Harvard Business review

How Network Effects Rule the World (with James Currier)

AZEEM AZHAR: Hi there. I’m Azeem Azhar, and you’re listening to the Exponential View podcast. Every week, I come together with a brilliant mind to explore how exponential technologies are shaping our near future. This is the first episode after our brief summer hiatus. I missed you. I hope you all missed us. I am so excited about the brilliant guests we’ve got lined up. Now, while we were away, my book, The Exponential Age, or for those of you outside of U.S. and Canada, it’s called Exponential, was published. In the book, I lay out a thesis for the exponential age, the new period of human affairs, catalyzed by accelerating technology platforms. Platforms are operating across the four broad domains, many of which we’ve talked about in this podcast: computing, energy, biology, and manufacturing. I loved writing it. I’m sure if you like this podcast, you will enjoy reading it–no, you will love reading it. Please get a copy wherever you get your favorite books, or you can find links to buy it at www.exponential-book.com. That’s www.exponential-book.com.

AZEEM AZHAR: Now onto my guest today, James Currier, is one of Silicon Valley’s best thinkers on growth and networks. He has founded seven successful businesses, including Tickle, one of the Internet’s first successful user-generated companies. He pioneered many techniques that are ubiquitous today, including viral marketing, AB testing, and crowdsourcing. Today, James is our partner at NFX, a seed stage venture capital firm focusing on companies that benefit from network effects.

AZEEM AZHAR: Network effects occur when a company’s products and services become more valuable when the number of people using them increases. In other words, companies can grow exponentially when they leverage those effects. Network effects themselves aren’t new, but massive increasing in computing power and the prevalence of smart phones connected to the internet have made network effects much more prevalent. Many of today’s most successful companies, and doubtless some of tomorrow’s too, have network effects at their core. And few people understand how they work better than today’s guest. James Currier, welcome to Exponential View.

JAMES CURRIER: Very happy to be here. Good to see you.

AZEEM AZHAR: James, I have learned so much from your work at NFX. What was your eureka moment when it came to network effects?

JAMES CURRIER: So we started a company in 1999 that was what we call a fresh produce business. We had to keep supplying new content to the website in order to keep people coming back and consuming new stuff. And when we were acquired in 2004 by Monster.com, they were a horribly run company, yet they were worth seven billion and we were only worth a few hundred million. And in the end, our management team ended up sort of showing them how to do things. We had a better team, but they had the network effect. And that’s when I first saw a network effect. They could go to sleep at night and wake up in the morning, and their business was bigger than having done nothing. And I realized, wow, going forward, I don’t want to build any business that doesn’t have a network effect.

AZEEM AZHAR: It’s such an amazing dynamic, isn’t it? This idea that as you just carry on your everyday affairs, your business gets stronger and stronger.

JAMES CURRIER: Yeah. It’s like having another 1,000 employees at your company. You set up this conditions and then the network grows, and then the network density increases, and the value gets greater and greater at all your users, and they get happier and happier with you, even though what’s adding value to you, your company, is actually the addition of new people.

AZEEM AZHAR: So the idea of network effects isn’t really, really new. I think people often point to Theodore Vail, who was running AT&T back at the turn of the 20th century. In 1908, in the shareholders letter, he makes a point that his service gets better as more people use it. So that’s over 100 years ago. We’ve known about them for a long time, but why do you think that it’s only recently that we’ve started to articulate them in the way that you do?

JAMES CURRIER: Because it’s now in software, and so it’s easy for us to see them, measure them, manipulate them, change them, demonstrate them. The time that we really started talking about them was in 1998, when the Department of Justice was looking at breaking up Microsoft, and they knew that Microsoft was suppressing competition, but they didn’t understand how they were doing it. And as they investigated the situation, they started calling on academicians to say, “What’s going on here? Why is this such a problem?” And the academicians were saying, “Well, they’re network effects.” And they say, “What are network effects?” So then the academicians realized there was a market for them to start studying this and start publishing about it. So we had about a three-year period where, because of this, there were people researching and studying it and publishing about it.

AZEEM AZHAR: As you rightly point out, the Microsoft case started to hinge on the power, the network effect of their operating system. You had software developers making software, which made the platform more attractive for customers, so more customers came, which created a bigger market for every new software developer.

JAMES CURRIER: That’s exactly right. And we call that a platform network effect. And some people try to use the word platform to describe all businesses with network effects, and I think that that’s misleading. Well, we have a two-sided marketplace network effect and something like a Monster, or an eBay, or a Craigslist. And then we have what we call a two-sided platform network effect, where it’s basically an operating system. So iOS would fall into that category. And essentially have developers or people or businesses building their business on top of your operating system. So they’re dependent on you, and then your customers are dependent on you in order to get to the app they want. And there are different playbooks for that network effect and for marketplaces, although they’re both two sided generally.

AZEEM AZHAR: I’m quite curious about the study of this subject. These things were literally live, they were in vivo out there in the wild growing, but we weren’t looking at them in any structured sense.

