Marc Andreessen：世界比你想象的更具可塑性

I wasn't expecting to start here. I want to talk about why you were consuming so much caffeine that you noticed your heart was skipping a beat.

I love caffeine. For a very long time, I always said the ultimate day—the perfect day—was 12 hours of caffeine followed by 4 hours of alcohol. That's just the ultimate. I did cut out—or at least for now, I've cut out—the 4 hours of alcohol.

Caffeine is one of nature's most marvelous things. It turns out you can't overdo it. A while ago, I was drinking so much coffee at work that I was sitting in a meeting a couple of years ago, and I started to feel a little bit off. Something felt off, and I took my pulse and realized I was skipping about every 10th heartbeat.

I had an existential crisis because I'm like, "All right, heart, do I need to call 911? Am I about to have a heart attack? Am I about to die?" I went under the table and Googled, "Is this a problem?" Dr. Google said, "No, it's okay. It's fine. You just might want to cut back a little bit on the caffeine."

You said something that I love, and I never hear other entrepreneurs talk about, but I think it's super important: You don't have any level of introspection.

Yes. Zero. As little as possible.

Why?

Move forward. Go. I've just found that people who dwell in the past get stuck in the past. It's a real problem, and it's a problem at work and at home.

I've read, obviously, 410 biographies of history's greatest entrepreneurs.

Yeah.

That was one of the most surprising things. When I ask what's the most surprising thing they've learned from this, they're like, "Oh, they have little or zero introspection."

Yeah.

Sam Walton didn't wake up thinking about his internal self. He just woke up and said, "I like building Walmart. I'm going to keep building Walmart. I'm going to make more Walmarts," and just kept doing it over and over again.

You probably know that if you go back before 100 years ago, it never would have occurred to anybody to be introspective. The whole idea of modern conceptions around introspection and therapy, and all the things that result from that, were manufactured in the 1910s and 1920s.

Say more about that.

Great men of history didn't sit around doing this stuff at any prior point, right? It's all a new construct. First, Western civilization had to invent the concept of the individual, which was a new concept several hundred years ago.

Then, for a long time, it was, all right, the individual runs and does all these things: builds things, builds empires, builds companies, builds technology, and does all these things. Then this guilt-based whammy showed up from Europe, a lot of it from Vienna, in the 1910s and 1920s—Freud and that entire movement.

It turned all that inward and basically said, "Okay, now we need to second-guess the individual. We need to criticize the individual. The individual needs to self-criticize." The individual needs to feel guilt, look backward, and dwell on the past. It never resonated with me.

Do you find that a lot of the greatest founders you've spent time with, backed, and partnered with have low introspection?

Generally. Although, in fairness, introspection is probably linked to the personality trait of neuroticism. A lot of the best founders are, I think, at 0% neuroticism. They just don't get emotionally fazed by things that happen, which is a superpower when you're an entrepreneur.

Having said that, some of the great entrepreneurs are, in fact, very neurotic. That's also the case. Maybe it's nice to have low neuroticism, but it's not necessary. There are some who get wrapped around the axle on personal issues.

As you know, these days, that sometimes turns into the use of psychedelics—different kinds of hallucinogenic drugs. That's one very interesting trajectory for the culture of the country and the culture of the world, and we'll see where that goes.

We've recorded, I don't know, around a dozen of these so far, most of them with some of the greatest living founders for the show. I can't believe how many times psychedelics have come up on almost every episode.

Yeah.

They're like, "You should try them." I'm like, "I'm not doing any drugs."

Just to be clear, I'm not. I never have, and I'm never going to. The problem is that I already have tons of horror stories from people I know, or know of, who came out the other side like...

Actually, my deepest conversation on this was with Huberman. I was describing this phenomenon we see in Silicon Valley, where these guys get under pressure and feel anxious or whatever, and somebody tells them about psychedelics, so they try them.

They come out the other end as changed people and much more at peace, but then they also tend to quit their companies. They move to Indonesia and become surf instructors. They're just like, "Peace out," and they're done.

There have been a whole bunch of examples of this, and I was complaining to Huberman about it. In true Huberman, wise-Yoda style, he's like, "Well, how do you know they're not happier?" Maybe that was the positive outcome.

Maybe the thing that was driving them to be great entrepreneurs was a fundamental level of insecurity—a dissatisfied, neurotic impulse. Now they're just satisfied. Their serotonin levels, or whatever, have been recalibrated, and they're satisfied sitting on the beach and being a surf instructor. Maybe they're better off.

I'm like, "Yeah, but their company is failing." So, anyway, there's a possibility that there's a better version of you or me on the other side of ayahuasca, but I'm not willing to find out.

I'm not either. Daniel Ek has the greatest way to put this. He thinks the best entrepreneurs are not optimizing for happiness; they're optimizing for impact.

I think that's true. I think it's certainly true for Daniel, who's a great case study of that.

Having said that, I always wonder about intrinsic versus extrinsic motivations. Impact strikes me as a little bit of an extrinsic motivation. It's impact, money, fame, and so on. By the way, I think extrinsic motivations are fantastic, and I think they can be very motivating. The people who get great rewards for building great things deserve them.

At least what I've found is that intrinsic motivations are what actually get people up in the morning. That's where you're dangerously close to straying into introspection, but it's about what causes somebody who's now extremely materially wealthy and extremely successful to get up in the morning and continue to punch away at the world. I think those motivations tend to be interior.

What's that for you?

That would require introspection. I'll let other people speculate.

No, you have to have—

It's a lot more fun to speculate about other people's motivations.

But I am curious about you because you have a series of quotes that I absolutely love. I save them on my phone and reread them from time to time. One of them—and I'll butcher it—is, "The world is way more malleable than you think."

Mm-hmm.

"If you pursue something with maximum effort, drive, and energy, the world will recalibrate around you more easily than you think."

Yeah.

I reread that this morning before I came over here, and I thought, "What is that for Marc? Today, what are you waking up trying to change in the world?"

There's a lot that we're actually trying to do. I'm suspicious that that's my actual underlying motivation.

Why?

I don't think external impact is enough to keep people going. At least, I've seen way too many people who had a high level of external impact and then, at some point, just stopped.

Okay.

Here's the problem with external impact: It's 4:00 in the morning, you're staring at the ceiling—is that enough? External impact is stuff that's happening to other people.

Yeah.

What is it about you? The story I like to tell myself is that I'm competing with myself. I'm getting up in the morning because I'm trying to become a better version of myself. I'm trying to become smarter and better informed, reach better conclusions, be better at what I do, and continue to expand my skills.

But again, to actually analyze that properly would require a level of therapy that I'm not willing to engage in. So, anyway, the much more comfortable conversation is, "What are you trying to do in the world?"—which I would love to talk about.

I have almost no introspection either, so I understand that.

Right. Yes.

All right, so tell me what you're trying to do in the world, then.

We have this fundamental view. It's actually fairly amazing that it's become controversial, but we believe that technology is, on balance, an enormously powerful force in the world. Basically, the big problem with the world is that there's not enough technology, not enough information, and not enough intelligence.

We have this opportunity. We have these special sets of technologies that let us fundamentally improve things. Then there's this very special personality type of the entrepreneur who's able to build the product, build the company, build a phenomenon, and really make an impact on things.

And so when I look at the world, I'm like, okay. The world we live in is a very primitive and crude place compared to what it should be and what it could be. And so the whole thing that we've been trying to do for 17 years at our firm is build the ideal partner to the founders who are trying to do that, based on our own experiences of having been founders who were trying to do that.

Overall, the world, especially the Western world, is stagnant. The overall theme of things—

Mm-hmm.

Everything is stagnant, and we could talk a lot about that. But every once in a while, you have somebody who comes along, and it's like, all right, no, I actually have an idea of how to make things fundamentally better, and I have a way to build a business around that, build a company, build an empire around that. And those people include ourselves in this, but those of us who are trying to do that are basically a rough movement against stagnation. Without us, there's nothing but stagnation.

Mm-hmm.

But it's actually really funny. There's always this criticism that you get from, as I say, the corporate press or outside critics, which is like, “Oh, you VCs are funding the wrong things,” or, “You entrepreneurs are building the wrong things.” It's like, well, nobody licensed us to do any of this. We didn't apply for a permit to get judged by somebody ahead of time and told, “Yes, you get to do this; you don't get to do this.”

Many people could be trying to do this. Anybody can do this. Anybody can start, build a product, start a company, or even try to be a VC. These are all completely open fields. And it's shocking to me how few people actually give it a shot. The fate of the world over the next 1500 years is riding on the people who actually want to give it a shot.

I want to tell you about the presenting sponsor of this podcast, Ramp. I have been reading a lot about SpaceX lately. SpaceX is one of the most valuable private businesses in the world, and one of the main themes in the history of SpaceX is constantly attacking and questioning your cost. Ramp helps many of the most innovative businesses in the world do exactly that. The median company running on Ramp cuts their expenses by 5%. And one thing SpaceX has demonstrated is that a religious dedication to controlling costs can help actually increase revenue, because you can pursue opportunities you couldn't otherwise. And we see that in the Ramp data too. The median company running on Ramp also grows their revenue by 16%. So, when you're running your business on Ramp and your competitors are not, you have a massive competitive advantage that compounds over time. Ramp is the only platform designed to make your finance team faster and happier. Many of the top founders and CEOs I know run their business on Ramp. I run my business on Ramp, and you should too. Go to ramp.com to learn how they can help your business save time, save money, and grow revenue. That is ramp.com.

So, when you started the firm 17 years ago, was your thesis exactly the same as it is today?

I'd say the core thesis is the same. The specifics have varied; they have changed enormously. We can talk about both parts of that. But, yeah, no, the core thesis was that the startup, the entrepreneur, the founder is going to be the core engine of progress in the world, and I think that's more true than ever.

In fact, when we started, it was still controversial: the idea that a founder would run their own company.

Even in 2008, 2009?

Yeah. Well, it was still very controversial. In fact, there were high-profile companies at the time that were getting heavily criticized for having these little kids running these companies.

Okay, so you have this encyclopedic knowledge of the history of Silicon Valley in your head. I probably read 30 to 40 books on it, so I have some level of knowledge, but not the level that you do. I remember reading a book on Nolan Bushnell, founder of Atari, who was 27 at the time, and it was exceedingly rare. It talks about that. In his story, it was exceedingly rare for him not to be replaced by an older CEO once Atari started growing.

Yeah.

Were there other examples before him?

Well, so Christopher Columbus. Alexander the Great.

Yeah. Right? So, throughout history—

Yeah. Thomas Jefferson. Throughout history, most of the great things that have been built have been built by this super-charismatic founder type, this will-to-power founder type, who basically built and ran something.

Okay, hold on.

Henry Ford.

Hold on. I love that you went here because you don't remember this—

Yeah.

—but we had dinner in Miami—

Mm-hmm.

—with Jared Kushner a year ago or something, and you and I would wrestle because I was so excited to talk to you. I was trying to get out of you, because I think about how history's greatest entrepreneurs all day. This is what I do seven days a week. Who are these entrepreneurs from history that you like? You just started naming country founders.

Yes, true. Exactly. There's this recency bias, right? The world that we live in today is the normal state of the world, and everything that happened in the past is weird and different, and those people were dumber than we are and all screwed up. And it's like, well, maybe.

Or maybe the world worked a certain way for thousands of years, and we're in the weird time. Maybe we're in a time that's just really unusual from a historical standpoint. I think this is one of those dimensions in which that is true. It never would have occurred to anybody 100, 200, or 300 years ago that if somebody was going to start something, they were going to be the person who ran it. It was obviously the case. It was just obviously the case.

The book that I always recommend on this topic is called The Machiavellians, which is this famous book from the 1940s by James Burnham, who's one of the great geniuses of the 20th century. The way he describes it is basically this: He says, “Look, there have been 2 fundamental modes of business organization over the course of the history of capitalism.”

There's what he calls bourgeois capitalism, which is basically the founder runs the company, name on the door. The classic archetype for bourgeois capitalism was Henry Ford in the 1920s, and today it's Elon Musk, right? It's like, that's you. And by the way, in the old days, it was Ford Motor Company; it's not Musk Motor Company. But everybody knows Tesla and SpaceX—these are Elon.

