Marc Andreessen: Deep Vs. Broad Founders, AI in America, & the New Face of a16z

Just close your eyes and just imagine two states of the world. One is which the entire world runs an American open source LLM, and the other is where the entire world, including the U.S., runs on all Chinese software. And I don't know, for me, that's very straightforward. in this episode. We're sharing Mark and Drees PN recorded live. wide-ranging conversation. an open source model. along the way. branding, how startups are scaling Let's get into it.

As a reminder, the content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any A16Z fund. Please note that A16Z and its affiliates may also maintain investments in the company's podcast for more details including a link to our a16z.com forward slash disclose

Welcome. How are you guys? Hey guys, what's happening? What's happening? It's been a great day. I'm going to hit this real quick. Thanks so much for joining us. And thanks for wearing us. Thanks for wearing a suit. Yeah, exactly. Yeah, you look fantastic. Yeah, special occasion. yeah congratulations on the podcast i just want to start out by saying i've been watching it's just it's tremendous

Thank you so much. I have a funny story. The moment that I realized that you were maybe paying a little bit of attention to what we were doing and it gave me some conviction that we were on the right track as we did a reply guy of the week to this guy, Baldo. And it was like this inside joke. And I was like, Holy shit, Mark just followed Baldo. Very talented guy by that. But yeah, so thank you for following along and it's been awesome talking to your whole team today. Yeah. Fantastic.

So let's go through some hot topics. I want to start with we're in the AI era. You obviously lived through the dot-com era. There's some comparisons. What have you learned and how is this time different? companies are making more money, valuations are high, but how are you thinking about teaching the next generation what they should learn or what they should ignore from the people that might say, this is the dot-com boom 2.0.

Yeah, so look, I guess I'd say, I think it was Mark Twain who once said, history doesn't repeat, but it rhymes. And so, you know, I think people looking for, like, a direct compare or contrast, sometimes you see people, like, drawing charts or the stock market's going to do the same thing or whatever. Like, I don't think that stuff ever quite...

happens that way but you know it does rhyme um and you know the line that john door had in 1995 i remember was that the internet is a cream that you rub on investors to get them all excited John Doar said that. That's hilarious. I have many, many, many John Doar. Oh, yeah, exactly. You know, I ask like that now. You know, look, the internet went through phases. People actually forget, but there were even the phases on the way up from 95 to 2000.

There were these phases of skepticism and panic. It looked like the whole thing was going to fall apart in 98. And then, you know, in 2001, you know, after the dot-com crash, you know, like all of the companies just like completely wrote the internet off and they just said, you know, thank God that's over. And then the Internet itself just kept growing, right? And then, you know, by 2005, it was back to, you know, kind of back to back to where it had been before. And we, you know, the rest of us

was history. And so, you know, people kind of go, people are kind of bipolar on these things. They kind of get overly excited. They get overly depressed. I just always thought then, and I think now the substance matters, you know, the substance overwhelmingly matters. And so, you know, is the technology great? Like, are the products great? Are people using them? You know, AI is like off to the races so far. It's just like unprecedented rate of growth of actual use.

Yeah. Right. And you see that in all the numbers. And then, you know, look, the businesses, I, you know, I just, I talked to another conversation yesterday with like, you know, companies raise seed money. They're already over 10 million ARR. with this incredible you know like it's just and you know there's like there's a lot of those like that you know they're all over the place and so

The things that are working, the products are fantastic. The users are getting tremendous value out of them. And as you know, the base technology is moving really fast. So I feel really good about it. Yeah, I mean, it can move faster because we have the internet. Like, these apps can get to hundreds of millions of daily active users because there's hundreds of millions of people. There's billions of people on the internet.

uh very interesting uh not to linger on the dot com stuff too much but i'd love to know a little bit of a history lesson around the way the browser wars played out. And is there anything you can learn from that that applies to the open source versus closed source AI debate? You've been very opinionated on that. I'd love to know, kind of, are there any... previous eras of tech that you're mapping to when you think about open source AI.