JAMES CURRIER: No, and we weren’t because the mechanism to make it easy for many entrepreneurs to build businesses with network effects weren’t there until ’94, when we finally got the internet. Once we got Mosaic browsers on top of TCPIP, that’s when it became accessible enough so that people would choose to build businesses there, rather than build another shoe manufacturer, another concrete manufacturer, another investment bank.

AZEEM AZHAR: And of course, up until that point, all of the networks that we had were extremely expensive and often politically difficult to get into. It was building a gas supply network, or a water supply network, or a sewage network, or a telephone network. I mean, you needed to have deep pockets and great political connections to be able to do it.

JAMES CURRIER: Your cellular network starting in ’83, your cable network starting in ‘7–we call those physical direct network effects, and they are very pernicious.

AZEEM AZHAR: So the NPS score, the net promoter score, is a very good barometer of whether customers love a company or not, or how well they’re being treated. And Comcast has a network effect, has a terrible NPS score. But if you look at companies with these modern digital network effects, like the Apples of this world, I think they have really high NPS scores. Why does one type of network effect company give us a terrible set of customer services and yet the other one super satisfies us?

JAMES CURRIER: Well, I think there’s going to be a bell curve on this actually. I don’t know that there’s a huge correlation between whether you have a network effect and what your NPS score is. I think some people would say Facebook works really well, and others, I think would say that they’re doing a horrible job of managing the world. So you might have a bifurcation of NPS scores on that particular product as well.

AZEEM AZHAR: My run in with network effects was with the fax machine. So about ’86, ’87, my parents got a fax machine and it didn’t do much for the first three months. And then there was a moment, I came back after being at school on the next vacation, and the fax machine was always busy. Do we still see takeoff points like that with networks?

JAMES CURRIER: We do. Absolutely. I mean, in 2001, you saw a takeoff of matchmaking sites—


JAMES CURRIER: Which are an interesting marketplace, where every seller is a buyer and every buyer is a seller. But you had online matchmaking sites for probably four years prior to that time, but there wasn’t digital photography. So as soon as the digital photographs started being easier to upload to the internet, you had this acceleration in 2001, mid-2002 for the matchmaking sites. And now 40 or 50 percent of people who get married in the Western world find each other on these matchmaking sites.

AZEEM AZHAR: Is there a commonality between those takeoff moments across different platforms, products, or industries?

JAMES CURRIER: Yes and no. I mean, you can point to lots of different attributes. So for instance, in a network like a Clubhouse, which is this online audio, the audio app, there was a takeoff moment at about 500 high status people when suddenly the next 2,000 people wanted to join in with those cool 500 people. And then another 200,000 and then another eight million wanted to join in with that group. And you could feel that moment when it tipped. That has to do more with status than size. You will see a tipping moment in a B2B marketplace when the second largest player in a market starts using a marketplace, then it’s really safe for number three through 100 to now join.


JAMES CURRIER: And so sometimes status or liquidity in a market–B2B marketplace, is more about liquidity and trust or safety or compliance. So in every different network effect type of business, you see tipping points generated by different elements. But every company that wants to build a network effect needs to move from that chicken or egg problem to a tipping point. And we call those bonding curves, where you’re trying bond enough people, nodes, buyers, sellers to the network until you reach this liquidity point where you hit that tipping point and off you go. And the science of how you do that is just beginning to take place and we’re going to see a real emergence of that study over the next 10 years.

AZEEM AZHAR: It’s not simply a new way of doing business, it’s a new way of analyzing problems. And we can pick up some of the science that you’ve alluded to. What is a network? And a network is made up of people or entities that we might call nodes, and the relationships between them that we might call edges. And I guess the question that one might have is the configuration of those nodes and edges, how important is that to understand what’s going on in a network, what will make it successful? Now, a simple person like me might say, don’t you just need sort of a critical mass of the thing that’s most wanted? So in the empty disco, the critical mass is young ,cool people because then all the oldies will show up. And that gets its started. But it sounds to me when you start talking to me about bonding curves and network topologies that it’s more sophisticated, more nuanced than that.

JAMES CURRIER: It can be. It hasn’t been in the past. We have been operating these businesses in a quite unsophisticated way. And if you were building a network effect business in the ’90s, like Monster was, you just did it, you didn’t even know what you were doing.


JAMES CURRIER: And I would say the same thing was true about Facebook, he just copied Friendster and copied the idea of using real names from the Winklevoss brothers and he just did it. But I think going forward, if you want to compete in these spaces, you’re going to have to get more sophisticated ,so you can get there faster. Because whoever gets there first, typically with the tipping point ends up owning that network effect. It’s very hard to compete with you once you get there.

AZEEM AZHAR: We look at the companies that have network effects at their heart, and they are the biggest companies in the world really. There’s Alphabet and Facebook, two multi-trillion-dollar companies. Trillions of dollars of value later, we still seem to be figuring out the tools, figuring out the know-how, and at the start of this journey.