Again, that maps to this historical thing: that's also how countries ran, and that's also how cities ran, and all these things. Religions, by the way. Basically everything: founders led the way. That's the historical norm.

And then what he basically says in this book is that there's this new model that is an artifact of this weird period of time between the 1880s and 1920s, where the modern world as we know it today formed. He said there's a new philosophy of leadership and management, which is called managerialism—the rise of the concept of a manager.

Specifically, a manager as contrasted with a leader. Therefore, the manager. Therefore, the idea of a management school. Right? Therefore, Harvard and Stanford business schools. Therefore, the idea of the manager who replaces the founder running a company. Therefore, the idea of management as a skill set that can be used to run many different kinds of businesses.

In the 1970s, this then turned into the conglomerate, which was the idea that it doesn't matter what the company does. If you have a good manager, the company should do 30 different things. Managerialism is this idea that you have this interchangeable management skill, and that can basically run anything.

Actually, what Burnham says is, “Look, people are going to try to draw a value judgment on this, and they're going to try to say this is better or worse than the old name-on-the-door model.” But he said, “The reality of the modern world is everything is big.”

Mm-hmm.

For the electrical power grid to get big, or the road network to get big, or the car industry to get big, large-scale systems need to be run by people who are trained in how to run large-scale systems. And the same thing with countries. Large-scale countries are going to need to be run by people who are good at running large-scale things, right?

The founding personality type is not the manager personality type. Those are different. And so there's going to be a handoff when things get big and complicated. That's the model that Nolan Bushnell talks about, and that's the model that dominated Silicon Valley for 50 years.

The problem with his argument is that it assumes the managers are going to do a good job.

Mm-hmm.

Right? And I think if there's one dominant theme that we're seeing in the last 30 years—in the West, for sure—is that managers, generally, writ large, are not doing a great job. Or another way to put it is that managers may be good at managing something that's going to be status quo for a long time.

If it doesn't change, maybe they can run the banks for a long time, or they can run the power company for a long time, or the car company. As long as the car is the car is the car, or soup is soup is soup, it kind of doesn't matter. But the minute things change, the manager personality type, because it's not the founder personality type, doesn't know how to deal with change.

Not everything is changing. A lot of things aren't changing, but for the things that are changing, they're changing really, really quickly. I mean, SpaceX is the classic example of this.

Imagine being a professionally trained manager, trained at a top management school, working for a rocket launch company competing with SpaceX. The assumption of the entire rocket industry for the last 100 years had been that rockets were used once, and then that was it. The economics of launch were dominated by having to build a new rocket every time.

Then this crazy guy in California comes up with this thing where the rockets land on their butt, and you can’t replicate it. What good are your management skills at that point? I think there are a whole bunch of interesting areas of human activity where that shift is happening. This is where Burnham’s thesis collapses: the managers actually can’t do it.

Yes, there’s a need to run things at scale, but no, the managers actually can’t do it because they can’t adapt.

The founder can just learn how to run things at scale.

Well, that’s the theory, and that’s a big part of our theory: the founders can actually learn how to do this. And look, this is still a controversial topic. This still comes up because founders—especially founders on day 1—aren’t necessarily good at doing this.

In tech, let’s talk about tech specifically. The founder tends to have been in a lab, literally or metaphorically, for 20 years before they start their company. They’ve probably been working by themselves with a small team. They’ve been building technology. They haven’t been running things, managing large organizations, or running public companies.

There is a missing skill set. On day 1, they don’t know how to do that. They do need to be willing to learn how to do that, and they need to be capable of doing it, because some of them can and some of them can’t.

So this may be the core thesis behind our firm: you’re much more likely to build something important in the 21st century if you start with the founder and train them on management than if you start with the manager and try to train them to be a founder, to create new things.

I think this trend is intensifying. All the old edifices, all the old incumbent institutions of the last 100 years that are run by managers, are in some state of fundamental collapse. They’re all collapsing in trust and credibility because they can’t adapt. This issue is becoming more and more acute: the system that we thought was necessary and sufficient actually just does not work.

If anything good is going to happen, it’s going to have to be somebody—it’s going to have to be a Henry Ford or Elon Musk type—who actually does it.

You think it's in a vast minority of people agree with you?

Look, it’s becoming more common. When you get an Elon Musk and a Steve Jobs, when you get these archetypal examples of it, it’s a lot easier to sell it. Mark Zuckerberg, who we were talking about earlier, is now a great case study of this.

When Mark started Facebook, he had never had a job before.

Yeah.

Not only had he not managed people, he had not worked for anybody. He started with zero, and his learning curve—which happened fully in the public eye—was vertical. By the way, it’s still vertical. He spends an enormous amount of time learning how to become better at running these things at a large scale.

He’s still the founder, and he’s still the innovator, and he’s still a fountain of ideas about what to do. He’s the classic example of the double threat. What happens is that other founders look at that and they’re like, “Oh, I could do that.”

Which is exactly what Steve Jobs said when he saw Nolan Bushnell: “I can run my company.”

Exactly.

“I can do that.”

“I can do that.” Yeah, exactly.

It’s amazing how fast this stuff shifted, because Steve famously had this short period of time when he worked for Hewlett-Packard. I don’t know if it’s true, but the legend is that Jobs pitched his manager at Hewlett-Packard—

No, Wozniak pitched him.

Was it Wozniak? Okay, all right.

Yeah, Wozniak.

Okay, Wozniak pitched him.

Yeah. There was some other story where Jobs went into a meeting with some manager trying to pitch the thing, and the line from the manager was, “Absolutely not. This is the dumbest idea I’ve ever heard. Get your feet off my desk and get out of here.” You can just imagine Steve with his—

And they had to be barefoot at that time.

Oh, probably.

My favorite Apple lore is that the first sale in Apple’s history was made barefoot. When he walked into the Byte Shop, he was barefoot.

What’s amazing about that is that Wozniak for sure worked for Hewlett-Packard. Everything I’m describing was Hewlett-Packard in the 1940s, 1950s, and 1960s. Dave Packard and Bill Hewlett were also that founder type.

Yeah. Dave Packard and Bill Hewlett ran their company, between the 2 of them, for 50 years.

By the way, Silicon Valley was built, in large part, on HP. HP was the original Silicon Valley company.

Okay, that’s the next question.

It was run by its founders for 50 years, and yet people concluded that founders shouldn’t run companies. It’s one of those things where it’s so obvious—it was staring everybody in the face—and people had to construct these elaborate lattices of theories to get around the fundamental fact that you need somebody who knows what to do actually running the thing.

Do you think HP might have been the most influential company in Silicon Valley history?

It was, for sure, the most influential company from 1940 to 1980, and then probably, after that, Intel.

Well, go to the founders of Intel, and you read biographies of them, and they talk about modeling themselves off HP.

Yeah.

And then how many founders modeled themselves off Bob Noyce and Intel after the fact, including Steve Jobs, who would go to Bob Noyce’s house for dinner?

Yeah. That’s right. That’s right.

By the way, that’s another great example, because Bob Noyce, at least—if you look at photos of Bob Noyce, you’re like, “Wow, this guy’s a pillar of society.” He’s very well-dressed and very adult, and he’s famously the leader of the Traitorous Eight.

You know, they’re the group that left Shockley to start Fairchild—

—and then left Fairchild to start Intel.

And then left Fairchild to start Intel.

Yeah, exactly.

Bob Noyce was 100% the Steve Jobs of his time, just in the short-sleeve white dress shirt and the skinny black tie. I unfortunately never met Bob Noyce, but I could easily imagine Bob Noyce and Steve Jobs sitting down and being able to talk for 3 hours, completely understanding each other despite the fact that their look and feel were completely unlike.

He was almost like a disciplinarian to Steve because Steve was wild and reckless: “I was also wild and reckless. You need to mature.” I think Bob’s wife may have gone to work at Apple early on, too. He talked about this in his biography. There are a few great biographies of Bob Noyce, but he said that the reason he spent so much time, after he was really successful, with young entrepreneurs was that it was “restocking the stream in which I fished from.”

Amazing.

He thought it was really important. He was like, “I learned from all the guys before me. I need to take that knowledge I’ve built up over multiple decades and push it down the generation.”

I want to go back to starting the firm, though. This is interesting. What was occurring in your life, either at that time or before that, that led you to the observation that this had to be done?

We’ve got all these elaborate theories. The practical reality of it was that my partner, Ben, and I had become very active angel investors. I’d been an angel investor since the mid-1990s, but then Ben and I started doing it as a real thing, putting significant time into it, probably starting in 2003 or 2004. It’s hard to remember now, but if you go back to 2003 or 2004, there weren’t thousands of angel investors. There were about 8. It was Ron Conway and a handful of people.

Ben and I were running around doing it. This was very significant in the evolution of the venture capital industry, because this was the point at which traditional VCs got disintermediated by angels and seed investors, who inserted themselves before the VCs arrived. That was a fundamental change that changed the whole industry, and we were part of that.

As a consequence, we were investing in all these new companies, basically at the point of formation. We were playing amateur early-stage VC. We were always like, “We’re not going on the board. You’re going to raise money from a real venture firm later. They’re going to go on your board and work with you.”

What we found, over and over and over again, was that we ended up getting pulled into these companies, either because there were issues the other people they were working with couldn’t help them with, because they hadn’t raised venture yet, or because the VCs they’d raised from couldn’t help them.

We just got pulled in. The reason was that we had been running companies for 20 years at that point. We at least had some idea of what we were doing. The other reason was that we kept getting brought in to do conflict resolution between the founders and the VCs.

That was especially common at that time when the VC’s fundamental point of view was that the founder wasn’t going to run the company, and that they needed to replace the founder with a professional manager as fast as possible.

The founders are not necessarily going to like that, and they might resist it. And by the way, even if they're on board with that idea, they might not like the person the VC wants to bring in. And so, we kept ending up as arbitrators in this sort of situation.

In theory, we were trusted intermediaries because we knew the founders, we knew the VCs, and we could help bridge between them. But what happened was, after a while, we were spending 8 hours a day just doing this. And we're like, “All right.” It's weird: you're writing a $100,000 check, and you're spending all this time doing it, basically arbitrating with somebody who wrote a $10 million check.

And it's just like, all right, we should probably just write the $10 million check. So I always think that founders—one of my theories of great founders is that they tend to be able to operate at a strategic, conceptual level and at the practical level at the same time.

Please do.

Yeah. But underneath that was just the lived experience of what was actually happening on the ground. The big theory of the firm that we had at that time was linked to this idea of founders running the show, but it was also a structural observation of what was happening in the venture industry.

Basically, what we did was, in line with your philosophy, we went back and studied a lot of other businesses that had similarities to the venture business. We studied private equity, hedge funds, investment banks, law firms, management consulting firms, ad agencies, and accounting firms—basically anything where the product is fundamentally a relationship, knowledge work, and a relationship, as compared to something that gets manufactured.

What we observed was basically this. Hollywood talent agencies are actually the example we've probably talked publicly about the most, and that was a great case study.

Ovitz.

He was in this studio a few months ago.

Oh, the Ovitz story.

Fantastic.

Yeah.

And by the way, we make a point of crediting him. He gave us a lot of this theory, so a lot of this comes from him. Actually, I'll tell it through his experience.

When he started his agency in 1975—

In the 70s, I think, yeah.

In the 70s, in the mid-70s, it was actually very similar structurally to when we started a16z in 2009. The configuration of the industry at that point was basically a bunch of service firms, a bunch of talent agencies, none of which were at a very high scale. Each of them was a tribe of solo operators and lone wolves.

The concept in Hollywood was that you had an agent, and that was your guy. That agent knew whoever that agent knew and had whatever relationships that agent had. But the other agents at your agency were not available to you, and there was no collective benefit to the fact that you were at an agency that had not just your guy, but 100 other guys. There was no collective payoff to that.

They ran it that way for a very specific reason, which was this “eat what you kill” professional-services mentality, where everybody had to build their own book of business. But you end up just dealing with a guy as opposed to a firm. There's no firm, no collective thing.

That was basically the condition of venture capital in 2009. At this point, we knew all the VCs really well, we'd raised venture, and we'd worked with all these other companies that had raised venture. Basically, all of the legacy venture firms at that point were like that. They were all just tribes of lone wolves.

The thing that we knew that was not publicly known was that, generally speaking, inside the firms, they didn't even like each other.

Oh, I hear stories like this all the time.

