Current Issue
This Month's Print Issue

Follow Fast Company

We’ll come to you.

long read


The Humans Who Dream Of Companies That Won't Need Us

How would Ethereum's network autonomously run transportation apps, delivery services, and other companies? (And would we even want that?)

[Photo: Flickr user Derrick Tyson]

Like other serious crypto-anarchist visionaries, the 21-year-old Russian Canadian entrepreneur Vitalik Buterin sees the blockchain—the public ledger underlying bitcoin—as a way to liberate us all from inefficiency, cut out the middleman, and make big government and business bow at the free market's feet. What makes his vision different is its scope. Where bitcoin aims to disrupt banks, Buterin's Switzerland-based company, Ethereum, aims to become what he calls "the foundational platform for everything."

At the heart of this open-source software platform is a currency, Ether, named after the pre-relativity hypothesis to explain the propagation of light waves. Though it has yet to make a public release, Ethereum has already recruited a strong roster of programmers and last year raised $18.4 million, in the third most well-funded crowdfunding effort in history.

Vitalik Buterin, Ethereum's founder

Bitcoin's blockchain model has been proposed as the backbone for a wide range of applications, from asset trading to real estate transactions, from escrow services to even a "national income distribution" system. What Ethereum proposes, in effect, is a global computer that could not only handle those transactions but also eventually emulate many of the functions of companies like Uber, Airbnb, Dropbox, Amazon, and Kickstarter—but without the "inefficient" bureaucracies and the other intermediaries who take a slice of the pie. That is to say, companies that, once started, can run themselves.

If the blockchain is a giant ledger, Buterin's goal—first articulated in a January 2014 article in Bitcoin Magazine, which he cofounded—is to build the army of robot accountants working on top—what are sometimes known as "smart contracts." Nick Szabo, the cryptographer who is credited with coining that term and is speculated to have been involved in bitcoin's creation, described blockchain-based autonomous organizations last December on his blog as armies of accountant robots:

Instead of the cashier and ticket-ripper of the movie theater, the block chain consists of thousands of computers that can process digital tickets, money, and many other fiduciary objects in digital form. Think of thousands of robots wearing green eye shades, all checking each other's accounting. Individually the robots (or their owners) are not very trustworthy, but collectively, coordinated by mathematics, they produce results of high reliability and security.

Bitcoin as a whole is a step in that direction, but it's only one application. Ethereum, on the other hand, is "Turing complete," a system in which a program can be written to find an answer—or to execute a smart contract that can buy something, sell something, or do something. In aggregate, a group of smart contracts could run what is known in Ethereum-speak as a "decentralized autonomous organization" (DAO) or a "distributed autonomous corporation" (DAC)—in other words, a corporation distilled to its most basic tasks, and operated by little more than code and the logic of if this, then that.

An overview of how the three components of the full Ethereum system work together. Alongside the blockchain, two additional protocols are being developed by the company: Swarm, to manage distributed file sharing, and Whisper, for messaging.

For all of his bluster and dedication, however, Buterin—one of three co-founders of the project, and its most prominent spokesman—doesn't project a Jobsian ego. Sometimes, he seems like the inverse of the Apple founder: Quiet, nerdy, and unequivocal in his belief that that ownership of large systems should be distributed across the web rather than consolidated under governments or corporations. While it overlaps with the popular libertarian philosophy of Silicon Valley, this decentralized ethos is also in some ways the diametric opposite of another Silicon Valley model: the startup economy, which must ultimately enrich its developers and investors more than its end users.

Buterin's crypto-utopia of cloud companies distributed across the global network is also a counterpoint to Washington, with its ossified bureaucracy, red tape, and taxes. A DAO or DAC, operated by servers around the world, would exist within no borders and, in theory, be very difficult to shut down. While some critics of Facebook and Uber have suggested they should be regulated more like public utilities, an Ethereum-based ride-sharing service or social network offers a parallel path, say its proponents—in theory, functioning more equitably than a standard corporation, taking in the fees necessary to maintain the network.

Gavin Wood, an Ethereum co-founder and its chief technology officer, believes that the new digital bazaar could battle and eventually triumph over Silicon Valley’s old cathedral. As he wrote on the company's blog in April,

The idea of a rigid organisation or corporation will evaporate and left will be the true essence of human interaction patterns, policed only by openness and information-theoretic mathematics.. strict legality of the emergent behaviour will become increasingly less relevant as it becomes drastically pluralistic and unpoliceable with no entity, legal or otherwise, coordinating it or profiting from it.

And if Ethereum fails, they insist, some future blockchain layer will accomplish the task.

Working in parallel to Ethereum, but at a much more modest scale, has been the New Economy Movement (NEM). Like Ethereum, the Nemcoin developers don’t see it as a cryptocurrency but a vibrant decentralized ecosystem upon which any number of services and apps can be built. Elsewhere, Nxt and Mastercoin are attempting to map out some of the same decentralized app building and smart contract services as Ethereum is.