JAMES CURRIER: We are. And we went through this earlier, when we finally figured out how to get things viral back in sort of 2000, 2001, after doing a lot of AB testing and learning how to do AB testing. By 2005, we looked at ourselves and said, “Look, everyone’s going to understand how to make things go viral in the next year.” And still people aren’t very good at making things go viral. I think it’s just because it’s complicated, and the combination of things you need. And I see the same thing sort of happening with network effects. We’ve been explaining it now for years, and it’s still incredible to me how few people understand it. It seems to be hard for people to get, despite the fact that there’s so much wealth that springs from it, that the prize for understanding it is so massive, it still seems to be tough for folks.

AZEEM AZHAR: NFX, which is your investment firm, specializes just on businesses that have network effects in them. But it’s a very wide set of companies. And I think quite surprising. Will you give us a flavor of the diversity in this ecosystem?

JAMES CURRIER: Out of the last 100 companies that we’ve invested in about 35 were marketplaces. You also then have things like Radical where it’s basically GitHub for crypto, where you have open-source developers who are all working together on a project to then also build a network on top of their network effect of the open-source project. We’ve got companies that are the networks between people buying and selling, but this is really a SaaS product like [unintelligible], which is providing payment solutions for Fortune 500 companies. They will have, let’s say, 100,000 small companies they pay and they have no way of doing that efficiently. And the small companies have no way of receiving that money efficiently—it takes 90 days, 180 days. So you’ve got so many pieces of software that are revealing the networks that exist already in various configurations and the diversity of opportunities to build these network effects is big. I’ll give you another example of network effects people don’t think of, but they’re tribal network effects.


JAMES CURRIER: So you can even look at PSG [Paris Saint-Germain, a professional football club based in Paris, France], which brought [Lionel] Messi on, paying Messi partially with these fan coins, which they just made up last year and people decide they want to own them. But if you’re PSG as a company, you are a tribal network effect business. The more fans you have of PSG, the more fun it is for the other PSG fans. The more fans you have, the more a Messi wants to bond to your network. And once he’s there, you get more fans, and then the coin gets more valuable. The more valuable the coin is, you’re making money so you can buy your beers while you watch PSG. So even a soccer team can be a network effect if you see it as a tribal network effect, and then you manage it as such. And now that we have fan tokens for PSG, we can now start to see the reality of the network effects around these businesses, which we couldn’t see before.

AZEEM AZHAR: As someone who fell down the whole of the internet in about 1991, what you describe is exactly what I felt when I got on the internet, which is wait, we’re all part of this thing where we can read and we can write, we can participate, we cannot participate, we can earn status. And the thing that’s fascinating to me about your example of PSG is that not only is there a network effect, which drives more Messis as well, there’s the ability for PSG to analyze its network structure, and they can start to be scientific about how they segment what you used to call the customer base, but you now call the fan network.

JAMES CURRIER: And we’re gonna see more and more of that. And I’ll give you an example. Imagine if you have a group of particularly attractive PSG fans that like to come and wear scantily-clad clothes on the sidelines. Being able to show photos of that to fans will get them more energized about buying tickets. Should PSG be paying those fans to come? Can they start to measure the impact of those particular fans and their oversized impact on the value of the network to others, just as they would evaluate Messi’s oversized impact on his value to the network to others? So you’re going to start evaluating every node of the network, and those nodes are going to wake up just like Messi has and say, “Well, I want compensation for the value I’m adding to your network effect.”

AZEEM AZHAR: Marketers have for a long time had this idea of the CLV, the customer lifetime value, how much is a customer worth to me based on their ongoing business. There’s also the idea of the CRV, the customer referral value. So can we start to segment our customers and look at which ones refer people more? But the thing that strikes me about referral values is that they’re really hard to measure. The second is that, they’re often very faint. They’re kind of soft referrals, so it’s hard to attribute what really happens. And the last thing is that the referral value seems not to follow the normal distribution, but rather follows a power law distribution. And we don’t need to run complex network maths to say, “Hey, it’s the 10 really kind of well-presented guys and girls at the front who are driving all the excitement,” right? We can just see that. So to what extent can we just let the VIP hosting team figure it out?

JAMES CURRIER: Today we’re managing with approximations and like with everything it’s just going to get more and more accurate. And today nodes on your network don’t ask for fan tokens, but in the future, they probably will. That awareness that they are adding value to the network doesn’t exist yet in many cases. They just think, “Oh, I’m getting VIP.” They don’t think of an actual transaction. So I think that is going to transform how we think about it.

AZEEM AZHAR: If we think about something like a Google, what’s very interesting is that of all the value that Google has created, and it’s created a ton of societal value, but of all of the corporate value that it’s created has been held by its founders and the employees. And I suppose your suggestion is that, “well, there should be a point at which the rest of us who contributed to its success, whether it was webmaster or Uber drivers on Uber, should also benefit.” But I guess the argument that the companies would come back to is say, “Well, you do benefit, right? Because you get the better service, you’re getting more customers to your barber shop, why do you need a piece of the pie?”

