a16z Podcast | Talent, Tech Trends, and Culture -- with Ben, Marc, and Tyler Cowen

1. 0:01 – 2:23

   How Ben and Marc Met at Netscape

   1. SPSpeaker0:01

      Hi, everyone. Welcome to the a16z Podcast. Today's episode features a16z co-founders Marc Andreessen and Ben Horowitz in conversation with economist Tyler Cowen, who is Chair of Economics at George Mason University and Chairman and General Director of the Mercatus Center there, and also has a podcast, Conversations with Tyler. The discussion originally took place at our most recent annual innovation Summit and covers everything from talent and tech trends to software eating culture, or the other way around. You can also find other podcasts and videos from this event at a16z.com/summit.

   2. TCTyler Cowen0:34

      Thank you all for coming. I'd like to start with the two of you as a couple.

   3. BHBen Horowitz0:39

      Oh, yeah. [audience laughing]

   4. TCTyler Cowen0:40

      How was it you met at Netscape in nineteen ninety-five?

   5. BHBen Horowitz0:44

      Yeah. Well, Marc interviewed me, you know, way back then. And he interviewed me, I think, I was interviewing for a product management role, and, you know, he was the founder of the company. And I had worked on a product called Lotus Notes. And Marc had a fascination with Lotus Notes for a couple of reasons. One was it was kind of sort of the closest proprietary thing to the Internet, and then they had email in it, and Netscape was looking at doing email. So he had all kinds of questions for me, and I remember him just being absolutely shocked and flabbergasted that like fifty percent of our code base in Lotus Notes was just making like all the LAN protocols work together, IPX and AppleTalk, and NetBEUI, NetBIOS. [laughs] It was, uh... You guys don't even know what any of that is anymore, but-

   6. TCTyler Cowen1:27

      [laughs]

   7. BHBen Horowitz1:28

      It was like literally just making the network talk to each other, which he thought was like ridiculous given there was TCP/IP. That, that was the very first conversation we had.

   8. TCTyler Cowen1:38

      Marc, how did he do in the interview?

   9. MAMarc Andreessen1:40

      He did very good. He did very good. So-

   10. BHBen Horowitz1:41

       [laughs]

   11. MAMarc Andreessen1:41

       Well, he had, he, he had a giant asset. He was the first employee from Lotus, um, so he looked great in comparison to all the ones we hadn't seen before. [laughs]

   12. TCTyler Cowen1:48

       [laughs]

   13. MAMarc Andreessen1:49

       No. So... No, he, he did great. He was, he was super knowledgeable. And actually, it was, it was actually a really big deal at the time because Lotus Notes was a big thing at the time and, and the, what, what Ben just alluded to in the architecture, like it just... It basically assumed that the internet was not going to work, which was the dominant assumption at that time. It's kind of how the whole thing was built, and then most of the people working on it, I think, probably agreed with that. Like, and then, and then-

   14. BHBen Horowitz2:08

       Yeah

   15. MAMarc Andreessen2:08

       ... there were many, many arguments that the internet couldn't possibly do what a system like Lotus Notes did. And so Ben was, I would say, young enough and smart enough and knowledgeable enough to figure out, you know, very early among his cohort of kind of the professionals in the space that it actually was gonna happen a different way.

   16. BHBen Horowitz2:20

       [laughs] Yeah.

   17. MAMarc Andreessen2:21

       Um, and so... And, you know, look, we were a little fly-by-night startup,

2. 2:23 – 4:04

   Loudcloud and the Idea of the Cloud

   1. MAMarc Andreessen2:23

      so.

   2. TCTyler Cowen2:24

      So we fast-forward to nineteen ninety-nine. You two do Loudcloud together. Why? How'd that happen?

   3. MAMarc Andreessen2:31

      Well, so the big thing basically that happened was, so [laughs] it turns out the internet worked, which turned out to be a big deal. Go figure. And then basically what happened was, basically, people were unprepared for the internet to work at, at some fundamental level. And then we saw a very specific kind of version of people being unprepared for that, which is we sold Netscape in nineteen ninety-eight to AOL. And so B-Ben and I and a bunch of us were working at AOL, and then AOL had this thing at the time. AOL was a little bit like the Google of the era or something, where it was just a absolute... Th-they had like a fire hose of traffic that they could basically steer wherever they wanted to steer it. And so if you went on AOL and you typed in buy clothes-

   4. BHBen Horowitz3:00

      Half the traffic on the internet went through AOL at the time.

   5. MAMarc Andreessen3:03

      I mean-

   6. TCTyler Cowen3:03

      I remember that

   7. MAMarc Andreessen3:03

      ... it was a really big deal, so.

   8. TCTyler Cowen3:04

      Not my half, by the way.

   9. BHBen Horowitz3:05

      [laughs] Yeah. But that's a lot still, half.

   10. MAMarc Andreessen3:07

       Yeah. No, that's right. And then by the way, almost all consumers, right? Almost all consumers are on AOL at the time. So you type in b- you know, buy clothes on AOL, and it would... You know, and then they would sell, you know, to like J.Crew or Gap or, you know, whatever that slot. And then basically what happened is the advertising business, and then AOL would turn on that ad, and it would... And then the, basically, the fire hose of traffic would just blast the website to bits, right? Of like if it's J.Crew, they just blast the J.Crew website to bits. And then the J.Crew people would spend weeks trying to get the website to work to actually take advantage of all this traffic. And so it was sort of, you know... It, it just seemed kind of obvious at the... And then we, you know, we were kind of in this business, and so you go talk to the people running these things, and they just, they didn't... They were just unprepared for this kind of sophistication. And so it, we sort of incubated this idea that, boy, what if there was a cloud, [laughs] right? What if, what if basically there was a system in which you could take your content and all your apps and you could run them and it could handle the load, and you could... You know, it could load balance and it could spike up in response to demand, and then all these companies, instead of being in the business of running all your own stuff, like why can't you just hand it over to the pros? Um, and so... And, and then this is in the full flush of the dot-com boom, and so it just seemed like an obvious opportunity.

