57 education, there is no measured growth, there is no measured results in the application of
0:23:01 technology in those fields. And in fact, probably it’s negative productivity growth, right?
0:23:04 Like the typical university has been going backwards in productivity, right? You just
0:23:07 look at the charts, the number of administrators that they hire, right, per student is just
0:23:13 skyrocketing and that is literally negative technological productivity. And so those industries
0:23:18 are extremely enticing to Silicon Valley, because they’re so big, they’re gigantic.
0:23:22 Healthcare, healthcare is a sixth of the American economy, right? And left unchecked, it will
0:23:26 become a fourth and then a third and then a half and then two thirds and then three quarters.
0:23:30 Like it’s just left unchecked, it’s just going to keep growing. And so it’s so much money.
0:23:35 It’s so big. It’s so important. It’s very enticing. And the incumbent structure of there’s many
0:23:38 smart companies in that industry, but the incumbent structure of how the industry works
0:23:42 is just, is wired to go the wrong direction. And so there’s this huge opportunity to insert
0:23:46 into it, which obviously we’re going after hard, but that’s still like super early.
0:23:47 Yeah.
0:23:51 And education, what, Ben, do you have thoughts on that front? I mean, there’s this interesting
0:23:57 point we’re at where there seems to be a growing backlash to the presence of screens, particularly
0:24:02 in younger kids’ school classrooms that it hasn’t lived up to the potential. And maybe
0:24:06 the kids already have too much software in their lives as it is.
0:24:13 So, you know, it’s funny, or it’s not funny. It’s sad that we’ve not applied technology
0:24:19 that well. And a lot of it has to do with the kind of structure of the kind of political
0:24:22 regulatory structure of schools. And we have a company, Udacity, that’s worked hard on
0:24:27 this. And their final conclusion was to kind of run outside of the school system, but it’s
0:24:32 very powerful. I’ll tell you a quick story about that. But, you know, obviously, very
0:24:37 obviously, if you could have like, any teacher or the best teacher in the world teaching a
0:24:44 math class, if students have to study and then be tested, like, when do you take a test outside
0:24:49 of school, like, ever in life? Like, what the hell skill is that does this create like tremendous
0:24:53 anxiety and like give people complexes. But you ought to, with technology, you ought to
0:24:57 be able to measure how people are learning every step of the way, give them harder problems,
0:25:02 if they’re going very fast, or get them help if they’re going slow. And there’s a lot of
0:25:07 things that ought to be able to be done. But then I think the more kind of pressing thing,
0:25:13 and the thing that Udacity really addresses is the four year education, general education,
0:25:18 doesn’t work that well in the modern economy because people are switching careers very,
0:25:23 very often every, you know, two, three years sometimes. And, you know, like four years,
0:25:26 and then you never go back to school for the rest of your life doesn’t make any sense at
0:25:31 all because people need to get retrained jobs get displaced. And so what Udacity has come
0:25:36 up with is this thing, the nano degree, which is two months, three months, you can learn
0:25:42 to program an Android phone or build a self driving car, or learn to do technical marketing.
0:25:47 And those degrees are connected right to the job market. So you can roll right in with
0:25:53 a skill and a certificate that says you understand the material and you’re ready to work. And
0:25:58 that is a great innovation and something that we’re really excited about. And just quick
0:26:03 story on that. So one of the huge problems we have in this country is prison and the
0:26:08 need for prison reform because we’ve got, you know, 75% recidivism rate where people
0:26:12 who go to jail and come out, go back to jail. And the reason they go back to jail, they
0:26:16 can’t get jobs. And the reason they can’t get jobs is because two things. One is we’ve
0:26:23 outlawed college in prison and then two, once they come out, their record follows them wherever
0:26:30 they go. So, you know, I’ve got a friend who came out of jail and I said, go to Udacity.
0:26:35 He goes to Udacity and he’s coming up on his technical marketing degree and he’s already
0:26:38 got job offers. And it’s like, that’s what we need.
0:26:43 Yeah. And I think it’s almost as if school, particularly high school, and I have two kids
0:26:47 in high school, so I think about this a lot, it’s kind of trapped in this middle zone that
0:26:50 doesn’t really work in a sense. It’s much more effective to have those kind of nano
0:26:55 skills, right, where you can actually kind of apply them or the skills should be broader,
0:26:59 right? I mean, when you read through, again, a book like the hard thing about hard things,
0:27:04 I just think about how there are so many skills in there that no one ever thought to teach
0:27:08 me in high school, right? I mean, the skills about decision making skills about kind of
0:27:11 emotional intelligence, dealing with, you know, difficult decisions. My kid actually
0:27:15 in his high school, to its credit, is doing a kind of design thinking class. And they’re
0:27:20 basically learning how to brainstorm ideas, interview a customer, think about different
0:27:24 possibilities, do mock-ups. And it was like, this should be the default. This should not
0:27:27 be an elective. This should be the thing you learn. And then if you want to go off and
0:27:33 do advanced chemistry or do advanced calculus, that’s fine. But those types of skills that
0:27:36 are just, everyone is going to have to know on some level, but it’s very rare to encounter
0:27:38 that. We’ve got a very dated curriculum. There’s
0:27:42 no question. I ran on the board of trustees at Columbia. And there are certainly people
0:27:47 who are going to go to like an elite school and become a scholar or a PhD. And I think
0:27:51 the system works reasonably well for them. But for, you know, the kind of bulk of the
0:27:57 population who goes to college to get into the workforce, it’s really difficult. It’s
0:27:59 exactly, as you say, it’s kind of neither here nor there.
0:28:04 Let’s talk a little bit then, kind of segueing a little bit to the job and automation question
0:28:09 anyway. In general, I think we all agree that there has been this growing and now kind of
0:28:16 reaching Crescendo backlash against big tech and the tech sector that the last year has
0:28:20 particularly brought to the fore. And I feel it very strongly going back because I live
0:28:24 part of the time in Bay Area and part of the time in New York, when I’m back in New York,
0:28:29 you know, nine out of 10 kind of opinion like pieces written in these media are negative
0:28:30 pieces.
0:28:31 It’s only nine out of 10.
0:28:34 I mean, so I want to get into some of the specifics about why that is happening, how
0:28:39 you guys feel about it. But how much in general do you and how much recently have you found,
0:28:42 do you find yourselves taking that seriously and how much do you feel that people just
0:28:44 don’t understand what’s going on here?
0:28:45 We might give two different answers.
0:28:50 Yeah. So I would first say there’s a huge difference between what gets written in opinion
0:28:56 pieces and the actual opinions of the public. So if you look at approval ratings of tech,
0:29:01 they’re incredibly high. Like they’re the highest of any industry. And like Amazon’s
0:29:05 approval raising, which is one of the biggest targets is like 80. Whereas Congress is like
0:29:10 20 and the press is like 20. And so like the guys at 20 are saying the guys at 80 need
0:29:14 to be stopped because everybody hates them. So there is that dynamic. And I think it’s
0:29:15 very real.
0:29:19 This is the concept of false consciousness, right? So literally the whole problem with
0:29:22 the communist revolution was the business weren’t signed up for it. And so the intellectual
0:29:24 leaders were like, well, but we got to take down the capitalist.
0:29:29 The other thing is, I think there’s something else going on that this is a side effect of.
0:29:34 And I think it’s the rise in the last several years. And in particular, after the 2008 crisis,
0:29:38 credit crisis crash, I actually think was the catalyst for a lot of this. It’s the rise
0:29:42 of zero sum thinking in both economics and in politics. Let’s say zero sum as opposed
0:29:46 to positive sum, right, which is this is sort of game theory, right? Zero sum game is I
0:29:50 win, you lose. And by the way, if I’m winning, it must mean that you’re losing because it’s
0:29:53 zero sum. It’s only a question of how we slice up the pie, right? Whereas positive sum is
0:29:57 we can all win together. It’s actually a great book called finite and infinite games that
0:30:01 actually goes through. If you go back historically, basically, economist philosophers and so forth
0:30:06 thought the politics and economics were zero sum. And there were huge battles over resources.