JAMES CURRIER: And they may, it may or may not change. If you can provide enough value that people are willing to do it for love, or for status, or for something that doesn’t cost the company any money, then that’s a better company. You’re going to have a higher margin. But in many cases I think going forward, there’s just going to be a competition between companies, as you’re seeing in the crypto space, to try to get developers to develop on their level-one coins, level-one tokens, right, level-one protocols. And I think those level-one protocols are going to start paying the engineers more and more to start to aggregate enough apps on top of their level one so that other people want to use it and it gets the flywheel going. And so I think in some markets you’re going to see enough competition where they will need to pay.

AZEEM AZHAR: How real is the defensibility of the network effect? I mean, Google has not been knocked off at search perch, even with an amazing company like Microsoft chucking billions of dollars at it. Are they really that resilient?

JAMES CURRIER: Yeah. They can absolutely be. And a good example of that would be a company like Craigslist, which is still massive and still growing, even though they haven’t changed the product since 2001. Look at Microsoft, they were down to $200 billion or value or something. They’ve been able to come back because their operating system was so sticky that they could just wait, they could be fallow for 10, 12 years and do nothing and be mismanaged. But then when they decided to get their act together—boom, they’re back to a trillion dollars.

AZEEM AZHAR: Microsoft took two kidney punches, right? They missed the internet and then they missed the smartphone. And now they’re back. But then I’m curious about this, which is that IBM also had some form of network effect, through both its PC business and its mainframe business. But it wasn’t able to really survive and respond to missing a couple of technology waves. What do you think was the distinction?

JAMES CURRIER: I think there was just big underlying technology shifts.


JAMES CURRIER: So you see the same thing with AT&T, they are no longer in the top 20 market capped companies, even though they’ve got this wonderful network effect because the internet and Datacom has sort of obliterated the voice technology that they were built on initially, and they’ve been supplanted by underlying technology shifts, where other network effect businesses have become more attractive, and so relatively their business diminishes. And so I don’t think that they didn’t have a network effect, it’s just that they’ve got a network effect on the technology that’s old and now has a smaller part of the market.

AZEEM AZHAR: You talked about two different classes of network effect so far: the marketplace company where a sense you are matching buyers to sellers and the other model, where I’m a platform marketplace. In other words, I want to bolt on some kind of operability. One of the things that strikes me about the former, the exception really being Craigslist, is that these marketplace companies find themselves, I call it thickening. They start to offer more services, verification, trust, insurance, they then might require real names. Then some of them offer financial product. I sometimes felt it got to the limit with a company like Opendoor, which was a real term marketplace matching buyers and sellers, and they realized that order for there to be liquidity, they better get sellers to sell their houses really quickly so they would buy the house themselves, give it a lick of paint, and then shop it off to a buyer. And that seems quite far away from the marketplace that is Craigslist. How are they both the same beast? Is that latter one, does it still retain the essential characteristics of a marketplace?

JAMES CURRIER: With all of these things we have to realize that we’re painting with different colors is the way I describe it. Yes, we’re painting with red, but there’s lots of different kinds of red. So these are flavors of these marketplaces, and an Opendoor needs a big balance sheet, and Zillow didn’t because they were just a listings marketplace. Now Zillow has decided to get into the business and compete with Opendoor. And their market cap has gone down quite a bit since they did so because it’s also varying in a different type of marketplace. It’s an asset heavy type where you actually take ownership of the assets. So some are better, some are worse, some are growing faster, some are slower. And so there are seasons to each of these companies. And one of the things that these marketplaces face is network pollution.


JAMES CURRIER: One of the reasons they thicken, as you say, is because of what we call network pollution, where people are figuring out the game, how can I tweak it here and take advantage of that? So then you have to close that door in order to protect the integrity of the marketplace. So it’s a culinary spots, it’s an evolving organism, we see this regularly.

AZEEM AZHAR: It’s a very different way of doing business. It’s not traditional business school—goods come in, go to manufacturing, add some value, get packaged, go to distribution, go to marketing and sales, and out it goes, a linear value chain. And those are the skills that you understand, and you compete for supply against another competitor. Is the world of business waking up in time for this?

JAMES CURRIER: Well, I remember in 1994, going to the Internet World Conference, which was 500 of us in New York city, I’m like, “Oh, this is going to be a big thing, but it is just getting going.” And we’re kind of at that point now with network effects, where we’re just starting to understand that we’re seeing them.

AZEEM AZHAR: So let’s talk about a few of the skills that you need. One of the things that I think that seems to be important is figuring out how to get what you call liquidity, right? How to get enough of the thing that one side wants into the hands of the side that wants it, right?

JAMES CURRIER: So when we’re talking about liquidity, we’re typically talking about marketplaces. And there’s a whole sort of list of techniques that we’ve seen or seen in the practitioner’s word about how you get to that point of liquidity in these marketplaces. And it’s everything from limiting the time. There was a great company called Tophatter that said, “Look, we don’t have anybody in our marketplace, so the marketplace on our mobile app is only going to be open between 8:00 and 9:00 PM Eastern Standard Time.” And so the 300 people who had downloaded the app all came on between 8:00 and 9:00, they all had a great time together seeing each other in real time, bidding on an objects. And there seemed to be energy for that hour and then it would stop.