3. 4:04 – 7:58

   Founding a16z and Betting on Talent

   1. TCTyler Cowen4:04

      And so two thousand and nine comes along a-and why is it, it seems to you both then that's the right time for a new venture capital firm? What was the thought?

   2. BHBen Horowitz4:12

      Well, you know, it was really one of those ideas that just came out of our experience. You know, we were customers of venture capital and, you know, we had a couple of observations. One was, you know, we, we had noticed over the years, and, and Marc really made the observation first, that most of the really great companies, like forget about successful, but like great companies, companies you admire in tech, were all run by their founders for a very long time. So going back to Thomas Watson at IBM or Dave Packard and Bill Hewlett at Hewlett-Packard or, you know, Bill Gates or who-whoever it was, all the really giant successes were run by their founders. And the conventional wisdom in venture capital was set up to replace the founder. And then in our own experiences as technical founders, we knew why that was true because, yeah, well, if you just took, you know, Loudcl- like nobody would've ever been able to fix Loudcloud other than us. Like, there was no way you could bring in a professional to do that. And so getting to that next product after you get the first product, getting to the next product market fit required an innovator. And so if you wanted subsequent product cycles, you needed the founder. And so we just thought there ought to be a firm that is designed to do that.

   3. TCTyler Cowen5:24

      And what is it that the two of you figured out? What is it you two understood that other people didn't, and how would you articulate that?

   4. BHBen Horowitz5:31

      You know, like we, we were just... Th-there was a bunch of things. One, differentiation. So like we came from a company, so we're like, first, we're gonna tell a very clear, sharp story about a real network that we've built out systematically. We've got people who are going on your board who know exactly how to build companies.

   5. TCTyler Cowen5:46

      So sometimes it feels to some people there's a lot of money running around, but talent is quite scarce. What is it that you two and the company as a whole have understood about talent search that other people have not?

   6. BHBen Horowitz5:57

      So we actually thought of it as a talent business, and a lot of the reason for that is, you know, a friend of ours, Michael Ovitz, who founded the talent agency CAA, was a board member of ours at Opsware and You know, he always thought about talent in amazing detail and, like, i- incredibly specific. And one of his biggest concepts about it was it was a network, and you had to run the... And it seems simple. Oh, it's a network. Okay, great. But, like, how do you build that network? How does it work? How do you take a long view of the relationships with the talent? So, you know, we really thought out very early on, we're going to invest heavily in building this network and having the time, the luxury of time, to build real relationships as opposed to transactional relationships with every single person that we run into, be they an engineer, an entrepreneur, a corporate partner, or whomever. And that all kind of came out of this original CAA concept. If you take a long view of relationships, you can build a network that's so powerful that nobody will ever be able to match it. And that was kind of the original inspiration, and, and we were fortunate enough to just hire a really astoundingly good team to help us do that.

   7. TCTyler Cowen7:09

      And Marc, how do you think about talent assessment?

   8. MAMarc Andreessen7:11

      So I think, you know, the other side of it is just, is the entrepreneur assessment, right? And so it's just one of those things where, you know, there is kind of this fundamental question of kind of how extreme are you willing to get, right, with some of these people, right?

   9. TCTyler Cowen7:20

      [laughs]

   10. MAMarc Andreessen7:20

       'Cause, 'cause, you know, the, the ki- the kinds of people who start these companies are not normal.

   11. TCTyler Cowen7:24

       [laughs]

   12. MAMarc Andreessen7:25

       Um, and we can say that speaking from experience, having done it ourselves. And so... And I, I would actually go so far as to argue, like, it may be that the founders are getting less normal as sort of society gets, let's just say, more interesting in recent years. And so, you know, the ideas are getting bigger. The technologies are getting more disruptive. The companies that win are getting much larger. Technology's more central in everybody's lives. By the way, there's more competition than ever. There are more tech startups than ever. And so the kind of very special person who's gonna conceive of an original idea and then be able to build a team and be able to prosecute the idea is gonna be a very ex-extreme person. And so a lot of it is sort of this, um, you know, the discovery and then, and then partnering with these really extreme people.

4. 7:58 – 10:21

   Blockchain as a New Computing Platform

   1. TCTyler Cowen7:58

      I'm gonna throw out a few questions about particular technologies, and either of you feel free to answer. Blockchain, what will be the breakthrough application for blockchain?

   2. BHBen Horowitz8:08

      I'll take that. So first of all, like, asking what the killer app is, nobody ever gets it right. Like, I remember the internet killer app. It was never Facebook, that's for sure. But the way that we think about blockchain technology is that it's a new computing platform. And like other new computing platforms that preceded it, it's worse in every way but really a very few ways. So if you think about the smartphone, when it came out, it was much worse than the PC. It had a tiny screen. It was far less powerful, et cetera, et cetera. People were like, "How am I gonna run my spreadsheet on that little ass screen?" Like, "There, there's no way it's gonna work."

   3. SPSpeaker8:43

      [laughs]

   4. BHBen Horowitz8:44

      But it had a couple of properties that you didn't have in the PC. It had a GPS. It had a camera. And so you can now build things like Lyft. You can now build things like Instagram that you could never build on a PC, and you still can't build on a PC, and it created a whole new world of applications. Blockchain is like that. It's slower. It's harder to use. It's harder to program. But it has a new feature, and that feature is trust. And trust is really, really an interesting feature because it means that you don't have to trust the government or a corporation or your lawyer. You just have to trust the mathematical properties and the game theoretic properties of the system, and then you can do things. And it opens up applications such as you can program money. You can program contracts. You can create digital property, which is, you know, if you just think about the art world, it's an amazing kind of world, and it's all virtual value. You know, Basquiat's $150 million, but, like, the canvas is, is probably less than $5. But because you know that's one of one, it's got value. Well, you can now do that digitally with blockchain. So there's those kinds of things. And then you don't have to trust companies. So if you're a developer and you're building an application, you, you know, you don't have to trust Facebook to not go, "Well, like, you know, we decided, like, we're changing our privacy policy. You can't run anymore." Or if you're a consumer, you don't have to say, "Okay, like I, I've got to trust you with my data." So that property of trust, we think, is very, very powerful, and there's a large set of applications that will come off of that. But, you know, like, it, it takes a while for developers to get used to it, for the technology to mature, and so forth. But it's one of the things we're most excited about.

