--:-- Jack Altman: Here's what I would encourage people to do. Here's the thought experiment to do. Write down on a piece of paper two lists: What are the things that I believe that I can't say? And then, what are the things that I don't believe that I must say? Just write them down.
--:-- [Intro music fades in]
--:-- Narrator: What happens when startups don't just sell the tools, but decide to take over the entire industry? On today's episode, Marc Andreessen, co-founder of a16z, sits down with Jack Altman, co-founder and CEO of Lattice, to unpack how the venture industry is changing—from small seed funds to multi-billion dollar barbell strategies—and what that means for founders, funders, and the future of innovation.
--:-- Marc explains how the classic playbook of picks and shovels investing gave way to full-stack startups like Uber and Airbnb, and why the biggest tech companies today are not just building tools, but replacing entire sectors. He also talks about the realities of fund size, venture returns, power laws, early-stage conflict dynamics, and why missing a great company matters far more than backing a bad one.
--:-- And then it gets even bigger. Marc dives into AI as the next computing paradigm, U.S.-China geopolitical risk, and why he thinks we're in a capital-T test for the future of civilization. This episode is about asymmetric bets, ambition at scale, and the deep forces reshaping tech and power.
--:-- [Intro music fades out]
--:-- Narrator: 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 companies discussed in this podcast. For more details, including a link to our investments, please see a16z.com/forward/disclosures.
--:-- Jack Altman: I am so excited to be here with Marc Andreessen. Marc, thank you so much for doing this with me today.
--:-- Marc Andreessen: Jack, it's a pleasure.
--:-- Jack Altman: So what I wanted to start with was the topic of small funds, big funds. We had Josh Koppelman on the podcast, and he made a point that resonated around fund size, the outcomes in venture, and just looking at the math of it all. I think as venture funds have grown, it spoke to a lot of people about what the plan is and how tech is going to go. To start, I'd be curious to hear your thoughts around that whole dynamic. Obviously, you've got a big venture firm. I just want to hear your perspective on this whole topic.
--:-- Marc Andreessen: Let's start by saying Josh is a longtime friend and a hero of the industry. I say that because he started First Round Ventures back in the very dark days—I forget the exact year—but back during the dark days after the 2000 crash. There was a period back then when the total number of angel investors or seed investors operating in tech was maybe eight total. Ben and I were two of them. This was the heyday of Ryan Conway, Reid Hoffman, and a very small group of people who were brave enough to invest in new companies at a point in time when basically everybody believed the internet was over. The whole thing was done.
--:-- I think that was an incredibly heroic, brave act. It obviously worked really well. It turns out buy low, sell high actually is a good strategy. It's very nerve-wracking when you're trying to do it, but it does work. He had brilliant timing for when he started, and the companies he supported have gone on to become incredibly successful. We've worked with him a lot, so we're a big fan of his.
--:-- Jack Altman: I didn't actually hear the discussion. As a rule, I never read or watch anything I'm involved in.
--:-- Marc Andreessen: That's good.
--:-- Jack Altman: Well, it wasn't about me. I totally missed it. To summarize, he coined this "venture arrogance score" idea. Basically, if you're going to own 10% of a company at exit and you want to have a 3x fund, and you're probably going to have a power law of outcomes, you basically need your big outcome to be really big. The question he was posing broadly is: Are the outcomes going to be much bigger? Are you going to own a lot more? Are you going to hit a lot more winners? It was that math question.
--:-- Marc Andreessen: I'll say a couple of things. First, venture is actually a customer service business in our view. There are two customers: the LPs and the founders. We think of them both as customers. At the end of the day, the market will figure this out. LP money will flow to where they think the opportunities are, and the best founders definitely pick who their investors are. It's actually a very unusual asset class—it's the only one in which the recipient of the capital cares where the money comes from and picks it. So the market will figure this out.
--:-- To respond to your general point, I think the big thing is the world has really changed. Modern venture capital in the form we understand it is basically a product of the 50s and 60s. Originally, Jock Whitney from the Whitney family created the model. George Doriot, an MIT professor, created a version of it. Then came the great heyday of the 1960s VCs—Arthur Rock, Don Valentine, Pierre Lamond, Tom Perkins, Gene Kleiner, all those guys. From that period, the 1960s through about 2010, there was a venture playbook that became very well-established. It consisted of two parts: a sense of what the companies were going to be like, and what the venture firm should be like.
--:-- The playbook was that the companies are basically tool companies—picks and shovel companies. Mainframe computers, desktop computers, smartphones, laptops, internet access, software, SaaS, databases, routers, switches, disk drives, word processors—all tools. You buy the tool, the customer buys the tool, they use the tool. But it's a general-purpose technology sold to lots of people.
--:-- Around 2010, I think the industry permanently changed. The big winners in tech became companies that go directly into an incumbent industry. The big turning point was Uber and Airbnb. Uber in 2000 would have been specialist software for taxi dispatch that you sell to taxi cab operators. Uber in 2010 was, "Screw it, we're doing the whole thing." Airbnb in 2000 would have been booking software for bed and breakfasts running on a Windows PC. Airbnb was just, "Screw it, we're doing the whole thing."
--:-- Chris Dixon came up with the term "full stack startup." The other way to think about it is the company is delivering the entire promise of the technology all the way through to the actual customer.
--:-- Jack Altman: Which is basically quicker to get there. Also, I suppose you get more margin capture when you do it that way. The technology gets seeped in rather than having to sell it through. Is that the idea?
--:-- Marc Andreessen: Prior to 2010, there were two kinds of tool companies: consumer tool companies and business tool companies—B2C, B2B, as we called them. The consumer side was great—selling video games and consumer software—but there was only so far that was going to go. On the B2B side, for things like taxi dispatch or bed and breakfast bookings, the problem is you're selling advanced technology into incumbents that are not themselves technology companies. Are they actually going to take those tools and build the thing that the technologists know should get built?
--:-- A more modern version is what you see now with cars. Who's going to build the self-driving electric car? Is it going to be an incumbent who's able to adjust, buying good components? Or is it going to be a Tesla or a Waymo?
--:-- Jack Altman: Same with SpaceX and NASA, I suppose.
--:-- Marc Andreessen: Exactly. There are many companies that sell technological components that go into rockets. But was any of that going to lead to the existing rocket companies making a rocket that lands on its butt and gets relaunched within 24 hours? Same thing with Airbnb or Uber. If you had sold the Uberized version of taxi dispatch software to the taxi operators, would it have resulted in the Uber customer experience?
--:-- As Peter says, these things are overdetermined—a bunch of things happened. The smartphone completed the diffusion challenge for getting computers in everybody's hands. Mobile broadband completed internet access in everybody's hands. The minute you had that, you had this ability to get directly to people in a way you never had before. You didn't need a giant marketing campaign or a giant established consumer brand. There was a way to get to market that didn't previously exist.
--:-- Consumers evolved too. Gen X and millennials were much more comfortable with technology than the boomers were. They were entering their consumer prime when this happened. Then you started having these big successes: Uber, Airbnb, Lyft, SpaceX, Tesla. You stack those up and at some point you realize there's a pattern here. That's what's been happening.
--:-- We're 15 years into that. That idea has now blown out across every industry. The tech industry used to be a relatively narrow picks and shovels business. Today, it's a much larger, broader, more complicated process of applying technology into every area of business activity. The result is that the companies are much bigger. When you're both the picks and the shovels of the whole company, you're much bigger.
--:-- Jack Altman: Yeah, you eat the market.
--:-- Marc Andreessen: That changes venture math. Tesla ends up being worth more—there have been points in the last five years when Tesla alone was more valuable than the entirety of the auto industry put together. SpaceX, Uber is worth far more than the totality of every black cab operator and taxi cab company that ever existed. Airbnb is worth far more than the bed and breakfast industry ever was.
--:-- It turns out some of these markets are much larger than people think. When we do retrospective analysis over 15 years, one of the hardest things has been market sizing. Sometimes we overestimate, but more often for the winners, it's the other way.
--:-- Jack Altman: More often it's the other way.
--:-- Marc Andreessen: Yeah. This goes to venture economics. The core thing about venture bets is that venture doesn't run on leverage. Nobody will bank a startup or a venture firm for leverage because there are no assets when these things start. It's asymmetric. You can only lose 1x, but you can potentially make 1,000x. That means there are two errors in venture: the error of commission, where you invest in the thing that fails, and the error of omission, where you don't invest in the thing that succeeds. Overwhelmingly, the error that matters is the error of omission.
--:-- If you run an analysis that says ride sharing is only ever going to be as big as taxi cabs, that leads you to the error of omission. You don't make the bet. That's the difficulty with market sizing.
--:-- Jack Altman: In your view, does this only apply up to a certain size? You look at some rounds that happen at huge valuations in companies that would otherwise be a large IPO. Let's say somebody's raising $10 billion at $100 billion. Does the power law still apply up there? How do you think about that type of round? Do you see venture capital turning into private equity at the higher end?
--:-- Marc Andreessen: There are two questions embedded in that. One is why aren't these companies public? The second is, even whether they're public or not, is it still the lose-one, win-20 dynamic?
--:-- The smartest public investors I've met with basically have the view that the public market works just like the private market with respect to this dispersion of returns. The extreme case I'll make is that there may be no such thing as a stock—only an option or a bond. The reason is there are fundamentally two ways to run a company. One is to try to shoot the moon, build for the future. The other is to try to harvest the legacy. If you're shooting for the moon, the big risk is you might fail. But if it works, you have this telescoping effect in the public market just as much as in the private market.
--:-- Historically, returns in the public market have been driven by a very small number of big winners, exactly the same way as in the private market. You see that playing out right now in the S&P 500. I've been saying for years that the S&P 500 is no longer the S&P 500. It's the S&P 492 and the S&P 8. There are 492 companies that have no desire to really charge hard at the future. They won't do it. Eight are all in. Everybody always knows who the eight are because it's completely obvious—they're the ones building all the new things.
--:-- If you disaggregate public market returns over the last 10 years, you see this dramatic explosion of value among the eight and relatively modest growth of the 492. Even the S&P 500 is having a portfolio of bonds and options. It's incredibly barbelled.
--:-- People get cynical and say, "If not for the eight, the stock market..." And you say, "Yeah, but that's the whole point." If you have a healthy functioning capitalist economy, the whole point is some number of these things are going to go non-linear. When someone says, "They're not a very good investor, but they invested in [name that $100 billion company], so they got lucky," you say, "Okay, yeah. That's the point. That's the job. That's the desired outcome."
--:-- The classic joke of venture is, "Isn't there just a way to invest in the good companies and not the bad companies?" For 60 years, we've been trying to figure that out. Here's a fun fact: Over the last 60 years, every one of the really great venture firms missed most of the great companies while they were investing. The best firms in the world—Kleiner Perkins in the 90s, Benchmark in the 2000s, Sequoia in the 2010s—they flat out missed most of the winners in each cohort. On one hand, you think, "Can't you do better than that?" But super geniuses have been trying for a very long time. We could have a whole separate conversation about why this is so difficult.
--:-- Jack Altman: The thing you said about companies building the whole stack—roll-ups are super popular. Is it fair to take from what you said that you're bullish on that strategy? Instead of building accounting software and selling it to accounting firms, just buy an accounting firm, become an accounting firm, AI yourself. I think that's becoming a more popular strategy. Do you like that? Or is there a nuance why it's different to buy something rather than build it from the beginning? What do you think of this whole roll-up thing?
--:-- Marc Andreessen: Let's come back to the venture question, because I was still winding up into that. But this is actually also relevant to that. There are a bunch of great, really good firms trying to do this roll-up thing. The opportunity is very obvious. The challenge is that cultural change of an incumbent, a legacy company, is just really difficult.
--:-- Charlie Munger was once asked a few years ago—I think GE was going through a big issue—and at a shareholder meeting he was asked, "How would you fix the culture at GE?" He said, "I have no idea. I don't even know how you would change the culture at a restaurant."
--:-- Jack Altman: That's funny.
--:-- Marc Andreessen: Right? How do you do that? It's really hard. People doing it have theories. I think we're much more oriented towards just trying to back...
--:-- Jack Altman: I think it gets a little into this private equity, venture-private equity blend. It's related not even just in dollar size, but in the mindset.
--:-- Marc Andreessen: This is where I go back to my bonds versus options thing. Fundamentally, the way I always describe venture is that we are buying long-dated, out-of-the-money call options. It seems completely insane, except when they pay off, they pay off spectacularly well. A lot of them expire out of the money. Statistically, top venture capital has a 50-plus percent failure rate.
--:-- Jack Altman: Yeah, yeah, yeah. I just want to get your hot take. I really wanted to hear about this, but we can go back to the venture math thing. I think there's a lot more in there.
--:-- Marc Andreessen: Okay, good. So anyway, what's happened is the world has changed. The number of companies being founded that are going to be important keeps expanding. The number of categories those companies are in keeps expanding. Those companies are more complicated now because they're full stack—they're in these incumbent industries. The winners are getting bigger.
--:-- Look at the S&P 8. They're all venture-backed. On any given day, any one of them is bigger than the entire national stock market of countries like Germany, Japan, and the UK. The telescoping effect of victory is just incredible.
--:-- What Ben and I did is we looked at this as it was happening. We said, "This is different. You could sit here and do things the old-fashioned way, but the world is moving on." This goes back to the customer service aspect. The founders starting these kinds of companies need something different. It's not sufficient anymore to have investors operating the way they did for the previous 50 years. That's not the value proposition they need. That's not the help they need. There's a different way to do it.
--:-- I think the venture industry had to restructure to accommodate the change in the market. Having said that, I don't think that's an argument that big firms win everything. That's definitely not my thesis. That's also not how I'm deploying my own money, which I want to talk about because I'm living what I'm about to say.
--:-- I think what happens is what Nassim Taleb calls the barbell. You have a continuum. On one side, you have high scale. On the other, you have high specialization. In industries that mature and develop this way—including many industries in the last 100 years—they often start with generalists that are neither subscale nor particularly specialized. Over time, they get disintermediated. Scale players emerge on one side, specialist players on the other.
--:-- The most obvious example in everybody's lives is retail. When I was a kid, there were department stores. Pretty good selection at pretty good price, but not a great selection and not a great price. Today, those are all out of business. They're gone.
--:-- Jack Altman: Crushed by Amazon on one end and amazing retail on the other.
--:-- Marc Andreessen: Exactly. Why do you go to Amazon or Walmart? When you go to Amazon or Walmart, you get an unbelievable selection of basically anything that's a commodity at super low prices. It's impossible to compete with that if you're subscale. Then the specialist retail experience is the Gucci store, the Apple store, the $15 candles. They give you Perrier when you walk in. They're so happy to see you. They'll do private showings, pour champagne. It's an entire experience.
--:-- What's happening is that the middle gaps way out and never comes back together. What the consumer does is build a portfolio of their experiences. They buy things at unbelievably cheap prices at Walmart and Amazon, which gives them more spending money to spend on the boutique.
--:-- Jack Altman: The bar that's in the middle is screwed.
--:-- Marc Andreessen: Yes.
--:-- Jack Altman: What is the mechanic by which they're in trouble? Is it because the customers go away?
--:-- Marc Andreessen: The founder customers go away. The office. They're neither getting size and scale value nor getting a special focus.
--:-- Jack Altman: Can you be a specialist with a $2 billion fund, let's say?
--:-- Marc Andreessen: Obviously we're at scale, but we do have a specialist approach inside the scale. We have investment verticals. They're discrete teams. In some cases, they have discrete funds. They have trigger-puller authority. They can make investment decisions. Ben and I don't sit and decide if this is a good investment or bad investment. Our specialists make those decisions.
--:-- Jack Altman: Do you basically determine that by "this is the size we think you can function"? This is the biggest you can function as a specialist in a highly successful way, and then we're just going to put a bunch of those together? Is that what defines the size?
--:-- Marc Andreessen: It's two parts. One is the external view: what's the size of the market opportunity? How much money does this strategy, this vertical need? How many companies are going to be? How complex is it? The other is the internal dynamic. If you're going to have a team, you need everybody around the table to be able to have a single discussion. That puts natural limits on how big that can get.
--:-- Jack Altman: What's your limiting reagent to building an even bigger firm? Is it the number of productive partners that can do this?
--:-- Marc Andreessen: Conflicts. Conflict policy.
--:-- Jack Altman: Conflicts.
--:-- Marc Andreessen: That's the single biggest issue by far. If you had all the great GPs wanting to work here, there would still be issues.
--:-- Jack Altman: So what's the conflicts thing?
--:-- Marc Andreessen: The conflicts thing is that mainline venture firms forever—the firms that do Series A, Series B, Series C, especially Series A and B—the relationship with the founder is just too deep. If you as a venture firm invest in a direct competitor, it's a giant issue. The founder you're already invested in will be extremely upset with you.
--:-- Jack Altman: Do you think that's practical? Do you think it's all emotions? Do you think it's correct that firms shouldn't do conflicts?
--:-- Marc Andreessen: When we were startup founders, we felt this very deeply. When you're a startup founder, the whole thing is so tenuous. Is this thing going to work? There are 18,000 things that are going to go wrong. People are telling you no every day. "No, I'm not going to come work for you. No, I'm not going to invest in you. No, I'm not going to..." Then your board member invests in a competitor and it's a dagger to the heart. What happens is the founder has to go into the all-hands meeting and explain why your investor has given up on you. You do some song and dance about it. Your employees look at you and think, "You, the founder, are so weak and lame. You can't even get your board member to not invest in a competitor."
--:-- Jack Altman: What about the marginal stuff, though? All these companies are near each other. They blend. They evolve over time. How does this play out on a practical level for firms?
--:-- Marc Andreessen: It almost never plays out the way that the founders think it's going to play out. And I say that in two dimensions. Number one, the companies—historically what we've seen is the founders who think that they're directly competing with each other generally end up not doing so. Because one or the other of them changes strategies and then they diverge. Which, by the way, is natural because it's specialization. The company specializes, they end up not competing.
--:-- Marc Andreessen: But the other thing that happens is two companies that were not competing that you're already invested in pivot into each other.
--:-- Jack Altman: Yeah, and then they're mad at you.
--:-- Marc Andreessen: Yes, and then they're very upset. And you have to remind them that you didn't know that was going to happen and it's not your fault. And then they're still upset. The founders are not... and also we have very low predictability in terms of where the conflicts are going to be. But that doesn't ameliorate any of the emotion at the time. It doesn't help for us to explain to the founder, "Oh, don't worry about this guy who you think is directly competitive because he won't be in a year."
--:-- Jack Altman: Yeah. Because you can't prove that. And the issue is in the moment. What does that leave for your strategy? Meaning, if you know conflicts are this huge issue, and you've got a big aggregate fund, and it's very important to catch winners, and then you invested in Blue Origin, which is really good, but SpaceX is bigger or whatever happens. What does that imply for your strategy when it comes to doing seeds and A's versus saying, "Let's just wait till the D. Let's have D be our early stage"?
--:-- Marc Andreessen: That's right. So the most obvious thing you do is you just say, "Oh, we just need to wait because we need clarity." Just don't deal with this whole issue. Keep delaying until it's obvious what the answer is. If it's big, it's going to be really big so we can buy later. But then the problem with that is, now you're out of the venture business. Because now you're doing serious D rounds. Now you're a pure growth investor. And by the way, there are very good pure growth investors. But our determination is to stay a venture investor.
--:-- Jack Altman: Why is it so important? Is it just because it's what you like? Or is there a strategic reason that it's important to stay doing early?
--:-- Marc Andreessen: We've always wanted to be the founder's best partner. To be the one who's closest in, the one that can really be relied upon, the one that's going to be around for the longest amount of time, the one they can really trust. And it only happens early.
--:-- Marc Andreessen: It's hard to insert after that. And look, the other thing is, there are great growth firms that invest later and have done very well, but we just think there's so much information at the early stage. For example, when we make a growth investment, because we have the active venture business that we have, by the time we make a growth investment, we have either invested in the company for several years, or at the very least, we've met with them repeatedly over time. We end up with enormous amounts of information.
--:-- Marc Andreessen: And then the other thing is time arbitrage. Sometimes the right answer is just to invest in SpaceX later on. But sometimes the answer is no, there's actually a new thing. Do you invest in the MySpace growth round at the Facebook seed round? If you're not in the early stage, you won't know that because you won't see the early things.
--:-- Marc Andreessen: And then the other thing I'll say is financially, one of the things people say that is inaccurate is they say if you're running a big fund, you won't have the time to spend on early stage opportunities because you can't justify it before you're putting the money. But that's actually not true in venture because the aggregate dollar return opportunity on early stage is just as high as any growth investment. If you get the right venture investment and you can make $10 billion on the upside case, it's definitely worth my time to spend with you. So I spend as much time as I can with the early stage founders for that reason.
--:-- Jack Altman: So the barbell—there's big on one end, and something sort of like me on the other end. Selfishly, I'd love to know, I would assume you think it's better to be the big version. But if you were conditioned on needing to be me at the small end of the barbell, how would you approach it?
--:-- Marc Andreessen: They're both good. This is the thing—they're both good. And if I were for some reason not doing this, I would immediately do what you're doing.
--:-- Jack Altman: That's good to hear.
--:-- Marc Andreessen: Yes, 100%. And I actually invest this way. My liquid assets are basically tied up in either a16z funds on the one side, or I run a very aggressive personal investment program in early, early angel and early stage seed funds. It's because I believe in the barbell. I believe in it so much.
--:-- Marc Andreessen: But the conflict thing I wanted to explain because that's the issue. For us, when we do seed investing, we have this problem every single time we're looking at a seed investment: are we really fully convicted that this is going to be the winner?
--:-- Jack Altman: Even at seed, it creates a conflict?
--:-- Marc Andreessen: For a board seat? There's debates. There's always debates on this. Do the seed ones care as much? Do the growth ones care as much? Do the crypto ones care as much? What I'll tell you is it's not a logical question. It's an emotional question. And we're just very sympathetic to the founder that needs to be able to justify their authority.
--:-- Marc Andreessen: You also definitely can't ask while you're making the investment. If somebody asked me while they were making an investment, "Hey, is it okay if we invest in a conflict in a couple of years?" I'd be like, "What are you talking about?" We used to have this separate branded thing called a16z Seed and we were like, "Well, we have a different conflict policy on this." It's great in theory, but it's like, no, it's a16z.
--:-- Marc Andreessen: The way I think about it is the more successful you are as a venture firm, the bigger the issue this is going to be because the more the people that you were investing in are going to care. It's the downside of success. The only investors you don't care about are if they literally don't matter at all and everybody knows that.
--:-- Marc Andreessen: So both of these things are true. Number one, we still do lots of early stage investing and we will make seed bets. But it's also true that we can't structurally, for this reason, do all of the seed investments that we would like to do. In fact, we can't even do a tiny fraction of them. Structurally, we just can't do it.
--:-- Marc Andreessen: This goes back to the barbell. It's the same reason why Amazon can't give you the champagne experience. They're not set up for it. It's not a scaled strategy. So what has to happen is the other side of the barbell: the specialization and intense focus and deep relationship thing. That's the role of the angel investor and the seed investor. In startups, that's incredibly important because that's the most formative, fraught time in the life of these companies. As you know, half the time, these are people who haven't started a company before, haven't run a company before. Some of them haven't had a job before.
--:-- Marc Andreessen: They need to learn a lot and they need people to work with them. There have to be, and there are, incredibly high-quality seed investors and angel investors on that side of the barbell. The big firms, if we succeed, we succeed by generating large numbers of aggregate dollars and a very good percentage return. The seed investors have this perpetual opportunity to absolutely shoot the lights out on upside. There are seed funds that generate 200x, 300x returns. These are both good strategies adapted to the current reality of the market.
--:-- Marc Andreessen: Two things fall out of that. One is the death of the middle. It just doesn't make sense to have the old-fashioned Series A, Series B, six GPs, $300 million funds, sitting on Sand Hill Road waiting for people to walk in the door. Those days are over and those funds are shutting down. That model is going away.
--:-- Marc Andreessen: The other thing that happens that causes some of the tension is this: what does a successful seed investor do? He raises more money and wants to become a venture investor.
--:-- Jack Altman: Right.
--:-- Marc Andreessen: But then you're going from one side of the barbell back to the middle and you're creating that same problem again. I think that's where the tension is coming from.
--:-- Jack Altman: I also feel like the mechanic that happens a lot of times is when you grow the fund, you raise a huge fund and then you start deploying it into things just because you've got to deploy at some pace. The threshold for—if I've got to deploy $400 million this year and I only see $700 million worth of investable things, I'm going to do four-sevenths of them. Versus, presumably, if you only had to do one-seventh of it, you'd pick better, hopefully.
--:-- Marc Andreessen: Yeah. I think that's part of it, but I think the related thing is your competitive set has changed. What we find with seed investors who migrate up and then regret it later is they didn't realize what their competitive set was.
--:-- Jack Altman: Right, because now they're going for bigger, more competitive rounds against you and Sequoia.
--:-- Marc Andreessen: Yeah. All of a sudden, now you're competing for $15 million bets. Good luck.
--:-- Jack Altman: Right, exactly.
--:-- Marc Andreessen: Look, market fundamentalist: if you have a better value proposition than Sequoia, you should go offer that. But I would not accidentally end up competing with Sequoia for Series A. That's a bad way to live. And I think that is what has happened to a bunch of the seed funds that have gotten larger.
--:-- Jack Altman: Why is it so rare for somebody to break through and get—I mean, you did it, and that's one that happened in the last 15 years. Maybe there's a couple others. Why is it as rare as it is? It seems like almost more rare than a new big company in a way.
--:-- Marc Andreessen: Yeah, that's true. In fact, our analysis when we started was there actually hadn't been—I think there had been two firms. Andy Rachleff actually... I mean Thrive also. Thrive was after us, they've done great. But in the 30 years before us, we think that there were only two new VCs that actually punched through to become top tier. In other words, VCs that were not either firms built in the 60s and 70s or derivations of those firms.
--:-- Jack Altman: Founders Fund?
--:-- Marc Andreessen: No, Founders Fund started around the same time we did. They were a little bit earlier, but around the same time. But I mean over the preceding 50 years.
--:-- Marc Andreessen: Sevin Rosen. You won't even remember.
--:-- Marc Andreessen: This is the thing. You won't even recognize these names.
--:-- Jack Altman: Geez, I need to read a book or something.
--:-- Marc Andreessen: Sevin Rosen was the venture firm that famously funded Compaq Computer, which was the big winner. And then they went on to become a successful firm. This guy, Ben Rosen. Early leader in the space. And then there was a firm called Hummer Winblad, which was a software specialist firm in the late 80s, early 90s. Those were the only two that punched into the top end while they were operating.
--:-- Marc Andreessen: Neither one of them sustained it. But they got there for a bit. That was the success case. So this is a little bit like Elon looking at the history of the car industry and saying, "Tucker Automotive in the 1950s." It's very, very rare.
--:-- Jack Altman: So why is it that rare?
--:-- Marc Andreessen: Two reasons. Number one, there's the intimate reason for it, and then a sort of macro reason. The intimate reason is you're going to have this incredibly intimate experience, a very close trust relationship with whoever you're working with. Can you reference them? Do they have a history and track record of the kinds of behavior and insight that you need? It's very hard to do that. It's very easy for an existing firm with a long track record of success to prove that. It's very hard if you don't.
--:-- Marc Andreessen: But the other reason goes back to the way the world is changing. We always believe the thing you want from your venture firm is power. As a startup, you want them to fill in all the missing pieces that you don't yet have when you're starting a company. You need power. You need the ability to go meet customers and have them take you seriously. You need the ability to get publicity and major channels—it used to be media, now it's podcast—and be taken seriously. You need to be taken seriously by recruits. There are thousands of startups recruiting for engineers. What makes yours stand out?
--:-- Jack Altman: I sometimes describe it as a venture firm providing a bridge loan of a brand. Until you have your own brand that's bigger for your own space than the VC, you're borrowing the VC's brand.
--:-- Marc Andreessen: Exactly. And that has been very effective for a long time. That was how we looked at it when we were founders.
--:-- Jack Altman: That's why you did media from the beginning.
--:-- Marc Andreessen: Yeah. That's one of the reasons. A very, very powerful one.
--:-- Marc Andreessen: And then, by the way, you also need the ability to raise downstream money. You're going to need to raise money again. So they either need a lot of money or they need to be connected to a lot of money.
--:-- Jack Altman: Yeah, exactly.
--:-- Marc Andreessen: Exactly. You just better if they just have it. It's like being full stack.
--:-- Marc Andreessen: Then, by the way, now you're getting into things like politics, global affairs, global events. How do you navigate the world? How do you navigate Washington? The regulators show up and they want to kill you. How do you navigate that? Or you get into some giant fight with the EU. Especially these full-stack companies, they're getting involved in very complicated macro, political, geopolitical situations much more early. In some cases, they have to escalate up to senior government officials, heads of state, major heads of sovereign wealth funds. They need to get to CEOs of major companies. How do you get to the CEOs? You're a new AI company trying to redefine visual production for movies. How do you get to the studio heads? The studio heads don't have time to meet with a thousand startups.
--:-- Marc Andreessen: So basically it's projection of power. This has been one of our theories in how we built our firm: you optimize for maximum amount of power in order to be able to give the startups access to it. Both the startups that are already in your portfolio, but also the startups that don't even exist yet. This goes to why the scale thing matters so much. There's a scale aspect of power. There's a big difference between being able to get to everybody who matters and not.
--:-- Jack Altman: Why is it rare for people to be able to accumulate power even if they were—let's say everybody was trying to do it. It's not like everybody could do it. What's the cause of the rarity to be able to build enough power in that sense?
--:-- Marc Andreessen: You have to want to. We met with all the GPs of all the top firms when we were starting out because we wanted to see who we could be friends with. It worked very well in some cases and not well in other cases. But one of them told us—this is a GP at a top firm in 2009. He said, "Yeah, the venture business is like going to the sushi boat restaurant."
--:-- Jack Altman: The sushi boat restaurant—it's a sushi restaurant where they've got the boats. Like a conveyor belt.
--:-- Marc Andreessen: Conveyor belt, right. The little sushi boat comes by. There's a tuna roll, a shrimp roll, this or that. He said, "Basically, you just sit on Sand Hill Road. You're like, 'We're going to crush these guys.' And the startups are going to come in. If you miss one, it doesn't matter because there's another sushi boat coming up right behind it." He said you just sit and watch the sushi go by, and every once in a while you reach in and pluck out a piece of sushi.
--:-- Jack Altman: [Laughs] That's funny.
--:-- Marc Andreessen: We walked out of that and said, "What the hell?" In what industry is that—2009, yeah. That was a very common mid-sized venture attitude. When I came to Silicon Valley in 1994, I had never heard the term venture capital. I didn't even know this thing existed. Then my business partner Jim Clark explained it to me, and I was like, "There are guys just sitting there waiting to give you money?" But you see this and you're like, "This is going to get eaten alive. The minute anybody takes this seriously, it's all going to change."
--:-- Marc Andreessen: It was this very clubby cartel kind of thing. It was fine as long as the ambitions of the industry were constrained. The tools companies didn't need all the power. They needed some of it, but they weren't dealing with governments or these big macro issues in the early years.
--:-- Marc Andreessen: Here's another thing that's happened: the world is globalized. Startups 30 years ago would spend their first decade just in the U.S. and then start to think about Europe and global expansion. Now you have to think about being a global company up front because if you don't, other people are going to do it.
--:-- Marc Andreessen: The expectations are much higher than they used to be.
--:-- Jack Altman: Maybe one final question on this topic of fund size, and then I want to go to AI. What do you think is the limiting factor for the creation of a lot more really big companies? Do you think it's founders? Do you think it's capital? Do you think it's market maturity? Do you think it's underlying tech stuff? If you had to pinpoint the one or two things that would allow for there to be way more big companies, what is it?
--:-- Marc Andreessen: There's the holy trinity of venture startups: people, market, and technology. I think the answer is all three. There's some limiting issue with markets—how many markets are there? How big are they? How ready is the market to take something new? Then there's the technology question. From the venture perspective, technology moves in stair steps. Things become possible in the world of smartphones that just weren't possible. You couldn't do Uber when everybody had a laptop. You had to wait until they had phones. Technology moves in stair steps. You get these paradigm shifts, platform shifts. They come when they come. Until they come, you can't do it.
--:-- Marc Andreessen: And then the people side—this is the one that vexes me the most. How do you get more great founders?
--:-- Marc Andreessen: I think there is definitely a training thing that is real. Getting people into the right scene in the right way. What Y Combinator does, what Thiel Fellows does—those are real things. They help a lot. But also, there are just not an infinite number of people running around who have the...
--:-- Jack Altman: You probably figure there's a lot of people who could have built big companies who haven't though.
--:-- Marc Andreessen: Hopefully. A few. I don't know, some number. But there must be people who are just in academia or government or education, doing something completely different, who if they were attracted to startups would have built a big company.
--:-- Jack Altman: So yes, but then the other question is, why didn't they? Why didn't they do the things required to get themselves in that position?
--:-- Marc Andreessen: Well, it could have been like 2001. Too many people were too scared to do it or didn't know about it.
--:-- Jack Altman: But what does that tell you about the people who didn't do it?
--:-- Marc Andreessen: Yeah. They were heard. I can tell you who didn't listen to that—it was Mark Zuckerberg.
--:-- Jack Altman: Are there more good—let's press this point harder. I always describe this as the test with capital T. If you're not in position to do the thing, the fact that you're not positioned to do the thing meant that you've already flunked the test.
--:-- Marc Andreessen: Well, I guess the question would be, is there a subset of people who could build Facebook who, other than being too scared to do it, would have had all the other ingredients? So when everybody's not scared, you get more Facebooks. There's a line in the movie—I actually never saw the movie, but there's a line: "If you could have built Facebook, you would have built Facebook."
--:-- Jack Altman: Yeah, yeah, yeah. That's right. That was a good line.
--:-- Marc Andreessen: Right. So this is the thing.
--:-- Jack Altman: Are there more great founders today than when you were, let's say, in net? Do you think there are more now than there were 20 years ago?
--:-- Marc Andreessen: I believe there are. But how many more? Is it five times more? Is it 50% more?
--:-- Jack Altman: Well, the number of wins is increasing. We used to talk about the 15 a year that matter. That number's probably, if you do the analytics, up like 10x. You think there's like 150 companies a year that really matter. The reason is because there are so many more sectors now. Industry maturation. So by inference, there kind of have to be—you're saying the markets are better more than you're saying the founders are better.
--:-- Marc Andreessen: Well, maybe a little bit of both. Look, I think the founders are getting better. Part of the founders getting better is they have better training. They're all online. When I showed up here in 1994, there were literally like three books in the bookstore, none of which were that great. It's not that the DNA is better. It's that the ecosystem has matured to teach people better. People come in and they've watched every video, every episode of your podcast. They walk in knowing all this stuff. Y Combinator didn't exist. Thiel Fellows didn't exist. That definitely helps.
--:-- Marc Andreessen: Brian has this great term: "scene plus genius." The individual genius on his own is always hard to get things done. Some people do, but it's difficult. More often in a profession where you're seeing creativity happen, there's almost always a scene. Silicon Valley is definitely a scene in that way. People come here and they just get better. They meet more people who are like them. They aggregate together. They learn from each other. So the founders are getting better. There are more of them.
--:-- Jack Altman: But does that mean there's now 10,000 as opposed to 1,000?
--:-- Marc Andreessen: Yeah. I don't know. And there are 8 billion people on planet Earth. Why are we debating whether it's 1,000 or 10,000?
--:-- Marc Andreessen: Right. I would hope over the next years and decades, we'll all figure out a way to make sure we get everybody who can do it and get them to do it.
--:-- Jack Altman: That's a good segue into AI. Do you feel that we're now at the beginning of what is the new next important paradigm? Is this cloud, but on steroids?
--:-- Marc Andreessen: Oh, yeah. I think much larger, and I'll explain why. I described the triangle: people, technology, market. But technology is ultimately the driver. For venture, the technological step function changes drive the industry. They always have. If you talk to the LPs, you can see this. When there's a giant new technology platform, it's an opportunity to reinvent a huge number of companies and products that have become obsolete and create a whole new generation of companies. They generally end up being bigger than the ones they replaced. The venture returns map this.
--:-- Marc Andreessen: They come in waves. The LPs will tell you: there was the PC wave, the internet wave, the mobile wave, the cloud wave. That was the thing. And by the way, in venture, when you get stuck between waves, it's actually very hard. You've seen this for the last five years. For the last five years, it's like, "How many more SaaS companies are there to found?" We're just out of ideas. We're out of categories. They're all done.
--:-- Marc Andreessen: It's when you have a fundamental technology paradigm shift that gives you an opportunity to rethink the entire industry. It would have been very sad if the AI breakthrough didn't happen. The state of venture would be sad, I think, three years ago. When we were talking to LPs three years ago, we were just basically saying we're in—Chris Dixon has this framing he uses. He calls it "search mode" or "hill climbing mode." In search mode, you're looking for the hill. Three years ago, we were all in search mode. That's how we described it to everybody. We're in search mode, and there are all these candidates for what the things could be. AI was one of the candidates. It was a known thing, but it hadn't broken out yet in the way that it has now.
--:-- Marc Andreessen: We were in search mode. Now we're in hill climbing mode.
--:-- Jack Altman: Thank goodness.
--:-- Marc Andreessen: Big time. And look, on the technology breakthrough itself, a year ago you could have made the argument that you don't know if this is really going to work because LLMs have hallucinations. It's great that they can write Shakespearean poetry and hip hop lyrics. Can they actually do math? Can they write code?
--:-- Marc Andreessen: And now it obviously is. And now they obviously can. For me, the turning point moment, the moment of certainty, was the release of O1 from OpenAI, the Reasoner, and then DeepSeek R1. Those happened back to back. The minute those popped out, you saw what was happening with the scaling law around that. You just think, all right, this is going to work because reasoning is going to work. And that is what's happening.
--:-- Marc Andreessen: Every day I'm seeing product capabilities, new technologies I never thought I would live to see. Really profound. I actually think the analogy isn't to the cloud or the internet. I think the analogy is to the invention of the microprocessor. This is a new kind of computer. Being a new kind of computer means essentially everything that computers do can get rebuilt.
--:-- Jack Altman: So you're investing against the thesis that basically all incumbents are going to get nuked.
--:-- Marc Andreessen: Yeah. And everything is going to get rebuilt. Just across the board.
--:-- Jack Altman: And you'll be wrong in a bunch of those cases because some incumbents will adapt. But the power law, the things that are right, will be super right.
--:-- Marc Andreessen: Exactly. And then AI makes things possible that were not possible before. So there's going to be entirely new categories.
--:-- Jack Altman: Is your mindset there that you should just bet on startups winning it all and go for the power law?
--:-- Marc Andreessen: Yeah, that's right. The reason is to remember two customer sets. The way the LPs think of us is as complementary to all their other investments. Our LPs all have major public market stock exposure. They don't need us to bet on incumbent healthcare companies. They need us to try to maximize alpha based on disruption. And then again, the basic math of venture: you can only lose 1x, you can make 1000x. You just slam that forward as hard as you can.
--:-- Jack Altman: When you have a moment in time worldview like this, do you give a directive that everybody needs to deploy in this kind of way? Or do you just build a system that's always picking winners from the bottom up?
--:-- Marc Andreessen: We don't do top-down investment decision making. Ben and I aren't sitting saying we need to invest in category X. We have a legal investment committee, but we don't run a process where they come to us to get approval. We let the leader of each group make those calls, and often it's delegated to the individual GP or check writer. The knowledge of what's going on and which one is likely to win is going to be focused in the mind of the person closest to the specific thing.
--:-- Jack Altman: But do you have a risk slider? Are you saying, "Hey guys, let's be at a nine right now"?
--:-- Marc Andreessen: Venture is the only asset class where the leaders of the firm are in the position of trying to get the firm to take more risk, not less risk. The natural organizational incentive is to try to reduce risk. So Ben and I are generally on the side of take more risk. One application of this is the old Sequoia adage: when in doubt, lean in.
--:-- Marc Andreessen: You see a company that is potentially very interesting, but there are issues. It's too early, the founder has a weird background, there's hair on the deal. The history of venture is when you see something very promising with a lot of hair on it, sometimes it goes to zero because the hair kills it. Sometimes it becomes the next big thing. It's much better than everything being fine.
--:-- Marc Andreessen: We describe this as invest in strength, not in lack of weakness. Differentiating good from great is straightforward. Differentiating very good from great is actually very hard. The risk-reducing way would be the checkbox thing: very good team, very good market. Then you have this other one where they've got six great things and nine horrible things. Usually the thing with the greater strengths is the better bet.
--:-- Marc Andreessen: Statistically, this shows up in the return data. The top decile firms have a higher loss rate than everybody else. In baseball, it's called the Babe Ruth effect: the home run hitter strikes out more often. The top performing venture firms statistically tend to have a higher loss rate than the mediocre firms. They're willing to invest in the thing that looks completely nuts but has that magic something.
--:-- Marc Andreessen: When Ben and I think about getting the team to take more risk, it's almost always that kind of thing. We tell the person closest to it, go with your gut. If your gut tells you there's something magical here, go ahead. It's okay because we're going to have some losses. The other form of risk I push constantly is go earlier. The natural inclination is to wait. We do want to make a lot of A bets. We're going to lose a bunch of those, but we have to get into some of these things early.
--:-- Jack Altman: There's risk of the flavor of doing things that are more asymmetric where there's hair but also brilliance. There's another flavor that's just being more aggressive: do a higher percentage of the ones where you're right on the line. Do you give that kind of guidance?
--:-- Marc Andreessen: We used to run a process called the anti-portfolio, the shadow portfolio. Every time we pull the trigger on an A round, we'd put in the shadow portfolio the other company we were looking at that we didn't end up pulling the trigger on. We'd build up the Earth 2 portfolio.
--:-- Jack Altman: I'm so curious.
--:-- Marc Andreessen: The good news is the main portfolio did better than the shadow portfolio. But the shadow portfolio was close. It did really well. So you realize you're not that smart. What does that mean? Do them both. If you had the opportunity to do both the portfolio and the shadow portfolio, you should do them both.
--:-- Marc Andreessen: In venture math, the thing that matters is, were you in the next big thing as early as you could get in, and buy as much as you could? That's the only thing that matters because if you don't do that, you miss out on the 1000x gain. The 1x losses wash right out. This idea that there's some virtue to being small and only making a few bets so you have a higher percentage... I'm glad people think that's a virtue they should shoot for.
--:-- Jack Altman: It seems very hard to assemble lots of very good productive GPs into the same firm. It's objectively rare. You've done it. Do you think of just finding greatness and you can't really teach it? Or do you think it's about creating the system and conditions in which people do great work?
--:-- Marc Andreessen: It only works if there's a reason why you would have the aggregation of GPs in the first place. Our answer to that is power. Our pitch to GPs is if you come here, you plug into this massively powerful engine. Everything you do will be blown completely out. It would be much more satisfying, you'll be able to help the companies a lot more, and you'll probably see more companies. Some people want to have colleagues. It's on us to keep proving that.
--:-- Marc Andreessen: Historically, we only hired GPs. We changed that like eight years ago. We now develop our own GPs. When we evaluate them, here's a big thing: it's fine to invest in a category five years early and something goes wrong. What's not fine is you invest in the wrong company when you could have invested in the right one. At the moment you made the investment, you made the wrong decision. Did you do the work to fully address the market?
--:-- Marc Andreessen: When we started, our friends in the business said two things. Number one, you don't know if somebody is a good GP for 10 years because you don't have the return data. Number two, nobody ever wants to admit they made a mistake, so they never actually fire anybody. They just keep them on the masthead and gently retire them out. One guy running a big firm told me they hired a partner in 1984 who was a big deal. He proceeded to nearly ruin the firm over the next 20 years. All his investments were bad, and he talked them out of all the other good investments. They couldn't get him out.
--:-- Marc Andreessen: A lot of these firms are partnerships. The problem of the partnership is it sounds good, but you end up with lots of internal dissension and then you can't make decisions. We evaluate the inputs just as much as the outputs. There's a subjective criteria for venture: are you good at it? Do you have taste? That's unquantifiable.
--:-- Jack Altman: Isn't one of the nice things about your model that somebody gets to make a call? In these partnerships, it would be very hard when nobody gets to make calls.
--:-- Marc Andreessen: Yeah, that's right. There's a taste thing. And there's a network cohort branding thing. Startups come in waves. It's not just new technology, it's also new people. These new scenes form. Are you in the scene or not? If you're not in the scene, I can't fix that for you.
--:-- Jack Altman: There's a ton of path dependence. You make an investment that gets you in the scene. Now other founders want to work with you because you invested in this really cool company. It snowballs. You can't go back and change history.
--:-- Marc Andreessen: This is the test with a capital T. There's the founder who says, "I could have done this, but I wasn't in a position to." That's your own fault. There's the anti-VC narrative: "These VCs are so arrogant, they don't see my unique genius." Raising money from venture capitalists is the easiest thing you will ever do as a startup founder. We are sitting here with checkbooks waiting to write checks. We are dying for the next person to walk in the door and be so great that they convince us to write the check. We don't care where they come from. We just want to know, do they know what they're doing? Are they going to be able to do it? Everybody else they're ever going to deal with—candidates, customers, downstream investors—is going to be much harder. If they can't pass the test of raising money, they're not going to be able to do it.
--:-- Jack Altman: There's going to be a lot of frustrated founders hearing that. When founders get a pass note, it's usually about the market or the product. A lot of times it's what you just said: they just want the founder to be great. But nobody says that to them. Is there a way for there to be a more honest, useful back and forth? Or is it just one of those impossible structural things?
--:-- Marc Andreessen: I know you think your baby's beautiful, but I think he's really ugly. It's really difficult. You embedded two things. One is, do they come across as good, which in theory is fixable. The other is, some people are better than other people at doing this. Some people should not be starting companies. They should be on a team.
--:-- Marc Andreessen: VCs make lots of mistakes. Even the great VCs are screwing up most of the time. The thing I always tell founders is, Steve Martin was asked about becoming a great standup comic. He wrote a book called *Standing Up*. He said the secret is, be so great they can't ignore you. If your business gets good enough and you prove you're really good, you don't have to show up in the one hour with a VC as very impressive. You just proved it on the field. We're dying for people to come in and be like, "I cannot believe how good this product is. I cannot believe how much the customers love it."
--:-- Jack Altman: The founders who really struggled to raise a round or two and then the business got working, there's a real strength that comes out of that. It's not the worst thing that ever happened.
--:-- Marc Andreessen: Having said that, there's breakage along the way. It sucks. It's really unpleasant. But having been a founder, it's an incredible privilege to be in an industry and in a world and in a country at a time when you can actually do this. In most of history, in most places, this kind of thing can't happen. We are genuinely trying to find the anomalies. Our business is defined by anomalies.
--:-- Jack Altman: On AI, I want to talk about the broader lens. The downside case people are very afraid of is AI embodied in humanoid robots. We have a Terminator situation. It gets agency. The happy path is it's the sickest software anybody's ever seen, a tool that humans use. Do you think about this?
--:-- Marc Andreessen: Any important new technology is dual use. It can be used for good things and bad things. The shovel can dig a well and save your life, or you can bash somebody over the head with it. Fire. The airplane can take you on a marvelous vacation or bomb Dresden. Atomic power could be unlimited clean energy or nuclear bombs. As it turns out, we got the bombs, not the unlimited clean energy. These things are double-edged swords.
--:-- Marc Andreessen: The nuclear example is really interesting. There was a big moral panic around nuclear power. The Green Movement in the 60s and 70s created the precautionary principle, which the same kinds of people are now trying to apply to AI. It says unless you can prove a technology is definitely harmless, you should not deploy it. That rules out everything. No fire, no shovels, no cars, no planes, no electricity. That's what happened to civilian nuclear power. They killed it.
--:-- Marc Andreessen: President Nixon in 1971 declared Project Independence. He said America needs to build 1,000 nuclear power plants by the year 2000, go completely clean, carbon zero, completely electric. Then he created the EPA and the Nuclear Regulatory Commission, which prevented that from happening. They absolutely killed the nuclear industry in the U.S. The Germans are going through the new version of that with Ukraine. They keep shutting down their nuclear plants, making them more dependent on Russian oil. They end up funding the Russian war machine that invades Ukraine.
--:-- Marc Andreessen: The history of moral panic and social engineering has been quite bad in its thinking and practical results. It would be a very, very big mistake to do that in AI. To regulate early and cut off the benefits.
--:-- Marc Andreessen: We're not alone in the world. We knew that before, but especially after DeepSeek, we really know that. There is a two-horse race. This is shaping up to be the equivalent of the Cold War against the Soviet Union. China has ambitions to imprint the world on their ideas of how society should be organized. The world 20 years from now is going to be running on Chinese AI or American AI.
--:-- Jack Altman: You think that's how it'll play out?
--:-- Marc Andreessen: Yeah. It's going to run on one or the other. AI is going to be the control layer for everything. It's going to be how you interface with education, healthcare, transportation, employment, government, law. AI lawyers, AI doctors, AI teachers. Do you want your kids to be taught by Chinese AI? They're really good at teaching you Marxism and Xi Jinping thought. The culture is in the weights. How these things are trained and who they're trained by deeply matters.
--:-- Marc Andreessen: This is already an issue in lots of countries. They may not want Chinese AI. But do they want super woke Northern California AI? That's another open question. If you had a choice between AI with American values versus Chinese Communist Party values, for me it's crystal clear where you'd want to go. There's also a direct military version of this. Do you want to live in a world of all CCP-controlled robots, drones, airplanes, and cars? Warfare and defense is going to fully go AI over the next 20 years.
--:-- Marc Andreessen: Robots plus AI, basically. There's a signal that you probably saw, the Ukrainian attack on the Russian airplanes. Those are autonomous drones. They were doing AI targeting of the right structural points to attack and destroy the planes.
--:-- Jack Altman: Right.
--:-- Marc Andreessen: This is a major issue with our defense doctrine, for example, regarding a potential invasion of Taiwan. Ukraine has been fielding AI-piloted jet skis. They take a jet ski, put an autonomous pilot on it, strap it with explosives, and you can send out 10,000 of those against an aircraft carrier. You just keep sending them because there's no loss of life, until you get through.
--:-- The entire supply chain, the entire defense industrial base, all the doctrine of warfare changes. The idea of human beings in planes or on submarines just doesn't make sense anymore. It's all going to change. The symmetry or asymmetry between defense and attack is going to change.
--:-- Jack Altman: You used the word "dual use." With previous technologies, they got used. At some point, does it blend from being used to being the user? A benign business example: if you could tell an AI, "Build me a software company, serve these users, run it for five years, and just wire me the money," and if that worked, in the middle of those five years, what is it doing? Are you telling it what to do? Does that happen at a warfare scale? That's where it turns into something scarier, especially with the embodied version in warfare. The prompt is just, "Fight this war for the next year."
--:-- Marc Andreessen: That's right. The domestic version is straightforward. US law and Western law have a concept of responsibility and accountability. If you use a machine to do something, it's legally your fault. If the machine goes wrong for reasons not related to you, it's a product liability issue—the manufacturer is liable. But if I buy a shovel and bash you over the head with it, the shovel killed you, but I'm to blame.
--:-- Your example of the autonomous corporation—the legal system is perfectly prepared to deal with that. That was your bot. You set it up. It's your fault. There's a natural constraint.
--:-- The most obvious military version is autonomous targeting and trigger pulling. This has been an issue in drone warfare for 15 years. Is there a human in the loop on pulling the trigger? A Predator flies overhead, sees the bad guy. How is the decision made to launch the missile? For a long time, it had to be an Air Force combat pilot who actually pulled the trigger, even if they weren't otherwise responsible for the drone's operations. You still had somebody whose job it was to make those decisions in the loop.
--:-- A lot of people in defense say it's absolutely mandatory that a human being makes the kill decision in all cases.
--:-- Jack Altman: Maybe that is the correct answer. There's a powerful argument for that—it's the biggest decision anyone can make. Even without Skynet scenarios, the idea of a human not being responsible for that decision sounds ethically and morally very scary.
--:-- Marc Andreessen: There is a counter argument: human beings are really, really bad at making those decisions. With self-driving cars, if it's safer than a human driver, there will be accidents, but there are fewer. Every post-analysis of any combat situation reveals shocking things. One is friendly fire—huge amounts of death caused by people shooting at their own troops because they're confused. Number two is the fog of war. Commanders have very little idea what's going on. Their battle plan immediately goes sideways. They don't have the information to make decisions. Number three is the physiological impact of stress. It's one thing to be on a shooting range; it's another to have a severe leg wound, adrenaline overload, and two hours of sleep. Is the highly trained person making the right decision?
--:-- There's a more basic thing. In World War II, it was estimated that only about 25% of soldiers even fired their rifles. A lot of people just don't act. The more you look at this, the more you realize the human being is actually really bad at this. Then there are issues around collateral damage—accidentally shooting civilians. You're back in the self-driving car situation. If you knew you could get better outcomes by having the machine make the decision—better, safer, less loss of life, less collateral damage—then what?
--:-- I don't have an answer to this, but I think it's a very fundamental question.
--:-- Jack Altman: This feeds into the next topic. Tech has gotten to a place where with government and politics, it's now undeniable. It used to be an underdog, but now it's too important to not be in the mix on the national stage. That has changed the dynamic insularly for Silicon Valley. People are looking at what people are doing, not just in tech, but pretty broadly now.
--:-- Marc Andreessen: That's right. I deeply agree with that. I believe it is mostly our fault in tech. There's an old Soviet joke: you may not be interested in politics, but politics is interested in you. I think we all got complacent between 1960 and 2010. We thought we could just sit out here, do our thing, talk about how important it is, but it would never become big social, cultural, or political issues. We could just get away with not being engaged.
--:-- Jack Altman: And then once it was undeniable, we weren't prepared?
--:-- Marc Andreessen: We weren't remotely prepared. It's the dog that caught the bus, being dragged behind it with the tailpipe in its mouth. Doesn't know what to do with the bus. Geography has a lot to do with this. We're 3,000 miles away. It's hard to get there. They don't come here very often.
--:-- We always wanted to build important things, and we are building important things. There are obvious political, cultural, and social consequences. If we don't engage, nobody will. The other thing I'll say is there's not unanimity in the industry on a lot of these issues. There are two giant divisions right now: big companies versus small companies. They often do not have aligned incentives or agendas. And on AI, there's a big dispersion of views even within the industry.
--:-- Jack Altman: That probably goes to why it's important for at least some VCs to have relationships with the government. Big tech has the resources to do that themselves. Small tech can't. If this is the state of the world, we as an industry need somebody to be doing it on behalf of little tech.
--:-- Marc Andreessen: That's exactly right. That's why we're doing what we're doing.
--:-- Jack Altman: On media in particular. Many years ago, biology started talking about the fracturing relationship between tech and media. From my read, it did continue to degrade. What's interesting recently is I've seen a little bit of life in tech publication stuff, but from the inside. Eric, who you just brought on as GP, is awesome at this. TBPN is really cool—I don't think I've seen something like that pop up inside tech before. What's your read on the tech media relationship and where it's been?
--:-- Marc Andreessen: My background is weird because of what happened in the 90s. I started dealing with the national press and tech press in 1993, 1994. I did an annual press tour to the East Coast, usually a week out of each year in the spring. You go around and meet with all the publishers, editors, and reporters.
--:-- From 1994 to 2016, I thought it was a quite healthy, normal, productive relationship. They would do investigative reporting and run stories they didn't like, but generally the major publications were trying to understand what was going on and be honest brokers. The meetings were super interesting. They always wanted to learn and had tons of questions. That was great until 2016.
--:-- In the spring of 2017, I went on the press tour and it was like somebody had flipped a light switch. Across the board, unbelievably hostile. Completely and across the board, a 100% sweep.
--:-- Jack Altman: Do you know why?
--:-- Marc Andreessen: Absolute hostility. The obvious answer is Trump. He got nominated and elected, and they blamed tech for both. There are other factors. On the business side, there was the fear that the internet would eat the news business in the 90s, which didn't actually happen. 2015 was the best year in history for newspaper revenues. Then after 2015, social networking went big, their businesses started to collapse, and they had lots of layoffs. That didn't help.
--:-- They would say, "Hey smart guy, that's also when you started doing all these things that actually matter more." Everything we've been discussing—the tech industry changed. You're going to get a different level of scrutiny because you deserve it. You're doing different things now. The political thing was a giant swamping factor. This whole thing ran in parallel with everything in Jake Tapper's book. They got locked into a mode of interaction. They became very polarized and lockstep. From the outside, you read it and think, "Wow, these people are really wrapping themselves around an axle."
--:-- Jack Altman: One of the hard things is as the truth has become more accessible, you more often see something in the news that you know about and think, "That's super backwards." Then somebody posts about how backwards it is. You see a clip from a major publication, here's the truth, and everybody can tell. So should we just believe the rest of it or not?
--:-- Marc Andreessen: Truth fact-checking went way up due to social media. The cliche has been that social media is where lies spread, and there's some truth to that. But the other side is what you're saying: the truth spreads on social media. Social media is an x-ray machine. Anytime there's evidence that something is not the way it's being portrayed, it shows up. People see it.
--:-- Martin Gurri wrote a book called *The Revolt of the Public* in 2015. He was a CIA analyst who did open source analysis for 30 years, studying newspapers and magazines for political forecasting. His prediction was that social media would completely destroy the authority of all incumbent institutions by revealing through this x-ray effect that none of them deserve the credibility.
--:-- Jack Altman: Do you think that's happened?
--:-- Marc Andreessen: That's exactly what's happening. There's statistical evidence. Gallup does an annual poll on trust in institutions, including the press. All the numbers are collapsing.
--:-- Jack Altman: In light of widespread social media, what would be the correct function or role of journalism?
--:-- Marc Andreessen: I'm a believer in the original idea. I'm a romantic. I like what journalism says it is. I would like it to be like that. I like what the universities say they are. I like what the government says it is. For journalism, it's just: tell us correctly and accurately what's happening.
--:-- There's a conflict at the heart of the journalism question. Journalists say two different things. One is to be fair and objective. The other is to hold power to account, or "comfort the afflicted and afflict the comfortable." There's an inherent conflict.
--:-- Jack Altman: That has nothing to do with the truth. It's just unrelated.
--:-- Marc Andreessen: Exactly. There's a selection process where people who go into journalism tend to be critical by nature. They want to be on the outside looking in. But do we need people to tell us the truth? Yes. Do we need people to hold power to account? Yes. I would like them to do that.
--:-- Jack Altman: Can they be for-profit corporations and have it work? They're getting all their distribution on social media. Eyeballs drive revenue. People want to stay in, which is unrelated to the truth. In fact, it's antithetical to the truth a lot of times.
--:-- Marc Andreessen: There are two mentalities that come out of that. One is that the profit incentive warps it, so you want it to not have a profit incentive so it can be true to itself. The other argument is if you don't like for-profits, you're really not going to like non-profits. At least for-profits have a market test, some discipline. Nonprofit just becomes somebody's agenda. "This is my agenda. I'm going to do what I feel like now." They can go arbitrarily crazy.
--:-- Jack Altman: That does sound worse.
--:-- Marc Andreessen: They're completely unaccountable. It's the opposite of accountability because of the tax break. You're actually paid as a donor to invest in the things that are the most unaccountable. They can spin into crazy land and they don't come back.
--:-- Jack Altman: The citizen journalism thing is a helpful fact check. It's good to have, but sometimes it doesn't feel sufficient to tell the full story on everything all the time. I think there's an important role. It still feels like it's very in limbo right now.
--:-- Marc Andreessen: Here's a theory that would be a reason for optimism. The last eight years were the human animal adapting to the existence of social media. It was the assembly of the brain, and you slam 8 billion people in a chat room together. We weren't wired for it. We're not adapted for it. Everything goes bananas.
--:-- Marshall McLuhan talked about this. He had a term, "the global village," which is what happens when everybody gets networked together. People miss that he didn't mean it in a good way. The nature of a village is gossip, innuendo, infighting, reputational destruction, and civil war. That functions at a certain size—up to 150 people you can kind of deal with it. At the size of New York City, it gets complicated. At the scale of the world, it's a disaster.
--:-- But you could say we went through this eight-year period where everybody went nuts in a thousand different ways. Maybe that was just us getting used to it. Maybe we just had to adapt. If you talk to young Zoomers now, a lot of the time they'll tell you, "Yeah, we don't take any of that stuff seriously. Of course you don't believe what you see on TikTok."
--:-- Jack Altman: That's wild. It's all ops. They're just adapted. I'm glad people know, but that's a crazy state of the world.
--:-- Marc Andreessen: So this is the thing on the news. Was the news ever as we were told it was? My favorite example is Walter Cronkite. People cite him as the great truth teller. He used to be on TV, and everybody would say, "Oh my God, he's going to tell you the truth." He built that reputation by going negative on the Vietnam War in 1968. He came out and said the war is unwinnable and we need to pull out. Everybody said he's the guy who told the truth, held power to account.
--:-- The problem is he went negative on the war in 1968. He was positive on it before that. What did he know the day before he said that that he wasn't sharing? And by the way, what else happened in 1968? The White House went from a Democrat to a Republican. The Vietnam War was created by Kennedy and Johnson, then inherited by Nixon. Isn't it convenient that he went negative on it when it became Nixon's war, as opposed to Kennedy's and Johnson's war? What was actually going on there?
--:-- I grew up in rural Wisconsin. We always thought the press was out to get us. We thought the press was the coast passing sneering judgment on the center of the country. We never believed the stuff to start with. People where I grew up were super resentful of how the media portrayed them.
--:-- Objective truth is a high bar. People have agendas. Maybe we just need to get all this out on the table. Particularly in politics, objective truth is not really how it works. People say, "Oh, that's a lie." I say, "Well, it's not a lie. It's just an interpretation of a situation that I wouldn't characterize that way." The ordering of society is a complicated topic. The functioning economy is a complicated topic.
--:-- The optimistic view would be humanity adapting to the global village by taking on a more humble attitude. We're not going to have a lot of objective truth-tellers running around. But we also don't want to be in a complete panic about everything all the time. We need to take a deep breath, touch grass, be a little bit more skeptical, be a little bit more open, be a little bit more understanding.
--:-- I think that's happening. Jake Tapper's book and the reaction to it—if you watch the interviews on YouTube and the crowd response—it feels like people are saying, "If we just take a step back from the intense partisanship, maybe we can get back a little bit more." That book is a very positive step forward towards a calmer approach. The other book I'd promote is Ezra Klein's book on abundance. Somebody who's supported Democrats for a long time, it's the most positive manifesto saying, "Whether you're on the right or the left, we need to actually build things." I think that's also a healthy moment.
--:-- Jack Altman: Related to this topic, a little bit adjacent, I saw you talking about preference falsification recently. I think this is a super interesting topic, particularly in the last five-ish years. A lot of preference falsification became made apparent. I'd be curious to hear what you think happened over the last some number of years where these changes happened.
--:-- Marc Andreessen: Preference falsification has two different elements. It's when people are required to say something in public that they don't actually believe, or they are prohibited from saying something in public that they do believe. Commission and omission issues.
--:-- There's a great book by Timur Kuran on it. The theory is it's easy to think about this for a single person—are you telling the truth? The thing that gets complicated is when that happens across a group or a society. If there's widespread preference falsification, you not only have people lying about what they think or hiding it, but everybody loses the ability to know what the distribution of views actually is.
--:-- If you look at the history of political revolutions, a revolution happens when a majority of the country realizes that a majority of the country actually agrees with them, and they didn't realize it. Whatever system they were in convinced them they were in a very small minority. Then there's a catalytic moment, a preference cascade, and everybody realizes it at once.
--:-- You can see this in a crowd with a controversial speaker. There's silence, then one brave person starts clapping. That person is at severe peril—if they're the only one, they might get killed. But if it cascades, a second person starts clapping, then a third, a fourth, a fifth. You get a snowballing effect. The entire auditorium is clapping. That's everybody realizing they're actually on the side of the majority, which they didn't realize before.
--:-- By the way, this is why comedy is fun. Comedy does this well because people can't control the involuntary response. When you get an entire group of people in a room laughing out loud at something they will individually swear isn't funny—they can't help it. The stress relief from that, because they all know they're part of it, they've rebonded the community. It's an incredibly powerful feeling.
--:-- It's very easy to apply this theory to the Soviet Union.
--:-- Marc Andreessen: Or like Eastern Europe in the Cold War, or Maoist China. It's a lot trickier to apply this theory to your current society. I believe we've lived in an era of intense preference falsification. I think the last five years, probably the last ten years, were way more intense than the preceding forty at least. Probably going back to the 60s, if not the 1920s or something, to find an analogous period.
--:-- Jack Altman: I think this period was characterized both by people saying things they didn't believe, but critically, not saying things they did believe. There are many reasons this happened. A lot of people want to say this is caused by social media. But when you phrase it the way you did, it actually makes a lot of sense as a prisoner's dilemma matrix. It doesn't really need an outside catalyst for people to get into their own box.
--:-- Marc Andreessen: That's true. Although there needs to be some kind of motivation for the cascade to have started where people end up in that box.
--:-- Jack Altman: It's social pressure. So specifically, I think the thing that happened in the last five years was... I guess it needs to be a high-stakes enough issue for it to matter. Otherwise, it's just like, who cares whether you think the clouds are pretty or not?
--:-- Marc Andreessen: Yeah, that's right. The way I think Timur Kuran would describe it is it needs to have political, social, cultural salience. It needs to get to something fundamental about how the community is organized. We call that politics, but it predates even the concept of politics. And by the way, you don't necessarily want to say all preference falsification is bad. I don't know that you want everybody telling the truth about everything. A lot of social graces come from people saying "It's great to meet you" when they didn't feel like it. "Your baby is very beautiful." Exactly. So some of it's good.
--:-- Jack Altman: So you get wedged in this box. The good news is preference falsification in a lot of totalitarian societies was administered at the point of a gun. You say the wrong thing, they shoot you.
--:-- Marc Andreessen: Yes.
--:-- Jack Altman: That for the most part is not what happens in our society. What happens is the nonviolent version: ostracized, canceled, reputation ruined, fired, unhirable, lose all your friends, lose all your family, can't ever work again. Still really bad.
--:-- Marc Andreessen: Still really bad. So you said it sounds pretty bad. Very bad. And it just turned out that part of adapting to the existence of social media was that social media became a very effective channel to destroy people reputationally. The social media mobbing effect we're all familiar with.
--:-- Jack Altman: And you think that helped create more false preferences?
--:-- Marc Andreessen: Big time. Big time.
--:-- Jack Altman: Do you think it also unwound them?
--:-- Marc Andreessen: Well, this is the thing. This is maybe what happened in the 2024 election. It was just like, "Oh, okay, we don't have to live this way anymore." Certain views become safer to say out loud. Also the censorship regime. Even in tech, for the 2024 election versus 2016, regardless of what you think and who you wanted, at least everybody can agree that it was taboo to support Trump in 2016, and it was not taboo to support Trump in 2024 in tech. Something changed there.
--:-- Peter Thiel had this great line in 2016. He was one of the only people in tech who was actually pro-Trump. He said, "This is the least controversial contrarian thing I've ever done. Half the country agrees with me. I've never had a point of view on anything else in my entire life where half the country agrees with me. And yet somehow this is such a heresy that I'm like the only one."
--:-- Jack Altman: That definitely changed. Optimistically, you could say there's a process of adaptation. We all just decide we're not going to live life by mobbing and scapegoating and personal destruction. Just because somebody is offended by something doesn't mean it's going to destroy their lives.
--:-- Marc Andreessen: Do you think it's basically been unwound now, or do you think there are still a lot of falsified preferences?
--:-- Jack Altman: I would say it's radically different than it was two years ago. I would say there's still a lot of falsified preferences. But in any healthy society, there's lots of falsified preferences.
--:-- Marc Andreessen: Do you have any guesses for something that is currently falsified that will become unfalsified?
--:-- Jack Altman: Sure. Yeah. But it's far too dangerous to say. We'll move on.
--:-- Marc Andreessen: Dang. Gosh. When you ask that, it's a very key question. Here's what I encourage. Break the fourth wall. Here's the thought experiment: just write down two lists in the middle of the night with nobody around, doors locked. Write it down on a piece of paper.
--:-- Jack Altman: And pull it out in ten years?
--:-- Marc Andreessen: Write down two lists. What are the things I believe that I can't say? And what are the things I don't believe that I must say? Just write them down. I bet if you're a reasonably introspective person, most of us probably have ten, twenty, thirty things on both sides of that ledger. Most of those are things where you don't want anybody to ever see that piece of paper. Maybe five or ten years from now, we'll be back and everybody can reopen their papers and it'll be safe to say whatever people wrote down.
--:-- Jack Altman: A few final topics. You're probably in a spot to give life or career advice to young people a lot now, both in general and specifically with AI and the current set of tech changes. What do you most often find yourself repeating to a really smart recent grad about what they should be doing with their career?
--:-- Marc Andreessen: To start with, I never took any advice.
--:-- Jack Altman: There's something there.
--:-- Marc Andreessen: Maybe the best people probably shouldn't take any advice.
--:-- Jack Altman: Okay, but for the rest of us.
--:-- Marc Andreessen: I would say, especially for young people, people are very different. Some people are very happy being in the middle of chaos. Some people are very unhappy and will get themselves out of a chaotic situation as fast as they can. Other people love chaos so much that if they don't have any, they will create it. Not everybody should be in a high-growth, high-risk tech company because it might just be too nuts. I don't think there's a one-size-fits-all thing.
--:-- Having said that, for the young person who wants to be in tech, run to the heat. Where are the interesting things happening? That's a conceptual question, and it's also a place question, a community question, a network question. Run to that as fast as you can. It doesn't mean running to the fads, but trying to get into those hot networks, ideas, or projects.
--:-- There's a geographic component to that, and I think we all wish it wasn't the case, but it really is. AI has very successfully unwound the geographic dispersion of what was happening in tech. It's kind of slammed everything back into Northern California. I don't think that's good for a lot of reasons, but I think it just is the case. If you're going to do AI, get here.
--:-- The other thing is the Steve Martin thing: be so good they can't ignore you. Time spent on the margin getting better at what you do is almost certainly better than most other uses of time. The old adage that you are the average of the five people you spend the most time with is also true. Pick that carefully.
--:-- When I talk to people about what kind of company to go to, there are certain people who should only be in a raw startup, and certain people who should only be at a big company. The general advice is the high-growth companies between Series C and Series E. They've hit product-market fit, they've hit the knee in the curve, and they're on the way up. On average, that's the best place to go because you won't have the downside risk of a complete wipeout. If you're talented, you can pick up new responsibility very quickly.
--:-- Jack Altman: Next is your Andrew Huberman thing that I see on Twitter. I actually can't completely parse what it is. What's going on with that?
--:-- Marc Andreessen: We have a completely fake beef. We're good friends. We're neighbors in Malibu, I've been on his podcast, we're very good friends.
--:-- Jack Altman: But you don't follow his protocol?
--:-- Marc Andreessen: I don't do a single thing he says, with one exception we'll talk about. I don't do any of it. He says maintain a regular sleep schedule. I'm all over the place. He says always get up and see sunlight. That's the last thing I want to do when I wake up. You don't drink caffeine for the first two hours of the day. It sounds like being in a North Korean concentration camp. I can't even imagine.
--:-- Jack Altman: You drink a lot of coffee?
--:-- Marc Andreessen: A lot of coffee. Cold plunge? I'm not doing any of that shit.
--:-- Jack Altman: You think it's good for you though?
--:-- Marc Andreessen: Oh, I'm sure it's good for you. I'm just not going to do any of it. It all sounds completely miserable.
--:-- Jack Altman: That's good.
--:-- Marc Andreessen: The one thing he says that I do is stop drinking alcohol. I am physically much better off as a result. But I'm very bitter and resentful.
--:-- Jack Altman: Towards him specifically?
--:-- Marc Andreessen: Yes.
--:-- Jack Altman: Why'd you do that one?
--:-- Marc Andreessen: Because it's much better for you physically. It fixes sleep and energy problems. It's the most tolerable of all of them.
--:-- Jack Altman: So you do one?
--:-- Marc Andreessen: No, no, it's completely intolerable. It's horrible. I don't recommend it. I think it's a horrible way to live. I'd much rather be drinking alcohol.
--:-- Jack Altman: Does he think even a glass of wine at night is bad?
--:-- Marc Andreessen: He does, yeah. All of it. He's actually had a big influence on the culture, and in seriousness, this is very positive, at least for health. There were all these fake alcohol studies. Red wine is heart protective and all this stuff. It turned out that really sick people either drink a lot or nothing, and healthy people tend to drink a little. One is healthy people tend to be very disciplined.
--:-- Jack Altman: Is that correlation or causation?
--:-- Marc Andreessen: It's all in the sample set. There are no health benefits to alcohol. That was all completely fake. Michael Crichton called this "wet streets cause rain." Unhealthy people stop drinking because they're in the hospital. Their doctor says, "If you keep drinking, you're going to die." Or by the way, they drink a lot. Healthy people tend to be very disciplined. People who are disciplined to drink moderate amounts of alcohol also do moderate amounts of exercise, experience moderate amounts of stress, go to the doctor regularly, take their medication.
--:-- Jack Altman: It feels like it should be a good thing that Andrew and other people have gotten so many more people interested in health.
--:-- Marc Andreessen: It's good physically. It's catastrophic emotionally. It's made me a much less happy person.
--:-- Jack Altman: Are you actually serious about that?
--:-- Marc Andreessen: Alcohol is something people have used for thousands of years to fundamentally relax. It's a very important social lubricant. The de-stressing could be healthy. The birth rate is crashing at the same time we all stop drinking. I don't think Andrew would argue you should live your life purely maximizing for physical health. That would be a miserable way to live. What are you going to do? Never leave the house? Never take a risk crossing the street?
--:-- He certainly doesn't judge people for drinking moderate amounts of alcohol. He just says, scientifically, you have to understand it is a poison. Having said that, speaking of scenes, the displacement thing happening is people in our world aren't drinking alcohol. Instead, they're doing hallucinogens.
--:-- Jack Altman: Why are you saying that?
--:-- Marc Andreessen: It's not necessarily an improvement. As you, Jack, know very well.
--:-- Jack Altman: Yes, tell us about your latest ayahuasca trial. You're so much different than you were last time I saw you. Your personality has clearly completely changed.
--:-- Marc Andreessen: I do feel different. The other theory is there's a law of conservation of drug use. Every society is going to pick some drug and abuse it. Apparently in our case, it's going to be LSD and mushrooms.
--:-- Jack Altman: It's a good one. Okay, my last question. When I tweeted out a request for questions, I almost got ratioed by one question. I'm going to ask this one nearly verbatim. It was by an anon named Signal. If you were frozen for 100 years and you woke back up and looked around, what would be the piece of data you'd want to know that would tell you whether your dominant worldview turned out to be correct in the fullness of time?
--:-- Marc Andreessen: I will pick a very unfashionable answer. I would say United States GDP. Just straight up U.S. GDP. Embedded in that is the question of technological progress. If you have rapid technological progress, you'll have rapid productivity growth, which means very rapid GDP growth. If you don't, you won't. You'll see that in the GDP numbers immediately.
--:-- Number two is that our market is a great way to organize, and the U.S. is the best market. Is that going to keep working? And third, is the U.S. going to be a great country?
--:-- Jack Altman: And you are long all of this?
--:-- Marc Andreessen: I am very long all three of those. I am very convicted on all three of those. If I'm wrong about something big, it's going to be something in there, and it will show up in that number.
--:-- Jack Altman: Marc, this is amazing. Thank you so much again.
--:-- Marc Andreessen: Good. Awesome. Thank you, Jack.
--:-- Narrator: Thanks for listening to the a16z podcast. If you enjoyed the episode, let us know by leaving a review at ratethispodcast.com slash a16z. We've got more great conversations coming your way. See you next time. Bye.