Right. So there's Joe and Mary, who are partners at a venture firm, and you're working with Joe, and Mary has a key connection that you need access to. You ask Joe, “Can Mary introduce me to so-and-so?” What you don't know is that they're having a brutal fight, where they're trying to destroy each other because, fundamentally, it's economics. They're going for a greater slice of the profit pool, and so they're really going at it.

We saw example after example of a venture firm that was basically either of 2 things. One was melting down due to internal strife and conflict. The other was generational succession.

A lot of the dominant venture firms in 2009 had been around for 30 or 40 years, and they were now on their third generation of partners going to their fourth generation of partners. Again, it's the same thing: they had been founded by dynamos, and the people in the later generations were not like that.

So we basically said, “Oh.” This is where the Ovitz thing comes in. We said, “Look, that's not going to last.” Our theory was what we call “death of the middle,” or sometimes—the negative way to frame it is “death of the middle.” The positive way is the “barbell,” which is what's happened in all these other industries.

Basically, the industry gets stretched apart like taffy. What you get is this barbell thing, and on one side of the barbell, you get early-stage angel and seed investors who are really first money in, staying very light on their feet, writing a relatively small check, but being involved in companies extremely early on and taking a lot of risk.

On the other side, you get scaled platforms. You get large-scale enterprises that have a lot of throw weight, a lot of access, very big networks, and access to a lot of money.

The other comparison we always make is to retail shopping. There used to be department stores like Sears and J.C. Penney, where the brand promise was a pretty good selection of products at pretty good prices. Now, those are dead.

What you have instead are boutiques like the Gucci store or the Apple Store, and then you've got these super-scale e-commerce companies like Walmart and Amazon. We're at the point where there's just no reason to ever go to a department store because it's got less selection than Walmart and Amazon, but it doesn't have the quality tier and special experience of a Gucci or Apple Store.

But you had that thought in mind—

Yeah.

—when you started a16z?

Yeah, 100%. Exactly.

Wow, okay.

It was a conceptual leap for venture capital at the time, but the exact same thing had happened to private equity. The exact same thing had happened in hedge funds. The exact same thing had happened in investment banking.

And you knew that by what? Just reading about the issue?

Reading. Investment banks are the classic example. If you read about the original investment banks in the U.S. between 1880 and 1920, they were all like boutique venture capital firms in the 1970s and 1980s in the U.S. They were 20 guys.

These are more like merchant bankers.

Yeah. Well, the classic stories, which I love so much—J.P. Morgan is one of my favorite historical figures. J.P. Morgan was an example of that. The J.P. Morgan investment bank was basically very important, but it was a tiny little operation. It fit in a single office. It was probably 20 principals and some office staff or something. It was not large.

Actually, the hidden secret to J.P. Morgan was that he was the son of Junius Morgan.

Okay. When you were talking, I was like, “Wait.” It was shocking that you would say to pick him, because I actually found his father a more formidable individual than him.

Yeah, he was, which is almost always the case with any famous public figure: the father is almost always a more interesting story. There are a lot of examples of that.

However, Junius Morgan and J.P. Morgan filled a specific economic role that's gotten lost in history. The Junius Morgan Bank was in London, and the J.P. Morgan bank was in New York. What the Morgan family was doing was funneling money from the old, slow-growth economy of Europe into the new, high-growth—

Yes.

—economy of the U.S. But again, it was exactly your point: it was this little boutique family operation.

The other great thing about that era of history is that they were all bifurcated by religion.

Yeah.

They were the Protestant investment banks and the Jewish investment banks.

Yeah.

And they did not mix.

No, not at all.

Completely different worlds. As a consequence, the Protestant banks like J.P. Morgan were able to fund the railroads, which were considered the real businesses at the time.

Yeah.

But then all the disreputable stuff, like movie companies and department stores, those were all the Jewish investment banks. By the way, they had almost entirely Jewish founders.

JPMorgan Chase is the great survivor of that today on the J.P. Morgan side. On the Jewish side, Goldman Sachs is the great survivor. But again, if you go back, there were—

You consider that the barbell in investment banking? You have J.P. Morgan, kind of like a family partnership, and then you have the complete scale of Goldman Sachs.

And so what happened was, both J.P. Morgan and Goldman Sachs started out 100 years ago actually in the middle. They were, again, boutiques, but they were, for their time—today, you'd call them mid-market. We sometimes call it bulge bracket—as opposed to just a solo operator or something.

Actually, the way JFK’s father got started was that he literally hung out a shingle in the 1920s: “Joseph P. Kennedy, banker.” He was a private banker, and he just did deals; he was an angel investor at the time. Then you had the big commercial banks, but they had no interest in issuing loans to these speculative, crazy entrepreneurs.

At that time, J.P. Morgan, Goldman Sachs, Kuhn, Loeb, Drexel, and all these other mid-market banks—Morgan Stanley, the banks that became Morgan Stanley—were kind of these mid-market things. Now, sitting here 100 years later, those are the scaled players. The ones that didn’t scale are kind of long forgotten.

Having said that, there’s one firm that survives in the old model, and that’s Allen & Company. There are other boutique investment banks today, but Allen & Company was founded in the 1920s and is uniquely the one that survived in the original model of being a boutique and deliberately being a boutique investment bank. It stayed that way for 100 years.

One way to think about it is that today, that’s the barbell in banking: Allen & Company on one side, and then J.P. Morgan and Goldman Sachs on the other side.

I found one of my all-time favorite quotes when I was reading the book "Zero to One." The quote says, "The single most powerful pattern I have noticed is that successful people find value in unexpected places, and they do this by thinking about business from first principles instead of formulas." That is exactly what AppLovin has done with their new advertising platform, Axon. Axon is the most powerful advertising platform in a generation. Axon allows you to capture undivided attention. Axon ads are full-screen videos that are watched for an average of 35 seconds, retention that blows other ad platforms out of the water. And you can launch in minutes. You set the goal, and Axon achieves it. No complex setup, no expertise needed. And Axon scales quickly. They can put your ads in front of over a billion potential customers. Other businesses have seen immediate results, scaled to hundreds of thousands of dollars of spend per day, and increased their revenue by millions. And most advertisers aren't even thinking about this channel yet. Less than 1% of advertisers have access to Axon, so you want to get started quickly. And you can do that by going to axon.ai. That is axon.ai.

So, are you reading about this while you’re founding the firm?

Yeah.

Before you’re founding the firm?

Yeah, both. Ben and I spent about a year and a half planning the firm, and part of it was that he was in—we called it—industrial servitude. He was working for Hewlett-Packard after we sold our company to HP, so he was running a big part of HP at the time. We couldn’t literally start a new full-time thing until he got free of that. So, we had a year and a half to study, think, and work.

And because you had this period from 2003 or 2002, when you’re doing angel investing a lot—

Yeah.

—until you start your company six or seven years later.

Yeah.

You’re observing all of the weaknesses in the model, and that’s where you have, “Hey, why don’t we take the CAA?” I think Ovitz calls it the phalanx, where it’s like, if you have—

Yeah, right.

—one agent at CAA, you have all of us.

That’s right.

And they would roll deep. I think he says in his book, it was like, “Oh, my agent’s coming to the premiere.” No, it’s like 20 agents are coming. I think they’d be dressed in the same kind of suit maker, and they were intentionally trying to intimidate their competition.

Armani suits. Sulka Shirts was a shirt maker in Beverly Hills, and all sober colors, white shirts. Then I think he had a bulk-purchase deal with the local Jaguar dealer. The legend, at least, has it that the license plates all said CAA 1, CAA 2, CAA 3. So, you’d go to a premiere, and there would be 20 Jags lined up, and then 20 guys in identical suits coming out.

Yeah, this is exactly the thing. That’s the Hollywood version, but just imagine the psychological impact of that if you’re an old-school agent. Michael is a very dear friend. He became very controversial over the years, and I think the reason he became so controversial is just because he smoked his competition so severely. He pounded them so hard. There was no response.

You’re just a guy working for an old agency, and you’ve got your clients, and these 20 CAA motherfuckers are showing up. It’s a force. By the way, a lot of his clients are still active today from that period. If you talk to them, it’s just like, yeah, it’s a no-brainer. Do you want to work with a guy, or do you want to work with a firm? It’s obvious.

I don’t know if he told you all these stories. Did he tell you about his morning schedule thing?

The getting on the bike, doing the karate call?

No, for the firm.

No, no, no, no. I don’t remember this.

This is, again, something specific to Hollywood, but it’s a great example. At the time he started CAA, the agency business was like 90 years old or something, right?

Yeah.

It started out doing vaudeville bookings and music halls, and it had been around for decades. The people involved in it had had decades to think about the best way to do it, and they had arrived at a set of practices.

One of the practices—I think I’m getting this right—was that at every agency, they would have their staff meeting in the morning at 9:00 a.m., and whatever information was going to get shared in the agency would get shared at that point. “This studio head wants a script to do. He wants to do a crime thriller, and here’s the script,” and whatever. This was the point where whatever minimal handoff existed to the other agency would happen. This is where everybody would get updated.

The staff meeting would go from 9:00 a.m. to 10:00 a.m., and then at 10:00 a.m., they would start calling their clients. They’d be like, “We heard there’s a whatever. There’s going to be a casting call for this great new role for this professional thief or whatever—”

Mm-hmm.

“And you should consider doing that.” So, of course, Michael’s like, “All right. We’ll have our staff meeting at 7:00 a.m. We’ll be done at 8:00.”

Yeah.

“Between 8:00 and 9:00, we’ll call the clients. By the way, we won’t just call our clients. We’ll call their clients.”

Right? So, imagine you’re Paul Newman, and you’ve got some agent you’ve been working with for 20 years. Your agent calls you at 11:00 and is like, “I’ve got this great role.” You say, “Oh, the guys at CAA called me about that three hours ago.” Your agent’s like, “They don’t represent you.” Paul’s like, “Yeah, isn’t it great? Isn’t that fantastic?”

You just rinse and repeat that 1,000 times, and to the client, it’s completely obvious what to do. The reason I go through this—the moral of the story—is, again, this idea of incumbency and the status quo. In any business, you just end up with all these embedded assumptions. Then 90 years later, right? The founders of the agencies 90 years ago weren’t involved anymore, so the people who were running competing agencies were managers—same thing, managers, not founders.

The thing a manager never does, unless they’re under duress, is reconsider fundamental assumptions. They hate that. The whole point of running something big is that you don’t have to do that. You get to run the big thing at scale. You don’t have to go in and reinvent it from scratch. That sounds like a nightmare, right?

But as a consequence of that, you end up with all these embedded assumptions that are basically unspoken. Nobody’s questioning them. It’s not happening. If you take the time, you can go in from first principles and say, “Okay, well, how did they arrive at that?”

What we’ve found in industry—this is what our founders do every day—is that, in industry after industry after industry, there are all these embedded assumptions that made sense in 1970, 1930, or 1880 that just don’t make sense anymore.

I love that you did that. I always say it’s not what you do, it’s how you do it. The idea that you could take—I’m not running a talent agency—

Right.

—but there are so many of these principles that I could apply to venture capital. In your blog archive, which I absolutely love and have read multiple times—I did episodes on it—you would give advice to young people. My advice is: go work in an industry where the founders of that industry are still working.

When I read Ovitz’s book, the way I would summarize his approach, because he was in this big, stodgy, slow-moving, very bureaucratic organization, is: mediocrity is always invisible until passion shows up and exposes it.

Oh, interesting. Yes, right.

And that’s what he did.

Yeah, that’s right. He’s just like, “There are so many things that you guys could be doing better here.”

And if I remember correctly, he took some of these ideas to his boss.

Oh, yeah, yeah, yeah.

Because that guy was his mentor. I can’t remember his name.

He famously worked for the CEO of William Morris.

Yeah.

Yeah, which was the biggest of the talent agencies at the time.

So, were you and Ben essentially just designing what you wished you had when you were founders?

Yeah, that’s right. Again, that may be a cheat code, but if you’ve been the customer, obviously all of this becomes a lot more obvious.

I don’t know if you want to answer this question or not, but in Warren Buffett’s shareholder letters, he has this great line about how it’s really important to play against weak competition.

Mm-hmm.

Did you feel, at that point in venture capital history, that you were going to be playing against weak or weaker competition?

