Joe Lubin was once less interested in changing the world than escaping it. Half a decade ago, he was living in Kingston, helping a Jamaican model break into the music business. Having worked on Wall Street, he thought the 2008 financial crisis was only the beginning of a system-wide global collapse to which he foresaw no solution and no alternative. He had considered battening down the hatches in South America or starting up an urban farm in Brooklyn before settling on Jamaica as the place to ride out the coming storm. In those recession years, his present life—as a co-founder of Ethereum, as the globe-trotting CEO of blockchain venture studio ConsenSys, as a cryptocurrency billionaire—was literally unfathomable.

Jamaica was only the latest stop in a varied career. After graduating from Princeton—where he played squash, studied electrical engineering and computer science, and roomed with Mike Novogratz, the sometime billionaire and Wall Street veteran who now runs a crypto merchant bank—Lubin spent years as a researcher and software engineer. Eventually he went from building robots to serving as a technology executive at Goldman Sachs, and later founded a hedge fund. By the time Wall Street was being occupied, though, Lubin had seemingly retired from the world of finance.

Then he met Vitalik Buterin. Only weeks before, in November 2013, the 19-year-old had released a white paper outlining his vision for a new blockchain network and cryptocurrency. As described, Ethereum would go beyond bitcoin in its ability to serve as the foundation for a host of other digital tokens, applications, and cryptographically binding agreements called smart contracts.

Lubin, who is in his mid-fifties, wanted to get involved. He found himself joining a small team that had assembled around Buterin to build and promote the technology he had envisioned. In contrast to Buterin and some of the others, Lubin was one of “the adults in the room,” according to fellow co-founder Anthony Di Iorio. Prior to the 2014 crowdsale, which raised $18.4 million for the project, Di Iorio says he and Lubin put up 95 percent of the funding for Ethereum.

It wasn’t long before philosophical differences divided the team. Buterin and a few others wanted to make Ethereum a nonprofit entity; Di Iorio and Lubin imagined a for-profit enterprise. Ultimately, Buterin decided to set up a Swiss foundation to oversee further development of the protocol, and in 2014, Lubin left to found ConsenSys, a company that would incubate crypto startups and help them raise capital and grow.

Since that initial conception, ConsenSys itself has not so much grown as metastasized. Fueled in part by the 2017 crypto boom and the concomitant flood of interest from software engineers, businesses, governments, and central banks, the company has launched or invested in a dizzying array of projects and services, from developer tools and decentralized apps to a consulting arm—Microsoft is a client—to ConsenSys Academy, which sells education and training. Many of these projects have since raised their own capital. (Full disclosure: Lubin is one of four co-founders of SingularDTV, BREAKER’s parent company.)

An employee showed me an interactive chart called TMNT—for Traditional Management Nullification Tool—which keeps track of the company's diverse ventures and project teams.

All of which has made Lubin one of the world’s most prominent crypto advocates. If he is not Ethereum’s high priest—that role, by popular consensus, is filled by Buterin—Lubin is nonetheless one of Ethereum’s chief proselytizers, a jet-setting Billy Graham of the decentralized future, preaching the gospel of Web 3.0. Although Lubin is said to be the world’s single largest holder of ether, and although this war chest is surely funding much of ConsenSys, he doesn’t care—or at least affects not to care—about its market price. (The crypto market recently took another nosedive in a year that has been full of them.) “The people who were in the space early,” he told The New Yorker, “were there for philosophical reasons, for political or economic reasons not tied to their personal wealth.” Of all the founders, says Di Iorio, Lubin was “the one guy who was like, ‘This is it. There’s nothing else.'”

He wanted his company to reflect Ethereum’s ethos of decentralization. Sam Cassatt, an early hire and ConsenSys’s chief strategy officer, worked with Lubin to develop a hub-and-spoke model that would permit each of the teams to operate more or less independently. The goal, says Cassatt, is to be “a company that builds companies”—analogous all on its own, albeit in microcosm, to the entire ecosystem of Silicon Valley and Sand Hill Road, where the top tech startups are incubated and the top venture capital firms are located. That storied ecosystem, of course, couldn’t function as a single monolithic company, and so the shaggy, sprawling nature of ConsenSys—the fact that it seems to be a dozen things at once—is seen as a feature, not a bug. To Lubin, even his company’s marketing, human resources, and legal teams are “proto-spokes,” which may one day become businesses in their own right.

