New platform Ethereum wants to bring about a "decentralisation singularity."
Today's internet looks nothing like expected. Back at its inception, it was hailed as a cyber-libertarian utopia where people could express themselves freely, interact directly and on an equal footing, with no honchos to run the game.
But then the web developed in a very centralized fashion: Most of our traffic runs through the servers of a handful of corporations such as Facebook, Google, Amazon, or Paypal. If you want to stay in touch with friends, send an email, or buy something online, you won't just deal with your counterpart, but also trust one of these middlemen. That comes at a price, as what we gain in effectiveness, we lose in terms of privacy, accountability, and sometimes security. But then again, it's just the way the internet works. Or is it?
One group is trying to bring around the original dream, and literally break down the internet. They're behind the Ethereum project: a platform, cryptocurrency, and programming language that aims to allow the creation of decentralised applications.
"We just want to take the internet to its logical conclusion: total decentralisation," Ethereum's spokesman Stephan Tual told me in a London cafe, where I met him to understand what the deal with Ethereum was ahead of its launch on 20 March. The deal is, apparently, that it might usher in an era where corporations run themselves.
Ethereum is the brainchild of 21-year-old Vitalik Buterin, a Toronto-based programmer who has earned a name as a Bitcoin sage. Bitcoin itself is a paragon of decentralisation: Its payment system is kept up by a distributed peer-to-peer network of users (the blockchain) rather than by a single authority. Ethereum builds on the blockchain technology to deliver its promise, but aspires to be about more than just money. It has been conceived as a "Turing complete" language, one which you can use to create any application. What does it mean, practically? Tual thinks that it could trigger a so-called decentralisation singularity.
"We would build a decentralised internet network where all of us would access documents and content without going through a server." he explained. "It means that you will need zero infrastructure to develop and distribute applications." In effect, like the blockchain, Ethereum will exist as a database shared among all its users.
My knee-jerk reaction was of how this could be used for less-than-reputable—or at least authority-defiant—ends. Cases in point: Silk Road and Wikileaks. "It'd be certainly possible to build something like Wikileaks on Ethereum. And if somebody wanted to take it down it wouldn't be able to shut the server, because there's no server," Tual said." The only thing they could do is chase every single user. Or shut down the internet—that would work."
"You'll have a machine making money for itself. It might even become an employer and hire humans"
But that's not really what Ethereum is all about. With the exception of its cryptographic messaging system, its main features are financial. People would directly engage in transactions using Ethereum's own Bitcoin-like money, Ether, according to scripted "contracts" whose enforcement would be guaranteed by the network. Contracts could be about whatever their creators want them to be—from escrow to investment services— and, more importantly, they could work independently from any human stewardship until they run out of Ether.
"A contract is, essentially, an autonomous machine. You can have a machine that negotiates all its financial contracts, for instance. It could make money for you, if you chose so," Tual explained. You could also decide to step down after setting off the contract, and it will keep operating on its own. "Then you'll have a machine making money for itself. Just a machine becoming richer. It might even become an employer and hire humans," Tual added with a grin.
Machines hiring humans is less transcendent than it sounds. I followed up with Buterin on Skype, and he told me that we won't be having job interviews with machines anytime soon. It's more that "anyone can join, and they get automatically paid when the work is done." For instance, a machine could "build a submission process" for a new website design: Once the ideas are submitted, users rate them and the winning proposal receives the money.
There's a name for these entities: decentralised autonomous organizations, or DAOs. Buterin himself described them as a sort of comeuppance for corporate top dogs."The industrial revolution allowed us, for the first time, to start replacing human labor with machines," he wrote in Bitcoin Magazine in 2013. "But this is only automating the bottom; removing the need for rank and file manual laborers, while the management of the company remains untouched. […] Can we remove the management from the equation, instead?"
DAOs should be particularly worrisome for those companies doing little more than providing a server and carrying out simple tasks. "Just think of Dropbox: it just stores stuff. An Ethereum contract can easily do the same thing. The same goes for many financial or betting services," Tual said. "You just won't need the middlemen with the servers anymore. You won't have to trust them anymore."
"If the entire planet was to use our stuff, then we got a problem"
The concept of "trustlessness" is key to Ethereum's creed. Right now, the argument goes, we have to trust internet giants to which we give our information or resources; we don't really know how they'll handle them, how safely, or if they'll yield them to corporations or governments. In general, all our transactions require that we trust our counterpart to stick to the agreement. But you don't have to "trust" an Ethereum contract, as it's publicly uploaded on the platform. "What DAOs are programmed to do is public," Tual said. "For instance, if one financial investment DAO decides to sell naked short, everyone will know and avoid that one. Or they might not avoid it and decide the risk is worth it."
The drawback is that, well, machines are still machines, and at least until roboethics develop to acceptable levels, they are unaccountable.
Harvard researcher Primavera De Filippi has raised the issue of how to deal with a corporation—possibly a vicious one—that can't be arraigned or forced to pay damages, let alone shut down altogether. De Filippi wondered whether Ethereum would become a "freenet," or a "Skynet," a reference to the murderous AI featured in the Terminator movies. Ironically, a Skynet scenario was conjured up by Buterin himself during a conference last year.
Ultimately, much of the success of the Ethereum venture will depend on how many people use it, and build stuff on it. But popularity could also be the force to bring it down prematurely, overburdening the peer-to-peer network with too much data. Servers, after all, exist for a reason.
"Obviously if the entire planet was to use our stuff, then we got a problem, because it would be just too much," Tual acknowledged. "One solution we're exploring is using a different language that allows everybody to keep just a small part of the blockchain while still being able to run and validate it. Another solution could be building multiple chains, multiple Ethereums. We're still working on it."
The decentralization singularity may be coming, but it won't happen overnight.