JAMES CURRIER: And then after they got 1,000 people, then they opened it up to an hour and a half. And then they… So that’s one technique. Or paying one side. So Uber will pay a driver a salary of $5,000 a month to just drive around and be available in a new city where they go in. And that helps to create liquidity on the supply side.


JAMES CURRIER: So there’s lots of different techniques that one can use.

AZEEM AZHAR: And you see founders go from one of these companies to the next and to the next, taking the knowledge with them because it hasn’t yet been codified. And then you wonder why is it that the second or third generation e-scooter business grows so much faster than the first generation, which itself grew much faster than Uber did, and Uber was the fastest growing company of all time when it was around. And it’s because they’re taking that knowledge with them and they’re building on it.

JAMES CURRIER: We’re getting more and more into these exponential curves that we’re all looking for.

AZEEM AZHAR: And one of the more interesting technologies that embeds network effects is the token economy—things that we can build on, blockchains in some way. I know this is an area that you have been exploring. I’m not sure if you’ve yet invested in, in any of them. But could you explain why that technology helps in the creation of network-based businesses?

JAMES CURRIER: So the short answer is because it allows us to not only see that we are a node and we have links, but we have ownership and we can see the benefit to ourselves for participating. My history with digital currencies and tokens goes all the way back to the ’90s when I was interested in cyber cash and e-cash. And then like was on the board of Second Life in the 2000s, and we had Linden dollars and 760 million GDP. And we were exchanging with British pounds and U.S. dollars, and we were actually thinking then about launching a global currency called blue.com. And so we bought the URL blue.com. We were going to create the world’s global currency called blue, and we’re going to launch it and base it off of Linden dollar. We had all these plans and then Bitcoin comes up. So we were very early on in doing that. And so we’re big investors in crypto, we’re big backers of Bitcoin and Ethereum and the whole thing. But by having these tokens, you make visible the appreciation and benefit that people get. And therefore it’s another color to play with. I can use my shares to pay my employees, but now I have tokens to pay the other nodes in my network. And maybe my employees are going to end up wanting more tokens than they want shares.


JAMES CURRIER: So I think that’s why you’re going to see the inclusion of tokens in a large number of businesses over the next 10 years. We have in venture capital GPs who are focused on crypto, crypto investors. We have crypto funds just like in ‘98, you had internet funds—

AZEEM AZHAR: Internet funds, right.

JAMES CURRIER: But by 2003 or four, everything was internet, and the same thing’s going to happen with crypto.

AZEEM AZHAR: So that is a bold assertion. And maybe what we should do is take that assertion and turn it into something that feels tangible for or us. So let’s take an example of ride hailing. Uber has created tens of billions of dollars of value, mostly for its investors and its founders and employees. And a very, very small amount of value for its drivers, who on average might make a few tens of thousands of dollars a year for working 13 hours a day. What would, let’s call it “Cuber,” Courier’s Uber, based on a token look like? How would that company operate, how would the tokens work, and why would that strengthen the network effect?

JAMES CURRIER: So the theory would be, and we’ve been looking for someone to do this for a bunch of years, the theory would be that you could eliminate Uber’s 30 percent take rate and give it back to the riders and the drivers if you could have a more decentralized token-based approach to this. The branding, the embeddedness, the network effects that Uber and Lyft do have are making it more difficult, but certainly not impossible. Because if I can get 500 drivers to multi-tenant on Uber and on the Cuber network, then I can provide a pretty good service there in London or in San Francisco.


JAMES CURRIER: So on the supply side, you should be able to hack into a few hundred with a good story. And then you would drop the transaction fee tom let’s say five percent, because five percent would be required for insurance and monitoring and other sorts of things that do come up—and maybe more. It might need to be 10 or 12 percent, but you could operate it at a near breakeven almost nonprofit. And then by service, meaning a demand side, if I decide to use Cuber all the time, I might start to accumulate tokens as well, and then there might be some sort of a market for those tokens where someone would buy them in an attempt to get a ride faster.


JAMES CURRIER: Or if I want to take an hour-long ride back to my home because I could pay some of these tokens to make sure that I get preference for that ride to take me home, so that’s convenience, so I make it home for dinner. So you could make a market for these tokens.

AZEEM AZHAR: You’d make a market for the tokens, but the tokens would need to be exchangeable in the real world, right? So you’d need to have a token dollar, a token Sterling exchange rate. But then of course, you’ve got this issue, which is that as a driver, I don’t necessarily want to take currency risk because the dollar may go somewhere. At the end of the day, I need to buy broccoli, and yogurt, and eggs for my family. So then someone has to come in and say, “You know what, Mr. Driver, I will guarantee you your exchange rate. I will provide you a hedge.” So now the Cuber network has actually got market makers involved who are saying, “This is how we make this thing healthy. We take out some of the risk. We say, “Listen, drivers don’t want to take the exchange rate risk, they want to know what they’re earning.” And so you’ve constructed another complimentary business on the network.