But Ethereum's supporters say that no other project has innovated on the blockchain the way that it has. "There is only Ethereum and projects that have forked Ethereum," Aeron Buchanan, an Ethereum developer, told me in an email. "Ethereum is the trailblazer. That's why it's exciting. Bitcoin introduced the blockchain, which was a huge step forward technologically, but the functionality of bitcoin is almost as basic as it could be on the landscape of possibilities: zero-sum account balances. Ethereum opens up the blockchain to endless possible functionality."

"I know this answer sounds rather dismissive," he added, "but it's like asking if, during the development of the aeroplane, there was something else similar. Well, Otto Lilienthal's gliders were an inspirational basis for the Wright brothers’ powered flying machines, but does it help to say that Lawrence Hargrave was flying kites at around the same time?"

How To Build A Robo Ride Company

What, exactly, would you need to build a blockchain-based Uber or Lyft? I asked Buterin to join my thought experiment. According to him, a ride-service app must satisfy four basic functions. First, it must allow passengers to publish ride requests. Then it must feature a driver mechanism that filters these requests for those nearest them or along their route. Finally, it must feature payment and rating systems.

An Ethereum-based transportation company would not perform all of these tasks. One developer might create the interface, another the payment system, while another designs the GPS system. Through smart contracts between user and developers, each of these programs would be building toward a whole that could be far cheaper than existing "ride-sharing" services.

"Long-term, there’s no reason all of these tasks need to be carried out by one company," says Ben Doernberg, a bitcoin expert and ‎research assistant at Harvard University’s Berkman Center for Internet & Society. "A designer in Brazil can build a lovely mobile app that sends your ride request to a matching engine based in San Francisco that pulls trust ratings from a blockchain-based decentralized identity system. When someone in Chicago makes a better matching engine, decentralized Uber switches over and doesn’t miss a beat."

When a driver is looking for a passenger, the driver's client software could propose a cost function, like "prefer rides that start closest to me," which in Buterin's eyes would be an "efficiently computable math formula." At that point, "the driver publishes a smart contract onto the blockchain which can automatically evaluate ride requests submitted to it, and after N seconds automatically pays a bounty of $0.10 to whoever provides the most favorable ride or set of rides," says Buterin.

Once the driver's contract receives the most favorable set of rides and the bounty is paid, the contract deletes itself, and the information is passed along to the driver. Any or all of this computation could be outsourced for efficiency. In the decentralized Uber example, he said, there would exist a "completely open-access industry" composed of "specialized decentralized Uber cloud computing nodes." Their sole purpose would be to look for contracts with search algorithms that determine which rides maximize the score with each other, and submit them. And Buterin says that in this system anyone could immediately start participating.

The payment system could be bitcoin, Ethereum, or any other cryptocurrency. "The rating system would probably be best done as a decentralized publishing platform like Whisper," Buterin said, referring to Ethereum's distributed messaging protocol. "[B]ut maybe with a few nodes trusted by default, where those nodes provide functions like criminal background checks, etc. From there, the whole process is basically just as it was before, but without the middlemen." (As for the rides themselves, Uber is already taking care of that, according to reports: The company has opened a robotics lab in Pittsburgh, reportedly with the aim of building self-driving vehicles.)

In this "partial" example of a DAO, said Buterin, Ethereum basically provides the software and an IT department in a decentralized fashion. Drivers directly pay for the results, allowing individual end users or groups of people to come together as a community to satisfy "the mechanism’s needs."

As fantastical as it sounds, something like this already exists. La’Zooz, a startup based in Israel, has built what it says is the first decentralized ride-sharing application atop Bitcoin’s blockchain. Similar to Buterin’s vision for an Ethereum competitor to Uber, the end users own it. They can also earn Zooz cryptocurrency by contributing to improve La’Zooz’s ecosystem. The difference between La’Zooz’s blockchain app, and a theoretical Ethereum-based one is that the former is run by three cofounders, while the latter wouldn’t need any humans at all.

In theory, at least. For a customer or developer to transact in an Uber or Airbnb killer, a transaction on Ethereum would have to run through all of the ecosystem's nodes. Another way of putting it is that the transaction will have to be processed by all of Ethereum's users, adding up to a very slow computer, one that's only as fast as its slowest node. The slowness and inefficiency of the blockchain will be a primary challenge for Ethereum's app developers, says Doernberg.

Ethereum isn't the only company to be thinking about turning the blockchain into a powerful, valuable platform for building global computers. Tech investor Marc Andressen has described the blockchain as an ideal medium of exchange for machines that need to talk to each other, and his firm, Andreessen Horowitz, has announced plans to invest "hundreds of millions" of dollars in bitcoin-related startups. Last year, Nick Szabo worked for Mirror, a bitcoin startup based in Palo Alto, and helped orient the company toward exploiting the blockchain for self-executing financial transactions.