5. 10:21 – 15:33

   The Future of Retail, Wearables, AR and VR

   1. TCTyler Cowen10:21

      Paint a picture for me 15 years from now. Retail and wearables, what will tech do for me in those areas that it's not doing right now?

   2. MAMarc Andreessen10:30

      So I think retail will basically be gone by then. I mean, so-

   3. TCTyler Cowen10:34

      Gone? I can't go to the sh-shopping mall anymore?

   4. MAMarc Andreessen10:36

      Who wants to go to the shopping mall?

   5. TCTyler Cowen10:37

      I wanna go to the shopping mall.

   6. MAMarc Andreessen10:38

      [laughs] Oh, you can go.

   7. TCTyler Cowen10:40

      Okay.

   8. MAMarc Andreessen10:40

      They'll, they'll have special preserved shopping malls-

   9. TCTyler Cowen10:42

      Okay. [laughs]

   10. MAMarc Andreessen10:43

       -uh, for the people-

   11. TCTyler Cowen10:44

       Near Washington, D.C. for me

   12. MAMarc Andreessen10:45

       ... the Smithsonian. You're in D.C.

   13. TCTyler Cowen10:46

       Yeah.

   14. MAMarc Andreessen10:46

       The Smithsonian will have a shopping mall, um, that, that you can visit. Um, you... In fact, uh, you, you can drive there on a special road for your non-self-driving car.

   15. TCTyler Cowen10:54

       [laughs]

   16. SPSpeaker10:55

       [laughs]

   17. MAMarc Andreessen10:55

       Um, so, so and tie your horse up out back.

   18. TCTyler Cowen11:00

       [laughs]

   19. MAMarc Andreessen11:00

       So, um, so I mean, look, uh, there, there will be... There ex- uh, so the term's experiential retail, so, like, experiences. Like, if it's... Look, I mean, if it's, like, a Gucci boutique and it's a whole experience to go there or it's an Apple store, it's a... You know, it's got something where it's like there's, like, real magnetic appeal to the experience, then fair enough, right? And there's, there's actually a bunch of companies we're involved in that are doing things like that. But like the, you know, the idea of we're gonna buy a bunch of stuff that other people make, and we're gonna put it in a big box, and then we're gonna make everybody drive to the big box. Like, the problems with that are kind of twofold. Number one is consumers don't actually want... What you want is you want the Star Trek replicator, right? Like, what you want is like you press the button and like, there's my stuff, right? Like it, we're great on Star Trek, we don't have that yet.

   20. TCTyler Cowen11:35

       [laughs]

   21. MAMarc Andreessen11:36

       Like we don't quite have that, we don't actually have the materializer yet, but we do have the ability to press the button and stuff gets dropped off. So, and the, the logistics infrastructure to support delivery is getting built out, right, very, very rapidly now. And so the, I think the con- the consumer behavior is cutting over quite quickly. And then the other problem is just kind of traditional, let's say, third-party retail, where you're not selling your own product, you're selling somebody else's product, is you just can't... Basically you're, you're levered as a retailer, right? You, you basically live on the basis of credit from the suppliers. And the problem with that... So you're kind of like an over-levered bank in a lot of ways. And so the problem with that is you lo- There, there's basically no retailer of other people's products where they could lose 30% of their revenue and then they stay in business. Which is why you see these retailers just going bankrupt like all the time, right? It's like Toys R Us was the, you know, the most recent big one. Like, it detonated, and then it detonated so hard that it actually went into full liquidation, right? Which everybody thought it... Obviously it's Toys R Us, like people want toys. Like kids want... Like obviously somebody was gonna buy Toys R Us, and they, there was just no way to make the math work. And so I just, I just think it's an over-levered business and it's, it's, that, that part is gonna go down. Now, what it's gonna do is very interesting. It's gonna open up all the space, right? And so there, there's a whole revitalization of physical environments that's gonna happen, right? Including, you know, you see that happening already in cities. But I think it's gonna happen all over the place because a lot of the space is gonna open up for, let's, let's say more interesting uses.

   22. TCTyler Cowen12:45

       So instead of going shopping, I'm gonna do something with my wearables, and what will that be? What's the potential in wearables?

   23. MAMarc Andreessen12:51

       I think the really big one right now is, I think audio. You know, audio's on the rise just, just generally, and particularly Apple, you know, with the AirPods has just, I think, hit an absolute home run. It's one of the most deceptive, you know, things, is it's just like this little product and like how important could it be? And I think it's like tremendously important. 'Cause it's basically just like a voice in your ear anytime you want. And so you have the... Well, I'll just give you one random example. There are now these YouTube, you know, there are these new kind of new kinds of YouTube celebrities, and everybody's kind of wondering like, what, you know, where are people getting all this spare time to like watch all these YouTube videos and listen to all these YouTube, you know, people, you know, in the tens and tens of millions? And the answer is like they're at work, right?