JAMES CURRIER: Correct. And when a new network comes to try to copy you and they don’t have those market makers, the value to the Uber driver will be higher with the Cuber network because we were the first ones to have these market makers to stabilize the currency.


JAMES CURRIER: And so this is a great description of how a network can build all sorts of different nodes that enhance the value to every other node on the network.

AZEEM AZHAR: So it sounds great because 30 percent take rate is a lot to be paying as Apple, as Discovery. And it would sound great that the people who do the work, the drivers, should be able to participate in being part of the network that is creating this social welfare. But the network still needs to be coded. Someone has to write the software. Someone has to pay the insurance, and that’s hard, right?

JAMES CURRIER: Someone has to vet the drivers to make sure that they’re not going to rape and murder people.

AZEEM AZHAR: So there still needs to be—I’m going to use my word, thick again, right? A thickness to the marketplace. There needs to be some central organization. So in this token economics world, what is the role of that central organization? Or is it a company? Is it a partnership? Is it just people hanging out in a Slack channel?

JAMES CURRIER: Well, this is one of the reasons we haven’t seen a distributed or a defi competitor to Uber and Lyft. Because there are real humans, there’s water moving around in those cars, and that water needs to be taken care of. And if most of the things like Ethereum are run by a DAO—

AZEEM AZHAR: And of course DAO’s are decentralized autonomous organizations.

JAMES CURRIER: Yeah. And that organization lives in the ether. And because Ethereum is only in the ether that works. Okay. It doesn’t work so well for Bitcoin. They’re a little less functional, but because they don’t have a figurehead, they don’t have a king, [unintelligible] sort of giving everyone direction the way the queen of England does. And so we haven’t seen a real attack on Uber through this mechanism because of what you’re talking about. If someone does get murdered, which is inevitable, someone does get raped, which is inevitable—these are billions of human beings, these things happen—who does the government then go to, to hold them accountable to make sure it doesn’t happen again so that the public feels safe to use Cuber?


JAMES CURRIER: And so these are difficult issues and they need to be addressed. And so it might not be even 10 percent, it might need to be 15 percent to even cover your cost. And then the question is how much more than 10 or 15 percent does it need to be in order to incentivize people who are talented enough to pull it off to do this rather than doing another project in the crypto area that’s in the consumer area? We haven’t seen the team to pull this off that has really dedicated their lives to do it, who are talented enough to pull it off.

AZEEM AZHAR: And as we start to understand how to do this, as the software becomes cheaper, then teams that have different talents can then go after the sort of the Uber and the Lyft model in the way that you propose.

JAMES CURRIER: That’s right. And we can look to the movie industry as a historical example of what will happen. So when we first had movie technology in the ’20s, you had entrepreneurs doing it. And then you got these studios, which had the knowledge about how to make movies—the technology, the audio, the distribution of the movie houses. And the studio model dominated until the ’60s, when enough people had been on enough great movies. So all the people who worked on Casablanca, all the people who worked on Gone with The Wind—they had seen what it looks like when you make a great movie, when you had enough of this operational expertise out there. Then you move to a world where the agents ran Hollywood because they could pull together a group of 12 and they could pretty reliably make a great movie. And we are getting to that point in tech, on business SaaS. And these network effect businesses are still not well understood, but they hopefully will be in the next 15 or 20 years. And we’ll move to more of an agent model where you can bring together a group of 12 people and they can attack a problem like this and solve a problem that is a little more intractable, like competing with Uber.

AZEEM AZHAR: And one of my favorite examples in the economy is a thing called Helium. So Helium is trying to build up a nationwide datacoms network. But instead of having to get regulated spectrum and put 5G towers all over the place, they’ll say to you or I, “Listen, give us a bit of your WIFI router and when people use your network, you’ll get paid in HNTs, helium network tokens.” And what’s kind of interesting about this is that Helium could build a nationwide datacoms network on WIFI without having to spend billions of dollars and having the participation of lots and lots of people who sit at the edges. I don’t know how well they’re doing, but the theory to me is appealing.

JAMES CURRIER: It’s very appealing and it’s just an exact replication of what Uber and Airbnb have been doing. I think the challenge for Helium is that a lot of people know how to lock a door and clean a rug, and a lot of people can drive a car at the speed limit and deliver you without killing you, but fewer people understand how to add software to their wifi and then open up their router. And so your density is harder, it’s harder to achieve the density you need.

AZEEM AZHAR: We also have a second issue, which all of these businesses will face, which is the one about the talent and the team and how they operate. Most people have been used to having an employer and being employed. So what’s the relationship between A, a world that might be dominated by token economies in the way that you said the world’s now dominated by internet economies or just somebody who wants to be able to work? Are they going to be able to get a salary by working for the core entity, or will they just have to accept this mishmash of sort of cryptocurrencies coming one way or another?

JAMES CURRIER: No, they’ll absolutely be able to pick a job where they’re working for a central authority and it’s very simplified. And we’re just going to increasingly simplify the interfaces to the crypto world. So when I go on my bank and I want to pay my bills, it’s pretty clear to me how to do it.