0:30:09 And this was colonization, all these other horrible things that happened over years were
0:30:13 fought through mercantilism, trade wars, right? All these things were fought based on zero
0:30:16 sum. And about, you know, 300 years ago, Adam Smith and a whole bunch of other really smart
0:30:20 thinkers figured out, no, you can actually gain from trade and you can actually interact
0:30:24 with more people and it’s good for everybody. And politics can be positive some just because
0:30:27 I’m doing well might mean that you’re also going to do well because again, we’re able
0:30:30 to culturally trade, we’re able to educate each other, we’re able to, you know, contribute
0:30:33 each other’s thoughts, and we’re all able to succeed. And so in the wake of the credit
0:30:36 crisis, I think zero sum thinking kind of came snapping back. And what’s interesting
0:30:40 is you see that on both the political left and on the right, right? For the anti attack,
0:30:42 the bloodite sometimes to come out of the left and Marx actually was shot through with
0:30:45 with leadism, like that’s one of the things he didn’t understand was the positive sum nature
0:30:49 of productivity growth. And anyway, so you get that on the left, you also get it on
0:30:53 the right, right? And you get it on the right, you get in the form of populism, right, which
0:30:59 in the form of opposition to trade and opposition to immigration, right? And so I just think
0:31:02 as a culture as an economy as a country right now, if you think that the formulation is
0:31:08 zero sum, you will then do things that will cause it to get worse. For example, on the
0:31:11 right, you’ll want trade barriers, right? And so you’ll want to cut trade under the
0:31:14 theory that that will make your people better. In reality, cutting international trade makes
0:31:19 your people worse. You’re dividing up a smaller pie. Yeah, you’re shrinking the economy for
0:31:21 everybody for no reason other than that you’re just mad at other people because you think
0:31:24 it’s their fault that you’re not doing well. And so it’s zero some thinking. And then on
0:31:28 the left right now, it’s this anti tech sentiment where like if those tech people are doing well,
0:31:31 then somebody else must be suffering, somebody else must be eating it. And it’s just it’s
0:31:35 the same sort of extremely reductionist thinking. And of course, the risk is as that sentiment
0:31:39 builds that at least a policies that actually impair the ability to be able to make progress,
0:31:43 make progress in the economy, make progress with productivity growth, make progress with
0:31:46 job creation, make progress with wage creation. And so there’s a pretty big risk that this
0:31:50 is all gonna go pretty seriously sideways for the wrong reason. Right. Let’s take the
0:31:55 tech backlash argument from a slightly more maybe sympathetic level, which is critiques
0:32:01 that have come from within the tech sector that the original vision of the web that inspired
0:32:07 so much of us, which was going to be this decentralized platform that was going to distribute
0:32:11 the kind of power of self publishing and voice to far more people. And it was going to kind
0:32:17 of topple this big, heavy, top heavy mass media model. That’s what inspired a lot of people
0:32:21 to get involved in it in the first place. At the end of that process, we’ve ended up
0:32:26 with, you know, four or five companies that in terms of their command over people’s attention
0:32:31 probably are the most powerful companies that have ever been on this planet and also some
0:32:37 of the greatest concentrations of wealth. So inside the tech sector, people say, re decentralized
0:32:41 the web and then we need to look at technologies that will enable us to have, you know, a more
0:32:46 even distribution in terms of the companies in terms of people’s attention and so on.
0:32:49 And blockchain is part of that. There’s some argument that people have been making along
0:32:54 those lines. How sympathetic are you to that side of the case, which does align with some
0:32:58 of the critiques that big tech is too big that are coming from people outside the tech
0:33:02 sector? Yeah. So there’s a technical argument for a decentralization. And then there’s
0:33:08 the kind of other thing that you’re getting at, which is should there be some like policy
0:33:14 answer to the big tech companies? And I think that, you know, you have to be very careful
0:33:19 there and look at specifically what’s going on. Well, are they kind of harming? Are they
0:33:26 suppressing innovation? So do people like us no longer want to fund anything because,
0:33:30 you know, Facebook or Amazon will wipe it out. And if you look at the numbers, there’s
0:33:35 probably more startups than there have ever been. And what we’re seeing and what we’re
0:33:41 funding is like super interesting. And, you know, for the most part, isn’t existentially
0:33:47 threatened all the time by those companies. Once you introduce policy, the potential side
0:33:54 effects are, you know, really scary, cronyism, corruption, the people who have the best relationship
0:33:58 get the best deal and these kinds of things. And that has knock on effects that are very
0:34:02 difficult. And, you know, if you compare it to the early nineties, when Microsoft was
0:34:07 super strong, that was really actually a far bigger suppression of innovation. There was
0:34:12 way less venture capital. There were far fewer companies being created. But like the technology
0:34:16 took care of it over time. And I think technology is changing at a faster rate now than it was
0:34:20 then. And there’s blockchain and there’s quantum computing. And there’s many technologies
0:34:25 on the horizon that could rejigger the playing field, you know, without a policy intervention.
0:34:30 Another question about the blockchain possibilities, you know, I’ve been really enjoying reading
0:34:36 Chris Dixon writing about this over the last year or two. And there is really an interesting
0:34:44 new way of incentivizing and compensating people both inside a technical organization associated
0:34:49 with an open protocol, early users of the service where all of those people are participating
0:34:54 in the value that’s created with it. And thinking back to the early stock option participation
0:34:58 of noise, you know, I wonder whether this, this suggests maybe that there’s a new model
0:35:03 here that might be as revolutionary as those kind of option plans were.
0:35:06 So the good news is the tech industry has had two models for making forward progress.
0:35:10 One has been what you might call pure capitalism, which is corporations, right, which is sort
0:35:14 of C corporations, employees, stock options, all the things we can take companies public
0:35:17 with that traditional structure. And then there’s been this other structure all the
0:35:21 way over on the ideological spectrum, right, which is open source, right, which is basically
0:35:24 the tribe, right, of developers that are interested in having something happen, coming together,
0:35:28 by the way, geographically distributed all over the world in a lot of cases, right, and
0:35:31 great examples, Linux and the web itself is an example of this and so forth. Actually,
0:35:36 the Internet, TCPIP was an example of this, right, or the new project MIT was an example
0:35:40 of this and people, technical people coming together and volunteering, right, literally
0:35:43 with metaphors like barn raising, right, it’s just like come together and make sort of breathe
0:35:47 life into these projects without a financial incentive and generally without, you know,
0:35:51 at least direct financial rewards. So sort of polar opposite of corporations you can
0:35:58 get. Blockchain is the first new third thing in, I don’t know, probably 40 years, right,
0:36:01 free software open source is like 40 years old. It’s the first new structure in 40 years
0:36:05 and it’s an interesting one because it’s a hybrid. It’s got the, it’s your point, it
0:36:08 has the decentralization of open source, right. These are protocols. These are things
0:36:12 that run Internet wide. These are things that are not necessarily developed by a team of,
0:36:15 you know, 100 people in a building in the Bay Area. They have that kind of open source
0:36:20 characteristic to them and they are decentralized. Like their protocols are inherently decentralized,
0:36:25 but they’ve got capitalism wired in. They’ve got money wired in, right, right, into the
0:36:29 protocol, right, in a way where there is a direct reward and incentive for the people
0:36:34 who actually create the thing. There’s a reward and incentive for the people who use the thing.
0:36:37 And then there was a reward and incentive for the so-called miners, the people who actually
0:36:40 run all the computers all over the Internet that make these things work. And it’s just
0:36:44 been so fascinating to watch because this is one of those kind of moments where people
0:36:47 walk up to this idea. And if they walk up to it from the right, they’re like, what on
0:36:52 earth is this decentralized hippy, like what on earth are you people doing? If they walk
0:36:55 up from the left, they’re like, Oh my God, it’s got money in it. It must be evil, right.
0:36:58 It’s sort of this weird, you got to kind of wrap your head around it. And so what we see
0:37:01 is like, it is fundamentally a third model for innovation. And I will also say this,
0:37:06 the thing that we see that I think maybe other people are missing, many of the smartest programmers
0:37:13 and mathematicians and economists and theorists and systems builders in the world and photographers
0:37:17 in the world are obsessed with this, like they’re just magnetically drawn to it, not
0:37:20 because of the money or this or that or the hype or whatever, because of the technical
0:37:23 innovations that are underneath this that are making this possible and what can come
0:37:26 out of this. And we just think like that’s the most positive sign you can possibly see.
0:37:30 We just have about five minutes left. So I want to just cover a couple of other giant
0:37:36 topics, artificial intelligence and the superintelligence debate. Can we solve that in about two minutes?
0:37:41 Can you give me, is this a legitimate concern? Is it appropriate to be worrying about the
0:37:47 threat from superintelligence now? Of the really scary things in technology. I would
0:37:51 have that one pretty low on my list. I mean, I think that one, like, I think it’s a little
0:37:56 bit of a miss, you know, intelligence is a funny word, right? Like, what is intelligence?
0:38:01 And it’s not one dimensional. And there are a lot of things that we have considered intelligence,
0:38:05 like doing hard math problems. Computers are already more intelligent, like playing chess.
0:38:09 Computers are already more intelligent. But there’s a lot of dimensions of intelligence
0:38:15 that computers are nowhere on. And AI, nobody is demonstrating anything in AI that says
0:38:20 like it’s going to get comprehensively more intelligent and certainly nothing along the
0:38:26 lines of free will yet. So yeah, maybe, maybe it’ll happen. But of all the things, it’s
0:38:30 a very theoretical. So I think it’s a little overblown. I do think also that there’s a
0:38:34 motivation of technologists to, it’s a very kind of, it makes you seem very intelligent
0:38:38 when you can talk about the robots taking over the world. So it’s a great thing to talk
0:38:42 about. The thing that drives you bananas is it’s the freaking physicists. And it’s like,
0:38:47 I’m a computer scientist. I don’t have like crazy conspiracy theories about black holes.