   24. TCTyler Cowen13:22

       [laughs]

   25. MAMarc Andreessen13:22

       And they've got like [laughs] they got, they've got like a Bluetooth, a Bluetooth thing in their ear and they've got a hat, right? And that's 10 hours on the forklift, right? 10 hours of Joe Rogan, right? And so like that's a big deal. It's a voice in your ear all the time. And then of course speech as a, as a UI is, is, is rapidly on the rise. And so I think audio is gonna be, you know, titanically important. I would say the second thing I nominate for wearables is just generally the concept of sensors on the body, right? And so here the Apple Watch is, is clearly out in the lead with what they're doing with the heartbeat sensor. But I think we'll have a full complement of medical-grade sensors on, you know, on our bodies in a way that we have chosen to over the next five or 10 years. And I think we're, I think we'll get to the point where we're gonna be able to do things like predict heart attacks and strokes before they happen. But I think it's like, I mean, talk about like a, talk about a killer app. Like I'm, I'm gonna have a-

   26. TCTyler Cowen14:03

       [laughs]

   27. MAMarc Andreessen14:04

       Beep, I'm gonna have a heart attack in four hours.

   28. TCTyler Cowen14:05

       [laughs]

   29. MAMarc Andreessen14:06

       Maybe I should drive to the hospital. The survival rate for a heart attack in the hospital is like 99%. The survival rate for a heart attack at home is like 50%. Like there's an opportunity for like a massive increase of, of, of quality of life with the sensor platforms people are gonna have. And then I think, I think optics are coming, right? And it's gonna be a long road, but I think AR and VR are both gonna work, and I think they're, I think we're gonna have heads-up displays that are, that honestly are gonna remove the need to, you know, what we have now, which is kind of this little pane of glass that we're expected to kind of experience the whole world through, right? The whole world's gonna open up around us.

   30. TCTyler Cowen14:33

       What are we, what are we gonna do with augmented reality and virtual reality?
6. 15:33 – 19:22

   Martin vs. Fresh Prince and TV Lessons for Founders

   1. TCTyler Cowen15:33

      Ben, there's a tweet I've been dying to ask you about. "There's two types of people in this world, Fresh Prince of Bel-Air people and Martin people. I'm a Martin person, for what it's worth." Who's Martin?

   2. BHBen Horowitz15:47

      That's funny. I think Nate wrote that tweet and I, I replied to it. So [laughs] there were two, uh, kind of major African American television shows out in the early '90s, um, the Fresh Prince of Bel-Air, which was based on a rapper known as the Fresh Prince and his DJ, DJ Jazzy Jeff. The, the Fresh Prince was actually Will Smith.

   3. TCTyler Cowen16:06

      Him I know, but the Martin is what puzzled me.

   4. BHBen Horowitz16:09

      Martin is, uh, Martin Lawrence, who is a comedian, a, a genius comedian who's also incredibly crazy. So crazy that his special was called You So Crazy. Um, he's just like a very, very crazy guy. But Fresh Prince of Bel-Air was kind of like the hood Beverly Hillbillies. It's kind of like, you know, you get the Black people from the inner city and you put them in Beverly Hills, and it's kind of funny and it's safe for, you know, everybody. There's nothing too edgy. [laughs] You know, they keep it kind of easy. And whereas Martin was like just full out. Like Martin was like actually the hood, and he was nuts, and like the whole thing was great, and so I was, that was my show.

   5. TCTyler Cowen16:48

      Now it makes sense. Could you recommend a rapper for people who think they do not like rap music?

   6. BHBen Horowitz16:55

      Will Smith.

   7. TCTyler Cowen16:56

      Will Smith. [laughs]

   8. BHBen Horowitz16:57

      The Fresh Prince of Bel-Air. [laughs]

   9. TCTyler Cowen16:59

      And does Marc like Will Smith?

   10. MAMarc Andreessen17:01

       [laughs] I'm gonna fire up Spotify tonight.

   11. TCTyler Cowen17:05

       Marc, if we think about television as presenting conceptual material to us- And every now and then y- you'll watch TV shows. Uh, what's a TV show you've been watching lately that has a lesson in it about venture capital, and what's that lesson?

   12. MAMarc Andreessen17:20

       Can I name three?

   13. TCTyler Cowen17:22

       Three, yes.

   14. MAMarc Andreessen17:22

       Uh, I watch a lot of TV. Halt and Catch Fire, which actually recently ended after four seasons. Halt and Catch Fire, when it came out, it came out right after Mad Men, and it came out as kind of... People were like, "Well, it's kind of like Mad Men, but it's, like, much more of, like, a pot boiler. It's, like, super, like, dramatic." And they were like, "It's just all the emotionality of it." Like, you know, it's about this creation of... Basically, it's about the creation of a, a compact. The, the PC company in the early '80s. It's a thinly veiled kind of... It starts out kind of with, with that, the birth of the PC. And so all the critics were like, "Well, this is, like, too dramatic." And, and, like, and, and Ben and I watch it and we're like, "You know, no, that's... It's about right."

   15. TCTyler Cowen17:51

       It's exactly- [laughs]

   16. MAMarc Andreessen17:52

       [laughs]

   17. TCTyler Cowen17:52

       It, it, it was, like, literally going back in time. It was like-

   18. MAMarc Andreessen17:55

       That's exactly right. And it really is, like, that dramatic and that stressful and that crazy.

   19. TCTyler Cowen17:58

       Yeah.

   20. MAMarc Andreessen17:59

       And so especially the, the first season of that show, I think i- it's, it's the most accurate portrayal of what a tech startup is actually like that's ever been aired. So that's one. Another one that I really like, um, for founders to watch, and they always think I'm crazy, but I, I really, really believe this. Um, there's a, there was a great show on USA years ago called Burn Notice. Um, uh, which is a very fun show. Um, it was about a spy who'd gotten burned, a CIA spy who'd gotten burned and had to hole up in Miami w- to try, try to clear his name, and he took on all these odd jobs. So fairly normal setup. The conceit of the show was, though, he had every conceivable skill you could possibly have, right? And so he knew how to make, like, explosives out of bleach. Like, he knew how to, like, you know, I don't know, like disarm somebody with a mop handle. Like w- whatever circumstance he was in, he had the... And then there was a voiceover where he would explain to you-

   21. TCTyler Cowen18:38

       And you haven't hired him yet.