I would say not exactly. We didn't view them as weak. We viewed them as running on a status quo set of ideas.

To be clear, part of why we think about it this way is that we had raised money from what were, at the time, probably the 2 best venture firms. Kleiner Perkins in the 1990s—I worked with John Doerr very closely for 5 years at Netscape and then at Loudcloud. We raised money from Benchmark when they were king of the hill, and Andy Rachleff, who was 1 of the founders of the firm and is a legendary, brilliant VC.

So, through an accident of history, we had worked with 2 of the top 5 people in the field for a long time. They were and are, by the way, brilliant at running on the model that existed. John was brilliant at that, and Andy's brilliant at that. They're still brilliant today.

It was less a competition of, "These people are soft," or "These people aren't smart." It was none of that. It was, "No, they're really good at executing against this particular playbook." And that's why it's okay. If we're going to do this, we need to be playing by a different playbook.

There was no such thing as scaled venture capital at the time.

No. At the time, no. The firms all fundamentally hit this limit. The idea of a partnership of equals, or even a hierarchical partnership, just breaks at some point because there's too much internal dissension. It's too hard to coordinate, and then everybody's fighting for slices of what was viewed at the time as a fixed-size pie.

Structurally, there was just no way to get to scale.

Where else did you take ideas from besides the agent business in Hollywood and the merchant banking and investment banking industries?

It was very obvious that it happened in private equity. This was around the time when KKR and firms like it were hitting their stride, actually building a lot of operational capabilities in-house. They were building their own investment banks in-house. One of the things we've never done, but that has always been on the idea list, is to have an in-house bank. KKR had actually done that—built a captive bank. They had done a bunch of things like that.

We saw it happening: the mid-tier private equity firms were collapsing, and you either needed a very light-on-your-feet kind of solo operator doing small deals, or you needed to have a scale platform like KKR. It happened in hedge funds.

It happened in advertising, too. The TV show Mad Men tells the structural story of this happening in the advertising field in the '60s and '70s. I will ruthlessly spoil Mad Men because it's been off the air for like 20 years at this point.

A big part of the arc of Mad Men is that Sterling Cooper is a classic mid-market ad agency. At the end of the 3rd season, they sell it to McCann, which was the scale player at the time. They show you all the pros and cons of working for McCann, because McCann's this giant machine. If Don Draper is used to making all the creative decisions, now he's just in this conference room arguing with people until he gets up and walks out.

But then Don Draper and Roger Sterling start their own startup. They start Sterling Cooper Draper Pryce. That's the second one, which starts out as a true startup, a true boutique startup. And then they have this year and a half of fucking hell. They can't get anywhere. They can't get clients because they're too small.

Yeah.

They're subscale.

And then I think in the end—I forget. It's been too long—but I think they end up selling it. No, sorry, I got it wrong. They sell the first one to the British ad agency—

Mm.

That just completely destroys it, and then they sell the second one to McCann. So, they actually show that process happening twice.

Again, if you go back in history, that is what happened to ad agencies basically between the '40s and the '70s. Television catalyzed that. When television emerged, advertising became a much bigger deal than it had been before, and it had to be professionalized in a different way.

The other thing that happened is, of course, the external environment changes, right? Everything we just talked about has to do with the internal mechanics of how these things run. But the other thing is that the external environment changes.

I think Michael would agree with this: part of what made CAA possible is that, at one point, Hollywood was just movies. There was a low-level TV division. By the '70s and '80s, Hollywood was becoming much bigger than just movies. It was movies and TV and advertising and music and sports and politics and culture and all kinds of things.

In fairness to our competitors, Silicon Valley from 1950 to 2010 was primarily in the tools business. The companies, starting with Hewlett-Packard, that we all backed and built were basically just building tools. You'd build a tool like an operating system or a disk drive or something, sell it to people, and they'd figure out what to do with it.

It was right around the time we started our firm that the Valley was going from being primarily tools businesses to actually building directly competitive companies in incumbent industries. Airbnb went directly into the hospitality industry. An alternate-universe Airbnb is just boutique booking-hotel software—a tiny little boutique business building basically little spreadsheet software. But no, Brian Chesky decided brilliantly, "We're just going to go into the hospitality business and compete with hotels directly."

Uber and Lyft in the old world were just taxi-dispatch software. In the new world, they're full transportation providers. Tesla in the old world would've just been software for self-driving cars. Tesla in the new world builds the entire car.

Facebook was the same thing. Prior to Facebook, if you built online ad software, you were selling it to the media companies. Mark was like, "No, we're just going to be the media company. We're just going to build the entire thing."

This was the other thing that happened for us: that was right around the pivot point when the Valley's ambitions went from just building tools to going directly into incumbent industries.

And then this goes back to the scale thing. Why do you need to scale a venture firm? Because the companies need to scale, right? AI now makes that crystal clear, because the winning AI companies are raising billions, tens of billions, and in some cases hundreds of billions of dollars. The old world of $10 million, $30 million, or $50 million checks, where VCs tap out, is just not relevant anymore.

But did you know the scale was changing at the time you founded the firm?

We had a pretty good idea. I'd been involved with Facebook informally since inception, and then formally on the board since 2007. That thing hit the knee in the curve. It was very clear to us that we didn't know how big it was going to get, but it was going to get much, much bigger than the Internet 1.0 companies had gotten.

It was also around the time Apple was directly entering the cellphone market, which was another great example of this. Silicon Valley didn't used to make cell phones. The original cell phones weren't made by Silicon Valley. They were made by giant industrial companies like Sony, Nokia, and Motorola in Illinois. Silicon Valley would make the chips that went into them or the software.

Of course, Steve was like, "Yeah, no, screw that. We're just going to make a phone." There were these signals that it was happening.

The other thing was that the Internet itself was maturing. At that point, the consumer Internet was 15 years in, and we had seen every part of that. I forget what the number was, but that was probably around the time global Internet penetration was crossing 1 billion users on its way to 5 billion.

Yeah, you have a very interesting lived experience where you were there at the very beginning of the Internet. One thing that I'm fascinated by—and this actually was going to be the first question for you, because I've never heard you speak about this, at least on a podcast—is your partnership and relationship with Jim Clark.

Yep.

You were, what, 20 when you met him? How old were you?

I was old-fashioned. I actually graduated from college and got my degree. It's a very Stone Age concept these days. That was in 1994, so I was probably 22.

So, there's this great book, I don't even think you like the book written by Michael Lewis, "Silicon Valley's Story."

I've skipped it.

I've read it twice, just because I don't know if anything in there is true, but the portrait he paints of this very eccentric character is wildly entertaining to me.

Yeah.

But what's shocking to me is when you talk to young founders, this guy started 3 separate billion-dollar technology companies. I think he was the 1st person in history to found 3 separate billion-dollar technology companies.

And almost no one knows who he is.

Can you just talk about how you met him? What was it like working with him?

I knew exactly who he was, and the reason was because his company, Silicon Graphics, his first company, was the company in the Valley between, call it, 1987 and 1994 or something. They were like whatever Google or OpenAI or whatever comparable company you want to make. They were the company. And by that I mean they were the company where the smartest people in the industry all wanted to work. They built the coolest products you could possibly imagine, and they had this incredibly young, vibrant, and dynamic culture.

Then they hit this cultural moment that was just incredible in, I think, 1992, which was the turning point in the movie business when computer graphics really kicked in. The 2 movies back to back were Jurassic Park and Terminator 2.

They ran on the machines they made. They were built on the machines they—

Yes.

It was the technology they made. The technology Jim invented was the technology that made that possible.

Those are still 2 of the great all-time movies. But at the time, I still remember the chills you get seeing dinosaurs on screen. It's just like, “This is…” And then there's this company that builds the machines that do this.

The Silicon Graphics computers are actually in the movie, by the way. There's a scene in Jurassic Park where the kids are navigating through UNIX.

Yeah.

And it was actually the 3D software. Those were actually the Silicon Graphics computers. That was this moment where they were just the absolute “it” company of all time.

Their legacy lives on in NVIDIA. NVIDIA is basically Silicon Graphics' Traitorous Eight thing. It had to be a new company, for reasons we could describe, to do the GPUs instead of the workstations and servers. NVIDIA is fundamentally based on Jim's ideas. That's where that stuff all comes from.

He was already legendary. Again, he was the full deal. He was legendary as an innovator in technology because he had a PhD in computer science, and he himself invented the original, I forget what they call it, I think it was the reality engine. The original interactive 3D-graphics-on-a-chip thing was actually him. I think it was his PhD thesis.

Then he started the company, and then he ran the company. The VCs brought in a professional manager. The reason we know about NVIDIA today and not SGI is because of this founder-manager issue, which we can talk about.

No, let's talk about that real quick.

Yeah.

Because I don't remember this part of the story.

There are 2 sides to the story. I wasn't there, so I just reflexively side with Jim Clark, but I'll try to at least represent both sides of the story.

Jim is a true Elon- or Steve Jobs-level guy: incredibly creative, incredibly bright, and incredibly charismatic, but volatile. He's exciting. Being around him is just incredibly exciting. There's always something new, and he always has new ideas.

Again, that was in that time when it was like, “Okay, that's the personality type that clearly can't run the company.” So the VCs brought in a guy out of Hewlett-Packard who had been trained at Hewlett-Packard. At the time, Jim wanted to hire a professional CEO. He went and hired a general manager out of either Hewlett-Packard or IBM, which were the 2 training grounds for this guy.

They brought in a really, really sharp guy. I don't really know—I think I met him once. By all accounts, he was a very good example of this kind of HP general-manager type who became a CEO. He took over the company, and in his defense, under him the company scaled enormously. I forget when he took it over, but it was like 1987 or 1988 or something. By the time I got to the Valley in 1994, the company had become huge, and whoever was running the company gets at least some credit for that.

But anyway, they got into this classic fight. The classic fight was the same story every time. The founder is talking to the CEO, like, “We need to do things completely differently.” And the CEO's like, “No, what we're doing is working. Stop fucking with the thing that's working.”

The founder is like, “No, it's working now, but it's not going to work in the future.” And the CEO's like, “Well, then we'll deal with it in the future.” And the founder is like, “You can't wait to deal with it in the future, because by the time the future arrives, it's going to be too late.”

And the manager is like, “Why are you in my pants? I'm making you all this money. The company's super successful. Get out of my shorts.” You see this, and that was exactly the deadlock they got into.

Jim Clark basically made 2 predictions as the founder of Silicon Graphics. At the time, Silicon Graphics was selling computers that basically started at a list price of $50,000 for a desktop workstation and then scaled up into the millions. Jim was like, “Look, 2 things are going to happen.” It's amazing that he figured this out by 1991 or something.

He said, “Number 1, everything that we sell today for $50,000 is going to go on a chip, and that's going to go on a card, and it's going to go on a PC, and it's going to cost $300. Either we're the company that's going to make that, or we're going to get destroyed.” Which, by the way, is what happened. That's NVIDIA. That's what actually happened. He was completely correct about that.

The other thing he had was, “Look, this idea of standalone computers is not going to be the thing. These computers are all going to get networked together, and the network is going to become the important thing.”

At the time, there were different terms. People were using terms like the information superhighway, video on demand, or 500 channels. You had all these concepts coalescing around what became the internet.

Even before the internet became a mainstream thing, Jim was just like, “Look, it's inevitable that this is all going to become connected. The function of a computer is no longer going to be mainly what the computer does. It's going to be the fact that it can talk to all the other computers. And we need to do that.”

To do that, he actually went to Japan. He got this incredible deal with Nintendo, which then and now is this giant video game company. Silicon Graphics actually built the original 3D graphics chip for a consumer game player, the Nintendo 64. He did that deal, and then he went to Time Warner, which at the time was this very important media company doing all kinds of things.

He struck a deal with them to do what's called interactive TV, which was basically pre-internet. It was like Netflix before Netflix in 1991. Amazing foresight—just amazing foresight.

But again, he and the CEO got into this conflict. The CEO was like, “Look, we just can't. We have to focus on the thing that we're doing. We're not going to do these things.”

So Jim did the classic founder thing, and he left. When I met him, that was basically the state he was in: “Okay, I, Jim, am in the prime of my life. I know I have all these ideas. I don't know exactly what to do with my next company, but I know it should be a software company, not a hardware company. I know it needs to be a company that is able to anticipate these changes that are happening in the world.”