ConsenSys today employs more than 1,100 people across 19 physical office locations and 29 countries. Since February, the workforce has doubled. When I paid a visit to the headquarters, housed in an industrial building in Brooklyn’s Bushwick neighborhood, in October, an employee showed me an interactive chart called TMNT—for Traditional Management Nullification Tool—which keeps track of the company’s diverse ventures and project teams. As he zoomed in on individual colored circles, all of them representing various ConsenSys divisions or initiatives, they expanded, revealing a profusion of other, smaller circles, which in turn disclosed yet smaller circles. Wheels within wheels. The employee compared working at ConsenSys to being in college again—constantly boning up on subjects with which he isn’t familiar, from finance to engineering. He meant it as a compliment.

Lubin delivering another keynote

Not everyone admires Lubin’s efforts. The bitcoin maximalist Jimmy Song has hounded him relentlessly over a bet Lubin proposed during a public debate with Song at Consensus, a big crypto conference in New York, six months ago. The nature of their disagreement, says Song, is that Lubin “thinks blockchain is this wonderful thing that will take over everyone’s life, and I think that it’s only useful for bitcoin.” (Onstage, Song dismissed dapps as “magical blockchain dust.”) The bet was supposed to settle, five years from now, which man was right—with the loser paying the winner in the cryptocurrency of his choice.

Lubin and Song appeared on Laura Shin’s podcast Unconfirmed in August to try to finalize their bet, but failed to come to terms. The matter is still unresolved, and Song doesn’t mince words about it. “The guy just does not want to bet. I think he did it all for show, as a way to save face—because he was more or less getting humiliated by me on stage,” says Song. “Yeah, you can be busy, but five months—something like that? It’s getting ridiculous.”

Song regularly emails and tweets at Lubin to remind him. I ask Song why it should matter so much. Last time I checked, reneging on a bet was rarely cause for a prolonged vendetta. Song says he wants Lubin to stand behind the technology he promotes. “The purpose of me pursuing this bet is to make sure that his name is dragged through the mud. If he’s too cowardly to actually bet, then he deserves everything that is coming to him,” he says. “And that’s a black mark on his record that should stay.”

Even Di Iorio, who runs his own crypto innovation hub, Decentral, in Toronto, has doubts about Lubin’s decision to focus exclusively on Ethereum. “I think it’s too early to put all your eggs in one basket,” says Di Iorio, whose company’s digital wallet, Jaxx, supports 85 crypto assets. “I’m not an Ethereum maximalist. I’m not an anything maximalist.”

And then, too, Di Iorio favors a more traditional, top-down approach to running a company, with employees gathered together under one roof. “All in all, I think [Lubin has] done an amazing job of bringing more maturity to Ethereum,” he says, “but I think they’re limited in what they can really achieve, due to the inefficiencies of the organizational structure.”

To Lubin and his lieutenants, though, efficiency is not the highest good. “A dictatorship is the most efficient form of government, but that doesn’t mean it’s the best form of government,” Cassatt says. “Certainly you sacrifice some efficiency [with the ConsenSys model], but efficiency in building the wrong thing is still failure.”

When we meet in Prague, on the last day of Devcon (a big Ethereum developers conference) Lubin is wearing a black ConsenSys T-shirt and sporting salt-and-pepper stubble evidently grown over the course of the four-day conference. Within ConsenSys he earned the nickname “white Morpheus”—after the inspirational figure in The Matrix—perhaps as much for his unflappable calm as for his shaved head. His gaze, as one might expect, is focused and intense, though his speaking voice, flat and uninflected, takes a little getting used to. He has a tendency, when giving a long answer, to get quieter as he goes, tapering off almost into a whisper by the end.

"It looked to me like there would just be a cascading collapse of financial systems."

Among other things, I want to ask Lubin about his vision for Web 3.0, the pros and cons of ConsenSys’s business model, and what drives him to live in perpetual fast-forward. (Characteristically, he is already thinking about 2019, a year in which, he says, he wants to travel less and read more.) While we talk, he munches on a banana and breaks off now and then to make arrangements for other meetings.

You had a successful career in finance before you ever heard the word “Ethereum”—
Did I?

It would seem so. But you had retired from that life, and were managing a singer down in Jamaica. What did you see in crypto that made you want to come out of retirement, so to speak?
Well, there’s no way for me to stop thinking, reading, working 10 or 12 hours a day. Because of my time in the finance space, I came to understand how the world worked a little better [than most], or thought I did. I grew disappointed over time with how monetary systems appeared as though they were moving toward end-of-life, and [were] corrupt and problematically architected. A lot of people were similarly disappointed in 2007, 2008—I think I saw it a couple of years before then—and during that period everybody was “occupying” everything, trying to get the word out, running around with their heads cut off because they were embedded in this crazy context, and it seemed super unhealthy and there wasn’t a clear way to get out of it or to heal it. It looked to me like there would just be a cascading collapse of financial systems.

I think we’re actually still in a slow cascading collapse. Our central bankers have been doing a good job of kicking the can down the road for quite a long time—and a long time ago I wasn’t a fan of that idea. Now I’m a big fan of the idea that they should keep kicking the can down the road because, [ever] since I read Satoshi [Nakamoto]’s white paper, it became clear to me that instead of just telling everybody that things are problematic, we could actually start getting busy building alternative infrastructure. But it’s going to take us a while to build scalable infrastructure—we’re going really fast, but it’s a big endeavor. So hopefully the legacy financial and governmental structures will hold together long enough for us to do all the experimenting and build better systems.

So basically you saw this transformative potential of the technology, and that’s what made you want to get involved?
Yeah. I saw it pretty immediately in the bitcoin white paper. But there was so little we really understood back then. It was [just] about money; we thought bitcoin would be adopted and that would be profound. But there were lots of groups in the world—starting in 2012 and later—who felt that we should use blockchain technology to build everything on this new trust infrastructure. And when I read Vitalik’s white paper, it was clear he’d come up with by far the most scalable and the most sophisticated version. Scalable in the sense that it was something software engineers could understand how to use.

What is Web 3.0, as you conceive it?
It’s exactly like Web 2.0, but instead of centralized servers that control our access to different services, the infrastructure is decentralized, so it’s not necessarily owned and controlled by central actors. It can be protocol-based—open business networks, open platforms—so that we can all participate in it from our decentralized, self-sovereign identity construct. We can plug into these networks, whether they’re music networks or supply-chain networks or healthcare networks, and there isn’t a central company in the middle; there are a lot of people and a lot of companies in different roles, all interoperating to make that ecosystem work and to create value for each other.

ConsenSys is a company that wants to be an ecosystem—or maybe already is an ecosystem. Why did you choose that model, and why do you think it’s superior to the old theory of the firm—or even to the big tech-campus model of Apple and Google?
It wasn’t the intention to build an ecosystem. It was the intention to build a much more decentralized organization. [I had] the thesis that it didn’t make sense to try to build an Apple or a Microsoft or a Google anymore, because access to information and the ability to start your own organization is just so easy [today]. We wanted to create a context where the best and the brightest entrepreneurs could come and feel like there’s no better place to be an entrepreneur, or even a serial entrepreneur—and that’s been reasonably successful, with some people building a few companies and spinning them out already. All of our different projects are empowered to take care of their own interests first, but we also have lots and lots of people who are working to facilitate communication, collaboration, and interoperation. So that’s how you end up getting a ConsenSys ecosystem.

Have you had to make any compromises at ConsenSys to achieve such rapid growth with a decentralized corporate structure?
Compromises? In terms of what? I compromise my sleep schedule…

In terms of efficiency, say. Or like maybe you haven’t had sufficient time to induct every new person into the company culture.
Yeah. Tons of growing pains, sure. We’re trying to get better with our on-boarding process. We’re growing too fast. We felt that was necessary because the alternative, growing too slow, would have been much more problematic for our product teams and our consulting arm, because we’re getting exponential growth in interest and working with people.

It’s not so difficult to sprinkle a bunch of talent across our many different projects, because they’re quite autonomous, almost all of them. So people can be absorbed and understand the local subculture pretty well and be really productive. We just have to work much harder on helping them understand where ConsenSys came from and what it is, philosophically. It keeps evolving. It’s been several different ConsenSyses since the start. For roughly four years we were working to be a big part of opening up an ecosystem, and now there are lots of entities pouring into the ecosystem, and we have to up our game and compete. It no longer is sufficient to show up and do something cool; now we have to do something excellent.

"It does get tiring at the end of 10 hours of meetings in a day, and sometimes I have to stay up till three in the morning to get through what I think are the important emails. But it's pretty much epic day after epic day."

One of your fellow Ethereum co-founders, Anthony Di Iorio, who has his own innovation hub, has embraced what he calls the “multi-coin future,” while you have gone all in on Ethereum. Can you explain to me your outlook?
We’re all in on by far the best decentralized application platform. We’ve also hired lots of people with significant Fabric expertise from our friends at IBM and lots of people from the R3 ecosystem. There are really smart people looking at developing next-generation systems, and there are good projects out there, so if a project came along that was as capable as Ethereum, then we’d like to support it—or jump to it, potentially. [But] Ethereum is orders of magnitude more capable, and bigger, and faster-growing than anything else, so [while] there are good projects out there, there’s nothing on the horizon that’s even remotely viable at this point.

Not EOS or anything like that?
I need to be careful. ConsenSys has built lots of applications, and to take an application and move it to a different platform, you’re basically preserving 95 percent of the work that’s been done. [Changing] just the smart-contract layer is not a heavy lift compared to identifying a niche and finding product-market fit and working with end users to make it all viable. That’s where most of the value is in those kinds of projects. The technology for these platforms is, similarly, [about] five percent of the issue. And there are great technology platforms being built out there. But the real heavy lifting is creating an ecosystem of hundreds of thousands of developers and millions of people who are paying attention.

Speaking of all the apps that ConsenSys has developed or funded, a critic of yours said to me recently, “As far as I know, none of [Lubin’s] companies produce a good or a service that’s actually useful to someone.” How would you respond to that?
Who was that? Who was that critic?

I can’t say.
You can’t say? Why not? You’re a journalist, or something like that?

The UPort identity protocol is a “spoke” in the Consensys “organism.”

Yeah, that falls under protecting one’s sources.
Our Balanc3 accounting system has revenues [and] a bunch of clients—they can’t onboard clients fast enough. We have 300 people working globally in our professional services consulting group, which has brought in millions of dollars. We have a diligence team that security-audits smart contracts; they turn away probably 99 percent of the [available] work, because we can’t onboard people fast enough with that sort of expertise. We’ve executed token launches—consumer utility token launches and one tokenized security launch for a film called Braid, which ended up in the TriBeCa Film Festival.

The token launches, if they’re for our own companies, end up bringing capital into those companies, and if they’re consumer utility tokens, you could consider that revenue. There are a bunch of those that have brought in many tens of millions of dollars. And we also help third parties do token launches. FOAM was a recent successful one. What else do we do? [ConsenSys] Academy sells education on Coursera or in person. Lots going on there. Infura is getting paid for offering services. That may be all of them.

Is ConsenSys profitable?
We’re still in the burn. We’ll probably be that way for a long time, just because we’ll continually invest rather than try to extract profits. If you [were to] consider some of our token launches revenue, then we’d actually be profitable, but a lot of those tokens are things that we’re going to keep frozen for a long time. If we calculated everything, then yes, we’re above zero.

Everything you just named makes ConsenSys seem intimidatingly large and fast-moving, like an Amazon. But someone who works with you recalls you saying, not that you want to take over the world—the goal of an Apple or an Amazon—but that instead, “We’re going to make sure that no one can take over the world.” Can you unpack that for me?
We’re trying to decentralize ourselves. Essentially, various technologies have enabled technologists to pile stacks of value higher than ever before. Those stacks of value [have been] contained within firms. [But] if we have these protocol-based open platforms, these network business models, then ownership will be much more decentralized. We’ll have companies [with] tokenized securities, and we’ll have these networks that offer products and services. [ConsenSys project] Ujo Music will be one of those things—last night they just onboarded their 1,000th artist into their content creator portal.

If all these networks are protocol-based, then ownership is sort of related to the token. It may be a [utility] token in some way. But if they get very popular, then there’ll be more demand for the token, and the tokens can rise in value and basically ownership would be broader. Also, with tokenized securities, hopefully they will be much more accessible to many more people [than traditional securities]. We’re tokenizing real estate as well, enabling fractional ownership—making that asset class much more accessible to anybody.

Somebody who might not be able to get exposed to real estate [today] might be able to buy $100 worth of real estate. By tokenizing real estate, we can create liquid and deep financial markets for [it]. So it becomes more transparent, price discovery is less manipulated, and financial instruments will enable you to do things like going short SoHo and going long Williamsburg in New York City.

So people will still start their own businesses, but now, rather than being complete siloes built from the ground up, like Microsoft, those businesses will exist on this common base layer of infrastructure. What will that change?
The idea is that governance and agency will accrue to many more people, and value will be more distributed. We’ll have much less hierarchy—or little or no hierarchy in many situations, because we’ll have better tools to organize ourselves into flatter [systems].

People can intuitively understand a dictator or a Steve Jobs, even if they don’t like them—the lust for power, the desire to win. That’s an ego trip. If you’re not on an ego trip, what are you on?

We’re at a great moment in human history. As a species, things are complicated and weird but, more than I’ve ever been aware of before, the ethos is about doing the right thing and collaboration and—my least favorite phrase—”Making the world a better place.” It sounds a little bit trite, but actually that’s what a lot of people are thinking about. In the Eighties and the Nineties, people just wanted to get rich as investment bankers.

What are you saying, that Michael Milken wasn’t trying to make the world a better place?
He is now. [laughs] It’s interesting that this technology is so married to that philosophy, that ethos of a smaller world working together. It’s a collaboration technology, and the collaboration comes out of the fact that we have this new trust infrastructure; we can trustfully interact with people we don’t know or who we’re competing with. We can build infrastructure that’s collaborative as sort of a foundational element, instead of siloed, walled-garden systems as in Web 2.0—and in the business world [generally]. To answer your question: That seems important. Why wouldn’t I want to pay attention to that?

That’s a pretty idealistic vision. Even some crypto advocates, like Jimmy Song, still argue that this innovation is primarily about money and replacing the existing financial system. Have you guys come to terms yet on this bet of yours?
I think we’re pretty close. I think we’re supposed to get on Laura Shin’s podcast…

What’s the sticking point?
I think Laura would be really upset if I told you what the sticking point was. I think we’re supposed to get back on [her podcast] and talk through the final terms pretty soon.

What do you say to the accusation that ConsenSys, by launching and funding so many token projects, is essentially printing money for itself?
That might end up being a legit mechanism in the evolving, natively digital economy. Companies currently print money for themselves: they issue shares and trade those shares to people for money. It’s quite similar, actually. Tokenized securities will be available to do that sort of thing. [As for] consumer utility tokens—if you stood up a centralized system on the internet [today], let’s say for a music industry platform, and you sold a token to consumers enabling those consumers to trade those tokens for songs, you’d kind of be manufacturing money, right? You’d be creating a means of exchange. The difference between doing that in the legacy economy and doing that with decentralized infrastructure is that you can trust the tokens more; there’s no ability to counterfeit or double-spend those things. So it’s a better infrastructure for building that sort of business model.

How about legal concerns around these token launches?
We have a very deep legal team that works with some of the top law firms in the world. We also work with legislators and regulators around the world. We’re not going to be coloring too far outside the lines; everything we do will be extremely well-vetted. Securities law applies, especially in the U.S., and we are well within the confines of securities law in everything we do.

So you would argue this is not that much different from what businesses have been doing for a long time, except that it’s actually better in some ways?

These are just better, cooler mechanisms. I’ve used the term “topical global villages.” Using these networks, we can organize ourselves in a decentralized fashion [and] we can govern those systems in a decentralized fashion, with identity and voting systems. We can do token launches and create capital pools, and these groups can define goals and fund those goals and execute towards those goals on these platforms.

Whether these topical global villages are supporting a rock band or attempting to cure a disease, they are focused ways of organizing for collective action. They should be able to acquire resources. You could call it printing money. I think that’s a reasonable way to think about it. But it’s not too different from what has been going on for a long time before blockchain.

You have said that we’re living in “exponential times,” that the decentralized future may actually be closer than we think. When do you think we’ll see the true advent of Web 3.0?
The decentralized future is already here; it’s just—

Unevenly distributed?
Unevenly distributed, exactly. Lots of people who understand the space well are already living to some extent in the decentralized future. We have huge numbers of components that are being built and matured, and they are interoperating with one another. We need to build out scalability, and we need to do better on privacy and confidentiality on these systems, but we are moving towards building for the consumer soon. In journalism, in music, certainly in gaming—I think there will be a lot more normal people, [fewer] crypto-heads, on-boarded into the ecosystem in 2019. We’ll probably have some breakthrough games, or maybe Civil will break through, or maybe Ujo Music will really gain some traction. So it’s happening. It’s a complicated technology—it’s got lots of sharp edges, and it takes time to figure out how to onboard people into a new paradigm.

Look at the iPhone. If I gave you this thing 15 years ago, you wouldn’t really understand how to use it. It would look like a device from the future, and you wouldn’t know the gestures. Apple did a really good job of on-boarding humanity into this new paradigm. We have a design team at ConsenSys of 50 designers, and they’re building a design system called Rimble. It’s sort of similar to Google Material, where it creates the language for depicting visual forms. This will be for Web 3.0. You’ll have all the Web 2.0 elements, but you’ll also have identity and reputation. And there’s a lot of gamification going on; you’ll have badges and levels and stuff like that.

"I haven't been very good at saying no. But in 2019, I will practice saying no. I need to get much more focused."

You employ dozens of Ethereum protocol engineers at ConsenSys. Can you explain the working relationship or symbiosis that exists between the Ethereum Foundation—or simply between Ethereum as an open permissionless network—and ConsenSys as a private, for-profit enterprise?
It’s in a really healthy state right now. Our protocol engineers recently put out Pantheon. It’s a Java client, and it’s connected to [the] mainnet. It will also have special enterprise-friendly extensions, so that corporations can use it or stand up their own private blockchains or consortium systems and still use the same node or codebase to be on the public network. We have a team that works on Quorum stuff, and another team that’s basically working on Ethereum 2.0—Serenity. We’ve had meetings every day all week. I may be missing one right now, actually. Our team is working closely with the other teams. All of this stuff is pretty open. You can go on the developer calls. They’re usually recorded on Reddit and you can listen to them on YouTube.

The worlds you come from—competitive sports, Wall Street—they’re about winning and losing…
Not necessarily. One could take a Zen approach to them.

But in squash, which you played—
I wasn’t very Zen when I was playing squash.

There’s a winner, there’s a loser. What’s your win condition now? When could you say that you’d achieved your objectives and are happy to go back to a beach in Jamaica?
I have no exit strategy. “Winning” is bringing forth technology that enables the world to build better systems for itself. That’s part of our vision and mission. Continuing to do that is pretty rewarding.

Is there any point—maybe the launch of Ethereum 2.0, maybe when ConsenSys grows to 5,000 people—at which you’d move into the chairman emeritus role?
No. I have no exit strategy. My role at ConsenSys has evolved pretty dramatically since day one, and we all keep evolving pretty dramatically in our roles as the company gets bigger, as the scope and vistas get bigger. I’m probably not going to be doing exactly what I’m doing [now] in 2019 or 2020. But I can’t script that. I don’t really know what’s going to happen.

What has 2018 looked like for you, and in what ways do you think it might change for 2019 or 2020?
Exponential growth. It feels like living in exponential times. It’s just an extreme onslaught of interest in our space and growth in the company.

Speaking of “extreme onslaught,” you are scheduled to the max—overscheduled constantly.
Yeah, I haven’t been very good at saying no. But in 2019, I will practice saying no. I need to get much more focused. I’m traveling too much. I think it was good for a while, just getting the message out, but I should [spend more time] getting more high-value work done, [spend] more time reading. I used to be able to be on top of everything in the bitcoin ecosystem and the Ethereum ecosystem and the ConsenSys ecosystem. Now nobody has a chance of being on top of everything in those ecosystems. I need to pay attention to the evolution of the technology and its intersection with other technologies. It will keep evolving.

What could possibly make it worthwhile to live in perpetual fast-forward, as you seem to be? You said it’s a big hit to your sleep schedule and so on.
I’m having a great time. It does get tiring at the end of 10 hours of meetings in a day, and sometimes I have to stay up till three in the morning to get through what I think are the important emails. But it’s pretty much epic day after epic day. It’s quite enjoyable.

This interview has been edited for length and clarity.