0:38:51 You know, I guess I could, you know, like in theory, a black hole could open up here in
0:38:55 this room and swallow us all. Like, I don’t have crazy theories about dark matter. Like
0:38:57 I’m not worried there’s dark matter in the glass. I’m not going to go around telling
0:39:00 everybody it’s going to, it’s just like, I don’t know why.
0:39:04 Yeah, it’s hard to find an AI expert who goes, Oh, yeah, this is a big problem.
0:39:07 Well, in fact, in the AI experts, of course, tend to be worried about the opposite, which
0:39:09 is they’re like, Oh, shit, expectations are getting set.
0:39:12 Like, we’re never going to build that. We’re never going to build the robot apocalypse.
0:39:16 I’m still trying to get the thing to play Mario Brothers, right? Like, oh.
0:39:21 Okay. So last question, I’d love to hear what you think, looking forward to the next
0:39:27 kind of 20 years, what’s the thing that you’re most curious to see how it turns out, right?
0:39:30 Where you think maybe it’s going this way, but you really are just dying to fast forward
0:39:33 20 years and be like, Ah, that’s what happened with that. Like, what’s the biggest kind of
0:39:36 question mark that you have over the next, say, two decades?
0:39:41 So the thing that makes my brain melt is this, now that we can program biology, so that kind
0:39:48 of, or we’re getting to the point where we can program biology, you know, the first step
0:39:53 is, you know, your one kind of dimension of that is, you know, solving disease, you know,
0:39:59 in a much, much better way, you know, another aspect of it is creating better humans. And
0:40:04 I’m very fascinated to see how that comes out and what it ends up meaning. And, you know,
0:40:09 whether it goes horribly wrong or incredibly right, and what does that even mean better
0:40:15 humans and how will, like, are humans even suited to, like, figure that out. So that,
0:40:18 from a curiosity standpoint, I would say that for me is probably it.
0:40:23 Yeah, yeah. The thing I think a lot about is, so through all of recorded history, and this
0:40:26 is why I just think that a lot of the tech credit assistants are just misguided, through
0:40:31 all of recorded history, most people have not been, I would say, most people have not been
0:40:35 plugged into what we would consider to be modern systems, right? So most people have
0:40:38 not been literate. Most people have not been healthy. Most people have not been fed well
0:40:42 enough to be able to reach full health maturity. Most people have not been educated and still
0:40:47 aren’t right to the level that we could consider modern. Most people don’t have access to economic
0:40:51 opportunity that we would consider to be, you know, modern jobs. Most people don’t have
0:40:55 access to what we consider to be high quality healthcare. Most people don’t have access to
0:40:58 high quality housing, transportation, you just go right down the list of all these things
0:41:01 that we’ve been lucky enough in this country to enjoy, you know, large percentage of the
0:41:05 population for a long time. Most people in the world have not had access to those things.
0:41:10 And I know that the existing systems, existing education system, the existing healthcare system,
0:41:15 the existing transportation system has had, you know, 50, 100, 200, 500 years to get to
0:41:18 the 7 billion people on the planet. It’s only gotten to every one of those systems has only
0:41:22 gotten a fraction of the people. And now we finally have the way to get right to everybody.
0:41:27 We get we’re at the point where 3 billion smartphones on its way to 6, 7 billion on the planet, we’re
0:41:30 going to be able to connect everybody. We’re going to be able to get over time, we’re going
0:41:32 to be able to get everybody all the things that I went through, right, starting by the
0:41:36 way with education, right, as sort of a foundational one. And so what is it going to mean for the
0:41:41 planet when everybody around the planet all of a sudden starts to, I would say, become
0:41:45 part of the systems that we know and understand. And we literally have 10, 20 times the number
0:41:48 of people around the planet who are contributing in all these different areas. And I just don’t
0:41:52 understand how people can be possibly pessimistic about the future knowing that that’s the potential.
0:41:54 And I think we’re going to see that. And I think our kids are going to see that. And
0:41:59 I think that’s very exciting. Yeah, that is. Okay, so we covered Knight Rider, Karl Marx
0:42:03 and universal education for the planet. I think we’ve done our job. Thank you guys. That
0:42:07 was great. Thank you. Thank you, everybody. Thank you. Thank you.
0:42:10 (audience applauding)
0:00:10 tax or investment advice, or be used to evaluate any investment or security, and is not directed
0:00:16 at any investors or potential investors in any A16Z fund. For more details, please see
0:00:17 a16z.com/disclosures.
0:00:23 Hi, everyone. Welcome to the A6NZ podcast. I’m Sonal. So this week, to continue our
0:00:28 10-year anniversary series since the founding of A6NZ, we’re actually resurfacing some
0:00:32 of our previous episodes featuring founders Mark Andresen and Ben Horwitz. If you haven’t
0:00:36 heard our latest episode with Stuart Butterfield turning the tables as the entrepreneur interviewing
0:00:43 them, please do check that out and other episodes in this series on our website at a6nz.com/10.
0:00:49 But this episode was recorded at our annual Innovation Summit in 2017 and features writer
0:00:53 Stephen B. Johnson interviewing them about everything from their relationship to creative
0:00:54 inspirations.
0:01:01 All right. I’m delighted and honored to be here with you. And we’ve got a lot to cover.
0:01:04 And what the kind of architecture for this conversation is, in a sense, we’re going to
0:01:09 kind of zoom out. We’re going to start on a more personal level and broaden out to think
0:01:13 a little bit about tech cultures inside a given organization, and then start thinking a little
0:01:18 bit more about broader social trends coming out of technology and looking into the future
0:01:21 a little bit. But I wanted to start with something actually just listening to your conversation
0:01:25 with JJ, who I don’t know at all, but I’m going to call JJ. He was talking about that
0:01:31 first kind of literally magical moment going and seeing Universal Studios and then getting
0:01:35 into magic and how that was so transformative as an eight-year-old. And it occurred to me,
0:01:39 do you guys have a memory of something like that with tech at any point where you really
0:01:44 saw something? For me, it was late. It was hypercard, sophomore year in college, where
0:01:49 I was just like, oh, there is this whole possibility that I hadn’t imagined could happen on a screen.
0:01:53 Do you have similar stories? It’s funny. This is an embarrassing question because I’m sitting
0:01:59 next to Mark, but one of the ones I remember most vividly was seeing Mosaic because for
0:02:04 years in tech, there were all these ideas about like if you were in computer science
0:02:10 about what was possible from all the things that you ought to be able to do, but you could
0:02:15 never actually quite get them to work. And hypercard was like that in that way, but Mosaic
0:02:19 was really it. It was all there on it. And when you downloaded it, you were like, oh,
0:02:25 my God, the whole world is like right there. I can reach the world. That’s the most craziest
0:02:29 thing ever. But I hate to say that with him sitting here because I go right to his head.
0:02:33 Well, it was a really striking point because up until, certainly for me, and I think for
0:02:38 a lot of people, there was discussion about hypertext that had been circulating through
0:02:43 different subcultures. But I would say probably 80% of the preceded received at that point
0:02:48 was strangely enough about hypertext fiction. It was people who were writing these nonlinear
0:02:54 stories. And when you saw Mosaic for the first time, you’re like, oh, this isn’t some obscure
0:02:58 avant-garde postmodern literary device. This is the future of media.
0:03:01 I have a much better answer than that. So I actually just mentioned on stage, but like
0:03:05 the early PCs really were the mystery box, the magic box and that really, that just,
0:03:08 you know, the flesh and cursor had me from go. So that sense of potential was a really
0:03:14 big deal. The other, I swear to God that this is true. Knight Rider, who remembers Knight
0:03:16 Rider, Knight Rider. There we go. Knight Rider outstanding.
0:03:17 You’re talking about kids?
0:03:22 Kids. Holy shit. So I was, I forget, I was, it was 82. So yeah, I was 10, right? And so
0:03:25 this shows on, and I don’t know, it’s this guy in the leather jacket. And I don’t know,
0:03:29 he seems cool, whatever. But they did this very clever thing, the mystery box thing.
0:03:31 And then there was no internet, no, nothing couldn’t find anything. You just saw a few
0:03:36 commercials. They did not tell you that the car was like that special. And if you go back
0:03:41 and watch the pilot, it’s like 45 minutes in. And like, it’s the whole thing has happened.
0:03:44 He’s been shot in the face. He’s had reconstructive surgery. He’s got the new name. He’s got
0:03:49 the mission. He’s got the car. He’s driving along 45 minutes in the car talks. And like,
0:03:52 I think I fell out of the couch. Like, I think I just like literally, I was like, the car
0:03:56 is talking. Right. And then I started to remember what that felt like. And then I have to remember
0:04:00 the screens, like the dash on that thing, right? It was like being in the space shuttle.