Yeah, so, you know, a lot of the browser wars, you know, there was kind of a big company, little company, kind of big company thing. And, you know, I think that's repeating itself. And, you know, I think one of the ways to think about what's happening right now is OpenAI is kind of growing up to become Google and Google's kind of, you know. reinventing itself to be open AI and so you know there's like a similar thing there

And then Microsoft and Apple and Google control the operating systems for end user devices, and they all have AI strategies and Amazon as well. So a lot of those issues are going to come back up. Two of these big companies are actually on federal trial right now. you know, in big FGC cases in Washington, D.C., actually in the same courtroom.

And so many of those issues will repeat or will be intense. The open source thing in tech has always been the wildcard. I think it's always been an incredibly positive wildcard. I've always been an enthusiastic supporter of it. My original work was all open source. And so, you know, the real thing that happened with open source is in operating systems, you know, basically Unix one and then, you know, specifically Linux one.

you know for everything on on the back end and i you know like you know i remember like in the 1990s there was like a furious operating system war and there were these companies that were making huge amounts of money on server operating systems you know companies like sun many others. And then, you know, Linux just like commoditized that entire thing.

I think it's plausible, you know, we'll see. But look, it's plausible that open source say, I may just be the standard. Like, you know, and whether that derives out of, you know, Deep Seeker Lama or, you know, any of the other new things that are coming out, you know, we'll see. But like...

i think that's entirely feasible and you know i think that would be obviously you know there are companies that would have to adjust to that if it happens but the other side is if kind of the world had ai for free

I think that would be a pretty magical result also. Yeah, totally. I mean, shifting to the geopolitics of all this, is it important not just that there's an open source champion, but that there's an American open source champion or a Western open source LLM. Yeah, I believe so. And I believe so for two reasons. One is just actually cultural reasons, which is just, you know, you know.

Open weights is great, but like the open weights, like they're baked, right? Like the training is in the weights and you can't really undo that. And so are you being trained by... a company or an organization or a set of people with American or Russian values? Are you being trained by a company with Chinese values? And look, there's issues on both sides. The American trained models have their issues, they have their weirdnesses.

you know but the chinese brand models you know look they like score really well you know they literally in the chinese benchmarks they literally have like line items for like marxism right and you know deep sick like it's like 100 out of 100 on marxism

And so it's here for DFC. Really knocked it out of the park. Congratulations. Yeah, exactly. It's fantastic, right? And so it's like, all right, so these are going to be, you know, this is going to be the technology that's going to intermediate the legal system, right? The courts, the education system, the medical system.

Like, you know, if you guys have kids, like, do you want your kids to, you know, your kids are going to be learning from these things their whole lives. And, you know, do they fundamentally have Western values or do they have, you know, sort of, you know, CCP values?

I think is really critical. And then the other thing, I would just say, just close your eyes and just imagine two states of the world. One is which the entire world runs an American open source LLM, and the other is where the entire world, including the U.S., runs on all Chinese. software. And I, you know, I don't know, for me, that's a very straightforward, it's a very important topic and a very straightforward answer. Yeah. I mean, in the American dynamism context, Is it...

kind of mission accomplished there? Or are there new territories that you're looking to expand into with the broader American dynamism thesis? We talked to you about energy and whatnot. What is exciting about you outside of artificial intelligence on the national interest investing side? Yeah, so look, we've made tremendous progress. I'm super proud of that team. And then, you know, in this new political environment, the new administration has embraced it, but also a lot of Democrats.

We're actually quite excited about it also. And our event that we hold has lots of people from both parties. So I think there's a... you know there's this unified view now i think in the american at least in the political world of like all right it's it's time you know it's time for the us to step up on a lot of these things um it's time for the us to build more um it's time for the us you know reinvent energy it's time for nuclear

You know, it's time for, you know, infrastructure housing, like all these things. And so I think there's a lot of momentum. I think we're at the very beginning of it. One of the ways to think about what we're doing in American Dynamism is we're going after all the sectors of GDP that are really big and not yet basically affected by tech.

And so education, health care, housing, defense, law, you know, just to pick five, you know, those are giant important slices of the economy that largely have not been transformed by technology in the last 50 years. You know, the U.S. Defense Department spends, you know, coming up on a trillion dollars. Very little of that money is going to technologies and companies that have been invented, frankly, in the last 20 years.