JAMES CURRIER: Three years ago, the banks couldn’t figure out how to make the interface simple enough for us to feel comfortable doing it, now it kind of works. Same thing will happen with crypto. You’ll have your wallet, it’ll be laid out with the right font. You’ll understand, it’ll all be down nominated in your pounds, or your dollars, or whatever it is that you’re comfortable using. And you can click a very clear button about what you want to look at and you’ll be able to navigate. And so we’re just going to keep including more and more people by reducing the complexity. I think the big picture here is that it does appear that we are at the inflection point in the explosion of money, because money is just a collective trance about confidence in the future.


JAMES CURRIER: If you have confidence in the future, money just kind of appears in your society. You go to a community where people don’t have confidence in the future, money just disappears. No matter how much money you give them, you give them as many Bitcoin, or dollars, or pounds as you want, and it’s just gonna disappear. It’ll disappear.


JAMES CURRIER: Because there’s no mental model, there’s no constructed trust in the future. And so what we’re seeing with all these networks, with the sort of growing number of things that have value—whether it’s a PSG coin, or whether it’s a board ape or whatever—all these things are now ownable and have value to a network that can see them and you can get into micro-collective trances, right?


JAMES CURRIER: So crypto punks. There’s a micro-collective trance that these things are worth a lot of money. How many people do you need to believe in them for them to maintain their value? I don’t know. You don’t need that big a fraction of the overall network in order to maintain more and more value. So we are going to see an explosion of money value over the next 100 years that I believe will lead us to a post capital world where the algorithms will allocate capital to the places where it’s needed, and we won’t talk so much about money anymore because it’ll just be there because money’s just distilled collective confidence in the future. And so I think we’re going there, I think it’s hard for most people who are working day-to-day to see, but when you look at it from a network perspective, this is inevitable.

AZEEM AZHAR: Listeners, James just put on his seven league boots there and took some very big strides into the future. I think there’s something really interesting in this idea that we can automate large parts of these underlying processes. Because underpinning the economy is essentially a large number of discovery systems that are figuring themselves out. But I don’t disagree with this idea that a lot of what we call the economy is effectively price discovery and resource allocation. And crypto seems to build more on top of that. But I guess the thing that I’m struck with is that there has been a challenge in networks where they start to fail, they start to have difficult, unintended consequences. They get big, and the pollution arises. And there’s a joke at the moment, which is that Facebook’s key skill is the ability to scale any network to the point at which it becomes unmanageable and horribly polluted—whether it’s Facebook marketplace or, you know, Facebook itself or Instagram. There’s another idea which is the tragedy of the commons. And we know that the oceans are suffering from the fact that no one takes responsibility for them. So in a network that’s highly governed, Facebook, you get this problem of growing pollution. In networks that are totally ungoverned, the oceans. In a world that you see where everything becomes networks, what does the governance look like? How do we avoid either the tragedy of the commons or the tragedy that is a company in East Palo Alto?

JAMES CURRIER: So that is a big question for the future of humanity going forward for the next 30 years. It’s such a big question, I don’t want to pretend I have the answers. But here’s some of the directions that I would point us in. And that is for us to become better practitioners at pollution management features. So it is the case that Facebook, if they had a particular mindset about reducing pollution, trading off growth for pollution reduction, they could actually be reducing their pollution quite effectively. It’s just a matter of AB testing different interfaces, different language, different modalities, different flagging systems, giving people different ratings and blah, blah, blah.

JAMES CURRIER: There’s lots of ways that we could go in there, the problem Facebook is having is that they’re culturally aligned perfectly on growth. There’s only one thing that matters and it’s more growth because Mark wants to connect the entire world. And he is only at three billion and there’s five more billion to go. And so until he gets all eight billion, he probably won’t switch his mindset to balance growth with pollution. And that’s the problem we’re having right now, is we’re still in the early innings of us all being able to see each other.


JAMES CURRIER: The second problem we’re having is that the pollution in the network is such because the interface is drawing the pollution out of our brains and putting it out there. We all have pollution in our brains. It’s just that these new interfaces are begging for that stuff to come out and then that stuff gets elevated because it gets more clicks, because it creates more emotional response, and that drives growth. So it’s that top-line metric, which I think is causing the pollution in a Facebook type culture and it’s not necessary. If you had a Reed Hastings running that company, you would see a very different result. You would still probably have a dominant network effect, but instead of being worth a trillion, it would be worth $600 billion, but it would be preparing to be worth four trillion in 2060 with a view toward the long term, which I don’t think Zuckerberg has yet evolved too. But hopefully he will. I mean, he’s young still, and so hopefully he will.

AZEEM AZHAR: Governance aside, networks end up having monopoly like characteristics within their sector. Google is number one at search, who is number three? Facebook is number one at what Facebook does, who is number three or four? And that in a sense might be problematic. And certainly in my book, what I argue is that the issue is not monopoly in the way that perhaps we thought about it with Standard Oil. It’s not about rising prices to consumers. It’s about the fact that increasingly our lived experience is intermediated by these technology platforms, whether it is finding a date or finding the news. And those platforms have their own incentive and they crowd out alternatives. So these things that are part of being a human are now controlled by a profit-seeking company that has dominant control within its thin slice of the market.