0:04:03 And to this day, when I get in a car, that, you know, the modern cars are like that, right?
0:04:06 They’ve got up to and excluding the fact that they talk to you now. But you know, they got
0:04:08 all the screens in the distance of that and the dash and the tuzzling, the whole thing.
0:04:14 It’s still, I always still feel like I’m getting behind the dash of kit. That is the best answer.
0:04:20 So there’s a great thing about your, the relationship that you guys have, it’s a long enduring one,
0:04:24 incredibly productive one. There’s a line in the hard thing about hard things in your book,
0:04:28 not to embarrass you, Mark, but I just wanted to quote it here. This is, you’re talking
0:04:34 about the relationship. And what you said is, even after 18 years, he upsets me almost
0:04:41 every day by finding something wrong in my thinking. And I do the same for him. It works.
0:04:48 So first off, is that true? But more than that, are you guys, is there something predictably
0:04:52 wrong? Are you guys wrong? And are you finding yourselves correcting each other in ways that
0:04:57 are kind of, are there patterns to the way in which you disagree? Do you tend to err
0:05:01 on the side, this side, where Mark errors on another side?
0:05:06 You know, I think it’s, you know, we’re close enough in personality, but different enough
0:05:11 kind of in skills that we often see things from different angles. And then a lot of it
0:05:16 is Mark himself, which is like, Mark always likes to take the other side of the argument,
0:05:21 whatever side, like he just enjoys taking the other side. That’s his thing. And so, you
0:05:26 know, it just kind of goes that way. I think that the real key to it is that we somehow
0:05:33 got to a level of trust where we can really go at it in a way that would, for most people,
0:05:38 you just go like, if you like, you can’t talk to me that way. Like how, you know, like so
0:05:42 disrespectful, like you’re stepping on me, you’re asking me these questions that hurt
0:05:47 my feelings. But you know, for us, you know, it has still like, you know, sometimes like,
0:05:51 get close to that, but not, not all the way.
0:05:54 I think the big thing is the thing I decided at a certain point, because we get asked a
0:05:58 version of this question by the founding teams that we work with, or if we bring a CEO into
0:06:01 a company, help a founder, bring in a CEO, and they’re going to have a partnership that
0:06:03 hopefully works something like this, you know, get kind of asked kind of, how do you make
0:06:07 it work? Because it is so easy for the conflict, for the emotion to, to drive people apart.
0:06:12 And so the way I think about it is, it’s more important to me that we have the successful
0:06:17 partnership than it is that I’m right on any particular issue. And I’m proud to say that
0:06:21 Ben, of course, is the exact opposite. It’s far more important for him to be right than
0:06:27 absolutely. And so it meshes perfectly right hand and glove. I’m joking. That was a joke.
0:06:29 And so we both will argue it all the way up, but each of us will defer to the other. At
0:06:32 the end of it, if it’s an argument, it’s over which one, which of us is going to defer to
0:06:37 the other one, with each of us volunteering to do it, say most of the time.
0:06:42 And that’s really like, sometimes the argument will not resolve, but we’ll kind of know
0:06:48 who knows more about that thing. And we’ll yield in that way. And that’s been super productive.
0:06:53 And there are ongoing disputes about where the technology world is heading. Are there
0:06:57 kind of senses like, oh, no, you think this thing is going to be huge, but this is the
0:06:59 old argument we’ve been having for five years. It’s never going to happen.
0:07:02 Well, we both believe a lot and disagree and commit, right? And so it’s important. Like
0:07:05 as an example, one version of the question you asked is like, what if we’re arguing about
0:07:08 some startup we funded? And whether it’s, you know, we’re going to have some argument
0:07:11 about like that was a mistake or not or whatever. Like we basically, I don’t think ever have
0:07:15 those arguments. And the reason is because we may argue whether, and this is true of
0:07:18 our partnership or broadly, we may argue about whether to make the investment, but once we
0:07:21 make it, we’re in. And then at that point, it’s important that it’s the dynamic sort
0:07:25 of implicit promise in the team and including between the two of us as we’re all in this,
0:07:29 we’ve all committed. And I think that’s really critically important because that’s how you
0:07:32 maintain, that’s how you don’t have, I told you so.
0:07:35 And backbiting and talking about people when they’re not in the room and that kind of thing.
0:07:36 That’s just bad.
0:07:40 Do you all have a, I’m actually in the middle of writing a book about long-term, complex
0:07:46 decision-making. So I have my own kind of bias in this question. But do you have, when
0:07:48 you’re confronting a decision to say, for instance, like should we fund this company
0:07:52 or should we follow in this round or other life decisions, do you find that you have
0:07:59 a process for that decision-making act that you go through and think about as a series
0:08:04 of stages? Or is it something that’s more fluid and conversational and intuitive?
0:08:09 Yeah. So it’s interesting. This business is different than our last. So running a company,
0:08:16 you try to be more structured in how you do this. In some ways, in that speed is really
0:08:22 important. So if you’re running a company, your output is decisions and you rate it
0:08:27 on quality and speed. And if you have to make the trade, which you always have to, you generally
0:08:32 go towards speed because you have a lot of decisions to make. And if you don’t make them
0:08:39 fast, then you freeze the entire organization. In our new business, basically quality is everything.
0:08:45 And so we’ll go around the horn 50,000 times if we have to to make sure that we’ve explored
0:08:51 every corner and every crevice of the discussion and we’ve not missed something. So I would
0:08:57 say in some ways, we have a lot of a framework in our minds about how we think of investments
0:09:03 and deals and so forth. But we’re willing to go in many loops where we would never do
0:09:04 that in a company.
0:09:08 One of the things that I love investigating and talking to people about is their kind
0:09:13 of creative workflow and where they find inspiration. There’s a lot of research out there that some
0:09:16 of which that I’ve done and other people have done about the importance of kind of diversity
0:09:22 of influences in your kind of worldview, leading to more creative thinking. So I’m just curious
0:09:27 about your kind of daily information diet in a sense, beyond the kind of the routine
0:09:32 of the meetings that you have with the founders and the pitch meetings and so on. Where do
0:09:36 you find that kind of outside influence in new ideas?
0:09:41 So we sort of cheat in a sense, which is we have, we see 2000 inbound startups a year.
0:09:44 These are by definition and 2000 are the smartest people in the world in all the domains that
0:09:48 they’re operating in. And so, I mean, honestly, after that, it’s just, it’s hard to pick up
0:09:52 like a magazine and open it with any level of enthusiasm because it’s like, you know,
0:09:54 you kind of have this, you know, you’re kind of seeing the stuff months or years before
0:09:58 it shows up in the press. And so that’s part of it. Personally, I’ve been running this
0:10:02 year a big experiment and I’ve always been a big reader and sort of information on the
0:10:05 board. And it just, you know, I’ve always tried to kind of balance short term, long term,
0:10:09 you know, different kinds of different time horizons of material, different kinds of material.
0:10:12 So I’ve been running a big experiment this year, which is I’ve been trying to do a bar
0:10:17 bell. I’ve been trying to polarize it. And so I’ve stopped completely reading newspapers,
0:10:22 magazines, basically anything that has a time horizon, basically greater than let’s say
0:10:26 five minutes to, you know, anything basically between five minutes and five years, which
0:10:30 is to say I basically only read social media on the one hand and then only books on the
0:10:34 other hand, right? And just polarize it and gap it way out. So what’s interesting about
0:10:38 that is of course, being on social media like that process, you know, necessarily you end
0:10:42 up consuming a lot of news and that a lot of what’s there notwithstanding the false
0:10:46 reports of the death of the web, a lot of what social media is, is links to things that
0:10:49 are interesting, right? People who you’re following are interested in. And so, you know, I do
0:10:52 end up reading basically everything. But one of the experiments was, does it matter? Like
0:10:55 if you don’t see the homepage of the newspaper, do you miss things? And it turns out if you
0:10:58 follow the right people, you really don’t, because they surface all the interesting
0:11:01 stuff anyway. And you get to see a lot of stuff that you wouldn’t necessarily see looking
0:11:04 at the homepage. But the other side, honestly, and you know, you’re accomplished book author,
0:11:08 the other side of it is just books, you know, books that probably become the great underestimated
0:11:12 source of information relevant to our daily lives that just gets, you know, as there is
0:11:17 just such a surplus of kind of near term information and consumption. And let’s just say, as the
0:11:21 real world is getting continuously more interesting in real time, you can spend all day long just
0:11:24 following the ins and outs of what’s happening in the political scene or what’s happening
0:11:26 in the sports scene or what’s happening in, you know, the business world or whatever.
0:11:29 And so you can really get, you know, let’s talk about myself, I can get really trapped
0:11:33 in the present. And so the ability to at least have some time to be able to go back and be
0:11:37 able to read things that were written five or 10 or 50 or 100 years ago, that have stood
0:11:40 the test of time in the form of books has been I think is very valuable.