He was very sad about this. “Silicon Graphics is not the company that's going to be able to do these things, and so I have to build the new company that's going to do it.”

I want to hear more about what it was like working with him, but there was a very astute observation you made in your blog archive. This post was essentially trying to educate founders: just like recruiting is the most important thing you're doing at the very beginning of a company, maybe forever, you're underestimating how difficult it is. You tell the story of Jim Clark in the blog archive. You're like, “This guy was a legend.”

He was, yeah.

The most famous person, the best entrepreneur, and he tried to recruit all these other people. I don't know, it was like 100 people, and you're like…

Well, you were one of 2 or 3 who actually followed through—

Yep.

—and took the chance, jumped, and started working with him.

Yeah, again, this is like Zuckerberg or Sergey Brin or Elon or whatever deciding to start a company. That was his candlepower wattage in the community at that time.

You would think the obvious thing people would say is, “Jim Clark wants us to start a company with you.” The obvious thing is that you just say yes. It was not happening.

I don’t know if I told this story, but the crystallized memory is of a dinner of 12 of us at a famous Italian restaurant in Palo Alto called Il Fornaio. It’s where a lot of these companies were formed. It was Jim’s favorite restaurant at the time.

Jim had a dozen of us there. We were people who were in existing companies, basically technical people whom he knew. This is another thing: He was constrained. He had a non-solicit agreement with Silicon Graphics, so he couldn’t just rip people out, and he didn’t want to violate that. He needed to reach out to the technical community—

Mm-hmm.

—and find new collaborators. There were about a dozen of us in there, and I remember that dinner very precisely for 2 reasons. The first is that I was the only one of the dozen to say yes. The other is that it was the first time in my life I drank red wine, and I didn’t know what to make of it.

I kept sipping it, trying to figure out whether I liked it or not. I didn’t realize that I was getting completely hammered because I had no idea how to calibrate red wine.

The true version of the story is that I leave the dinner and I’m thinking, “Wow, this is amazing. I’m going to say yes to this. We’re going to do this.” I go to my car in the parking garage in Palo Alto, across the street. It was my brand-new car, my first new car I had ever owned. My brand-new car.

I gun it, pull out, and rip the entire front end of the car off. It’s this screaming metal sound. The whole front end of my car is just hanging on the ground, and I’m like, “Oh, fuck me.” So I park the car, get out of the car, and walk home.

No Uber this time.

No Uber. It was a 3-mile walk at 11:00 at night with 6 glasses of red wine.

And you’re what, 22?

No problem. Twenty-two, exactly.

Twenty-two.

I’m thinking, “I probably won’t mention this to Jim.”

I don’t know. There are some wild stories in that book. He might have admired you even more.

He might have. Yes.

How many founders of the company? Is it just you and him?

Originally, it was him and me.

Okay.

We started the company, and it was one of these things where we had long conversations about what to do.

The problem he had was that there was the idea of doing the graphics chip, and that’s what NVIDIA did. But NVIDIA was essentially a spin-off of SGI. At that time, starting a new chip company from scratch would have been tough, and he didn’t want to compete with SGI by doing that.

Then the interactive television idea—it’s lost to history, but this interactive television wasn’t ready yet. There wasn't actually time for Netflix yet. It was going to be cost-prohibitive. Time Warner had rolled out this interactive television thing in Orlando, Florida, to 500 people.

Yeah, and Microsoft was involved in some way.

Yeah. They were doing a ton in it. At the time, all the big companies were—

It’s all these Bill Gates biographies.

Yeah, exactly. He talks about that a lot. But the capital expenditure per house was about $50,000 because you had to have a Silicon Graphics workstation in the house, and it wasn’t going to work.

He couldn’t figure that out, and then we cycled through a whole bunch of ideas. He actually went back to Nintendo, and we almost pulled the trigger on basically building what today you’d call Xbox Live—or what is it called, PlayStation Network or Xbox Live?

Yeah.

An online gaming service for the Nintendo 64 in 1994, which might have been a good idea. We thought it was too early. We almost did that, and then what happened, literally, was the internet.

I had worked on the internet in college, and fortunately, this was only a few months later, but the internet just kept growing.

Hold on, Marc.

Yes.

You had worked on the internet a little bit? That’s a little bit modest. A lot of people listening to this will know, but you should probably explain how you were working on the internet.

This is part of the story. At the time, it was not that big of a deal. It was not nearly as big of a deal at the time as it’s viewed now.

A group of us at Illinois did this thing called Mosaic, which was the first widely used web browser and the first one with graphics.

Explain what was different about what you made compared to what existed before.

The previous web browsers were text-based. There was this nascent concept of the web, but it was text-based terminals, and it didn’t have graphics. It wasn’t point-and-click. It didn’t work in the way you would expect this type of software to work.

By the way, it also didn’t have a scripting language or security—none of the actual capabilities that make a browser useful. There was this nascent idea, but it needed to be built into a full thing. So we built the original full thing, a full browser, at Illinois.

We also built the first mainstream web server—the first web server that had everything people needed. This had been a project at college, and at the time, the internet was not viewed as a consumer phenomenon.

Wasn’t it illegal to commercialize? Steve Case of AOL tells a story that—

Yeah.

—he had to lobby and get a law changed.

Yeah, that’s right.

What were the details there?

The internet as we know it today, in the 1980s, was called the NSFNET. NSF stands for National Science Foundation, which is a branch of the U.S. government that funds research, and the National Science Foundation funded the internet.

The reason I was able to do the work I did at Illinois is that the NSF had dumped a ton of money into 4 universities around the country to build what were called the supercomputer centers. Those were also the main hubs for the NSFNET.

The function of the NSFNET was fundamentally to connect the supercomputers to all the people who were going to use them. It was a government research and academic program, and it was very exciting in the technical field. But there was no conception that ordinary people were ever going to use any of this. Nobody thought this was something that normies were going to use.

Because it was taxpayer-funded, the government at least was not supposed to be funding businesses directly, although sometimes it does. There were formal legal restrictions on funding things with commercial applications.

There was something called the AUP, the Acceptable Use Policy. The Acceptable Use Policy said that basically the NSFNET—the internet, as it was then—was for academic and research use, and commercial activities were strictly prohibited, literally not allowed.

As a taxpayer, that makes total sense. I’m glad my tax money isn’t going to fund something like that. But as a user, you’re just like, “All right, that’s nuts.” That’s clearly crazy, right?

If you took the conceptual leap to say, “No, this is going to escape the lab, and this is going to be something that normal people are going to use,” then it became obvious that it would have to have commercial activities.

AOL was one of the early pre-internet online services that wanted to connect to the internet. I think they famously connected to the internet in 1993.

Do you know about the concept of Eternal September?

No.

There are 2 internets: the internet that existed before 1993 and the internet that existed after 1993.

People who were on the internet before 1993 often describe it in utopian terms because it literally was like taking the whatever million smartest people in the world, putting them on a network together with no commercial activity, no advertising, nothing, and just letting them talk to each other. It was amazing.

The old messaging system was called Usenet, and the discussions on Usenet were absolutely spectacular. It was the most pure, clean, intellectual, vibrant space since, I don’t know, Athens in 500 BC. It was just this amazing phenomenon.

AOL connected. AOL had a million or 2 million people at that point. They connected all the AOL users, who were just normal people, to the internet in September 1993, and that became Eternal September. That’s the day the internet changed.

By the way, I’m pro that. I’m glad that happened. But the pro and the con of that is that it took the internet from this ivory-tower kind of thing to basically a mainstream consumer, ordinary-people thing, which is, of course, just a fundamentally different thing.

The concept of Eternal September literally was when a new wave of college graduates graduated, got their first job, and then went online. September was when the new crop of internet users showed up for a long time, so the “September effect” didn’t just happen once. It happened over and over and over and over and over and over again.

Every cycle of internet users would basically be like, “Oh my God, this is great, but it’s all going to get ruined in September.”

Yeah.

The internet we live in today is the result of 30 Septembers.

If they could only see us now.

Right. By the way, there was controversy at the time about whether the Acceptable Use Policy should be revoked. There was controversy over whether normal people should be on it or not. There was controversy over whether the kind of content normal people wanted on it should be allowed on it.

We got quite a bit of flak at the time for putting images into web pages, under the theory that that would fundamentally make everything worse because you’d have normie content, and that would be bad. And it was the same about e-commerce. By the way, advertising—I remember there was actually a moment. There was a guy named Sanford Wallace, and he became known as “Spamford Wallace.”

He literally sent out the first spam message on the internet in 1992. It was the first internet ad, and it was spam for, I don’t know, legal services or something. He just dropped it on Usenet, and it was like a thermonuclear explosion because it was like, “Get this commercialized crap out of my newsfeed.” So all of these things were hot and controversial.

I was generally on the other side of all these arguments because I was like, “Look, this thing is great. Obviously, everybody should have access to this. Obviously, we need to connect everybody to this. Obviously, to do that, these need to be businesses. There needs to be commerce. There needs to be advertising. All these things obviously need to happen.”

So were those the discussions you and Jim were having, where you were like, “Okay, we’re going to start an actual company around this”?

Yeah. That’s how we got to the conversation Jim and I had, which was basically like, “Okay.” Because that was right at the pivot point. This was in early 1994, so the AUP had just been revoked, AOL had just done the first September, and the whole thing was just about to tip. And I knew that because I was personally tech support for the browser.

No, explain that.

Just me. Mosaic at the time was the browser everybody used, and if you used Mosaic, there was a “Submit a bug report” option or whatever: “You have a question? Submit it here.” That went to an email box, and that email box was me. So I became tech support for the internet for about 3 years.

How?

I got all the emails.

How many emails were you getting?

Well, there were actually 2 email boxes. That was one email box, and the other email box was for commercial licensing requests. Mosaic was actually created under, and also funded by, the National Science Foundation. It was not originally for commercial use. The original license said it couldn’t be used for commercial use; it was for academic, research, and individual use.

So we had this thing. We did a deliberately ambiguous license, and we said, “If you want to use the browser commercially, you need to email us to arrange terms.” Now, we had no concept at all of what those terms would be, but we just said we needed to create this incoming flow. So I was getting bombarded with tech support requests.

And, by the way, tech support for the internet means you’re tech support for everything. The old PCs had CD-ROM trays. You press the button, the CD-ROM tray comes out, you put the disc in, and the tray goes back in. The problem is, a lot of people thought those were cup holders. You press the button, the cup holder comes out, you put your cup of coffee down, and then 10 seconds later, the cup holder retracts back into the PC and spills your coffee all over the place. You’re like, “How the fuck do I keep the cup holder out?”

Let me email Marc.

Yeah, let me email Marc. It’s like, “Sir, that’s the CD-ROM drive.” So there was a lot of that.

One of the funnier things you can always do—in politics, they call this “focus groups”—is user testing. You see this over and over at tech companies: take whatever amazing new thing you have, put it in a room with normal people, and let them try to use it. You learn so much about how much of a bubble you’re in, and about the kinds of things that you’re familiar with, while normal people are just like, “I don’t know what the hell any of this stuff is.”

Yeah.

So there was a lot of that. But then I had this other email box, which was all the commercial licensing requests. I saw the consumer takeoff on the one side, and then the commercial requests hit 400 messages from people wanting to pay money for this thing. So I basically took those to Jim, and I was like—

There’s a business.

Yeah, this is going to happen. And then we actually went to Washington in 1993 to try to get NSF funding to staff a support desk so that it wasn’t me answering all of the emails.

The people at the National Science Foundation were very nice, and they were like, “Yes, the National Science Foundation is not in the business of funding customer support desks for your software.” I still have the denied NSF grant that would’ve kept the whole thing an academic project. But at that point, at least to Jim and me, it was just obvious that it was going to be a business.

By the way, again, it was very controversial. The original press coverage on Netscape for the first year was that these people would never make money. “This is ridiculous. Everybody knows the internet’s free. Everybody knows that none of this is going to work.” Even then, it was controversial.

What did you think the business model was, though? Was it just literally licensing it?

It was a combination of things. It was definitely software licensing, and we made this thing up where the browser was free, but the server software cost money.

Out of the gate, we started building all these—we called them—applications, server-side applications. We built the first publishing system for running a newspaper or magazine online, a content management system. We built the first e-commerce system for selling things online. This was pre-Amazon, so we built and sold a lot of that software.

And then we owned the main website that the browser had as its default homepage, so we built the original internet advertising business. Netscape was the largest internet advertising company until, I think, 1997.

That’s incredible. I didn’t know that.

That was when Yahoo passed us, yeah. People at the company invented the original ad formats around that time. I don’t know exactly who gets credit, but a lot of them rolled out on our site first. So it was literally advertising pre-Yahoo. It was e-commerce pre-Amazon.

We literally put The Wall Street Journal online. That was our software that did that, along with a lot of other newspapers, magazines, and all that stuff. So it was a lot of that, and then it was the web operation.

Again, it all looks obvious in retrospect, but when we started this, I don’t know what the total number was. We started the company in April 1994. There couldn’t have been more than 2 million people total online, right? Almost everybody was coming in over dial-up. Everybody was coming in on 14.4-kilobit modems, and we were hoping that people would upgrade to 56-kilobit modems, which would be super helpful.

Yeah.

Computers at that time did not come with TCP/IP installed, so to get your PC actually on the internet, you needed to buy a TCP/IP stack. Try explaining to a normal human being what a TCP/IP stack is. It makes no sense at all.

They’re going to ask if they can put it next to their cup holder.

Exactly. It was just like talking to Martians, right? Their talking to us was like talking to Martians. Monitors were 3 feet deep and just bathing you in radiation. You were kind of hoping that the radiation stayed up here and not everywhere else.

In retrospect, it was super early, and it was all very speculative. Again, it was just like, “Okay, e-commerce. Are people going to buy things online?” It was like, “I don’t know, maybe.” But the press at that time was just wall-to-wall: “If you put your credit card number online, hackers are going to steal it.”

I was going to say, if you read any books that were around this time, they’re like, “There’s no way in hell anybody’s ever going to put their credit card on the internet.”

But, by the way, the other thing you would never, ever do is put your real name online because it would mean immediate identity theft and your life would be ruined. So you would never, ever do that.

The other thing was that, right in the beginning, you had all the panic around kids: “This is going to destroy children. This is a huge risk to children.” So you had all that panic. And then there were immediately calls for censorship. There was clearly all this stuff that you had to take down.

The New York Times kept running stories talking about how the whole thing was fake anyway. They kept saying that all the numbers were made up, that there actually wasn’t anybody online, that it was a tiny little user base, and that we were all inflating the numbers and committing fraud. So it was just this—in retrospect, it’s all quaint and cute and sweet, but it was the precursor. It was all the moral panics around technology today. You could see nascent versions of them back then.

You pick up on something that, because you and I have read a bunch of the same books, is that humans’ reactions to something new are just consistent—

Yeah.

—throughout history.

Yeah.

And so I heard a podcast with you. I thought I was the only one who would tell the story in private about bicycle face.

Bicycle face.

Exactly. Do you want to say what bicycle face is?

Bicycle face. Bicycle face. Bicycle face, yes. It basically turns out every new technology is greeted with what they call a “moral panic,” right? A moral panic is basically, whatever this new technology or new form of media is, it’s going to ruin everything. Specifically, it’s going to ruin society, it’s going to ruin morality, and then, especially, it’s going to ruin the children.

The bicycle was pre-feminism, so it was also going to ruin the women very specifically. It was going to ruin the women because women, clearly, in 1880, could not be trusted to use a bicycle without getting into real trouble. I'll explain why. This is a persistent theme, and you can go all the way back to this famous idea that Plato and Socrates thought written language was a big mistake, that all information transmission should be oral. They had this whole thing back in 500 BC.

You just have to imagine. I always like to hypothesize that the first guy who brought fire down from the mountain—

They probably killed him.

Yeah. They're like, “What the f*** is this?” Exactly: “This thing is horrible. This thing could burn down the village. This is awful. This is going to destroy everything.”

And so it's been this consistent thing. There's this great website called Pessimist Archive, where they go back and find all these newspaper articles that were contemporaneous to these things. But it's everything. When I was a kid, it was heavy metal music, Dungeons & Dragons—all this stuff was awful. I remember the moral panic around the Walkman, the very first portable cassette player with headphones, because it was going to destroy society because everybody was going to be listening to their own music.

I remember the moral panic around the calculator. It was going to destroy education because kids were not going to learn how to do math anymore. Then you go back, and in the 1950s it was comic books and rock-and-roll music. Obviously, it was going to ruin everything. In the 1920s, jazz music was going to ruin everything, and playing cards were going to ruin everything. What else? Novels—paperback novels—were taking over, and kids were going to sit around and just read novels all day instead of doing any real work.

It's over and over and over again. It's this constant story. The bicycle one is the great one. The bicycle rolls out in 1870 or 1880, and the United States was still thinly populated compared with today, but the West had been settled. You had all these little towns and villages scattered all over the place, but to get from one town to the next was 5, 10, or 15 miles, and people didn't generally walk that.

Then the bicycle comes out, and all of a sudden it's easy to go 5 miles into the next town. Young people discover the bicycle, and they discover that there are young people they didn't grow up with in the next town over, and they head out to do it.

To do it.

Well, to do it, yes. To do everything. To do whatever it is that young people do, they're going to head to the—

What's over that hill?

Yes, exactly. I grew up in a small town. I can identify with that. Specifically, young men, but also young women, started riding bicycles. This was a big threat. If you're a guy in a town and all the attractive young women are heading over the hill to the next town on this bicycle thing, that's a big problem.

The press at the time created this thing called bicycle face. It was part of the moral lecture that was given to young women in the press at the time, which was basically that young women should not use bicycles because if you go on a bicycle, you have to exert yourself, and if you exert yourself on the bicycle, you're going to end up making an exertion face. The thing was, if you did that too much, your face would freeze into bicycle face.

They literally thought it would stay that way permanently.

It would stay that way permanently, and then you would never find a husband. That was the moral panic.

These things just rip through everything. Music is always a great one because, in the 1990s and 2000s, there was all this moral panic around hip-hop.

Dude, Jimmy Iovine, who is your neighbor—

Yeah.

He was in here 2 weeks ago, and he had to deal with— They called him a—

Yes.

—chemical gas or mustard gas. They compared what he was doing to genocide because he was funding hip-hop music—

Yes, that's right.

—and white kids were starting to listen to hip-hop music.

Hip-hop music in the late 1980s and early 1990s.

There were congressional hearings on this.

That's right.

The media was behind him.

That's right.

He was pushed out of a conglomerate.

Yes.

This wasn't a joke.

Yes, that's right. It's actually funny because I'm not in the music business, but hip-hop has become so normalized that today it would never even occur to you. In fact, hip-hop was a cultural phenomenon. It's even fading today. But that was super intense at the time. Rock and roll was super intense in the 1950s and 1960s.

Remember Elvis Presley? They wouldn't shoot him because he would shake his—

Yeah.

—hips, so they said, “No, it's waist-up on TV from now on.”

Yes, that's right. But here's the one that I love: jazz. They said all the same things about jazz in the 1920s and 1930s. It was the exact same thing: kids were going to get together, and they were going to dance to jazz, and then who knows what happens. There was a jazz musician who smoked pot, and that meant all the kids were going to start smoking pot. It's the same story over and over and over again.

I'll just say, in fairness, it's not that society doesn't change. Many of the technologies that we just described did cause society to change. Things are different before and after the bicycle. They're different before and after the car. They're different before and after the creation of modern culture—rock and roll, or whatever.

But this idea of the moral panic, this idea of outright panic, the end of the world, is just this repeated thing over and over again. What's happened is that this is the obvious way to sell newspapers. This is the meta-story of the press: “Whatever's happening is horrible and awful, and it's going to kill everything. Be sure to buy our newspaper tomorrow.”

Deel is the best company in the world at building infrastructure for global hiring. Deel will help your business hire, pay, and manage any worker anywhere in the world. Deel is one platform for payroll, HR, benefits, and device management across 150 countries. Deel gets you everything you need to run a high-performing global workforce on a single AI-native platform. From the first offer to final off-boarding, Deel handles the complexity so you can stay focused on your business. The best founders and operators in the world have one thing in common. They control as much of their business as possible. The founders of Deel do exactly this. When you use Deel, you aren't using a third-party payroll processor or a messy network of in-country providers. Deel built and owns their own rails. That means faster speed, better service, and total accountability. In fact, Marc Andreessen's firm, A16z, has been investing in Deel for more than half a decade. Deel has built AI teammates that take action. Most HR AI stops at chat. Deel puts AI inside the workflows themselves, so it can run real work across hiring, payroll, mobility, IT support, and reporting. Grounded in compliance-vetted knowledge, Deel AI takes action to help teams move faster without adding headcount. Deel is trusted by over 40,000 businesses. Learn how they can help your business today by going to deel.com. That is D-E-E-L dot com.

I want to go back to Jim Clark real quick. Is there anything that you learned working with him back then that you still use today? Jim Clark was probably 20 years older than you.

What a f***ing education you had to be able to work with that guy when you were in your early 20s.

Yeah. No, that's right.

Is there anything that you learned by working with him back then that you still use today?

A lot. It's easy to say it was very formative for me, so a lot of it. You mentioned that quote earlier about the world being a malleable place. Jim was the ultimate version of that. When he had an idea—and he was right; his ideas were correct almost all the time—he would just pound the world into adopting them, into believing them. He was a complete force of nature.

One thing that was malleable was himself. He has this great quote in that book—

Yeah.

—where he calls himself a “self-described loser” at 38 years old. The guy had 2 PhDs; he was a professor. But I think he had been through his second or third divorce, and he just snapped one day. He said, “I woke up one day with the undeniable urge to achieve something.” That's when he went from academic to founder and just ripped off company after company for a few decades.

Yeah. And of course, he did that not just by starting a company, but by inventing interactive computer graphics.

Yeah.

He completely changed the field and, indirectly, completely changed Hollywood.

Is there anything about recruiting or managing, or any other way that he ran his company?

No. My 2 mentors at that time were, in some ways, polar opposites. They always got along, but they were polar opposites.

They were both Jims: Jim Clark and Jim Barksdale. The Jim Clark side of my personality is like the will to power: “I’m just going to bludgeon the world into doing what I want.” And then there’s the idea of trying to be a fountain of creativity. There are many new ideas out there, and you just need to go find them.

I would say there’s also—I’m not sure how to put this—a sense of professional dissatisfaction. Like, “Okay, whatever.” Look, this is the other part of the story: a lot of founders would have had a successful sale of Silicon Graphics, and that would have been it. They would have spent the next whatever—whether they were totally happy with how it turned out or not—30 years just coasting on that, right? Having a great time, taking credit for it, and the whole thing.

Mm-hmm.

But Jim was always, at least in that part of his life, dissatisfied in the productive, positive sense of, “Okay, no, there’s something better. There’s something bigger. There’s something new that we should do.” So there’s that side of it.

And then Jim Barksdale was the other, who I literally was with yesterday in Jackson, Mississippi. Jim Barksdale is on the other side. Jim Barksdale is the manager of managers. Clark is the ultimate example of that bourgeois-capitalist thing I mentioned—the Henry Ford, Elon Musk type—and then Jim Barksdale is the ultimate example of the super-manager.

Jim had run big parts of IBM, AT&T, and Federal Express, and came in to run Netscape. What was interesting was that’s kind of where I got a lot of this from, and a lot of my skills from. I got trained by both of those guys, kind of both of those guys at the same time, and then was able to very clearly observe the difference between those mentalities.

But then the other question is, of course, how those concepts converge, right? Because just the fountain of creativity—you can’t build anything big with that alone. With just management, you don’t do new things.

Who’s a great example of that from history? Would it be Nikola Tesla—the fountain of creativity?

Oh, that’s right. Yeah.

He needed a George Westinghouse to commercialize his ideas.

Well, it’s Tesla versus Edison.

Okay.

Yeah, Tesla versus Edison. So I’m an Edison guy. That’s the thing. Elon’s a Tesla guy, obviously. But Elon, of course, has now become really outstanding—I mean, he’s obviously become an outstanding manager in his own way. In fact, to the point where I think he’s actually inventing an entirely new school of management, which we could talk about.

Let’s go there next.

Yeah. Okay. He’s maybe the greatest manager of our era, despite the fact that nobody thinks of him that way. So I actually think Elon’s more like Edison than he is like Tesla.

There was a big war, and it was kind of this thing because everything turns into these little morality plays. The basic story of Tesla and Edison was that Tesla had all these ideas but couldn’t commercialize them, couldn’t ultimately turn them into companies, couldn’t figure out how to make money on them, and couldn’t build big companies based on them.

And Edison—at least the way the legend goes—was more of a grinder. He was less incandescently brilliant and more of a grinder. He was just like, “We’re going to try 1,000 things.” When they invented the filament for the lightbulb, they tried about 1,000 different combinations of things to get to the filament. It was this brute-force approach.

But then he built General Electric, right? He built the national electric grid and built these giant companies.

Funded by—

Funded by J.P. Morgan—

There you go.

—as a venture capitalist in his spare time.

Yes, exactly. 100%.

And then Edison also invented the movie projector and literally spent years trying to enforce his patents, right?

And the phonograph.

Which—you tell the story, and I knew it because I read the book too.

Yeah. Yeah.

We should tell people what he thought the phonograph was going to be used for. This is a bit of a digression, but it gets to the personality type.

One of the things that people look for is, “Oh, what are the consequences of a new technology going to be? Let’s go ask the people who invented it, because obviously they know.” This is what happens when, for example, the pioneers of AI get interviewed in the press. It’s like, “Well, tell us the future of AI.”

The one I’ll pick on is Geoffrey Hinton, who is an actual self-declared socialist. He’s an actual card-carrying socialist, and people ask him, “What’s the future of AI?” Of course, he says, “It’s going to be rampant unemployment, and we’ll need to give UBI to everybody.”

It’s like, what a coincidence: the answer from—

Yeah.

—a socialist is communism. What an amazing coincidence. But people think that because he’s one of the inventors of AI, he must be the guy who knows.

Yeah.

The story I always tell is the Edison story. Thomas Edison was a very proper WASP. He was a WASPy personality type of that era.

Yeah. Extremely proper gentleman, always impeccably dressed, a very ethical, upstanding kind of citizen of that time, and very religiously devout.

For him, it was just obvious that the application of the record player was that everybody would buy a record player, and then everybody would buy a library of discs that would be the great sermons of all the great preachers of the time. You’d get home at night after a long day at work, turn on the record player, and listen to a sermon, with your adoring wife and kids—

Yeah—

—gathered around you.

And of course, the record player drops, and immediately it’s music. It’s obviously music. It’s ragtime, then swing, then jazz. Edison was completely horrified.

He didn’t know that if you put the phonograph in the window and played good music, you’d have all these girls on bicycles coming over with bicycle fists.

Exactly, exactly. And so this is what I always tell people. If Edison didn’t know what the phonograph was going to get used for, the idea that some Joe AI entrepreneur is going to be able to forecast the economic implications—you’re like, “No. No, that’s not going to happen.”

In fact, the people who invent the technology are often the least qualified people to understand the long-term implications, because they’re too buried in the specifics of the here and now. All these other questions are big cultural, social, and economic questions.

By the way, I don’t know if there’s anybody who can predict big cultural, economic, or social trends, but it’s certainly not somebody who’s been in the lab for 20 years—including myself.

So how this started—you think you greatly benefited from the two Jims essentially being polar opposites—

Yeah, basically.

—and showing you—

But also working very closely together.

Did they get along?

I don’t know if I’ve told this story publicly, so I should tell this story. They got along great and became very good friends. They both did great, and they’re both responsible for certainly everything that Netscape did and everything that I’ve done.

But it’s different disciplines and different worldviews, so there’s an oil-and-water kind of aspect to that. Clark ran the company for the first 9 months. At the time, this was Internet time, so it felt like much longer, but it was a highly compressed 9-month period.

We were doing all these new things. The company was just doing 100 new things. It was amazing. But nothing was being systematized, right? By default, it was not going to turn into a large company without the management part.

So Barksdale comes in, and he’s basically like, “Wow, this inventiveness is great, but we need to start having systems, schedules, and processes, and actually run this thing like a business.”

As founders do, Clark originally found that a little frustrating, because whatever the latest idea was, it wasn’t the thing that we were going to turn the entire company toward pursuing. This was when Clark was still coming to Jim Barksdale’s staff meetings.

Clark got up and had a negative reaction to Barksdale saying, “No, we’re not going to do this new thing. We’re going to keep doing the thing that’s already working.” It was one of those moments.

Barksdale was like, “Can I talk to you outside?” So they went out back. I heard the story from both of them later.

Clark said, “Look, this is the whole reason we’re here: because we do these new things. If we don’t do these new things, we’re going to destroy the company.”

Barksdale looked right at him and said, “Jim, I hear you. This is as serious as dick cancer.”

What? In the deep Mississippi drawl, right? Clark stared him right in the face and burst out laughing. They got along great ever since. They loved each other ever since.

It was first of all basically saying, “Look, we’re not going to make these decisions in a state of superheated passion. We’re not going to do that. We need to have the full version of this conversation, but we’re going to have it in a longer and maybe more dispassionate way.”

Yeah.

But it was to puncture the stress of the moment. And so I will say I have used that one a few times.

Yes. But I could see Clark—

And Clark thought it was hysterical. Nobody had ever talked to him that way before.

But I could see Clark thinking, “Oh, no, here we go. This is a replay of what happened at Silicon Graphics, though.”

I think he was probably afraid of that to a certain extent. But yeah, yeah. I don’t want to say anything negative about the SGI guy, but yeah.

As I said, Clark was just like... Barksdale was the manager of managers. He was so advanced on this. That story notwithstanding, Barksdale never took the position of, “No, it’s time for the new ideas to stop.” But it was always like, “Okay, we need to thread the new ideas into a business,” which is kind of the hybrid of the two.

So, I just had this thought while sitting here listening to you speak. Is there something about your partnership with Ben where he’s more Barksdale and you’re more Clark?

Yeah. Although we mix it up a little bit more, because he does have his own edge. But yeah, there is some of that.

For example, he runs the firm. I tend to come up with new ideas. He comes up with lots of new ideas, but I do tend to come up with new ideas, and then we have this kind of discussion frequently.

So, if I was to follow you around without you knowing, with a camera, what would your day look like then? Are you just a fountain of ideas? Are you this uncontrolled energy like Jim Clark back in the day?

But I’ve got both. This is the thing: because they both trained me, I’ve got both parts of it.

Okay. So, you’re not as uncontrollable or unmanageable as Clark?

Yeah. I believe Ben would tell you... Look, Ben’s been working with me now for 30 years, and so I think if this was a real issue, I think our partnership would not have lasted.

But I think he would say that I have a pretty strong internal edit function.

I want to see the unedited side.

Well, unedited is really fun. Unedited is very enjoyable. It is very disruptive, and so it has to be calibrated.

When do you show the unedited side?

I don’t tend to do it in the spur of the moment. This is the thing, and Elon threads this incredibly well, just incredibly well, as does Zuckerberg. Again, this goes back to the Edison–Tesla thing: when you’re responsible for an organization, when you’re responsible for a team of people that’s more than 5 or 10, if you’re going to have an organization that’s 100 or 1,000 or 10,000 or 100,000 people, you can’t change the plan every day.

You just can’t. You’ll burn everybody out, you’ll destroy everybody, there’ll just be mass confusion, and people will quit. You can’t do that. There has to be some calibrated middle ground.

There are a handful of examples of great business successes where it’s 1, 2, or 3 people, right? Maybe it’s Bitcoin, Minecraft, WhatsApp, and Instagram, and then I start running out of examples. But with AI, there will probably be more. There will probably be more single-person companies from here on out.

Yeah.

Or, by the way, an artist or a novelist. Let me say, there’s a difference between a novelist and a moviemaker. A novelist can put whatever the f*** you want in your novel. But if you’re a director of a movie, you can’t change the entire plot on Tuesday while you’re shooting the movie, or there are 300 people who are relying on you to complete a movie.

Anyway, the point being is, in tech, if you’re going to have an organization—or, by the way, in anything, in any field of activity—you do need to have some calibration, some titration process. Change does need to happen, but it needs to happen in a measured way. You can’t just blow it up every day.

To get the holy grail of a large-scale organization that’s still innovating, you either need 2 people involved who are able to balance each other. You could say Steve Jobs and Tim Cook would be a canonical example. Or, early on, Zuckerberg and Sheryl Sandberg, or early on, Bill Gates and Steve Ballmer. You can have that kind of configuration.

Or, every once in a while, you can get that in a single person, right? Which is very rare. Jensen Huang would be a single-person example of that. Every now and then, you get that.

I would say Ben and I have a version of the yin-and-yang kind of aspect to it. But like I said, he’s very creative on his own, and I have this—because I have the Barksdale training—this additional level of self-governance most of the time. I get it.

My big thing is just like, look, if I’m going to walk in and throw a fit and be like, “We have to change everything tomorrow,” and Ben’s going to be like, “F*** you, this f***ing sucks,” that leads nowhere good, right?

Yeah.

And so, I do a lot of self-editing.

You just said something. I think you said you believe Elon is inventing a new way to manage. I think he may have figured out the best way to reconcile the two—the fountain of ideas with the systematic builder. I think he might have cracked the code on how to do that for the next 100 years or something.

So, I’ll just start by saying, look, Elon’s method has been described by people before. I should say, I work with him, but from the outside. I’ve not worked in one of his companies, so I have 1 layer of indirection. But I work with him quite a bit now, and I study him very, very carefully.

It’s this extreme focus on substance. It’s this extreme focus on getting to the truth. One of the things you notice in any organization with multiple layers is that, basically, there are compounding lies.

I got this lesson early, because I worked for IBM at the point of its kind of maximum size and importance in the world.

Can you explain? I don’t think people understand just how big, powerful, and almost monopolistic IBM was.

Yeah. I worked for IBM at the very height of its power, right before it fell. It was my first job when I was in college. In the mid-1980s, they were 80% of the market capitalization of the entire tech industry.

There’s nothing even close to that today.

There’s nothing even close, right? So, this is like Google times 10 or Apple times 10. It’s a level of scale and importance that nobody had.

The TV show that does a great job of this is Halt and Catch Fire. In the first season, there’s a point where these guys are basically inventing the PC, effectively, and IBM shows up. It gives you a sense of the CAA story you told earlier. It’s like the phalanx. It’s like 20 people in blue suits who are just here to completely crush you. It was this overpowering kind of thing.

They invented all kinds of stuff, and the industry wouldn’t exist today without them. They were an incredible company for a very long time. The whole thing was run by their founder for 30 years, then run by the founder’s son for 30 years—this incredible company.

They’re not that anymore, but they’re still a big and important company today, whatever, from 1940, so 80 years later. How many companies survive in tech for 80 years?

My favorite IBM story is that Thomas Watson Sr. had been convicted of antitrust crimes before he started IBM.

Is this the cash register?

The cash register. He had previously run a company called NCR, National Cash Register, and he had been convicted by the federal government of monopolizing the cash register business before he even started IBM. Then, at IBM, he monopolized the mainframe business, and they convicted him again. He was a double dipper. He got very used to being in antitrust court.

There’s Kevin Maney, an old-school tech reporter, who wrote a book, a biography of Thomas Watson Sr., called The Maverick and His Machine: Thomas Watson Sr. and the Making of IBM. He went back—we’re talking about the 1940s, 1950s, and 1960s—and got the records.

At that time, they had a secretary transcribing, in real time, all of the executive staff meetings every Monday morning. He went back and got the archives of the transcripts of the executive staff meetings, and Thomas Watson was just cursing everybody out, a complete tyrannical psychopath, screaming at people. It’s all in the records.

How much of this stuff ever changes? I don’t know. Whatever Elon gets accused of, or whatever Steve Jobs gets accused of, it’s like, oh, no, that guy was... Whatever it is, it’s a pale version of what that guy was doing.

By the time I got involved in IBM, it was 50 years later. They were peaking in their power. But what happened was—I remember this because I was there as an intern, and I was trying to figure out whether I should work there after college—their internet was a mainframe app.

One of the functions was the org chart, and it calculated that there were 12 layers of management between me and the CEO. Which meant the following: my boss’s boss’s boss’s boss’s boss had a boss, boss, boss, boss, boss before it got to the CEO.

What really happened, and I saw this happen up close, was that each layer of management was lying to the one above it, because each layer wants to look good and wants to put a little spin on the ball. If 1 layer lies to the next layer above it, maybe that’s okay. But when that happens 2 or 3 times, the lies compound.

If that happens 6 times, the lies really compound. If that happens 12 times, the CEO has absolutely no idea what’s happening in the company, which was the state of play that IBM had.

They actually had a term. They had a whole vocabulary. This company was like a nation-state at the time. You could live your whole life in Austin, Texas, and never meet anybody who didn't work for IBM. It was just this incredible thing. They had this concept called “The Big Gray Cloud.”

It was literally the cloud of men in gray business suits who followed the CEO around and prevented him from ever talking to anybody who was actually doing the work. So, when he would come to visit, it was like a state visit.

Yeah.

It was like a visit from the king, and it was like the king and the traveling court.

Yeah.

And so, it was a completely impervious bubble to get information through. But I tell that story because that's the polar opposite of the Elon approach, right? And by the way, being the CEO of IBM in 1989 was a great way to live, right? It was just like, “Wow, everybody's bringing me good news all the time.” I wake up in the morning, and everything is great.

I'm famous, and I am rich, and I am successful, and I've got a chauffeur, and I've got a jet, and I've got these 80 guys in gray suits who are taking care of everything for me. And I don't ever have to talk to engineers, and this is great until it's like the turkey on Thanksgiving, until things change and there's a problem, and then you have no idea what to do about it, which is what happened to them.

Mm-hmm.

The Elon approach is the polar opposite of that. The polar-opposite approach is literally, “I'm only going to talk to engineers,” right? And so, when there's an issue, I am going to go straight to the source of truth, and the source of truth is the engineer who actually knows what's going on.

I've seen him do this, so he literally does this. When there's an issue in one of his companies, he goes to the engineer who's working on that problem, and he sits down with the engineer, and they solve that problem. I can just tell you, the number of CEOs in tech, even the great ones who do that—I mean, almost nobody ever does that.

Why does nobody ever do that? Well, first of all, it's just a giant pain in the ass, because your life consists of having to actually solve all these problems. The whole point of being big, powerful, and successful is that you pay people to do that, and now you're doing it. And you're in there at 2:00 in the morning doing it, right? It just sucks, right? And so, most people won't do it.

The other is that it means the CEO of the company has to have the skill set to be able to do that. So the CEO has to not just be a great CEO; they also have to be a great technologist. And not just have memories of having been a programmer at one point or a chip designer, but be able to actually sit down with the chip designer on a Thursday night at 2:00 a.m. in Austin and figure out what's wrong with the chip.

Elon has that ability, and he's encyclopedic on every area of technology. He's able to go hands-on with rocket designers, AI designers, and everything in between. Almost no CEO has that. But that's literally what he does.

Yeah.

I think this is why Tesla is smoking, has been so dramatically outperforming the rest of the auto industry: because he's fixing the critical production bottleneck at Tesla 52 times a year, himself.

Yeah.

I can tell you what the CEOs of the legacy automakers are doing. They're not doing that. That is not what's happening, right? And so, in contrast, in a normal company, it might take 6 months to solve these problems, and Elon's fixing it right now, tomorrow. “Let's go fix it right now.” And so, he just runs this loop over and over again.

He's absolutely indefatigable. Famously, for a while, he had sold all of his houses, and he was literally couch-surfing as one—

Yeah.

—of the most successful people on the planet. And so, they have a vacation house, and I offered him—I said, “If you want to take a week and use the vacation house, take the kids, feel free.” And he sent back, 5 minutes later, at 11:00 at night, the forwarded response: “I don't take vacations,” right?

Yeah.

Which, again, is like, there's no CEO like this. The whole point of being a CEO is you get to go jet around. So anyway, he's doing that, and then he turns this into a routine. He does that 1 day a week at each of his companies, and he'll basically do it all day. He'll do a 12-hour to 14-hour stretch where he'll do design reviews with—

He does it with 5 minutes per engineer, right? So he does 60 divided by 5—it's been way too long in this podcast. How much is that? 12?

12.

He can do 12 design reviews an hour.

Yeah.

And then he does it for 10 hours a day. So Elon will do 120 design reviews in the course of a day.

Are these one-on-one?

I have not actually sat in on these.

Okay.

I suspect there are other people around, including people who work for him, and probably some of the leaders of the companies are involved in different ways. But it's literally the thing I know. It's literally a rotating cast. It's the point engineer on each of the important things coming in and presenting for 5 minutes.

And then the question is, if it's going great, that's great. If it's not going, what's the problem? And then how does that problem rank, right? Is that the production bottleneck? And if it is the production bottleneck, then that's the thing that he fixes. And then that's when he's there from 8:00 until 2:00 a.m., working with that engineer to fix that problem.

One way to think about this is the velocity in military affairs. It's called maneuver warfare, right? The speed at which he operates is just—the cycle time is so much faster than anybody running in a traditional method. It's hard to even compare. It's like 4 hours versus 6 months. It's just this incredible gap.

And then, the other part of it is, somebody that I know once went to work for SpaceX, and they asked what it was like. He said, “It's like being dropped into a zone of shocking competence. It's like everybody is ultra-competent.”

The reason everybody's ultra-competent is because, number 1, if they're not, Elon sniffs it out and fires them. But he knows because he's talking to the people actually doing the work. So he knows at this point, having done this for 25 years, he can sniff this out really quickly now.

And then the other is, the best engineers in the world want to work for him because he's the one CEO like this who's able to work with them as a peer on whatever the technology is. And as an engineer, you're just like, “What would be better as an engineer than being able to design a rocket engine with Elon Musk as your engineering partner?” Right?

So he just has this incredible positive selection where the smartest people in the world want to work for him, and then anybody who can't cut it gets fired. The world sees this as raw aggression, but it's beyond that, right? It's a very systematic way of optimizing these companies to be able to take on these profound challenges, and then being able to actually solve all the problems and do these things at a speed that is just completely unmatched.

The challenge of all of this is, okay, that all works great if you've got Elon, right? One of my concepts is that I think we need a metric for founders in Silicon Valley called the “MilliElon,” right? How many MilliElons are you? Are you 10 MilliElons? That would be great.

Yeah.

Are you 100 MilliElons? That's 10% of an Elon.

Yeah. That'd be fantastic.

Five hundred MilliElons? I'm going to give you all the money, right? Most people are 1 MilliElon or 0.1 MilliElon. The question that falls out of this, which is a question that people ask, is, okay, you can't clone him. You can't bottle the essence. So what out of that can be transplanted to normal human beings?

And how much of it is predictable or knowable when he's much younger?

Because the famous example of this is Michael Moritz passing on Tesla.

Mm-hmm. He made all his money in PayPal with Elon. Obviously, there was contention there. He got kicked out and everything else. But then Elon pitched him Tesla.

Yep.

And he passed because he's like, “There's no way that you're ever going to surpass Toyota.”

Yep.

And then Moritz, to his credit, was just like, “I drastically underestimated the guy's determination and pain tolerance,” I think is the term he used.

Well, I wasn't there for that, so I don't know about that. I will say, the idea of having been a software entrepreneur and building a car company—

Okay, when Tesla started building, there had been no new successful car companies in the—

Yeah.

For 100 years, right? In the United States, for 100 years.

Founded from 1900 to 1910, and 3 that survived.

It was like 2,000 of them.

That's right.

And the previous real attempt to start a car company in the United States before Tesla, in the preceding decades, was?

Yeah, yeah, Tucker. Tucker or something? Tucker Automotive.

Which was such a disaster that they made a movie called Tucker, which is about what a disaster it was. So, obviously, you don't do that. Obviously, this is insane, and for a software guy to do this is insane. Oh, by the way, this is only one of the things he's doing. He also has the rocket company.

Yeah, which is also insane, right? And by the way, I didn't see it. I'm a software guy, and I was like, “I don't know. I guess he's going to go do cars. I don't know anything about cars.” It's not like I saw it, but I'm just saying, the level of incredulity that he was greeted with at the time was, I think, almost uniform.

And there's that famous photo. The most famous Elon photo, I think—the most powerful one—is the one where it's a young Elon, probably 2005 or whatever, and he's in the—

He's in shorts and a polo—

Yeah.

—and he's crouched down, and there's nothing but the explosion remains of the third rocket, the second or third rocket.

Yes.

The one he had been funding personally. Did you ever read Eric Berger's book, Liftoff?

No, I didn't.

I'm surprised you haven't.

Oh, you've got to read it.

I like these company histories that focus on the first 6 years.

Yep.

And it just stops.

Yep. It's a first-6-years history of SpaceX.

Yep.

And there's just nothing good in the book. It's just reading one failure after another, and one catastrophe after another. It's a good read.

Yeah. When my kid was 5, he loved rockets, and so his favorite rocket video was a compilation of all the SpaceX rocket explosions.

Geez. Elon talks about this. before his friends, after he sold them to... I think he had 180, I think the story tells 180 million after taxes, and he's like, “I'm going to do this rocket company.” One of his friends—I think Adeo Ressi, or I forgot which friend—sat him down, and they made him watch all the rocket explosions. There was a compilation, probably pre-YouTube, of just rockets blowing up over and over again. It was like, “No, you're literally going to light your fortune on fire. It's going to explode in the sky.”

Yep, exactly. Exactly.

So, obviously, it's working, right? His method is working, and it's working far better than—it's certainly working far better than anybody else's method in cars, and it's certainly working better than anybody else's method in rockets. And then in a bunch of other areas, also. So it's clearly working.

And because of just who he is, what he's doing, and how he does it, he draws so much heat. There's just so much—the environment is just full of criticism and attacks, nonstop. We all get sucked into these narratives. But I think the key thing, for me, is just that there is a method there that he has been working on and refining for coming on 30 years, and it has worked better than anybody else's method.

Like I said, I don't know how many people can do it. Maybe there's just a fundamental limitation, which is that you can do it if you're Elon, and you can't do it if you're somebody else. Or maybe you need to be above 30 MilliElons, but not below, or something like that, right? Maybe there's some threshold where you break through on this. But it is clearly the best method. It clearly generates the best results.

And, again, conceptually, I like it because it's this bridging of the founder mentality with the manager mentality. He's not just doing one-offs. These are not just one-offs. He's scaling. Everything is scaling. What is it? Starlink just hit—what was the number? Starlink just hit 10 million subscribers?

I'm one of them.

Yeah, exactly, right. You probably have read about Iridium and Teledesic.

No.

Oh, okay. So Elon is not the first guy who said, “We're going to do satellite-based internet access.” There was Bill Gates and Craig McCaw. When Microsoft went on top of the world, and Craig McCaw had basically built cellular telephony in the United States—he built what became AT&T Mobile—those guys teamed up in the early ’90s and did this thing called Teledesic, where they put up satellite-based voice, and then it was going to be internet access. Complete catastrophe, total bankruptcy, complete disaster.

And Motorola, which used to make all the cell phones in the United States, had another system called Iridium that's actually still up. Again, it's just this classic business-school case study of complete disaster, capital destruction.

And so Elon was like, “I know. I'm going to do number 3 of those with Starlink as a side project at SpaceX.”

Yeah, right?

And, you know, in retrospect, it's total genius because he's like, “We're going to be putting up—if the rockets are reusable, we're going to be launching them all the time.” And then the question becomes, “What's going to go in the rockets?” He's like, “I could wait for the customers to come to me with more stuff to put in the rockets, or I could just put up my own satellites. What would be the satellite to put up? Oh, it would be consumer-grade, consumer-priced internet access.”

And it's just like, okay, anybody who knew anything about the history of satellites knew that that was the craziest idea in the world. And, of course, it's this giant success. It's like the side project.

There's clearly a method. It clearly incorporates invention. It clearly incorporates scale. It does a brilliant job at both of those. It's clearly, in part, the Henry Ford–Alexander the Great method. But there's also real scale and heft to it. SpaceX now is building—they've got their own city down in Texas, right?

And so it's a formula that captures both sides of it. It may be the least studied and understood thing I know of in the world right now.

It's incredible. Marc, we're running out of time. When I started the show, you were at the top of my list of guests I wanted to talk to. Thank you so much for doing this. I hope you come back in a few months, because there's a million other things we need to talk about.

Good. Awesome. Fantastic. Thanks.