Ethereum's founders Vitalik Buterin, Jeffrey Wilcke, and Gavin Wood

Not All Humans Are In Agreement Here

There is a whiff of sci-fi dystopia to all of this. Could a robotic or algorithmic corporation, completely decentralized, become a real-life capitalist Skynet? Should, someday, a network of computers come to control 51% of the Ethereum blockchain, could it override the human element in the network, and wreak havoc on securities, financial markets, home appliances, and infrastructure? Probably not, given the inherent slowness of blockchain technology, and the fact that a DAO still requires humans to develop its various Voltron-esque parts. And, after all, DAOs are not sentient but narrow-AI autonomous agents executing very specific tasks.

Emin Gün Sirer, a Cornell professor and systems builder highly active in the cryptocurrency community, has more practical concerns. He said that while blockchain enthusiasts are eager to see a Uber and Airbnb middlemen cut out, he's not so sure there is demand for it. For him, there is no technological reason why an Uber or Airbnb would benefit from being built on an Ethereum-like blockhain. He also believes the highly rational, deregulated approach wouldn't necessarily be good for users.

"It turns out you can build some portion of these applications peer-to-peer, but often people misunderstand what these centralized services [Uber, Airbnb] really do behind the scenes," Sirer says. "To a naive observer it looks like Uber and Airbnb are just about matching people; [but] they’re all about doing that security check, that incredible vetting [of drivers and sublessors or landlords], and it’s all about having that presence and identity."

"Vitalik and others will say, ‘Oh, there will be these services that vet drivers for you and they will compete for legitimacy, and we will pay these third-party services, and if there is a third-party service that isn’t doing its job, well, the invisible hand will correct them," Sirer says. "But, if you think about it, no, that is not how things work in the real world, because what we’re talking about here is life and death and rape. This is not something I want to trust to market forces."

"I find these arguments that Ethereum and its descendants will somehow solve this just incredibly naive," he added.

Still, Sirer believes that more simpler smart contracts are the prime candidate for an Ethereum-type application—cryptographic clearinghouses for private transactions in which people decide to follow a set of defined rules.

"Computer games would be a catchy, cool, and killer application. Financial derivatives would be a fine thing to do [on Ethereum], too. Also, future contracts where a farmer wants to hedge against a harsh winter that might effect crops. Those kinds of contracts I understand."

At CES in January, IBM and Samsung announced another practical purpose for Ethereum: simpler, crude, repetitive tasks in the so-called Internet of things, like monitoring the temperature or turning lightbulbs on and off. Under a project called ADEPT, or Autonomous Decentralized Peer-to-Peer Telemetry, Ethereum wouldn't just serve as a global computer: with connected devices around the world contributing to the global network—and mining Ether in the process—they'd be generating money, and "revenue-generating opportunities in their own right," IBM researcher Veena Pureswaran told Fast Company last month.

Under an Ether-backed Internet of things, perhaps, home appliances wouldn't be sold but freely leased to consumers, subsidized partly by the minting of Ether. "The reason that they need this stuff is that they don’t want to pay," Stephan Tual, Ethereum’s chief communications officer, said about the IBM-Samsung project. "It’s not altruism. They don’t want to pay for its maintenance."

Rick Dudley, an IT consultant who is trying to develop games for blockchains and has been digging into the Ethereum source code, believes it will make sense to launch blockchain-based competitors to platforms like Uber and Airbnb, but is measured in his expectations about the time frame.

"I think the Ethereum community, which as far as I can tell is by far the furthest along [among various virtual currency projects] has a long way to go before they will provide an obvious benefit to existing organizations in those spaces," he said. While Dudley says he sees "huge demand" for Ethereum-like services in industries like online gaming, he sees less of a need for the blockchain in other realms, like providing Internet service. "I am more skeptical of its value in the case of more physical things."

Before we can see Uber or Lyft competitors, Dudley said users will have to wait for more of the OS-level surface to be exposed. This won't happen overnight. "It will take a lot of work to get to that point from where we are today or what is even being discussed in depth. $15 million is nowhere near enough."

Next, Dudley hopes "userland tools" will emerge that allow for development and deployment by anyone experienced with contemporary web development. In Ethereum's ecosystem, the currency is Ether, and one's Ether wallet is also a browser, which will function as one of many web interfaces, or portals, into Ethereum.

"The Ethereum client is literally a fork of Chromium's webkit backend," Buterin says. "The idea is that users can build their own interfaces with HTML/JavaScript just like websites, and they will be viewable with the browser much like websites are viewable with the web browser."

"That being said, no new discoveries will need to be made, all the theoretical heavy lifting has been done, arguably before bitcoin was invented," Buterin added. "The technical side of Ethereum’s efficacy is 100% an engineering exercise."

In Doernberg’s opinion, the future of decentralized robot companies all comes down to this: Are services like Uber more than the sum of their parts? And, if so, by how much? He believes that eventually Ethereum and its successors will answer these questions for "every service in the world." The catch, though, is that it will "take years to build enough technical and social infrastructure for this to be a viable alternative."

"Ethereum and decentralized applications are a perfect example of Amara’s Law," Doernberg says. "We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run. I don't know how happy Ethereum investors will be in 18 months, but I think Vitalik will feel very at home in 2025."