0:11:43 It has been very interesting. I mean, the book business is actually quite healthy and
0:11:47 people are reading, you know, reading print books, there’s a kind of return to print books.
0:11:51 And it does feel as if I think one of the things you don’t realize until you write them,
0:11:54 particularly with nonfiction books, but it’s true fiction as well that when you meet someone
0:12:00 who’s read one of your books, they have been living inside your mind for 12 hours, 20 hours,
0:12:05 depending how long the book is. And so it is still an unrivaled way to get complicated
0:12:09 ideas into other people’s minds. And so it’s been, I think a sign of health in the culture
0:12:13 of that books are actually thriving in the midst of all this kind of minute by minute
0:12:14 social media.
0:12:16 And also, by the way, as you well know, like audio books, right, I think there’s a renaissance
0:12:21 in audio books, which is just having the smartphone and now the wireless, you know, ear pods makes
0:12:26 it so much more convenient for your content, long form audio content. And podcasts, obviously
0:12:29 are a big part of that. But audio books in the course is drive time and wait time and
0:12:32 this time and, you know, morning time and so forth completely fit into my life in a
0:12:34 way that books didn’t use to.
0:12:38 I also wanted to ask you, Ben, about music, can you talk a little bit about that in terms
0:12:40 of your own kind of creative view of the world?
0:12:44 Yes. Well, it’s interesting and it’s very specific to hip hop for me and hip hop is an
0:12:49 unusual music form in that it’s a very kind of capitalistic form of music, which is completely
0:12:55 kind of unheard of in popular music. And that the main theme of hip hop, if you go through
0:12:59 all the great rappers is like, how do you build something out of nothing? You know, how do
0:13:03 you compete these kinds of things as opposed to R&B, which was maybe love songs and like
0:13:09 rock and roll, which is more communist. But it’s perfect. It’s a perfect analog to entrepreneurship.
0:13:13 It’s kind of the exact kind of motivational soundtrack for entrepreneurs. And that’s really
0:13:19 how I started with it, because any theme I wanted to write about, like it was a great
0:13:26 way to find inspiration. But it led to, if you say I made a contribution to the management
0:13:33 literature, it actually came out of rap music in that the big thing that was different in
0:13:38 my book was that the logic of management is not very complicated. Then you can understand
0:13:44 all the management theory. It’s just not that hard. But the emotional, psychological complexity
0:13:49 of doing it is incredibly difficult. And you know, we see tremendous fallout from brilliant,
0:13:55 brilliant people who can never get over that. And so the big challenge for me was like,
0:14:01 how do you communicate the emotional part of the lesson? And hip hop is great for that
0:14:06 because it carries the emotion. And it’s all about kind of the capitalism. So I wrote
0:14:12 a post, how do you handle politics in a company? And I went through all the things that cause
0:14:16 politics and the subtle things, like how somebody asking for a raise can do it and how you deal
0:14:22 with that technique and so forth. But a lot of it is the attitude of the manager. And
0:14:27 so the rap quote that I used was Rick Ross, who do you think you’re fucking with? I’m
0:14:33 the fucking boss. And like, once you get that, then you know how to do it. That’s great.
0:14:39 Okay, so let’s zoom out a little bit now. You were asking JJ Abrams about LA as the
0:14:43 kind of epicenter of the movie business. So with all the changes that we’ve seen in the
0:14:50 tech sector and all the volatility, the one constant really for half a century has been
0:14:55 that the Bay Area and Silicon Valley have been the epicenter of the technology world
0:15:00 really without any near arrival, probably for 50 years, I think it’d probably be fair
0:15:03 to say, despite the fact that it has gone through all these different revolutions and
0:15:08 you had big computers and then personal computers and then the web and then social media. So
0:15:13 really two questions I think, why, why they are, like what was it about that particular
0:15:18 configuration that rooted tech in that world? And do you think we’re going to look back
0:15:21 in 30 or 40 years and it’s going to have the same concentration?
0:15:25 Yeah. So the why, so the why is I think it’s history, right? And so just the fact that
0:15:28 it’s been a network effect, right? It’s been a snowball rolling down the hill, picking
0:15:32 up momentum now for 56, actually turns out 50, 60, 70, 80 years. A lot of ways it goes
0:15:37 back to the 1920s, 1930s, the early defense contractors. Steve Blank has a whole series
0:15:41 of videos called the secret history of Silicon Valley. He traces it all the way back almost
0:15:42 a hundred years.
0:15:43 Fantastic.
0:15:46 Fantastic shares. And the point of it is, it’s just, it’s this kind of network effect
0:15:49 that’s just kept rolling, right? And so it’s been this place where it’s just like, it’s
0:15:53 the place where the next really smart engineer programmer or, you know, equivalently salesperson,
0:15:58 marketing person, west door contact, whoever they are, finance person on the margin, right,
0:16:01 is more tempted to move to the valley than many other places, which isn’t to say that
0:16:04 there aren’t many capable people all over the world. It’s just on the margin. Many of
0:16:07 the ones who are super ambitious end up at the valley. And of course, I’m an example
0:16:10 of that. And as a consequence, right, it’s a story of imports, right? And so another
0:16:14 thing just to read, I’m sure if people are interested, Tom Wolf, the great novelist,
0:16:18 journalist wrote a piece in the 80s in Esquire about literally Bob Noyce, who was the original
0:16:22 CEO of Intel, one of the fathers of Silicon Valley and literally grew up in Iowa, grew
0:16:26 up in the Midwest and was the Silicon Valley import. And actually Wolf ascribes a lot of
0:16:31 modern Valley culture to literally Bob Noyce importing, interestingly, Midwestern culture,
0:16:35 right, including, by the way, egalitarianism, right. So the whole open floor plan thing,
0:16:39 stock option ownership, everybody owns a share in the company. He traces that actually back
0:16:43 to Midwestern culture. And so it just got established and it developed this ethic and
0:16:46 it’s probably not an accident that it’s the frontier, right? It’s probably not an accident
0:16:49 that this sort of gold rush happened, right? It’s just kind of this frontier out the mentality
0:16:53 has continued. So that’s the good news, right? The bad news is, as I discussed with JJ, like
0:16:56 it’s just number one, we’re just bursting at the seams, like it’s just become a hard
0:16:58 place to do business. And the number two is there’s great people all over the world and
0:17:02 like why on earth? So the joke in the valley is, you know, help wanted, right? Software
0:17:06 company puts up Silicon Valley, software company puts up a help wanted out on the internet
0:17:09 or whatever and says, you know, help wanted, you know, software engineer to work on new
0:17:14 collaboration software tool, online collaboration software tool that will enable people to work
0:17:19 together independent of geography all over the world. So in real time, PS must relocate
0:17:22 to San Francisco to apply. And so it’s this weird incongruity, which is we’re building
0:17:26 the technologies that in theory should let this stuff spread. And yet for some reason
0:17:30 in the last 10, 20 years, it’s actually been concentrating more and more. And so I’ve come
0:17:33 to believe it’s a maybe this is obvious to some people, but I would come to believe it’s
0:17:38 a human dynamics question. It’s a psychology, sociology question, not a technology question
0:17:42 in a lot of ways, which is just like how do people best work together, right? And it just
0:17:46 so happens that at least for the form of traditional companies, which you just see over and over
0:17:50 again is just when you can get everybody in the same room physically in the same room,
0:17:53 right, with the level of, say, fidelity of communication interaction where we’re sitting,
0:17:57 you know, it’s why, by the way, it’s why we’re all physically here. And there are a few successful
0:18:00 distributed companies, but there really aren’t very many as a consequence of that. And so
0:18:05 my hope is that we’re going to get there in the next, you know, let’s say 10 or 20 years,
0:18:09 my hope is that we’re going to get telepresence, right, in the form of video conferencing and
0:18:14 telepresence robots and VR and AR and all these things to collaboration software and
0:18:17 work group software and Slack and GitHub and all these amazing technologies are building
0:18:20 for collaboration. My hope is we’re going to get it to the point where it’s just going
0:18:23 to be obvious that we don’t all have to be in the same place. If that happens, you could
0:18:27 say it’s quote bad for the Valley in the sense of like maybe Silicon Valley is not central
0:18:31 anymore, but it would be so good for the world for that to be the case and we would all benefit
0:18:34 so much from that. I think it’s a very worthwhile thing to pursue and something I’m very fired
0:18:35 up about.
0:18:39 How much do you think, just to go back to the point about noise in the early days of
0:18:43 Silicon Valley and the history of it, to have written about this a little bit as well, how
0:18:51 much do you think that the participatory option granting culture, which is very different,
0:18:54 there were very few kind of East Coast firms that were doing that. So you had much more
0:19:00 traditional kind of top down equity systems in those corporate entities. How much do you
0:19:03 think that is part of the success of Silicon Valley? This is something I think that would
0:19:06 be interesting to go back and look at just economically.
0:19:10 So I think it ends up being very important because of the nature of technology companies.