Most military hardware in the Department of Defense, the F-16 is from the 1970s. Right, like the U.S. plane that does high-altitude photography, the high-altitude spy plane, it's still the U-2, which is literally a 70-year-old plane.

Like there's just like enormous changes, you know, that really have to happen. And I think everybody kind of in the system knows that, but it's hard to get there. And I think, you know, a big part of that is, you know, these new companies, you know, that we try to support, you know, they need to show up and make their case. The entrepreneurs are fantastic. And then we see more receptivity coming from the system than we ever have.

Talk about the position that the firm is in today. In many ways, it feels like you and Ben and the team had somewhat of a crystal ball to see how the private markets were evolving, companies staying private longer. AI being this massive platform shift that's going to upend every industry. And then now the sort of political environment that enables industries to kind of get a second or third life. Did you just get lucky or did you see a lot of this coming?

I mean, I think we got a few things right. We got a bunch of stuff wrong. You know, we buried those ideas out back behind the shed. You know, we made a bunch of mistakes along the way or changed a bunch of things. But, you know, we got a few big things right. You know, I think the things I think we got right.

you know what is like the venture ecosystem you guys know has just evolved enormously um in a very positive way and the kind of old model of having a sort of a bunch of kind of mid-sized firms they kind of sit on sandhill road and wait for the founders to you know to come in like that that you know those days are kind of over

You have the rise of the high scale firms like us, but you also have the rise of all these seed and angel investors that are really first money in and then really working closely with founders. from the very beginning and i think that that ecosystem is healthier than ever and doing really well

And so, you know, that's been a big change. You know, on the ITOs, like, I don't know, I would say we have changed on that, which is I used to complain a lot. Like, I used to give these interviews years ago and, you know, like early 2010s. and just complain about that it was basically becoming impossible for companies to go public and the things that had caused that to happen and what needed to change.

And then, of course, nothing happened. There were no reforms. There were no improvements. If anything, everything got worse. um and so you know we and others adapted to that by you know by basically putting you know putting ourselves in a position to fund companies you know later in their life cycle at higher higher you know bigger larger amounts of money you know later growth stages um It's now much more common for companies tenders and these kind of private sales even for their employees.

You know, so that's been a big change. You know, but look, the best directional bet we had was just technology was going to keep becoming important. It's just like in every area of our life, the bet that, you know, in sector X, tech is going to be more important. There are going to be big tech companies built. They're going to really matter. That's a very good bet. It's hard for me to ever see that ending.

What are you, you know, you guys are with your LPs today, tomorrow. What are you hoping they take away from the event? Oh, so this is not a John Dora quote, but I'll give you another quote. When we were raising money originally, we were going around. We raised our first fund in 2009, and the world had completely collapsed after the financial crisis. And we went around and we sort of briefed all the venture VCs we respected to just give them a heads up what we're doing.

And one of the GPs took us aside at the end and said, you know, you guys have been dealing with like these, you know, very smart public market investors. He said, you know, part of the job you're going to hate the most is working with these LPs. You know, they're just terrible. And he said the key is you need to treat them like mushrooms. You put them in a cardboard box, you put the lid on the cardboard box, and you put the box under the bed, and you don't open it for two years.

What's the markup today? You know, look, the view Ben and I always had was the exact opposite of that, you know, which is our investors are our partners. You know, look, that's how we want to be treated by the companies we invest in. That's how we always viewed it when we were running public companies. Our investors, our partners, they're in it with us together. They're making a big bet on us and trusting us. And then look, correspondingly.

the promise that we made to them is like look we're going to try to like deliver you know excellent returns but in this business it's just going to take a long time And, you know, you guys know this for any company that really succeeds, it takes, you know, 10, 20, 30 years, you know, to really have it succeed. And like, and along the way, along the way, like a lot of things happen, you know, everything from negative headlines to, you know, lawsuits, disasters, chaos.

All kinds of crazy twists and turns, changes happen along the way. And so I just think it's really important for us as a firm. We really always need to make sure that we're treating our LPs as full partners. And so what we're trying to do is just really open the kimono all the way, show them everything, kind of expose them to everything that we're doing, make sure they really understand it.

go as deep as they want to go on the technology and on the companies. And I, you know, this is like year 15 of this for us and so far that I think they would say that that's gone well.