JAMES CURRIER: I will point out though that we are only in this situation because [Mark] Zuckerberg and Sergey [Brin] and Larry [Page] were so talented.


JAMES CURRIER: In other markets where you have network effects, you don’t always see a winner takes all. This is a common trope that as people are approaching the understanding network effects, they say, “Oh, it’s a winner takes all thing.” But in fact, in most markets, it’s not a winner takes all. It’s literally the fact that Google just out executed everyone so badly, and Facebook did as well. I think Facebook was like the 150th social network with the same five features. It’s just that they really did a great job. I don’t think we should live in fear that because of network effects, everything becomes a winner takes all market and we’re always going to face these problems.


JAMES CURRIER: We’re literally just in this situation right now, and maybe 20 years from now, we just won’t be. But you’re right, these organisms need to start to think about their stewardship of the people on the edges of their network, which is the users. And they don’t do it well yet, and they’re trying to. But it’s against their monetary and growth interests, and so they’re hesitant. So it’s going slower than we would all hope—much, much slower actually.

AZEEM AZHAR: What instigates that change and that change in value?

JAMES CURRIER: Three ways, one, competition, if we could actually figure out if there could be a team that could compete effectively with an approach. Crypto might provide that window to open that up. There might be some other technology shift that would allow us to make a change there. Number two, collective action. We can all create sort of a union to demand changes to the platform and get enough people involved so that Facebook realizes that their market cap will go down further if they don’t make changes. Applying pressure from below. And then there’s of course, siccing the hierarchies onto the networks. When we look at China’s recent actions against the network effect business, they figured out that these network effect businesses are actual threats to the government hierarchical power because they’re coming at it with a whole different approach, which is a network bottoms up approach. So the hierarchy is defending itself against the network structure, and we always say that the SEC in the United States, the Securities Exchange Commission, they come with briefcases and ties, and the FTC, the Federal Trade Commission, they come with the guns. So you’d prefer to have the SEC knocking on your door, not the FTC. But often it’s the FTC that will knock on Facebook’s door and say, “Okay, it’s time for a change. And the hierarchy is going to impose itself on your network, and here’s how it’s going to go.” Now, the problem with that, the hierarchies don’t understand the networks and they will get it wrong. They don’t have the tools, they don’t have the understanding to actually implement the changes that would actually work in the new world, which is dominated by networks—not by hierarchies.

AZEEM AZHAR: I suppose the challenge is about understanding what is the lesser of the two evils? And we may not like the governance mechanism, the lack of democratic accountability and so on in China. But you can’t fault the fact that the authorities articulated that there was a problem, and they’ve articulated a set of interventions that will be effective. The unwinding of Ant Financial in the way that they are doing, and its data will attenuate whatever risk Ant Financial was going to pose to the system more broadly. So in the U.S., you have Gary Gensler at the SEC, and he’s just put out a statement when we recorded this discussion that I thought was quite sensible, around how he was thinking about cryptocurrencies and the risks to the consumer. And then we have Lina Khan, who is running the FTC and three years ago, wrote a very, very insightful post on the way in which Amazon affects markets negatively and sort of escapes the traditional antitrust lens. So when you look at her and what she might do with the FTC, do you think there’s somebody who understands the network nature of this?

JAMES CURRIER: She does. She understands it, however, from a hierarchical perspective and she wants to protect the hierarchy. She believes in the hierarchy. She grew up in the hierarchy. She understands the hierarchy. She can navigate well in the hierarchy. She wants the hierarchy to be the way it works. And I would argue that what we need is someone who not only understands networks, but understands how networks will be the future, how networks will have winners and losers, but they will be a lot more winners with networks than we will if we just keep it hierarchy. Now, the challenge we have is the hierarchies are easier to understand


JAMES CURRIER: So if you want to explain something to a mob, you’re going to end up with a hierarchy every time, as opposed to understanding the sort of dynamic nature of these self-organizing networks.

AZEEM AZHAR: I mean, networks are beautiful. They are also complicated. James, you and NFX have done an incredible job over the past few years in teaching the world about them, and their nuances, and their power, and their beauty, and their ugly underbelly. I do recommend that people search out your blog, and thank you so much for taking the time to talk to me today.

JAMES CURRIER: Oh, it’s been a pleasure.

AZEEM AZHAR: I hope you enjoyed this discussion. As we’ve been back from the break now is a great time to go back to the podcast feed and listen to some of the fantastic discussions we’ve had over the past few years. Please feel free, go and check them out, find your favorites. To become a premium subscriber of my newsletter, go to www.exponentialview.co. To stay in touch, you can follow me on Twitter, in the U.S., that’s @Azeem, A-Z-E-E-M, and elsewhere in the world, it’s @Azeem, A-Z-E-E-M. This podcast was produced by Fred Casella, Marija Gavrilov, and Mischa Frankl-Duval. Bojan Sabioncello is our sound editor.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *