Auren Hoffman (00:02.338)

All right. Hello, fellow data nerds. My guest today is Ben Horowitz. Ben is the co-founder of Andreessen Horowitz, one of the most influential venture capital firms in Silicon Valley. Prior to his VC career, he founded Opsware slash LoudCloud, which was acquired by Hewlett Packard for $1.6 billion. Ben's also the author of the bestselling book, Hard Thing About Hard Things, which is personally one of my top three favorite business books. And also what you do is, what you do is who you are. Ben, welcome to World of Daz.

‍

Ben Horowitz (00:31.677)

Thank you, thank you. It's great to be here.

‍

Auren Hoffman (00:34.654)

Yeah, I'm really excited. Um, now, uh, we've been talking a lot about AI and you guys are famous for the fact that software is going to eat the world. Do you think AI is going to eat the world the way software did over in a slightly different way or.

‍

Ben Horowitz (00:52.165)

Probably in a little different way. You know, it's kind of in a way, it's a continuation of self-reads the world. And in a way, it's different in that kind of when we talk about software historically, we've kind of had this deterministic, very factual, very literal kind of software. And, you know, with AI, we have this probabilistic, non-deterministic software.

‍

And so one, you can solve a whole class of problems that we could never touch with traditional software, but also the nature of it is quite a bit different than traditional software in the way it does things and potentially makes mistakes and what not.

‍

Auren Hoffman (01:42.482)

Is there is this is kind of like a fundamentally new moment though, right? Or you disagree?

‍

Ben Horowitz (01:49.333)

Oh, I think so. I think it's a fundamentally new moment. And that, look, we're going after a class of things that we could never go after before. So that makes it, it's almost like, oh, we have a microprocessor. It's kind of that level of change.

‍

Auren Hoffman (02:05.858)

The narrative is that AI will be, you know, there'll be some like heavy advantages to these large incumbent players who have lots of data who can build these like very huge models and afford these massive compute costs. Do you buy that narrative?

‍

Ben Horowitz (02:21.505)

Well, I think that it's a little bit TBD. So I think on the current, you know, course and speed that's been where it's been going in the sense that, you know, basically the algorithm change advantages, particularly for LLMs haven't seemed to be as significant as, you know, just bigger models. However,

‍

I think that it's really early on and we don't know how far the scaling goes on both the data side and the compute side. And we don't know what the open source real counter attack is yet. So I think meta sort of shocked the world.

‍

changed the landscape by putting out, you know, llama. And llama is really good. And one could imagine if the whole world rallied around that and, you know, Metta found a partner or came up with a way to not only give out the open source but to serve llama. That could change things a lot on the big side. And then, you know, one of the things we're also saying is just...

‍

For certain kinds of tasks, a lot of what businesses want to do, a large model can be, for interpreting English or that kind of writing code, there's no substitute for it. But for a lot of other kinds of tasks, you might want a smaller, faster, cheaper model. So I think there's a lot to learn still.

‍

Auren Hoffman (04:11.538)

One of the things is that you have these, a lot of this is based on just like crawling the internet. And obviously most sites allow Google and Microsoft to crawl them, but many of them, most of the biggest ones don't allow really anyone else. So these two companies have a bit of an advantage when it comes to just like doing the crawl itself.

‍

Ben Horowitz (04:33.197)

Right, they kind of have the data, which is one thing, and then they have the money to buy more data, you know, kind of buy proprietary data and so forth. So that's, you know, certainly definitely part of it, but one could imagine, you know, there's always a, the internet was like this, by the way, where I think if you go back to the early, if you go back to 94, almost nobody in the

‍

win, that the open version would win. Everybody thought that it would be controlled by a single company, probably Microsoft. And in fact, when Bill Gates wrote The Road Ahead, his big book about the future, yeah, he didn't mention the internet in the first edition. So it wasn't a fait accompli that the internet would be open. And I don't think it's a fait accompli yet that the large models are gonna be proprietary.

‍

Auren Hoffman (05:14.782)

Yeah, I don't think he even mentioned the internet really, right?

‍

Ben Horowitz (05:32.193)

You know, like I think for the good of humanity, it would be nice if it wasn't proprietary. And I think that if everyone, I mean, the thing that happened in the internet is the internet got at least, you know, in many ways, it wasn't as functional as Microsoft network originally, but it was functional enough and then had the entire world working on it and contributing to it. Whereas

‍

you know, only Microsoft is contributing to MSN. And, you know, hopefully that'll happen and it won't, you know, the world can defeat Google and OpenAI and some of these guys.

‍

Auren Hoffman (06:11.663)

Now, we, you know, one of the things we're looking at ways to kind of like improve our workflows and kind of try to hand them off to AIs. Like I'd be, I'm really excited to have these like agents working for me. How does that look like for like either the average person and the average worker over the next decade?

‍

Ben Horowitz (06:31.069)

Wow, you know, that I would say it's a lot unpredictable. But, yeah, well, I think that everything that, you know, is no fun to do, you ought to be able to hand off to an AI and, you know, this is everything from, you know, building your personal network to doing mundane tasks.

‍

Auren Hoffman (06:37.638)

What are you most excited about there?

‍

Auren Hoffman (06:47.905)

Yeah.

‍

Ben Horowitz (06:56.549)

like, you know, kind of figuring out how to order food and all these kinds of things or book a flight. You know, those are certainly kind of coming down. I think it'll get, though, you know, much more aggressive than that. And one could imagine all the things like, you know, just on like, if you take something simple, like, okay, now to book a flight, I've got to pull down all these windows and this and that and the other. Can I just say, hey, I'm going to Paris, can you get me there?

‍

And then I was talking to my friend Ariel, who's the CEO of Navon, which does travel stuff. And he said, like, why do you have to tell it you're going to Paris? It already knows that. And so, maybe you just have like the most helpful friend in the world who can do anything and they know you and you just kind of have conversations with it from time to time and give it access to your calendar and your email and it just does its thing.

‍

Auren Hoffman (07:53.326)

And even if you think of like a UI path, like UI path was like, you have to bring in like Accenture and it was like, you have to have, you spend millions of dollars on it. So it really wasn't applicable to like most work situations. You could see a scenario where we all have like at least our own mini UI path for us.

‍

Ben Horowitz (08:06.013)

Yeah.

‍

Ben Horowitz (08:10.385)

Oh yeah, no, I think so. And then, you know, like every kind of software tool that you use, you probably need to completely rethink. So like, you know, we have all these salespeople using like Salesforce, but Salesforce seems like the complete wrong idea for, you know, an AI world in that. Like, why am I entering like what I think happened in a call I made, you know, into a system and incomplete. And then, you know, I have to worry about my job too. And.

‍

Auren Hoffman (08:34.219)

right.

‍

Ben Horowitz (08:40.317)

you know, whether I'm going to get fired because my forecast isn't high enough and this and that. And then a manager looks at it and adjusts what I said. And then some other manager looks at what that manager did and adjusts what he said and all that kind of thing. We like, why not have the AI just track what I'm doing, the emails I sent, the contracts I exchanged, the phone calls I make, the meetings I go to and make the forecast for me based on like real information. And then, yeah.

‍

Auren Hoffman (09:03.553)

Yeah.

‍

And update it too. Like, I mean, just updating the CRM is such a pain in the butt, especially Salesforce, cause it's like 17 clicks to enter one thing or something.

‍

Ben Horowitz (09:10.822)

Yeah.

‍

Ben Horowitz (09:15.613)

Oh, yeah, yeah. It's like literally technology. It's like we used to call that a 4GL back in like the 80s when I started my career. It's just like completely absurd. But, you know, that's where it is. And so, yeah, I think that, you know, everything is good. I guess my big assumption is that everything is going to change and in very, very unpredictable ways.

‍

Auren Hoffman (09:40.458)

Now there's been like some of the other big waves, the internet, the cloud, mobile, like AI seems like something that's actually going to impact all of our lives. Crypto which was like kind of built as a great, like didn't really seem like it had the potential to help everyone. Maybe there's this idea of remittances or something or am I underselling crypto?

‍

Ben Horowitz (10:02.501)

I think you're underselling crypto. I mean, I think that crypto has run into this weird regulatory morass in the US due to, I would just say, a very weird specific agenda from the administration. But as a technology, the way I would think about it is the history of computing is sort of the history of computers and networks and going back to microwaves and integrated circuits.

‍

and all these kinds of things. And they kind of evolve at different speeds and in different ways. And the network, the value of the network is always harder to see than the value of the computer. Because the computer, it's just like, ask it a question, it tells you the answer. It's like very obvious, it's amazing. But even with the internet, like in the beginning, people like businesses were like, the internet's nothing, we're not gonna even use it.

‍

Auren Hoffman (10:46.943)

Yeah.

‍

Ben Horowitz (10:58.225)

Maybe we'll build an intranet, but we're definitely not going out on the internet. And then some micro systems ban the use of the internet for employees. And like, I think 1995, pretty amazing. But nobody, you know, with networks, because they're a network effect, it's really hard to see the importance until, you know, there's enough people on it where it's, it's a critical mass. And I think with crypto.

‍

Auren Hoffman (11:10.26)

Oh wow, I had no idea.

‍

Ben Horowitz (11:27.761)

You know, it solves all the problems we have had with networks since the internet. Um, and, you know, particularly the kind of concentration of power over really important things like the truth. Um, yeah, companies control what's true now. And I think that is probably the single most dystopian part of what's happened with social networking and, you know, how it combines with the government and so on.

‍

And so if you think about an AI world, figuring out what's true just got way harder. You have deep fakes that are going to be such high quality by the time, by the way, the election cycle hits, that it's going to be impossible for anybody to determine what's what. So, you know, really, the answer is like, if you or I say something, we ought to assign it and kind of put it on the network.

‍

Auren Hoffman (12:04.811)

Yeah.

‍

Ben Horowitz (12:25.449)

Do we want to sign it and put it on a public network that is controlled by nobody? And if we say it's true and came from us, that wins. Or do we want to have a single company decide what's true and have whatever political agenda or whatever political pressure they have to make that determination? And I think that, you know, the world with the crypto answer is such a better world than the world without it.

‍

So I think that technology is gonna prove to be exceptionally important. 

Auren Hoffman (13:04.886)

What I hear you saying, though, is that maybe AI can be like crypto savior, like AI, like there's like a good, better reason for crypto, because there's like a safety. And you would think with this administration, like safety would be something they would maybe jump on or something.

‍

Ben Horowitz (13:14.789)

Well, he's right, Faye.

‍

Ben Horowitz (13:20.285)

Yeah, I certainly hope so. I mean, I think it is the it is in some ways, maybe the killer use case for the technology. And by the way, like, the current blockchain technology would already work brilliantly for that idea. You don't even need to really enhance the infrastructure to make that happen.

‍

Auren Hoffman (13:34.09)

Yeah. Okay. Interesting.

‍

Auren Hoffman (13:38.922)

Now, one of your most influential management concepts is this idea of like a war time versus peacetime CEO and that there's these different leadership styles that are optimal in different times of crisis versus times of growth. And they're basically I think we're now in a situation where there's all these amazing peacetime CEOs that find themselves operating in wartime. Like, what's your advice to them to make that transition?

‍

Ben Horowitz (14:08.165)

Yeah, so the way I would think about it is, you know, imagine, you know, you're kind of, you've built an army to fight the Cold War against the Soviet Union. And you've hired you, you've profiled all the people you've hired to fit into that. You've created a management structure. The deal, they speak Russian, they understand that part of the world. You have goals and objectives and KPIs around that and so forth.

‍

Auren Hoffman (14:26.874)

I'll speak Russian.

‍

Ben Horowitz (14:36.601)

And one day you wake up and you're fighting ISIS. Probably all that apparatus that you set up is incorrect. So the way that you've been managing, let's look at our KPIs. Let's see that, you know, isn't going to work. You have to like quickly change the army in pretty radical ways. You probably have to change out a lot of your leadership. You've probably got to, you know, find a different kind of soldier. You probably.

‍

you know, your KPIs are all irrelevant. And the only way to get from here to there, you know, you can't kind of recreate that whole apparatus that you've set up to fight that war. You've got to move faster than that. And that involves, you know, the CEO making very, very sharp, fast decisions. And, you know, sometimes going low in the organization to make them.

‍

so that people get the object lessons of, oh, no, we're setting up over here. And once you get the army to move, you can do that. A really kind of great recent example of somebody not making the transition was the Silicon Valley Bank crisis, right? Where you had Gary Becker who, okay, they did a dumb thing that kind of sent them towards war, which is they bought a lot of long dated, you know,

‍

Auren Hoffman (15:48.854)

Yep.

‍

Ben Horowitz (16:01.097)

treasury bill. It wasn't the dumbest mistake in the world, but like, it just turned out when interest rates went up that having a bunch of 30 year T bills was going to make your balance sheet look really bad. And then, you know, the Financial Times article hit that pointed that out. And then kind of at that point, the race was on. So what they needed to do to kind of build confidence was they needed to turn their balance sheet upside right, which required, you know, whatever, $3 billion.

‍

not an unsolvable problem, but he stayed in peacetime. So the way he went, the right way to do it was to switch to wartime. Okay, everything we thought about raising money, like, fuck that, that's no longer true. Like we're gonna do, we know the playbook, Goldman Sachs ran it, we're calling Warren Buffett. We're going to give him the deal of a lifetime, warrants, a discount, coupon, whatever he needs.

‍

Auren Hoffman (16:53.025)

Yeah.

‍

Ben Horowitz (16:59.497)

And he's going to get money in and he's going to scare off the short sellers. And then like, we're going to stop the run on equity and there's never going to be a run on the bank. That was the right one. What he did was he did the exact same thing that he would have done. If he wasn't at war, which is, Oh, I'll call my friends at general Atlantic. You never mind that they're having trouble raising their own fund or whatever. But like, you know, and I'm going to give them a market deal. And so of course the deal never goes through.

‍

the run on the equity turns into a run on the bank and like that's the end of the company. And it was very fast, you know, in his defense, that whole thing changed lightning quick. But I think that's generally the case for all these CEOs is like everything changed overnight. And now they've got to get the company going in the other direction. And the problem that everybody faces at CEO when they're in that situation is even their own team doesn't understand what the hell is going on.

‍

Like there's one person who understands it and you have a limited time to get the whole organization to understand it and make all the very kind of almost absurd changes if you think about it in the old world to get you to the new place.

‍

Auren Hoffman (18:11.722)

Is that even possible for now a non-founder CEO to have the authority to even make to do that?

‍

Ben Horowitz (18:19.105)

You have all... I think that it's very hard for a professional CEO to do that, to come in. Because the problem is, there's all these assumptions in a business, all these kind of things, principles, culture, product ideas, so forth. And as founder, you made all those. And so you know which ones you can get rid of safely, because they were just wrong, like they were based on

‍

you have to keep. If you're coming in from the outside, I don't know how you do that. And it's funny, because, yeah.

‍

Auren Hoffman (18:53.994)

And do you even have the authority? Cause you, you know, you, you don't own enough of the company or you don't have the moral authority necessarily to, you know, you have the antibodies in the company that are going to rebel against you. You know, you mentioned the defense department. I doubt like a secretary of defense could do that.

‍

Ben Horowitz (19:07.961)

Yeah, no, I think it's a very hard thing for a... Well, I'll give you a funny story on that. So like after I sold Opsware to HP, I got a call from a private equity firm. And they said, Ben, like, we want you to join us. Like they just like that, like no interview, no nothing, just like a cold call. We want you to join the firm. And I said, well, like why? And they said, well, because we think that the loud cloud

‍

opsware turnaround was the greatest tech turnaround last 20 years. And I said, well, you are aware that I'm the one who fucked it up in the first place, right? You know, I don't want to pick somebody else's company.

‍

Auren Hoffman (19:47.982)

Right. Maybe easier to fix your own mistakes. Right. Now, a lot of what you cover in the hard things about hard things is about dealing with some of like the toughest things that CEO face, layoffs, firing people, surviving downturns.

‍

Ben Horowitz (19:51.046)

Yeah, for sure. For sure.

‍

Auren Hoffman (20:04.062)

In some ways, you're now in your moment in the last couple of years. Right? Yeah. It's like, it's like, it's like, we can like bring back then, like, what's your, what's your broad advice to CEOs at this time?

‍

Ben Horowitz (20:07.466)

Yeah, this is my town!

‍

Ben Horowitz (20:13.618)

Yeah

‍

Ben Horowitz (20:17.413)

I'm relevant again. So the most, I would say that the quote that I like the best for CEOs right now is from Al Davis, who used to own the Raiders before he passed away. And he used to say this thing, I'd rather be right than consistent. And I think that's the key. Like the need to hang on to consistency, to make whatever you said before is true.

‍

Auren Hoffman (20:18.998)

Heheheheheheh

‍

Ben Horowitz (20:46.362)

is the thing that will kill you right now.

‍

Auren Hoffman (20:49.758)

And how do you, how does one, I mean, is there a way to, it's hard to disagree with yourself, maybe from yourself 20 years ago, it's one thing, but yourself from one year ago, it's much, it's much more difficult or two months ago, right. Especially if you like written like 10 blogs and all these internal memos about it and you've tweeted about it. And now all of a sudden you're like, ah, you know, I'm that guy who looks like me and has the same name as me is completely wrong.

‍

Ben Horowitz (21:03.795)

Or two nuts ago.

‍

Auren Hoffman (21:19.198)

Like, how does one get the fortitude to do that?

‍

Ben Horowitz (21:24.541)

Well, I mean, look, I think that you have to, it's a measure of humility that you're okay with people, one, realizing that you're wrong, and then the confidence to like continue to be the leader. Because if you were wrong before, why should we believe you now? That's the fear, right? And actually, hilariously, you know, when I was at LoudCloud, when I was running LoudCloud, I said at some point, you know, Opsworld will never be on a CD.

‍

because that's how we deliver software in the old days. And as soon as we kind of split off Opsware and made that the company, every engineer was like, you know, quoting me on that and so forth. I was like, welcome to me, and I'm like, got that one wrong. But, you know, I think what you have to be willing to say is, look, you know, the facts changed, so I've got to change my opinion.

‍

Auren Hoffman (22:10.446)

Sorry, I was wrong.

‍

Ben Horowitz (22:23.225)

It's just like you have to deal with reality and like, why me? Well, like I now have the kind of, I've got more experience in this market and with this technology than anybody. And so I've learned a lot, you know, um, the old me wasn't as good as the new me. And that's why the new me is changing that bad decision. The old me made. And you know, like that's the best you can do, but the worst thing is to try and kind of stay with the old idea.

‍

Auren Hoffman (22:39.969)

Yeah.

‍

Auren Hoffman (22:51.126)

really admire like traders who, you know, they had a view of the world, they made a trade. And for whatever reason, that view turned out to be wrong. And they unwind the trade, they take their loss, and they move on and do something else like, but I think it's a very special person who can do that consistently.

‍

Ben Horowitz (23:05.211)

Yeah.

‍

Ben Horowitz (23:09.945)

Yeah, no, that is true. Look, it requires that you don't attach your success to like you personally, you know. Your success is a result of the things you do and the decisions you make. And just cause it's you doesn't mean that's always gonna be correct. You know, you can make a lot of mistakes. And I would say that's the psychological error that, you know,

‍

people have great success early often make is that, oh, I did it, and it worked because I did it, and so the next thing I do is gonna work, and that's not at all how the world is.

‍

Auren Hoffman (23:47.594)

Paul Graham has a saying like, keep your identity small. But I think that's easier said than done.

‍

Ben Horowitz (23:54.377)

very much easy because look everybody around you is like pumping it into your head, you know, once you're doing it.

‍

Auren Hoffman (23:59.18)

Yeah.

‍

Now I love the anecdote in Hard Things about Freaky Friday, where you basically like swap two of your execs to running each other's functions. Have you ever actually seen that done since?

‍

Ben Horowitz (24:13.949)

Yeah, no, actually a lot of entrepreneurs have told me they've done it. So the key to it, which I probably, I don't know if I made it clear enough in the original post is you have to have the skillset on both sides to swap it. So, you know, I ended up, I swapped like the head of professional services and SEs for the head of support and they both kind of had the same skillset. So that worked fine. I think if you tried to swap the head of engineering with the head of HR, you'd have a real problem.

‍

Auren Hoffman (24:36.404)

Yeah.

‍

Auren Hoffman (24:41.738)

Yeah, yep.

‍

Ben Horowitz (24:43.385)

But the people who clash tend to be the people who think they can do the other person's job. So it does work in that sense.

‍

Auren Hoffman (24:50.986)

Okay, got it. So there's the more people are alike. I'm just like the Gerardian thing, the more they clash or something like that.

‍

Ben Horowitz (24:57.569)

Exactly, it's exactly the Gerardian thing. That's a great kind of analogy.

‍

Auren Hoffman (25:03.654)

Now building a strong culture is kind of a recurring theme in almost all your work. Like, how does, like, how does, what's the biggest way that people do it right? I feel like there's lots of easy ways to say people do it wrong, but what's the way people do it right?

‍

Ben Horowitz (25:20.697)

Okay, so then, and I got this from the samurai, which by the way, you know, samurai culture, you know, had a lot of issues, you know, in terms of like what the culture actually was, but in terms of the consistency, and the conformity to the culture was maybe the most amazing example in human history. And the thing that they really got right was a culture is not a set of beliefs, it's a set of actions.

‍

And I think that anybody who defines their culture in terms of beliefs, we do the right thing, we have integrity, blah, is gonna feel that never works, right? Like it's what are the behaviors that you have to have here in order to stay here? Like that's the thing, like what do you have to do? Which is why the name of the second book is What You Do Is Who You Are. It's like, what do you have to do to be in this culture?

‍

Auren Hoffman (26:01.386)

customer comes first or whatever.

‍

Ben Horowitz (26:19.377)

not what do you have to believe. Because what do you have to believe that like it's nothing. Like what you tweet, what you say, what you blah, nothing. Like how do you conduct yourself?

‍

Auren Hoffman (26:23.24)

Interesting.

‍

Auren Hoffman (26:29.634)

So this is why like religions have these very, very odd rules. Like you can't eat shellfish or something like that. Um, cause it's really about what you do.

‍

Ben Horowitz (26:35.729)

Right, exactly. Like rules like that create a culture, absolutely. Yeah, yeah.

‍

Auren Hoffman (26:40.574)

Yeah. Okay. I never really thought about that. That's really interesting. Um, there's some questions on, uh, on, on just building a VC firm. I heard you say recently that, um, that compared to the rest of venture entries in Horowitz is most similar to YC. Like why do you make that comparison?

‍

Ben Horowitz (26:58.593)

I think we were most similar in that, you know, both of us were like, wow, the way VC works is weird in that it's all about them in that like, okay, are we good at picking investments, da da, but nothing about the product that they offer to entrepreneurs. And I think the breakthrough for both Andreessen Horowitz and NYC was like,

‍

we're going to start with what does the entrepreneur get? And then like, we'll worry about like deal picking second. And I think that, you know, kind of has been the biggest change in VC over the last 15 years has been kind of us and them saying like, what is the offering? Like, you know, as like, what is my product?

‍

And is it competitive as opposed to just am I like really smart and everybody wants to be around me? Cause I'm so f*****g smart and da da. Um, we're, which is like, we just haven't seen it work over yet.

‍

Auren Hoffman (28:03.918)

And GPs don't scale either, right? Like GPs don't scale. So you need to have something. Most people don't even know who the CEO of YC is, right? But it's amazing. Now, VCs give startups advice to create a monopoly, build a differentiated product, do dilution to accelerate growth, have one CEO, eventually go public.

‍

Ben Horowitz (28:11.306)

Yeah, yeah.

‍

Ben Horowitz (28:16.241)

Right. Here it was. I do. There it is. Yeah.

‍

Auren Hoffman (28:32.802)

Probably none of these things like most VCs take like, why are they so different?

‍

Ben Horowitz (28:39.385)

Yeah, so I think it kind of goes back to what you said, which is scale. And if you look at venture capital, historically, the market, the market for software was just smaller and the market for VC was smaller. So there used to be a thing when we started Andreessen Horowitz that all the VC said, which I think was correct at the time was like, there are 15 companies a year that will get to a hundred million in revenue.

‍

And your job as a VC is to invest in as many of those 15 as you can. So that was the like the scope of the industry. And I think like there's no way anybody would say that's true now. It's much closer to 150 or 200. And so now like scale actually is interesting. And then all those things that you refer to are kind of scale ideas. And which is why, by the way,

‍

You know, if you look at how our firm works, there is a CEO, that's me. There is like, you know, there is dilution in that, you know, we have like lots of capabilities to build a product and so forth. So we kind of follow that model a bit more, you know, that we advise people on, but it didn't make sense for the traditional VC, like a benchmark to do that because they're chasing 15 deals.

‍

And so they never needed to scale six partners, $400 million, 15 deals. Like the whole thing was that math.

‍

Auren Hoffman (30:12.51)

And in some ways, I mean, both obviously can work, right?

‍

Ben Horowitz (30:15.441)

Yeah, yeah, yeah. Look, I think the current world is better suited towards us and the old world was better suited towards them. And that's just how it is. But that's my opinion.

‍

Auren Hoffman (30:28.622)

Well, yeah. I mean, the scale of venture capital has changed massively over the last decade. I believe you guys are now at like 35 billion AUM or somewhere. Um, I mean, which is just kind of like mind boggling. If you would have just thought about venture capital a couple of decades ago or something. Um, and the scope of the, what you, of the, what the firms are looking at doing has changed. Um, you know, you guys are expanding internationally. There's like wealth management solutions. Sequoia has got a hedge fund. They've got, you know,

‍

Ben Horowitz (30:37.699)

Yeah, something.

‍

Auren Hoffman (30:58.515)

And so how do you, do you see like VC and financial services somewhat converging?

‍

Ben Horowitz (31:05.365)

Yeah, like I think they're certainly getting like much closer together. And the way we think about it, like there's different kinds of things. Um, and certain things are closer together than others, but the way we think about it is like, look, if you're somebody who's going wants to build a company, we want to be the best partner for you from the day you start doing that, um, to the day you become a philanthropist, uh, and we want to even help you with that. So like all the way through your life cycle. Um,

‍

And so what are all the financial services needed to accomplish that? It's a big set. Now, I think, you know, for us to not be on the builder side, to be on the kind of more like optimized, like something that Vista would do, that would be a bigger departure. But in terms of anything that the entrepreneur building a company needs, I think that's kind of very much what people want from us.

‍

Auren Hoffman (33:18.186)

Now, do you think we're going to see in the 2000s, I think we saw a couple of VCs go public. Do you think we'll see a scenario where like VCs go public and maybe acquire other VCs or

‍

Ben Horowitz (33:33.385)

Well, so the fundamental, so the value of going public is you kind of monetize or advertise the franchise itself. So you now have another way to compensate people and all those kinds of things that go with, you know, having the actual firm be worth something. And you've seen a little like, I think, Thrive just sold like a...

‍

piece of the GP, you know, the value of a certain level and all that kind of thing.

‍

Auren Hoffman (34:06.606)

And obviously you have like from the management fees, you have revenue streams and stuff.

‍

Ben Horowitz (34:10.309)

Yeah, you've got revenue streams and then you've got the investing income and all that. I think the challenge for a VC going public though is if you look at how say Apollo is valued or BlackRock is valued or the kind of similarly structured things. Yeah, KKR, their value, Wall Street values the fee stream much higher than the investment returns. And I think that's a dangerous incentive set for a

‍

Auren Hoffman (34:25.858)

or even like a KKR or something.

‍

Ben Horowitz (34:40.349)

VC like if all you were doing was venture capital and you were Massively incentivized to get more fees and not incentivized to get more returns That probably Yeah, that that's probably not gonna end well in venture capital However, like if a firm expanded to like if you look at Apollo people think of them as a private equity firm

‍

Auren Hoffman (34:53.846)

Then it's just an A-U-M game, right?

‍

Ben Horowitz (35:09.225)

But their other businesses are actually way larger than the private equity business. So like, it's not that big a problem for them the way Wall Street values them because they've got like insurance companies and lending companies and all this kind of thing where fee streams actually are a fine incentive. And so I think for us in our current configuration or for the, our kind of current competitive set, it doesn't make any sense.

‍

Auren Hoffman (35:12.446)

like their lending business or.

‍

Ben Horowitz (35:38.973)

But one could imagine like VC involving into something where that made sense.

‍

Auren Hoffman (35:43.234)

Like if you think of a big bank or something like that, like there's the investment bank side and that, that would be a very hard to go public on its own because it's, it's very like, they can have one quarter where they have this like massive deal, they could probably have three quarters where they get nothing or something. Right. Um, but then it's married with all these other products that they have, which are much more consistent, like you could see a scenario where, you know,

‍

Ben Horowitz (35:46.003)

Yeah.

‍

Ben Horowitz (35:55.325)

Yes. Right.

‍

Ben Horowitz (36:06.865)

Yeah, well look, JP Morgan, the original JP Morgan was a VC. I mean, that's what they did, right? They financed railroads and this and that the other. And they did it, and it was equity investing. They got the equity on a deal by deal basis, or they got the money on, they kind of phrased it, deal by deal as opposed to fund by fund, but they were VCs. And then that evolved into them being.

‍

What we now know is, you know, JP Morgan Chase, the loss is that it is. So yeah, you could imagine a VC like flipping into that again. It's a little harder to become JP Morgan Chase these days because the original one achieved like quite an amazing degree of regulatory capture as we've seen. Yeah. You know, the big banks, we have this huge big, too big to fail kind of legislation.

‍

Auren Hoffman (36:38.559)

Yeah.

‍

Ben Horowitz (37:03.421)

you know, following the 2008 crisis. And since then, every big bank has gotten bigger and no new banks have been formed. So it had the exact opposite effect that was intended.

‍

Auren Hoffman (37:15.542)

Now there's these companies who are like slightly on the side of venture that I think like over time could play a very big role. There's Angelus, there's Carta, which is an A16Z company. Like how do you see these companies interacting in the world? And they could have potentially a lot of power that they could leverage, um, in discovery and transactions and secondaries and all these other things that haven't yet happened.

‍

Ben Horowitz (37:42.961)

Yeah, so I think that, I mean, this is also kind of like this weird question of the way the private markets work. So currently you have, you know, the public markets, which are massively regulated, you know, to the point where, like, every public company has been sued. Like, and then the number of public companies, I think, is like half what it was 20 years ago in the U.S. Because it's just so onerous to go public.

‍

And then you have the private markets, which are, I don't want to say they're not unregulated, but like very light regulation and almost no enforcement. You know, so you've got closely held companies and this and that.

‍

Auren Hoffman (38:29.134)

And you guys yourself only recently became an RIA, and I presume it's because of like the crypto stuff you're doing, right?

‍

Ben Horowitz (38:35.089)

Yeah, no, that's exactly right. Crypto stuff and just the ability to do whatever we want with our public stuff and this kind of thing. But it's a weird dichotomy. And I think that what's happening now is because of that kind of bipolar world we live in.

‍

Auren Hoffman (38:40.886)

Yep.

‍

Ben Horowitz (38:59.261)

the people are building, reconstructing the entire public markets infrastructure in the private markets, which is, I think, good. But it does, you know, it's a little hard to see where it goes because, you know, it's just so weird. I think we've overregulated the public markets quite a bit. And then, you know, it just creates this, you know, not so smooth path.

‍

to going public, which has also had the really unfortunate kind of side effect of really that plus easy money exacerbating the wealth gap because rich people buy stocks in companies in the private markets and they own them and they have all this stuff and they have houses and this and that and the other. And so we massively increase the money supply and all these kinds of things.

‍

assets go way up in value and people holding cash lose, you know, lots of money. We have very high inflation now and all that kind of thing. And so then the rich get richer and the poor get poorer. And I think the lack of access to new high growth companies and public markets is certainly a large factor in that process.

‍

Auren Hoffman (40:20.479)

One of the things that's happened in the venture capital firm is this rise of these solo GPs that have happened. What do you think of that? Well, I think that's a good question.

‍

Ben Horowitz (40:29.597)

Well, you know, like I think the big thing is, you know, was that a boom market phenomenon or is it kind of a lasting thing? So look, what happens in the boom market and every entrepreneur like entrepreneurs, it's just like you have to be positive person to be an entrepreneur. So if you're only thinking about what will go right, which is all anybody thinks about in a boom market, then like, oh, there's one person who put money into the company and like,

‍

They've got 30 other companies and like, that's all fine. Great, I just need the money. Like, I don't need any fucking help. Or like, you know, I was one of the first three and I've talked to the guy or this and that. You know, in a market like this, I think, one, like, it's harder to be a solo entrepreneur because you don't have any money coming in or you have a lot less money coming in. And then, you know, for the entrepreneurs, they're like, okay, I need somebody who's gonna actually help me.

‍

you know, when things go wrong, like how do I do a layoff? Like how do I, can you help me place my people? Can you help me sell the company? Like all the kinds of things that you actually need a financial partner to help with, like they can't possibly do that. They're a solo, they're literally a solo entrepreneur. And so I think the idea is kind of built for good weather, but like that remains to be seen. Like maybe it continues to work. And look.

‍

The solo entrepreneurs or the solo VCs that we work with, you know, many of them are really good and we like working with them and they're a good compliment to us. So they certainly have no problem with it, but I think the sheer number of them will probably shrink down to the best of the best in a market like this.

‍

Auren Hoffman (42:15.466)

Now in the world of like a CEO handoff, we've seen that recently in Amazon, with Google, with Apple, with Microsoft, all seemed like they went really well. And one of the reasons I admire like Sequoia and YC is they've executed multiple CEO handoffs. Like how do you think about that from your context? Like how do you design a company? Think of like, if you think of A16Z as a company where one can actually do a handle.

‍

Ben Horowitz (42:23.035)

Yeah.

‍

Ben Horowitz (42:43.061)

So I think Google and Microsoft and Amazon are a little misleading in that they're monopoly businesses. Well, the fun is around, but they're, you know, if you're a monopoly, the, so the really difficult thing to do if you're the successor CEO is kind of get to the new product. Like, can you innovate?

‍

Auren Hoffman (42:50.83)

Because the founder is still there too. Or no, okay, they're Monopoly.

‍

Ben Horowitz (43:13.205)

There's a lot of people who can take over and run the thing that somebody else, that Jeff Bezos built, right? And I like, I think Andy Jassy and Satya and Sundar are all great, but that's like a smaller test to take over like the monopoly business and like get the money out and do the thing and deal with the regulators and what you do in that. Then, you know, what like I think...

‍

I think the Sequoia example is a harder example because they had to navigate across technological changes. So if you're in the VC business, probably the trickiest thing is, okay, we were great at social mobile. There's nothing to invest in that anymore. Are you going to be great at AI and do you have the right talent and did you turn over the right people? And do you?

‍

Auren Hoffman (43:59.693)

Yeah.

‍

Ben Horowitz (44:09.277)

built the right capabilities to compete in that market. And that's something that, you know, you have to be really good to do. Like that's a harder, a much more difficult thing than just taking over and continuing to invest in social mobile. And, you know, in companies, I think it's even more difficult. Like if you have a company that you're trying to transition to a professional CEO, and then that CEO has to innovate. And I think, look,

‍

I think we're starting to see that with Snowflake right now, where like, Frank Slootman is probably the greatest professional CEO ever in enterprise software, but like now that they really do need to get new products and they're seriously challenged by Databricks, it's tough. Like that's not his skill set. And so, so it's different, I think in different scenarios.

‍

Auren Hoffman (44:45.398)

He's incredible, yeah.

‍

Auren Hoffman (44:59.758)

It's the zero to one versus the one to N type of thing.

‍

Ben Horowitz (45:03.877)

Yeah, exactly. Zero to one is, there are very few people who haven't done zero to one that can do zero to one. That's a very kind of special skill. And then it's actually in some ways harder when the company is bigger. Because like the resistance that you get is, you know, fairly remarkable.

‍

Auren Hoffman (45:12.98)

Yeah.

‍

Auren Hoffman (45:30.478)

Um, what you've been involved in tech for a long time, over 30 years, like what, what are the most important things that have been true the whole time?

‍

Ben Horowitz (45:44.624)

Well, look, I think that...

‍

Ben Horowitz (45:49.609)

Probably, the product really has got to be a breakthrough to create a company. So it's very hard to, and right now we're seeing a lot of this where people are going to the obvious idea. So, okay, AI is here. So like, let me kind of build something that helps you edit your documents on top of one of the large models that somebody's built.

‍

Like, of course, somebody's going to do that, but there's probably going to be like 80 people who do that. And, you know, it's very hard to kind of run into a space that's that competitive and so forth. It's not really a breakthrough idea. It's just like, okay, a simple application of a breakthrough technology that somebody else made. And I think those have always been tough to get a company out of that. Although like you might, it might kind of

‍

get you enough capital and enough customers to be in position to create a breakthrough. And then you can do that. But that, you know, that I think has been true probably the entire time.

‍

Auren Hoffman (46:53.622)

Mm-hmm.

‍

Auren Hoffman (47:02.114)

Just to push back on that, it does seem like a lot of these companies that win, they're the second, third, fourth mover. They've got a product that's pretty similar, maybe like a slightly better UI, a better sales model, maybe they're better operating at it, and they seem to win. They just move faster than the other. They see what's working and they iterate on it. Or you don't agree. You think they're having real innovations?

‍

Ben Horowitz (47:13.653)

Thank you.

‍

Ben Horowitz (47:31.037)

Well, I think that, you know, it's a little bit how you interpret the history. So like the most famous of this was like MySpace Facebook. But I think when you get into the MySpace product, it was pretty different product than Facebook. So like the network effects got into competition, but architecturally, they took radically different approaches. And I think that

‍

Auren Hoffman (47:35.991)

Yeah.

‍

Ben Horowitz (47:59.097)

Yeah, the Facebook people would always say, okay, like execution is all that matters and da da. But you know, I'm not so sure about that. Like, you know, as we witnessed from the threads Twitter contest that, you know, recently came out, I think threads like from a software perspective was an awfully good competitor. But you know, the network effect was kind of baked at that point.

‍

Auren Hoffman (48:26.218)

Sometimes the innovation is very slight and it's like the door dash innovation, it's like, okay, it's like, well, we're in the suburbs. It's like these very, very slight innovations. As an investor, how do you even know that's an innovation as opposed to like, oh, this is a good operator going after a good market.

‍

Ben Horowitz (48:42.741)

Yeah, you know, I think that's really hard to tell. I think that you have to start with the entrepreneur. So like, is this an entrepreneur who you really want to back because they are, you know, you think that they won't even be doing it if it wasn't like way harder than it looked. This was also the case with Dropbox, right? Dropbox didn't look.

‍

really like a breakthrough in a lot of ways in that it's a very simple idea. There were many other people doing kind of variations of the theme, but I think that with Drew, he was really an unbelievable engineer himself, and he had dug into that problem and knew what was hard about it, what was hard about making that experience great. And that was sort of what he did, you know, that was their kind of contribution.

‍

Auren Hoffman (49:41.278)

Like I remember the first time I used Dropbox, like it was, it was, it was amazing. I mean, I was, I was doing this like with a hacky thing I used with SVN. Um, and like trying to do this myself and it was like terrible. And then all of a sudden this came out and it was like almost free. It was, it was amazing. Like I couldn't believe it. Yeah.

‍

Ben Horowitz (49:41.559)

And I think that happens a lot.

‍

Ben Horowitz (49:51.475)

Yeah.

‍

Ben Horowitz (49:59.621)

Yeah, it's like a sigh of relief, right? But just looking at it from a market map standpoint, it would have been very confusing. Yeah.

‍

Auren Hoffman (50:06.558)

Yeah, okay. That's fair. Now, a couple of personal questions. Your dad is a conservative pundit. You're probably more on the center left. How has his intellectual journey taught you about public discourse, about civility, life in general?

‍

Ben Horowitz (50:23.493)

Yeah, you know, it's funny how, well, it's taught me that like how misinterpreted everybody is, you know, particularly when you get into politics, it's so tribal and so weird. And actually, I'll give you a couple of stories. So there was a story that was done on my father and me and the New York Times a while back. And it was, you know, probably something I should have.

‍

never done because being associated with my father is super dangerous for me. But the very first question out of the, it was David Streitfeld, who by the way, did a very honest job of that piece. But the very first question he asked me, he said, sits in my office and he goes, Ben, your father said Black Lives Matter is a criminal organization. What do you think of that? That was his first question. And so I'm like, okay.

‍

Auren Hoffman (51:18.803)

Oh my gosh.

‍

Ben Horowitz (51:20.497)

So do you want me to attack my father? You know, like, what do you want here? And I said, well, why are you asking me that question? And he goes, what do you mean? Why am I asking you that question? I was like, well, you're asking me, like that's very specifically directed at me, probably because you're like, Ben, like you're very like involved with the black community. Like, how do you feel about your father saying something like that? And he goes, well, yeah. And I said, well.

‍

Have you ever thought about like, why am I so involved in the black community? I said, might it be because when I was like five years old, my father took me to the black Panther church every Sunday, son of man temple, and I know all the fucking words to all the songs. And I like drink cognac now, cause that's what all the Panthers drunk and it like stuck in my fucking head. And like, so like before you assume that guy said that because of some racist thing, maybe you ought to consider the whole picture that you're pulling.

‍

Auren Hoffman (52:14.716)

Mmm, yeah.

‍

Ben Horowitz (52:16.085)

And then like I went on to explain it, but it was like that whole and to his credit, he kind of adjusted his narrative on the story and all that kind of thing. But those are the things that come up. But, you know, like politics gets really, really kind of stupid when it comes to relationships and judgments and all these kinds of things. And, you know, people think, oh, that person's evil and that person's evil because they had that opinion on this thing that I disagree with. But it's almost.

‍

You know, it's a lot of people misinterpreting each other. And, you know, and of course it's much easier to do in 140 characters.

‍

Auren Hoffman (52:53.026)

One of the things in politics is that everyone has this like public persona in politics and maybe this is true in general, but certainly in politics. And then, um, and in some ways they need to nowadays, especially you need to like play into that public persona and really need to live it up. And then when you meet these people, like in real life and you actually have like, uh, dinner with them or you, you have a relationship with this person. You realize like, they're really not like their public persona. Um, yeah.

‍

Ben Horowitz (53:20.849)

Oh yeah, it's kind of shocking, isn't it? You're like, what the? I remember the first time I met Karl Rove and I couldn't even, my jaw was on the floor of the entire conversation. Like I couldn't believe I was talking to that Karl Rove and he was saying like these things. It just didn't make any sense to me.

‍

Auren Hoffman (53:23.022)

It's very, very different. Yeah.

‍

Auren Hoffman (53:32.375)

Hehehehe

‍

Auren Hoffman (53:39.53)

Yeah. Okay. That's really, I find that very interesting as well. Um, all right. What, what do we like to ask is what is a conspiracy theory that you believe?

‍

Ben Horowitz (53:42.83)

Yeah, that's normal politics.

‍

Ben Horowitz (53:49.497)

Okay, so I 100% believe that COVID started in the Wuhan lab, you know, and escaped the lab accidentally. And the research that they were doing... Yeah, however it happened, like, they disappeared all the people in the lab. Like, that's a first, like, clue, like, why did all the scientists disappear? And like...

‍

Auren Hoffman (54:00.95)

Like on someone's shoe or something or... Like on somebody's shoe or something like that or...

‍

Yeah.

‍

Ben Horowitz (54:18.289)

By the way, and then that lab and that research was funded by the US government, the Gates Foundation, and all the people ID'd by the conspiracy theorists. That all turned out to be true, as I've read it in detail. So maybe now it's not that controversial anymore.

‍

Auren Hoffman (58:21.386)

Last question, we ask all of our guests, what conventional wisdom or advice do you think is generally bad advice?

‍

Ben Horowitz (58:33.852)

There's so much that's bad advice.

‍

Auren Hoffman (58:35.938)

Hehehehe

‍

Ben Horowitz (58:38.613)

So one that higher, slow and fire fast. And let me explain why. So the truth is nobody who's not a psychopath fires fast. That's never how it works because you thought about it a lot when you brought the person on, you have a relationship with them, there's huge consequences to firing them. I've never met...

‍

Auren Hoffman (58:48.undefined)

Okay.

‍

Ben Horowitz (59:08.573)

I've never spoken to any CEO who is like, I fired that person too fast. But, but, uh, like, uh, for 99.9% of the CEOs, they never fired too fast. So when you say hire slow, what you're basically doing is giving them an excuse to not really focus on the recruiting process, the hiring process.

‍

being the correct amount of prepared to like make a decision that look hiring particularly executives is fraught with peril like generally like half of them make it so by hiring slow all you're doing like if you take a year to bring in a candidate that doesn't actually help you that doesn't like improve the quality what improves the quality is do a good job of preparing

‍

meet with enough people who understand what it means to be a CFO, what a great CFO looks like versus like a mediocre CFO and then interview candidates for what you need in your company, what they need to be world-class at, not like some generic idea, some guy who comes in wearing a suit you think will talk good to Wall Street, that kind of dumb stuff. Like you can do that slowly too.

‍

So it's hire correctly and then, you know, make the hard decision when you have to is actually the right advice. Like hire slow, fire fast is generally advice given to people who have never hired executives, I find, or certainly never fired executives because they don't know. They just like heard it. It sounds good. I'm passing it on. Every VC says that. It's a bunch of bullshit.

‍

Auren Hoffman (01:01:00.85)

For certain types of executives, like I imagine for most entrepreneurs, like hiring a head of engineering is something they could really dive into and do well. I certainly would do well in hiring a head of engineering, whereas hiring a CFO might be a much, much more difficult task. They may have never met a good CFO in their entire life.

‍

Ben Horowitz (01:01:12.561)

Yeah, that's a good one.

‍

Ben Horowitz (01:01:24.581)

Yeah, right. It's like hiring a Japanese interpreter and you don't know Japanese. Like, they're all going to sound like they know Japanese, aren't they?

‍

Auren Hoffman (01:01:29.45)

Yeah, yeah.

‍

Auren Hoffman (01:01:33.994)

Exactly. Yeah, exactly. So, and I guess that's where you use trusted advisors, but then, you know, sometimes you don't know how much to trust them for this particular thing, right?

‍

Ben Horowitz (01:01:44.393)

Well, trustee advisers don't actually work because, look, most executives fail not because they're not competent, but because they don't work in your organization. Right? So what I need for right now, like the CFO of General Motors is not going to get the job done. Like, because what I need is somebody to put the basic control structure in place and this and that and the other. So kind of...

‍

Auren Hoffman (01:01:56.298)

Right. Don't fit your culture.

‍

Ben Horowitz (01:02:12.917)

the right way to approach that is you have to learn enough about the job. And there's kind of a variety of ways to do that. So, you know, one kind of way is to just like act. If you've got junior people on that team act as CFO, just run the meeting, see what problems are running into understand the job so that you know something about it before you hire. Another thing you can do is say, okay, like, I just want to meet.

‍

five CFOs that people consider good and ask them, you know, ask them the right questions. What is the job of the CFO? What are the parts to it? What's the difference between a good CFO and a great CFO? Like, you know, tell me some examples of like how you would test that in an interview and kind of build up a view of all the things that a CFO might be good at.

‍

and then say, well, what do I need in my company for them to be good at right now for the next 24 months? And then match that and look for that person as opposed to like, you know, trust somebody who knows a CFO from another company who did well in that context, but you know, may not be at all relevant for you. And so it's, you know, it's that, that's the right advice on like hiring executives is,

‍

the right kind of preparation, the right kind of process. And then look, firing is a different event. You can hire slow and still fire slow and you can hire the wrong person even if you go, takes you two years to find the guy. Like I've seen that many times. So that's my...

‍

Auren Hoffman (01:04:03.539)

That's great.

‍

Ben Horowitz (01:04:04.157)

It's your conventional wisdom that drives me crazy. Because it's like not enough, like there's some truth in it, but it's not true.

‍

Auren Hoffman (01:04:10.562)

This is amazing. Well, thank you Ben Horowitz for joining us at World of Daz. I follow you at B Horowitz on Twitter or X or whatever they call it. And I definitely encourage our listeners to engage with you there. This has been great.

‍

Ben Horowitz (01:04:21.789)

All right, great to see you, Orin, and that was fun. Talk soon. Bye.

‍

Auren Hoffman (01:04:25.618)

Awesome, thank you.

‍