   22. MAMarc Andreessen18:39

       Well, we, we would, we would love to hire him. But, but basically I look at him and it's like, that's the kind of... That's, that's a good founder. Like a, a, a good founder has to basically have every conceivable skill. Like you basically have to be good at product development and at marketing and at sales and at finance [laughs] and at legal and at HR and management and, you know, on and on and on and on. And, and there really is no substitute for actually being good at all these things. And so I like that one. And then the third one is Succession, which I just finished, um-

   23. TCTyler Cowen19:02

       [laughs]

   24. MAMarc Andreessen19:03

       ... which is one of the darker, um, and funniest, uh, things I've ever seen. And let's just say, um, it's a, it's-- I would say it's inspiring for founders 'cause it, it, I think, pretty accurately shows the dysfunction at, um, let's say certain kinds of larger companies. It's a show about a succession battle at a major media company, and I can't recommend it highly enough if you've got the stomach for bad words.

7. 19:22 – 24:18

   Culture, Hollywood and Seeing Different Kinds of Talent

   1. TCTyler Cowen19:22

      So Ben, the company is starting something called Cultural Leadership Fund. What are the strengths and weaknesses of applying the venture capital model to culture and entertainment?

   2. BHBen Horowitz19:32

      Well, I think we're trying to apply culture to the venture [laughs] capital model, so i- it's a little bit the opposite. Look, you know, you know, back to your earlier question, you know, w- we really pride ourselves on being able to understand talent and talent of all kinds. And, you know, one of the things we did very early is we built a lot of relationships with geniuses at Moving Culture, and we thought this was important because as tech was moving into a much more kind of consumer-oriented field, uh, when we started the firm, that the people who really knew how to change and create new, uh, consumer behaviors would be interesting. So we, you know, had relationships with all these cultural geniuses like Quincy Smith and Sean "Puffy" Combs and Nas and, and so forth. But we were doing it kind of fairly one-off, and we thought, you know, it would be really great, you know... And, and it was a great advantage for us, and Oprah had one of our, uh, entrepreneurs on her Favorite Things show. But we thought, you know, it was a good thing to share with the rest of the industry, and so we would have these cultural geniuses but, you know, geniuses, but who didn't look like the geniuses our guys were used to, like, you know, Mark Zuckerberg or Brian Chesky. They kind of felt different. But our guys were interested in working with them, so we put them together. They get to know each other, which has got value on both sides. And it also gives a lot of value to our CEOs because not only do they get to, uh, kind of learn how to move culture, but they also get to learn about how a different kind of talent looks like, uh, which is very, very valuable when you're kind of, uh, in the war for talent. And then we invested back in kind of young African Americans who are wanting to come into tech. So we create a talent pipeline with the fund, and we have straight access to the pipeline. So I, I would just say we get a lot of credit for being nice, but we're really [laughs] just winning. So it's gone great, and I think it works well. And, and look, you know, the, the main thesis is, you know, if you've got like a very small group of people that created every new musical art form in the last century, from jazz to blues to hip-hop to rock and roll, you know, like that's a real thing, like to be able to do that, and that's a real talent base that, you know, we need to figure out how to get to.

   3. TCTyler Cowen21:45

      And we're here in Los Angeles. We're very close to Hollywood. What is it conceptually that Hollywood grasps about venture capital and talent identification where Silicon Valley lags behind?

   4. BHBen Horowitz21:57

      Well, I think that, you know, I just think of there are different kinds of talent. So... A- and this is the thing that I think people make a mistake on when they think about, you know, how diversity works or how inclusion works and so forth, is th- there's talent that looks different, and then if you don't have that talent, you might not be able to see it. And so, you know, in Hollywood, they see certain kinds of talent that they're used to because they know what that is. They know how it pops on screen. They know how, like, people emotionally connect to it. And then in Silicon Valley, we know another kind of talent. You know, a talent for like systems thinking and, uh, engineering and this kind of thing. But both are very valuable when you put them in together in a company. And so I think that, you know, in these endeavors, what we try to do is to see talent that we're not. And it's not an easy thing to do. And it's a story I tell that Margaret had in her profile, one of the things she looked at in her employees was helpfulness. And when I saw that, it shocked me because I had been managing for like nearly 30 years at the time, and I'd never interviewed anybody on that. I couldn't even see it Like there is a thing that's an important talent, very important talent to a services firm like ours that I couldn't even see. So how am I gonna hire it if I can't see it? And so we spend a lot of time at the firm trying to see talent that's not like us.

   5. MAMarc Andreessen23:16

      I'd also offer, could, the, the, you, this firm, t- the LA, the fact we're down here. So, um, I think it's also a very interesting time because, you know, for basically from when I entered tech in the early '90s up through call it maybe 2012, 2013, it was just kind of taken as a given that the Silicon Valley companies were never gonna figure out culture and never gonna figure out content. And it was also taken as a given that the media companies were never gonna figure out tech. And there were tons of attempts to kind of cross over, but like they basically didn't... none of them worked. And it really is in the last like three years it feels like. It feels like a bunch of the Valley companies are really starting to decode culture, uh, a bunch of the big... I mean, Netflix being the-

   6. BHBen Horowitz23:47

      Yeah, Netflix

   7. MAMarc Andreessen23:48

      ... Netflix sort of, you know, s- sort of setting a new model, but now being, you know, replicated by other companies, Amazon being the most notable example. You know, and becoming big forces in the cr- in the formation of culture and entertainment and media. And then, and then also, by the way, the other is, the flip side is the, uh, uh, a bunch of the big media companies now have gotten to the point where they now take tech incredibly seriously and have, you know, really sharp people working for them, working on very interesting projects. And then there's a whole tech thing obviously now happening down here in LA that's, that's of a new level of magnitude than before. Um, and so it does feel like both of the kind of central hubs of California are developing and crossing over, you know, quite nicely now.

8. 24:18 – 26:09

   Will Location Matter More or Less in 20 Years?

   1. TCTyler Cowen24:18

      A general question, 20 years from now, will location and being in the Bay Area matter more or less?

   2. MAMarc Andreessen24:24

      Yes. Clearly both. So on the one hand, it is absolutely true, I mean, in 20 years, you know, with the telepr- basically like telepresence technologies, right, so video conferencing and VR and all these things. Like 20 years from now, there's no question it's gonna be like much easier to run large distributed organizations, large distributed efforts than it is today, right? We're, we're gonna have such high fidelity video conferencing everywhere. Y- you can actually see this today. If you see the super high-end video conferencing systems, it really is like you are there. And then we have these robots that we love in our office. Some of you have seen the, the beams, which are a prototype of what I think... I think telepresence robots are actually gonna be a very big deal 'cause they give you a sense of physical presence of somebody in a room that's even different than a screen. And so like th- those technologies and then collaboration technologies like Slack and GitHub, right, are becoming really good today, and in 20 years, they're gonna be, you know, just spectacularly amazing. And so it's gonna be easier to run all these companies and be able to run all these efforts on a broad basis, and then that's just gonna make it much more straightforward for people all over the world to participate. So that's on the one hand, but the other thing though is just like, okay, that's gonna be a world that's much more connected, right, much more networked, right? It's gonna be, you know, past 5G. We're gonna be at like 9 or 10 or 11G, right? It's gonna be bandwidth everywhere. It's gonna be everybody's gonna be online all the time. We're gonna have all these, you know, wearables. Being online is gonna be part of everybody's daily experience all the time. You know, the economy's gonna reform itself around software network effects. And so the winning companies, you know, the winning entrepreneurial companies 20 years from now that win are going to be staggeringly large. Like, they're gonna be like, I don't know, 10 or 100 times the size of Google and Facebook. And so the prize to have a startup that scales and wins is gonna become so large that you're gonna wanna hyper optimize every possible thing you could possibly do to have that extra chance that you're gonna be the one that wins, right? And, and that's gonna mean, like, people in the same room together, right? And so I, I think the Valley's actually gonna become more central, not less central, even though the technologies that we're building are making it possible for the world to distribute.

9. 26:09 – 28:48

   Managerial Intelligence and Emergent Platform Behavior

   1. TCTyler Cowen26:09

      Ben, I love your book on management. It's the only book on management I've ever given to my daughter.

   2. BHBen Horowitz26:14

      [laughs] Oh, I appreciate that.

   3. TCTyler Cowen26:14

      I knew I had one chance. I picked yours. In your view, what is the best predictor, not of innovation, we've talked about that, but of simple managerial intelligence? How do you spot that, and what does it consist of?

   4. BHBen Horowitz26:26

      Well, you know, it's interesting. It's two skills that don't normally go together. So it's, it's systems thinking, uh, which is, you know... I hadn't even n-noticed-- Mark actually pointed this out to me many years ago, which is most people are not systems thinkers, meaning they cannot think about, "Okay, if I change this here, then it's gonna affect that over there." And you know, you know, as an economist, [laughs] people always make these dumb mistakes like, "Okay, well, move the minimum wage and nothing else will happen." It's like, well, no, no, it's a system. You have to think of it in terms of the system. And so that's kind of part of it, uh, and a big part of it. And but the other part, which is can you actually see the people in your organization, like do you know who they are as opposed to you're talking to them like they're you? And meaning do you understand their motivation? Do you understand what they would think about something if they weren't in the room and you're making a decision? Can you interpret them well enough so that it's as though they're there? And can you understand the implications through the eyes of the people who work for you? And if you have those two things together, those are the people who are really great, but it's a rare thing. And you can kind of see it because you'll be talking to them and, like, you might not be able to articulate something and they can articulate it for you the way you would've done it, better. [laughs] Um, like somebody who's that perceptive on people plus a systems thinker is really the-- tho-those are the people who are gifted.

   5. TCTyler Cowen27:54

      And Mark, did you really invent the tweet storm? And if so, what's the general lesson about innovation we should learn from that?

   6. MAMarc Andreessen28:01

      Uh...

   7. BHBen Horowitz28:02

      [laughs] What is the general lesson?

   8. MAMarc Andreessen28:03

      What is the general lesson? Um, inability to shut up, um-

   9. TCTyler Cowen28:07

      [laughs]

   10. MAMarc Andreessen28:07

       ... I think might have had a lot to do with it. Um.

   11. TCTyler Cowen28:09

       But you did invent it.

   12. MAMarc Andreessen28:10

       I think there might have been sequences of tweets, but literally I couldn't shut up, so like I think it, it kind of catalyzed. Well, look, I, I mean, the big lesson from it has been the, the big lesson actually of a lot of the, a lot of the internet platforms, which is emergent behavior is incredibly important, right? The really successful platforms let the users surface the behaviors, right, that the, that the creators of the platform could have never thought of. And, you know, Twitter's had all kinds of issues over the years, but, like, it always has been amazing. Most of the compelling ways in which people use Twitter have been invented by the users. I mean, I, I think it's true retweets were invented by users, right? Like very, very fundamental features. And, and that's not just Twitter. That's also been true... It was true of actually personal computers. It was true of smartphones. It's been true of, you know, many of these platforms. And so it's, uh, it's a useful principle of product design, which is let the users innovate.

10. 28:48 – 30:14

    Barbecue, Luxury and Non-Profit Sushi

    1. TCTyler Cowen28:48

       And Ben, you're famous actually for your barbecue cooking. View it as a problem of management and also innovation. What makes for the difference between good and truly excellent barbecue?

    2. BHBen Horowitz28:59

       Time.

    3. TCTyler Cowen29:00

       Say more.

    4. BHBen Horowitz29:01

       Time. So, you know, it's funny, I had, uh- Uh, an interesting conversation years ago with my wife's cousin, Atley, and Kanye West, which was just, like, a weird thing 'cause Atley's from Baton Rouge, Louisiana, and I happened to be in New Orleans, and Kanye was there, and we're all out to dinner. And Kanye asked Atley, he says, "What's the definition of luxury?" Which is like, if you're from Baton Rouge, you just don't think of that word, like you never hear the word luxury. And so Atley says, "I just like to cook." [laughs] And Kanye says, "Well, how do you cook?" He says, "Well, like, I like to make red beans and rice." And he's like, "Well, how do you do that?" He's like, "Well, I take my time. You know, I cut the onions very slowly. I boil it for a long time. I make sure it simmers." And Kanye says, "Exactly. Time is luxury. Like, that's why I make luxury records. I take my time." [laughs] And I was like, "Yeah, that's it." So... [laughs]

    5. TCTyler Cowen29:56

       Marc, do you prefer to eat for-profit sushi or non-profit sushi?

    6. SPSpeaker30:00

       [laughs]

    7. MAMarc Andreessen30:02

       This is Tyler's favorite question to suss out whether people are actually pro-government-

    8. TCTyler Cowen30:05

       [laughs]

    9. MAMarc Andreessen30:05

       ... um, pro increased government services. The idea of non-profit sushi makes me so nauseous-

    10. TCTyler Cowen30:10

        [laughs]

    11. MAMarc Andreessen30:11

        ... um, that I think I wanna throw up on stage.

11. 30:14 – 32:21

    What Wall Street and Government Misunderstand About Tech

    1. TCTyler Cowen30:14

       General question.

    2. BHBen Horowitz30:15

       [laughs]

    3. TCTyler Cowen30:16

       What's the one thing that Wall Street does not understand about technology that you would change if you could?

    4. MAMarc Andreessen30:23

       I think part of it is it's a 3,000 mile, like, thing. I think a big part of it's culture. There's just a delay, and there always has been. And so, you know, if I, if I wanted to fix that, I would say, like, we, you know, we need to spend a lot more time... B- and by the way, the tech industry does need to spend a lot more time, you know, trying to tell people what we're doing. But at the same time, people from outside the tech industry need to spend more time in the Valley and understand what's happening here. And a lot of that is happening as well. Um, that said, I'm not sure I wanna fix it. [laughs]

    5. BHBen Horowitz30:46

       Yeah. I don't think so.

    6. MAMarc Andreessen30:46

       The question assumes I wanna fix it. Like, I don't think I wanna fix it, 'cause that's a big part of the opportunity.

    7. TCTyler Cowen30:50

       What's the one thing the US government does not understand about tech that you would change if you could?

    8. BHBen Horowitz30:55

       So I think the first thing is something that Andy Grove said many years ago, which is, "It's inevitable." Um, and so, you know, somebody had asked him, you know, "Is the microprocessor good or bad?" He said, "Well, that's like asking is steel good or bad?" He's like, "We gotta deal with it. Like, it's here." And I think that, you know, the biggest mistakes the government makes are assuming it's not. So, you know, stem cells is a great one where the US government, uh, went to really hold that back. They didn't hold back stem cell, like, development or research at all. They just made it very inconvenient for people in the United States, and a lot of people died and, you know, missed out on cures and all kinds of things because of that. And, uh, so I think that's number one. And then I think the other thing is that technology has always had and always will have good and bad implications, uh, going back in f- to the cotton gin, the printing press. Uh, you know, certainly the internet, uh, has had good and bad. But if you look at it overall, it's overwhelmingly positive. And more than that, we have to go back to our population levels in like 1750 if you're gonna take away technology and take away technological advancement because the way the human population is growing, there's no way we can, y- you know, live the way we wanna live and have the lives we wanna live without technology. So the... getting into these debates of whether we should hold it back is just... You know, if there's one thing I would change, it's like, let's not have that debate. Let's have the debate how to make it great.

12. 32:21 – 34:37

    The Last Thing Software Will Eat: Picking Small Teams

    1. TCTyler Cowen32:21

       Marc, what's the last thing software will eat? [laughs] Other than sushi. [laughs]

    2. MAMarc Andreessen32:28

       Other than sushi. So I think it's fundamentally... The, the term that you used actually, you called it, I think, project selection. And so basically, it's this question of like, okay, how do you organize a small number of people to do something new, right? And, and by the way, that could be a startup company. That could be many other kinds of efforts where you need a small number of people to do something new. It could be a new political campaign, a new activist movement, whatever, but a small number of people to do something new, right? And then how do you pick, if you're gonna finance or donate or fund those things, how do you pick which ones to donate to? And then how are those things actually gonna run, right? And, and, and the, the new part there is really important. Like, the, uh, little known fact, for example, about venture capital is the, the, the, there's a term in venture capital and hedge funds called carry, which is, uh, basically called carried interest, which is the sort of profit participation that the VCs or hedge fund managers make. The term carry actually comes from whaling in the 1600s off, like, you know, the co- the, uh, in the Atlantic Ocean. Like literally, the, the term carry was the people who would finance the captain and the crew of the boat, but the boat actually would run as a startup. There was actually, like, equity participation for all the people on the boat, and then they would... You'd pick a captain. You'd raise money in town. You'd raise capital, and then the boat would go off and try to take down a whale, right? And about, you know, three quarters of the time, the boat would come back. The other, [laughs] the other quarter of the time-

    3. BHBen Horowitz33:37

       The whale would win

    4. MAMarc Andreessen33:38

       ... the whale would win. Moby-Dick was, was not a joke. Um, and, and so the, the boat comes back about 75% of the time. And then, and then literally carry was the portion of the whale that the investors got, right? And so if you think about it, like the process, if you're like in, you know, I don't know, it's whatever, Boston or wherever in like, you know, 1675, and you're trying to say, "Okay, this ship, this captain, this crew, this mission into these waters, right? With, like, these weather patterns. Like, and how are they gonna behave under pressure? And what's gonna happen when things go wrong? And is the crew gonna mutiny? And like, are we gonna make any money doing this?" Like, the whole thing is just such an intricate kind of puzzle, and it revolves around people. And so th- taking... starting at the whaling expedition, you know, s- you know, the, the Nina, the Pinta, and the Santa Maria, like it's the same kind of thing. Tech startups are the same way. By the way, you know, green-lighting a movie or a TV show in Hollywood is the exact same kind of process. And it's just, it's so intangible, and it's so much based on the interaction of a small number of people who are gonna be under extreme pressure. Like if we, if we could figure out a way to automate that, like we'd fund that company and then retire, but at least we don't know how to do that.

13. 34:37 – 38:24

    Book Recommendations and Closing Misconceptions

    1. TCTyler Cowen34:37

       For each of you, what's an interesting book you've read lately?

    2. BHBen Horowitz34:40

       Yeah. So the... one of the most interesting books I, I read lately is, uh, Genghis Khan and the Making of the Modern World by Jack Weatherford. And it turns out to be very unexpectedly the most interesting book on the topic of how you think about inclusion that I've ever read. And Genghis Khan is not known for his thoughts on inclusion-

    3. TCTyler Cowen34:59

       [laughs]

    4. BHBen Horowitz34:59

       ... because he's mostly known for, like, being just ruthless. But he really... You, you know, he was a guy who came from kind of the border of Northern Mongolia and Siberia, very bad part of the world. He had a very, very hard life growing up. He was from kind of the lower, uh... They had white bones and black bone, kind of the higher and lower caste. He was a lower caste person. And You know, a lot of his experience growing up led him to this idea that he should choose for kind of talent, not the caste system, and not even the tribe that he was in, which was a huge breakthrough at the time. You know, nobody had done that. And the way he thought about it and the techniques that he used for doing it were breakthroughs today, [laughs] you know. And, and so I, I just found it to be like a super interesting book. Definitely a great management book for anybody who's interested in that topic.

    5. TCTyler Cowen35:49

       Marc?

    6. MAMarc Andreessen35:50

       So, uh, the best book, the b- book that's had the biggest impact on me, it's an incredibly well-written book. It's, of course, out of print. It's by, uh, actually, a, a guy I think, uh, uh, Tyler, you know, uh, Timur Kuran.

    7. TCTyler Cowen35:58

       Sure.

    8. MAMarc Andreessen35:59

       Uh, you, you know quite well, s- uh, who's a economics? I think eco-

    9. TCTyler Cowen36:02

       Economist at Duke.

    10. MAMarc Andreessen36:03

        At, at Duke. So it's a book about 20 years ago, it's called, uh, Private Truths Public Lies. And it basically tells the story, the theory that he basically has, he calls preference falsification, and it's basically the idea... It's, it's a situation in which people believe something in their own head, and then they feel for social reasons that they can't say it out loud. And so it starts kind of with that as kind of a, a, an idea. Then it kind of extrapolates out kind of what happens as a society becomes the kind of society in which people feel like they can't speak the things that they believe. And it turns out to be quite an elaborate process because basically, right, you, you can have these very interesting situations where you can have a majority of people who believe something, but then they all believe they can't say it. Uh, but then as a consequence, they all come to believe that there are many fewer people who believe it than there actually are. And so you can kind of suppress a point of view artificially for quite a long time. But then he describes in the book how there-- you can then kind of get the reverse process. Kind of you can kind of get the whole thing to, you know, kind of the spring to expand, which is if a few brave people start to speak up, then a lot of other people who have had that secret belief all of a sudden realize they're not alone, and then you start a cascade, right, in which a lot of people start to speak up. And he basically models in the book like that's where revolutions come from. It's basically like an explana- it's like an explanation for, for the fall of the Berlin Wall. It's an explanation for, you know, for, for political revolution. It so happens to be, I think, highly relevant to what's happening both on the left and the right in the US like right now. Like I think it's... As you read the book, you're just like, "Okay, that's the cleanest explanation of the Trump phenomenon I've ever seen." And furthermore, it's also the cleanest explanation of the Bernie phenomenon I've ever seen. Like, I think it actually describes both quite accurately.

    11. TCTyler Cowen37:28

        For the last question, I'd like to return to Marc and Ben as a couple. Ben, what's Marc's biggest misconception about you? And Marc, what's Ben's biggest misconception about you?

    12. BHBen Horowitz37:39

        So this is the sad thing, is that he knows me so well that I wish he had misperceptions about me.

    13. MAMarc Andreessen37:45

        [laughs]

    14. BHBen Horowitz37:46

        But, like, he actually knows who I am. [laughs] And so this is... And, and, and it manifests itself the worst, like, if something's going wrong at the firm, it's always some combination of his and my fault. Um, and he'll know exactly the flaws that I have that have led us to that situation, and, like-

    15. TCTyler Cowen38:02

        [laughs]

    16. BHBen Horowitz38:03

        ... it's, it's unbearable. And, and, you know, and vice versa. But I'll, I'll let him answer.

    17. TCTyler Cowen38:08

        And Marc, pl- you have the last word.

    18. MAMarc Andreessen38:10

        Ben's biggest misconception of me is he thinks I'm gonna go to my room tonight and listen to Will Smith. [laughs]

    19. TCTyler Cowen38:14

        [laughs] I thank you both very much for this dialogue.

    20. BHBen Horowitz38:19

        Thanks a lot.

    21. MAMarc Andreessen38:19

        Great.

    22. BHBen Horowitz38:19

        Thank you.

    23. MAMarc Andreessen38:19

        Thanks, everybody. [clapping]

Episode duration: 38:24