That's great. Last question from me. Just general advice for the next generation of entrepreneurs. Are you hoping more college, less college, more coding, less coding? There's a ton of technological change right now. How can the next generation of entrepreneurs set themselves up for success?

Yeah, so there's an old Steve Martin line. He wrote this great book on how to be a successful stand-up comedian. It's just a short little book, really well written, and every inspired comedian in the world opened it up. says Dave Martin's, you know, advice for being a great stand-up comedian, and it was

be so good they can't ignore you. Right? It's like, okay, thank you, Mr. Martin. You know, I'll get right on that. But like that, that remains the best advice, I think, which is just like quality, like we said, like quality bears out, like quality shows. And so, you know, really excellent thinking coupled with a high degree of energy and courage. You know, team building a great team and then a great product, you know, and a real insight into a market with customers.

figure out how to design you know not just design a product but also design a business like that's basically like my view of that as an entrepreneur always like that's my job as an entrepreneur um and like if i you know if i put a lot of time and effort into that i'm very serious about it i can hopefully do a good job

And I think time spent improving the quality of the thing, the product and the business is almost always better than time spent trying to, you know, networking or, you know, publishing a presentation or, you know, trying to get positive press coverage or kind of all the external stuff.

And so it's really it's really the quality of the thing. And then, you know, again, you know, my hope would be like we as a firm and others like us like that. We really, you know, that we really recognize that and see it, you know, see it, you know, kind of see it when it shows up. i think that's the thing and then i think the other thing would be like like ai is superpowers right and like if you as a founder

have access to, say, ChatGPT deep research. It's just like, oh my god, I have a PhD level. It used to be so hard to get information, right? And then it's been so hard, you know, even still to figure out what to do with it. And so to now have these tools for thinking through everything and increasingly being able to do things like write code.

We ought to see a generation of founders here that are just incredibly capable relative to what we used to be like. And we're starting to see those. We're starting to see a handful of those. you know that's what i'm still waiting for like the real conceptual breakthrough is like i'm i'm still waiting for the you know the you know the company where it's like you know it's got a thousand employees 999 or bots like and i haven't seen that yet

And somebody's going to really figure out how to use this technology to really not just bring a product to market, but fundamentally change how companies operate. And I think that's probably the next big unlock.

How do you think about an individual's edge around their own proprietary knowledge? There have been talk with Warren Buffett. you know retiring he read something like a hundred thousand plus earnings transcripts he had this incredible recall you're often credited with having this incredible recall and it certainly gives

can give you an edge in thinking through, you know, net new opportunities and decisions. Do you think that edge remains in a world where all human knowledge is accessible in a single query or what's your kind of mental model there? Yeah, so I think it's going to be, I mean, look, I think there's basically like two ways to really have a differentiated edge, like in general, right? There's sort of, you know, there's kind of go deeper, go broader.

GoDeep has kind of become a more and more specialized expert over time, and there are domains in which that really matters, biotech and things like, or by the way, designing AI, working on AI foundation models, that stuff really matters. The deeper you are, the better. So there's certain fields where that really matters. I think for most fields though now with these new tools, I would probably bet more on basically people who are able to be broad.

Which is to say, basically, can you know something about a lot of different aspects of life? and how the world works and then you can use the tool you can use the ai to go deep whenever you need to but then your job as the human is to basically then basically cross the domains across the disciplines. And look, you see this, if you talk to any of the great CEOs, you kind of see this, which is like,

They're really great tech CEOs. They're great product people, and they're great sales people, and they're great marketing people, and they're great legal thinkers, and they're great finance people, and they're great with investors, and they're great with the press. So it's this sort of multidisciplinary kind of approach.

uh and being able to cross domains and i you know i don't know we'll see we'll see how good you know we'll see how good the ai gets at that hopefully it gets good at that also but That in a lot of ways, the entrepreneur has always kind of had the burden for somebody who wants to do something serious. They have to be like that, at least to some extent.

And, you know, the best entrepreneurs in the future, I think, will probably be like quite skilled at like six or eight things and then able to kind of cross and combine them.

How do you think about the responsibility of the venture industry broadly to try to influence outcomes in the world? I was sort of laughing. Less than a month ago, a lot of investors were clearly shifting capital out of the U.S. and just YOLOing into French bonds and things like that. Venture capitalists obviously didn't do that, right? But at the same time, there was a high-profile story of a venture capital firm investing in a Chinese AI company that we won't mention here.

And it got me thinking around, if you're operating a hedge fund, you can go invest wherever you want without a ton of scrutiny, right? People don't look at that as... if you move assets into some... It's not as impactful. It's less of a... Yeah, it doesn't feel like you're betting against your own team in the same way. So I'm curious how you think of the generalized kind of...

responsibility of venture capitalists, especially as the AUM has scaled dramatically. It's actually been the arrow of progress. Yeah, look, the big thing I think I would say is we always thought and aspired that we were building things that matter and that we're going to have an impact on the world. And for a long time, we were trying to convince people that we were doing that and they never quite took us seriously.

because it's like you know it's like all right databases routers like okay nice you know nice guys i don't even know what those things are um you know even personal computers like a thing on my desk i write letters on it you know whatever like they're just like tech was never like relevant to politics really

I'm talking about like between the modern era, 1950s, call it through like the 2000s. And so generally speaking, like our press coverage was generally people were like, wow, startups are exciting. And when companies succeed, it's great.

And then, you know, in the last 10 or 15 years, you know, like if questions like yours, you know, have started to come up and then, you know, correspondingly just as you know, like just enormous amounts of kind of criticism attacks, you know, kind of indictments of, you know, people with all kinds of points of view on the industry.

i would say you know in the beginning that kind of irritated me because i had gotten used to the environment where either nobody cared about us or people just said nice things about us what i came to realize is

we're the dog that caught the bus like we actually now are building things that matter right and so you know tech now interfaces direct you know when elon goes into the car market like you just you know it's the repercussions of that you know to your point like on many countries are just gigantic and there's

you know, a thousand examples like that that we could give and, you know, AI probably being the biggest of all. And so, you know, look, it worked. Like the stuff that we do now, the companies that we fund, they really matter. They really have really big implications for society and for policy and politics and geopolitics. And so mainly I think the responsibility is on us as a firm as well as our companies. We need to go explain ourselves.

We need to go be present in the policy debates. As you know, we have a huge push now into policy and politics that I never imagined we would have. By the way, that's something I got wrong. I never thought we would have that, and then it became very clear that we need that. kind of for this reason. So I just think that's part and parcel of it working. We're going to have to do that. And then the geopolitics, the China thing, I think is probably the most intense version of that.

Maybe that's when we got right. We just never really did anything in China for a bunch of reasons. And by the way, we'll see. This is a fluid situation, and it's been publicly reported that there are talks underway with China, so the relationship with China could be better or worse. you know i think that's what everyone wants yeah yeah everyone wants like cooperation in general Yeah, like, look, it's possible. Look, I mean, I'll give you my favorite news. Here's my favorite story.

I think it was like 2011, at the time the Obama administration was trying to reestablish a relationship with Russia. And there was a famous moment where the Secretary of State, who was this woman named Hillary Clinton, I don't know whatever,

happened to her. I hear rumors that she then went on to, I don't know, did something in politics, but, um, She was the Secretary of State, and there's this famous photo where she was on stage with Sergey Lavrov, who was the Foreign Minister of Russia, her counterpart, and she brought like a big plastic red button, which was the reset button.

to reset warm relations with Russia. And then what happened was Medvedev, who was the showman. The reset air horn. Exactly. And then Medvedev, who was Putin's number two, actually came to Silicon Valley.

And the Secretary of State's office called all around to Silicon Valley saying, please, you guys, roll off the red carpet to Russia. There are new friends. Do everything with them. Do whatever they want. Please go invest in Russian companies. Russia's building this new Silicon Valley. See what you can do to support them. And so it was like, you know, like it was just like a love fest. And then, you know, fast forward four years later and, you know, Putin became, you know, Butler.

as they say, and, you know, just, you know, and then, you know, Russiagate and like everything that followed. And so I would just say part of being citizens, I think is, you know, we're just going to have to navigate through shifting geopolitics. Our view, and quite honestly, like our view is just we're going to put ourselves squarely on the side of the United States. I should say our foreign policy is affirmed as the United States foreign policy.

If they don't want us to do things in China, they don't. By the way, if they come to us and they say we should do more, we'll look at it at that point. But we're trying to be good citizens among everything else.

Yeah, that makes sense. Talk about the evolution of the A16Z brand. You guys came out with a new logo and a website. And as always, you're good at getting the attention of... the internet. We talked about it on the show.

sure we liked it yeah i i think it's something that's different and uh and will age very well but i wasn't exactly sure like what are you trying to say with it what are the references and even what was the process like did you pick the color i don't know or even that conversation i want to know more about the brand Yeah, so first of all, I know we got some Twitter controversy on it, and I just want to thank Kanye West.

Drowning that outcome entirely. Yeah. At the moment. So yeah, so that worked out well. So anyway. We have this incredible designer, Greg Truesdell. He's got a team. They did a fantastic job. I worked closely with him on it. I have no design skills, but I provided a lot of input.

You know, I think the big thing about it is what it's intended to sort of reflect is like an embrace of what I think is a broad set of cultural changes that are happening right now. You know, frankly, you guys are a part of as well, which is this.

what i would describe as an i don't know 15 year era of i don't know like cultural almost shrinkage everything getting you know and you see in the design world it's like everything getting like ultra clean minimalistic and culturally everybody getting like very cautious and careful about what they say

with all this kind of censorship and speech control and then this whole thing of everybody needs to feel bad about everything all the time and everybody needs to feel bad about the country and bad about their ancestors and bad about this and everybody's just miserable all the time. um like i i just think like that whole thing got kind of as you know i don't know you might call it like you know neopuritanism or something um like that thing kind of crested in i don't know 2021 or 2022.

And basically in the last three years, I think there's this renewed sense of energy, enthusiasm, ambition. you know, achievement and dynamism. But like, you know, and again, it's like literally, yeah, we should have nuclear energy. Yeah, we should have rockets going to Mars. WE SHOULD HAVE THESE THINGS. and that it's good to succeed, it's good to build businesses, it's good to make new things in the world.

You know, yeah, we should have the thousand, you know, yard tall, you know, shining Colossus on Alcatraz Island. Yeah, yeah, I was going to ask. My only request is to, like, make it real, you know? Like, make it the Statue of Liberty of the West Coast.

Yeah, I'm imagining a massive instantiation of that coin in the office and maybe physical versions. Not the coin, the figure. Oh, yeah, the actual figure. True, yeah. You know, she has a name. She has a name. She has a name. Well, no, what's the name? She has a name, Technomedes. Technomedes, I love it. That's fantastic. Amazing. That's great. All right, well, Technomedes. Bye, plot of land.

somewhere visible from the Golden Gate and just, you know, build it. A mile high, please. Mile high. Make it happen. Make it happen. Thank you so much for joining. This is fantastic. Thank you for coming on. Really enjoyed it. Have fun with the team and the LPs. Cheers. We'll talk to you soon. Thank you.

to the A16Z podcast. If you enjoyed the episode, let us know by leaving a review We've got more great conversations coming your way. Hello, A16Z Podcast listeners. This is Seth. If you're a longtime listener here, we have spent three years, hundreds of conversations, and the equivalent of numerous nonstop days together. That's the beauty of podcasting. And the goal was always to bring you a dose of the future. So I hope I successfully delivered that in some way.

But all good things do come to an end and for me that means stepping away from the mic to I am so excited to share that I'm joining Thrawn. That is Grok with a Q, not AK, to lead growth there. So now my new goal is to get as many developers as possible to accelerate through our cloud, built on custom chips for inference. So if that sounds like you, well, you know what to do. For now, I leave you in excellent hands. Over the next month you will hear from me a little bit here and there.

But you'll also hear from our new host and my good friend, Eric Tornberg. Eric is A16Z's newest general partner. And he's also put in his hours behind the mic hosting several shows at... podcast and newsletter network from the ground up. Thank you so much for listening all these years, and I truly can't wait to join you on the other side. It's been a fun ride, A16C, but as Marc Andreessen said...