0:19:18 So if you look at, there are other kinds of companies where the people are much more interchangeable
0:19:23 and this kind of gets into why the network effect is so important and so forth. And in
0:19:29 like a tech company, there’s lots of people who are extremely valuable and that innovation
0:19:36 as a way to get them their kind of proper compensation for their contribution, the great
0:19:39 conversation with Mark and Charles Koch, where he talked about like, you have to be
0:19:43 rewarded for what you contribute to others. And that really is key to any business and
0:19:47 any incentive system. And particularly in technology, because there are so many people
0:19:53 in the company who are so valuable and so fundamentally critical to the company’s success,
0:19:58 it really is one of a very few kinds of compensation systems that would work. And certainly, you
0:20:02 know, a lot of the systems on the East Coast would never work for tech companies to be
0:20:03 kind of world-class competitive.
0:20:11 So it’s been six years since Mark, you wrote the software eats the world essay. I went
0:20:15 back and looked at it and reread it. It was a great piece. It reminded me of, I’m sure
0:20:18 a lot of people have seen this. There was a great thing that was circulating on social
0:20:24 media a couple of years ago. It was an old kind of single page flyer for Radio Shack from
0:20:29 like 1988 or something like that. It was a list of like 30 products that Radio Shack sold.
0:20:36 And the answer machine was, you know, a VCR, an alarm clock, like a TRS 80 kind of descendant,
0:20:39 you know, a game console, something like that. And literally without exception, every single
0:20:43 one of them is now an app on your phone, right? The whole thing had gotten swallowed up by
0:20:48 software, which is of course a measuring productivity problem because all those things
0:20:55 in aggregate cost $30,000 in 1988. And now they’re free on a phone that costs $600, which
0:20:57 is actually progress, but doesn’t sometimes look like it.
0:21:03 So obviously I think that that was a very prescient forecast to make. Has anything kind
0:21:08 of surprised you six years later looking back on it? I mean, in it, you say the next big
0:21:12 stages are health and education. And I’m wondering, you know, particularly on those fronts, has
0:21:14 it lived up to the kind of promise you saw back then?
0:21:16 Yeah, they’re sort of the overall concept of software eats the world. But then there
0:21:20 was a specific framework that I proposed in the piece, which is sort of a weak form of
0:21:23 semi strong form and a strong form of this hypothesis, right? And so the weak form was
0:21:28 every product that can write every physical product will become a software product, right?
0:21:31 And that’s that’s exactly your radar check example. Things go from being physical products
0:21:37 to being apps. The second sort of semi strong version of that was therefore any company
0:21:41 that makes a product that can be turned into software will itself therefore have to become
0:21:44 a software company. Right. And in fact, I was thinking you could you could see this thing
0:21:47 for example, playing out right now in the car industry, right, where all the car companies
0:21:50 are spinning up software efforts, they’re buying software companies are spinning up
0:21:53 software and as fast as they possibly can because they see what’s coming with autonomy
0:21:58 and all these other software advances. And then the strong and sort of audacious slash
0:22:03 ambitious slash arrogant hubristic version of the thesis is in any industry as a result
0:22:07 of this dynamic in the long run, the winning company in the industry will be the best software
0:22:12 company, right, which is a provocative statement, right? Because in a lot of these industries,
0:22:16 and again, cars are a great example. You have incumbents who are really good at making cars
0:22:18 trying to become great software companies. And then you have great software companies
0:22:22 that have no idea how to build a car, right, who are going to start who are going to start
0:22:25 making cars, right? And then you’re going to have basically, right, this giant collision
0:22:29 between companies coming from two totally different backgrounds. And so I think that
0:22:34 you’re seeing lots of that first stage that week stage, lots of products transitioning,
0:22:37 you’re seeing lots of companies becoming software companies. I think we’re just entering in a
0:22:41 lot of industries were entering that Thursday’s where there’s this very interesting structural
0:22:44 battle that’s forming up. The other thing I says, yeah, I think you exactly nailed it
0:22:48 with healthcare and education, right, which is there are these giant sectors of the economy
0:22:53 in which not only is there no productivity growth, like overall in both healthcare and
0:22:57 education, there is no measured growth, there is no measured results in the application of
0:23:01 technology in those fields. And in fact, probably it’s negative productivity growth, right?
0:23:04 Like the typical university has been going backwards in productivity, right? You just
0:23:07 look at the charts, the number of administrators that they hire, right, per student is just
0:23:13 skyrocketing and that is literally negative technological productivity. And so those industries
0:23:18 are extremely enticing to Silicon Valley, because they’re so big, they’re gigantic.
0:23:22 Healthcare, healthcare is a sixth of the American economy, right? And left unchecked, it will
0:23:26 become a fourth and then a third and then a half and then two thirds and then three quarters.
0:23:30 Like it’s just left unchecked, it’s just going to keep growing. And so it’s so much money.
0:23:35 It’s so big. It’s so important. It’s very enticing. And the incumbent structure of there’s many
0:23:38 smart companies in that industry, but the incumbent structure of how the industry works
0:23:42 is just, is wired to go the wrong direction. And so there’s this huge opportunity to insert
0:23:46 into it, which obviously we’re going after hard, but that’s still like super early.
0:23:47 Yeah.
0:23:51 And education, what, Ben, do you have thoughts on that front? I mean, there’s this interesting
0:23:57 point we’re at where there seems to be a growing backlash to the presence of screens, particularly
0:24:02 in younger kids’ school classrooms that it hasn’t lived up to the potential. And maybe
0:24:06 the kids already have too much software in their lives as it is.
0:24:13 So, you know, it’s funny, or it’s not funny. It’s sad that we’ve not applied technology
0:24:19 that well. And a lot of it has to do with the kind of structure of the kind of political
0:24:22 regulatory structure of schools. And we have a company, Udacity, that’s worked hard on
0:24:27 this. And their final conclusion was to kind of run outside of the school system, but it’s
0:24:32 very powerful. I’ll tell you a quick story about that. But, you know, obviously, very
0:24:37 obviously, if you could have like, any teacher or the best teacher in the world teaching a
0:24:44 math class, if students have to study and then be tested, like, when do you take a test outside
0:24:49 of school, like, ever in life? Like, what the hell skill is that does this create like tremendous
0:24:53 anxiety and like give people complexes. But you ought to, with technology, you ought to
0:24:57 be able to measure how people are learning every step of the way, give them harder problems,
0:25:02 if they’re going very fast, or get them help if they’re going slow. And there’s a lot of
0:25:07 things that ought to be able to be done. But then I think the more kind of pressing thing,
0:25:13 and the thing that Udacity really addresses is the four year education, general education,
0:25:18 doesn’t work that well in the modern economy because people are switching careers very,
0:25:23 very often every, you know, two, three years sometimes. And, you know, like four years,
0:25:26 and then you never go back to school for the rest of your life doesn’t make any sense at
0:25:31 all because people need to get retrained jobs get displaced. And so what Udacity has come
0:25:36 up with is this thing, the nano degree, which is two months, three months, you can learn
0:25:42 to program an Android phone or build a self driving car, or learn to do technical marketing.
0:25:47 And those degrees are connected right to the job market. So you can roll right in with
0:25:53 a skill and a certificate that says you understand the material and you’re ready to work. And
0:25:58 that is a great innovation and something that we’re really excited about. And just quick
0:26:03 story on that. So one of the huge problems we have in this country is prison and the
0:26:08 need for prison reform because we’ve got, you know, 75% recidivism rate where people
0:26:12 who go to jail and come out, go back to jail. And the reason they go back to jail, they
0:26:16 can’t get jobs. And the reason they can’t get jobs is because two things. One is we’ve
0:26:23 outlawed college in prison and then two, once they come out, their record follows them wherever
0:26:30 they go. So, you know, I’ve got a friend who came out of jail and I said, go to Udacity.
0:26:35 He goes to Udacity and he’s coming up on his technical marketing degree and he’s already
0:26:38 got job offers. And it’s like, that’s what we need.
0:26:43 Yeah. And I think it’s almost as if school, particularly high school, and I have two kids
0:26:47 in high school, so I think about this a lot, it’s kind of trapped in this middle zone that
0:26:50 doesn’t really work in a sense. It’s much more effective to have those kind of nano
0:26:55 skills, right, where you can actually kind of apply them or the skills should be broader,
0:26:59 right? I mean, when you read through, again, a book like the hard thing about hard things,
0:27:04 I just think about how there are so many skills in there that no one ever thought to teach
0:27:08 me in high school, right? I mean, the skills about decision making skills about kind of
0:27:11 emotional intelligence, dealing with, you know, difficult decisions. My kid actually
0:27:15 in his high school, to its credit, is doing a kind of design thinking class. And they’re
0:27:20 basically learning how to brainstorm ideas, interview a customer, think about different
0:27:24 possibilities, do mock-ups. And it was like, this should be the default. This should not
0:27:27 be an elective. This should be the thing you learn. And then if you want to go off and
0:27:33 do advanced chemistry or do advanced calculus, that’s fine. But those types of skills that
0:27:36 are just, everyone is going to have to know on some level, but it’s very rare to encounter
0:27:38 that. We’ve got a very dated curriculum. There’s
0:27:42 no question. I ran on the board of trustees at Columbia. And there are certainly people
0:27:47 who are going to go to like an elite school and become a scholar or a PhD. And I think
0:27:51 the system works reasonably well for them. But for, you know, the kind of bulk of the
0:27:57 population who goes to college to get into the workforce, it’s really difficult. It’s
0:27:59 exactly, as you say, it’s kind of neither here nor there.
0:28:04 Let’s talk a little bit then, kind of segueing a little bit to the job and automation question
0:28:09 anyway. In general, I think we all agree that there has been this growing and now kind of
0:28:16 reaching Crescendo backlash against big tech and the tech sector that the last year has
0:28:20 particularly brought to the fore. And I feel it very strongly going back because I live
0:28:24 part of the time in Bay Area and part of the time in New York, when I’m back in New York,
0:28:29 you know, nine out of 10 kind of opinion like pieces written in these media are negative
0:28:30 pieces.
0:28:31 It’s only nine out of 10.
0:28:34 I mean, so I want to get into some of the specifics about why that is happening, how
0:28:39 you guys feel about it. But how much in general do you and how much recently have you found,
0:28:42 do you find yourselves taking that seriously and how much do you feel that people just
0:28:44 don’t understand what’s going on here?
0:28:45 We might give two different answers.
0:28:50 Yeah. So I would first say there’s a huge difference between what gets written in opinion
0:28:56 pieces and the actual opinions of the public. So if you look at approval ratings of tech,
0:29:01 they’re incredibly high. Like they’re the highest of any industry. And like Amazon’s
0:29:05 approval raising, which is one of the biggest targets is like 80. Whereas Congress is like
0:29:10 20 and the press is like 20. And so like the guys at 20 are saying the guys at 80 need
0:29:14 to be stopped because everybody hates them. So there is that dynamic. And I think it’s
0:29:15 very real.
0:29:19 This is the concept of false consciousness, right? So literally the whole problem with
0:29:22 the communist revolution was the business weren’t signed up for it. And so the intellectual
0:29:24 leaders were like, well, but we got to take down the capitalist.
0:29:29 The other thing is, I think there’s something else going on that this is a side effect of.
0:29:34 And I think it’s the rise in the last several years. And in particular, after the 2008 crisis,
0:29:38 credit crisis crash, I actually think was the catalyst for a lot of this. It’s the rise
0:29:42 of zero sum thinking in both economics and in politics. Let’s say zero sum as opposed
0:29:46 to positive sum, right, which is this is sort of game theory, right? Zero sum game is I
0:29:50 win, you lose. And by the way, if I’m winning, it must mean that you’re losing because it’s
0:29:53 zero sum. It’s only a question of how we slice up the pie, right? Whereas positive sum is
0:29:57 we can all win together. It’s actually a great book called finite and infinite games that
0:30:01 actually goes through. If you go back historically, basically, economist philosophers and so forth
0:30:06 thought the politics and economics were zero sum. And there were huge battles over resources.
0:30:09 And this was colonization, all these other horrible things that happened over years were
0:30:13 fought through mercantilism, trade wars, right? All these things were fought based on zero
0:30:16 sum. And about, you know, 300 years ago, Adam Smith and a whole bunch of other really smart
0:30:20 thinkers figured out, no, you can actually gain from trade and you can actually interact
0:30:24 with more people and it’s good for everybody. And politics can be positive some just because
0:30:27 I’m doing well might mean that you’re also going to do well because again, we’re able
0:30:30 to culturally trade, we’re able to educate each other, we’re able to, you know, contribute
0:30:33 each other’s thoughts, and we’re all able to succeed. And so in the wake of the credit
0:30:36 crisis, I think zero sum thinking kind of came snapping back. And what’s interesting
0:30:40 is you see that on both the political left and on the right, right? For the anti attack,
0:30:42 the bloodite sometimes to come out of the left and Marx actually was shot through with
0:30:45 with leadism, like that’s one of the things he didn’t understand was the positive sum nature
0:30:49 of productivity growth. And anyway, so you get that on the left, you also get it on
0:30:53 the right, right? And you get it on the right, you get in the form of populism, right, which
0:30:59 in the form of opposition to trade and opposition to immigration, right? And so I just think
0:31:02 as a culture as an economy as a country right now, if you think that the formulation is
0:31:08 zero sum, you will then do things that will cause it to get worse. For example, on the
0:31:11 right, you’ll want trade barriers, right? And so you’ll want to cut trade under the
0:31:14 theory that that will make your people better. In reality, cutting international trade makes
0:31:19 your people worse. You’re dividing up a smaller pie. Yeah, you’re shrinking the economy for
0:31:21 everybody for no reason other than that you’re just mad at other people because you think
0:31:24 it’s their fault that you’re not doing well. And so it’s zero some thinking. And then on
0:31:28 the left right now, it’s this anti tech sentiment where like if those tech people are doing well,
0:31:31 then somebody else must be suffering, somebody else must be eating it. And it’s just it’s
0:31:35 the same sort of extremely reductionist thinking. And of course, the risk is as that sentiment
0:31:39 builds that at least a policies that actually impair the ability to be able to make progress,
0:31:43 make progress in the economy, make progress with productivity growth, make progress with
0:31:46 job creation, make progress with wage creation. And so there’s a pretty big risk that this
0:31:50 is all gonna go pretty seriously sideways for the wrong reason. Right. Let’s take the
0:31:55 tech backlash argument from a slightly more maybe sympathetic level, which is critiques
0:32:01 that have come from within the tech sector that the original vision of the web that inspired
0:32:07 so much of us, which was going to be this decentralized platform that was going to distribute
0:32:11 the kind of power of self publishing and voice to far more people. And it was going to kind
0:32:17 of topple this big, heavy, top heavy mass media model. That’s what inspired a lot of people
0:32:21 to get involved in it in the first place. At the end of that process, we’ve ended up
0:32:26 with, you know, four or five companies that in terms of their command over people’s attention
0:32:31 probably are the most powerful companies that have ever been on this planet and also some
0:32:37 of the greatest concentrations of wealth. So inside the tech sector, people say, re decentralized
0:32:41 the web and then we need to look at technologies that will enable us to have, you know, a more
0:32:46 even distribution in terms of the companies in terms of people’s attention and so on.
0:32:49 And blockchain is part of that. There’s some argument that people have been making along
0:32:54 those lines. How sympathetic are you to that side of the case, which does align with some
0:32:58 of the critiques that big tech is too big that are coming from people outside the tech
0:33:02 sector? Yeah. So there’s a technical argument for a decentralization. And then there’s
0:33:08 the kind of other thing that you’re getting at, which is should there be some like policy
0:33:14 answer to the big tech companies? And I think that, you know, you have to be very careful
0:33:19 there and look at specifically what’s going on. Well, are they kind of harming? Are they
0:33:26 suppressing innovation? So do people like us no longer want to fund anything because,
0:33:30 you know, Facebook or Amazon will wipe it out. And if you look at the numbers, there’s
0:33:35 probably more startups than there have ever been. And what we’re seeing and what we’re
0:33:41 funding is like super interesting. And, you know, for the most part, isn’t existentially
0:33:47 threatened all the time by those companies. Once you introduce policy, the potential side
0:33:54 effects are, you know, really scary, cronyism, corruption, the people who have the best relationship
0:33:58 get the best deal and these kinds of things. And that has knock on effects that are very
0:34:02 difficult. And, you know, if you compare it to the early nineties, when Microsoft was
0:34:07 super strong, that was really actually a far bigger suppression of innovation. There was
0:34:12 way less venture capital. There were far fewer companies being created. But like the technology
0:34:16 took care of it over time. And I think technology is changing at a faster rate now than it was
0:34:20 then. And there’s blockchain and there’s quantum computing. And there’s many technologies
0:34:25 on the horizon that could rejigger the playing field, you know, without a policy intervention.
0:34:30 Another question about the blockchain possibilities, you know, I’ve been really enjoying reading
0:34:36 Chris Dixon writing about this over the last year or two. And there is really an interesting
0:34:44 new way of incentivizing and compensating people both inside a technical organization associated
0:34:49 with an open protocol, early users of the service where all of those people are participating
0:34:54 in the value that’s created with it. And thinking back to the early stock option participation
0:34:58 of noise, you know, I wonder whether this, this suggests maybe that there’s a new model
0:35:03 here that might be as revolutionary as those kind of option plans were.
0:35:06 So the good news is the tech industry has had two models for making forward progress.
0:35:10 One has been what you might call pure capitalism, which is corporations, right, which is sort
0:35:14 of C corporations, employees, stock options, all the things we can take companies public
0:35:17 with that traditional structure. And then there’s been this other structure all the
0:35:21 way over on the ideological spectrum, right, which is open source, right, which is basically
0:35:24 the tribe, right, of developers that are interested in having something happen, coming together,
0:35:28 by the way, geographically distributed all over the world in a lot of cases, right, and
0:35:31 great examples, Linux and the web itself is an example of this and so forth. Actually,
0:35:36 the Internet, TCPIP was an example of this, right, or the new project MIT was an example
0:35:40 of this and people, technical people coming together and volunteering, right, literally
0:35:43 with metaphors like barn raising, right, it’s just like come together and make sort of breathe
0:35:47 life into these projects without a financial incentive and generally without, you know,
0:35:51 at least direct financial rewards. So sort of polar opposite of corporations you can
0:35:58 get. Blockchain is the first new third thing in, I don’t know, probably 40 years, right,
0:36:01 free software open source is like 40 years old. It’s the first new structure in 40 years
0:36:05 and it’s an interesting one because it’s a hybrid. It’s got the, it’s your point, it
0:36:08 has the decentralization of open source, right. These are protocols. These are things
0:36:12 that run Internet wide. These are things that are not necessarily developed by a team of,
0:36:15 you know, 100 people in a building in the Bay Area. They have that kind of open source
0:36:20 characteristic to them and they are decentralized. Like their protocols are inherently decentralized,
0:36:25 but they’ve got capitalism wired in. They’ve got money wired in, right, right, into the
0:36:29 protocol, right, in a way where there is a direct reward and incentive for the people
0:36:34 who actually create the thing. There’s a reward and incentive for the people who use the thing.
0:36:37 And then there was a reward and incentive for the so-called miners, the people who actually
0:36:40 run all the computers all over the Internet that make these things work. And it’s just
0:36:44 been so fascinating to watch because this is one of those kind of moments where people
0:36:47 walk up to this idea. And if they walk up to it from the right, they’re like, what on
0:36:52 earth is this decentralized hippy, like what on earth are you people doing? If they walk
0:36:55 up from the left, they’re like, Oh my God, it’s got money in it. It must be evil, right.
0:36:58 It’s sort of this weird, you got to kind of wrap your head around it. And so what we see
0:37:01 is like, it is fundamentally a third model for innovation. And I will also say this,
0:37:06 the thing that we see that I think maybe other people are missing, many of the smartest programmers
0:37:13 and mathematicians and economists and theorists and systems builders in the world and photographers
0:37:17 in the world are obsessed with this, like they’re just magnetically drawn to it, not
0:37:20 because of the money or this or that or the hype or whatever, because of the technical
0:37:23 innovations that are underneath this that are making this possible and what can come
0:37:26 out of this. And we just think like that’s the most positive sign you can possibly see.
0:37:30 We just have about five minutes left. So I want to just cover a couple of other giant
0:37:36 topics, artificial intelligence and the superintelligence debate. Can we solve that in about two minutes?
0:37:41 Can you give me, is this a legitimate concern? Is it appropriate to be worrying about the
0:37:47 threat from superintelligence now? Of the really scary things in technology. I would
0:37:51 have that one pretty low on my list. I mean, I think that one, like, I think it’s a little
0:37:56 bit of a miss, you know, intelligence is a funny word, right? Like, what is intelligence?
0:38:01 And it’s not one dimensional. And there are a lot of things that we have considered intelligence,
0:38:05 like doing hard math problems. Computers are already more intelligent, like playing chess.
0:38:09 Computers are already more intelligent. But there’s a lot of dimensions of intelligence
0:38:15 that computers are nowhere on. And AI, nobody is demonstrating anything in AI that says
0:38:20 like it’s going to get comprehensively more intelligent and certainly nothing along the
0:38:26 lines of free will yet. So yeah, maybe, maybe it’ll happen. But of all the things, it’s
0:38:30 a very theoretical. So I think it’s a little overblown. I do think also that there’s a
0:38:34 motivation of technologists to, it’s a very kind of, it makes you seem very intelligent
0:38:38 when you can talk about the robots taking over the world. So it’s a great thing to talk
0:38:42 about. The thing that drives you bananas is it’s the freaking physicists. And it’s like,
0:38:47 I’m a computer scientist. I don’t have like crazy conspiracy theories about black holes.
0:38:51 You know, I guess I could, you know, like in theory, a black hole could open up here in
0:38:55 this room and swallow us all. Like, I don’t have crazy theories about dark matter. Like
0:38:57 I’m not worried there’s dark matter in the glass. I’m not going to go around telling
0:39:00 everybody it’s going to, it’s just like, I don’t know why.
0:39:04 Yeah, it’s hard to find an AI expert who goes, Oh, yeah, this is a big problem.
0:39:07 Well, in fact, in the AI experts, of course, tend to be worried about the opposite, which
0:39:09 is they’re like, Oh, shit, expectations are getting set.
0:39:12 Like, we’re never going to build that. We’re never going to build the robot apocalypse.
0:39:16 I’m still trying to get the thing to play Mario Brothers, right? Like, oh.
0:39:21 Okay. So last question, I’d love to hear what you think, looking forward to the next
0:39:27 kind of 20 years, what’s the thing that you’re most curious to see how it turns out, right?
0:39:30 Where you think maybe it’s going this way, but you really are just dying to fast forward
0:39:33 20 years and be like, Ah, that’s what happened with that. Like, what’s the biggest kind of
0:39:36 question mark that you have over the next, say, two decades?
0:39:41 So the thing that makes my brain melt is this, now that we can program biology, so that kind
0:39:48 of, or we’re getting to the point where we can program biology, you know, the first step
0:39:53 is, you know, your one kind of dimension of that is, you know, solving disease, you know,
0:39:59 in a much, much better way, you know, another aspect of it is creating better humans. And
0:40:04 I’m very fascinated to see how that comes out and what it ends up meaning. And, you know,
0:40:09 whether it goes horribly wrong or incredibly right, and what does that even mean better
0:40:15 humans and how will, like, are humans even suited to, like, figure that out. So that,
0:40:18 from a curiosity standpoint, I would say that for me is probably it.
0:40:23 Yeah, yeah. The thing I think a lot about is, so through all of recorded history, and this
0:40:26 is why I just think that a lot of the tech credit assistants are just misguided, through
0:40:31 all of recorded history, most people have not been, I would say, most people have not been
0:40:35 plugged into what we would consider to be modern systems, right? So most people have
0:40:38 not been literate. Most people have not been healthy. Most people have not been fed well
0:40:42 enough to be able to reach full health maturity. Most people have not been educated and still
0:40:47 aren’t right to the level that we could consider modern. Most people don’t have access to economic
0:40:51 opportunity that we would consider to be, you know, modern jobs. Most people don’t have
0:40:55 access to what we consider to be high quality healthcare. Most people don’t have access to
0:40:58 high quality housing, transportation, you just go right down the list of all these things
0:41:01 that we’ve been lucky enough in this country to enjoy, you know, large percentage of the
0:41:05 population for a long time. Most people in the world have not had access to those things.
0:41:10 And I know that the existing systems, existing education system, the existing healthcare system,
0:41:15 the existing transportation system has had, you know, 50, 100, 200, 500 years to get to
0:41:18 the 7 billion people on the planet. It’s only gotten to every one of those systems has only
0:41:22 gotten a fraction of the people. And now we finally have the way to get right to everybody.
0:41:27 We get we’re at the point where 3 billion smartphones on its way to 6, 7 billion on the planet, we’re
0:41:30 going to be able to connect everybody. We’re going to be able to get over time, we’re going
0:41:32 to be able to get everybody all the things that I went through, right, starting by the
0:41:36 way with education, right, as sort of a foundational one. And so what is it going to mean for the
0:41:41 planet when everybody around the planet all of a sudden starts to, I would say, become
0:41:45 part of the systems that we know and understand. And we literally have 10, 20 times the number
0:41:48 of people around the planet who are contributing in all these different areas. And I just don’t
0:41:52 understand how people can be possibly pessimistic about the future knowing that that’s the potential.
0:41:54 And I think we’re going to see that. And I think our kids are going to see that. And
0:41:59 I think that’s very exciting. Yeah, that is. Okay, so we covered Knight Rider, Karl Marx
0:42:03 and universal education for the planet. I think we’ve done our job. Thank you guys. That
0:42:07 was great. Thank you. Thank you, everybody. Thank you. Thank you.
0:42:10 (audience applauding)
with Marc Andreessen (@pmarca), Ben Horowitz (@bhorowitz), and Steven Johnson (@stevenbjohnson)
Continuing our 10-year anniversary series since the founding of Andreessen Horowitz (aka ”a16z”), we’re resurfacing some of our previous episodes featuring Andreessen Horowitz founders Marc Andreessen and Ben Horowitz.
This episode was actually recorded in 2017 at our annual innovation Summit, and features technology writer Steven Johnson interviewing Ben and Marc about everything from their relationship to creative inspirations.
You can find other episodes in this series at a16z.com/10.