What is Ethereum? Can this technology actually support the establishment of a utopian, free, and decentralized society? Or could it instead promote a more dystopian vision of society – or even a Skynet? Before we can understand anything about Ethereum, we must first understand Bitcoin: what it is, and how it works.
The Bitcoin story begins in 2009, when Satoshi Nakamoto (whose identity is still unknown, in spite of recent rumors published in Newsweek) released the first implementation of Bitcoin, which he described as a “a decentralized currency based on a cryptographic ledger” – but what does that mean, exactly?
- First, there is the notion of a decentralized database, known as the blockchain. To understand the concept of the blockchain, imagine a public ledger that records every transaction made on the Bitcoin network, and that a copy of this ledger is distributed to every user connected to the network. These users all agree that once a transaction has been recorded, it can never be deleted.
- Second, there is the concept of digital tokens, which are the basic unit of transaction. Combining this with a transaction system on the decentralized database creates a decentralized currency or payment system.
- Finally, there is encryption, which serves to ensure the integrity and security of the overall system by preventing anyone from performing a transaction unless they have the proper authority to do so.
And so, we have Bitcoin: a decentralized cryptocurrency that operates on a peer-to-peer network, which is not regulated by any central bank or other governmental institution but only – and exclusively – by code.
I will not go into the technical (and actually quite complex) details of the Bitcoin protocol, but it is important to know that its code is entirely open source, so anyone may examine it and try to find flaws in the implementation.
After a few years of struggle, testing, and experimentation, Bitcoin’s value rose from less than a dollar in 2009 to over $1,000 in recent months. Nowadays, it is accepted by many commercial actors, such as Foodler or Eurostock, and a growing number of ATMs are being deployed all over the world. Bitcoin has proven that it is possible to have a working decentralized currency system. However, the really interesting point about Bitcoin is not the currency itself, but rather the fact that its protocol – the blockchain – can be used to implement many other applications which have nothing to do with money.
Now let us consider Ethereum, which builds upon the technology of Bitcoin in order to create a next generation “Smart Contract and Decentralized Application” platform. What does that mean, exactly? Well, if Bitcoin is a decentralized cryptocurrency, Ethereum is the platform upon which a decentralized cryptocurrency can be built. Some have defined it as “cryptocurrency 2.0”, but actually, it is much more than that.
Just like Bitcoin, Ethereum implements a decentralized database, a system of digital tokens, and an encryption scheme. But it also implements a Turing-complete scripting language, which makes it possible for anyone to deploy an application directly on the blockchain. So, instead of adding new features to the Bitcoin protocol, Ethereum took a step back and actually removed all features from the blockchain, in order to make it easier for users to build their own applications by implementing only the features they need as an extra layer on top of the blockchain.
Therefore, just as Bitcoin marked the establishment of a decentralized cryptocurrency that subsists independently of any government or financial institution, Ethereum could potentially lead to the deployment of decentralized applications that operate autonomously on the blockchain.
In fact, Ethereum not only makes it very easy to deploy alternative cryptocurrencies, but also to set up decentralized communications systems (like BitMessage), alternative social media (like Twister), or online storage (like Dropbox) in a completely decentralized way, therefore not controlled by any third party. Given that there is no centralized third party to interact with, the interactions between applications and users are regulated by the code of these applications. In Ethereum lingo, this code is actually a “smart contract”, which establishes the rules and procedures that everyone must abide by.
So, what’s a smart contract? Well, a contract is an agreement between two or more individuals who agree to do something, in exchange for receiving something else. The problem is that each party has to trust that the other party will, in fact, fulfill its side of the contract. Smart contracts eliminate the need for trust between parties to the extent that they are self-enforceable. In fact, the code and the contract have been merged into one, given that the contract is both defined and enforced in the same way – by the code.
In the analog world, the most common example we could use is that of a vending machine.
You put money into the machine, and – assuming you inserted the correct amount of money – the machine will deliver the product you ordered. You don’t need to trust the machine, and the machine doesn’t need to trust you, either. But what’s even more interesting is that in Ethereum, you can also have decentralized applications interacting directly with other decentralized applications, essentially eliminating the need for any human interaction.
This leads to the most interesting aspect of Ethereum, which is the concept of Decentralized Autonomous Organizations. Basically, these are a more sophisticated kind of smart contract, with a constitution that stipulates the rules of governance for the organization, and with a system of equity allowing users to invest in the organization by purchasing shares.
Take ICANN, for example. Instead of trusting an organization to operate according to a certain set of principles, we can encode these principles into the protocol of a decentralized application (like Namecoin), or even incorporate them into the constitution of a decentralized autonomous organisation.
But, to come back to the legal issues: what’s so special about Distributed Autonomous Organisations? And why do they raise so many interesting legal challenges?
- First of all, they are autonomous in the sense that once they’ve been created on the blockchain, they no longer need their creators, nor are they under any obligation to respond to, or be responsible of any requests made by them.
- Secondly, they are self-sufficient in that they charge users for the services they provide in order to pay others for the resources they need (such as bandwidth and processing power).
- Finally, they are decentralized, since they do not subsist on a specific server, but instead are encoded into the blockchain (which is distributed to the entire network), and their code is executed in a decentralized manner by every node of the network.
These characteristics make them extremely difficult to regulate because there is no single entity which has control over them. In addition, given the self-enforcing properties of their code, they might actually challenge some of the most basic principles of our legal system. In fact, there are many legal challenges raised by Ethereum, but I will focus here only on the three that seem most interesting to me.
Let’s begin with Contract Law. As previously stated, the particularities of smart contracts are that they are transparent (their code is open source; anyone can examine them) and self-enforcing (trust between parties is unnecessary; contracts are executed automatically, independent of their will).
In traditional contracts, each party is free to decide whether to fulfill the contract, whether to only partially implement the contract (by leaving out some obligations), or whether to breach the contract (and pay instead for damages or compensation). By contrast, in the case of smart contracts, parties have no choice but to implement the contract, because the contract has been encoded, written into the code. It cannot be breached unless one actually manages to break into the code.
This raises the question of what is legally binding vs. what is technically binding.
For instance, there are many situations in contract law that might either invalidate the contract (if it was agreed to under undue influence, for example) or limit its enforceability (to the extent that it goes against the interests of consumers). But smart contracts are not affected by these provisions as they operate within their own closed technological framework, which does not necessarily implement any of these legal safeguards. In this sense, smart contracts could effectively bypass the legal framework of contract law.
When it comes property law, the situation is quite similar, in that Ethereum implements its own technical framework which operates outside of the legal framework of property law. In particular, Ethereum introduces two important features that significantly differ from traditional property rights.
The first is the concept of smart property, which relies on smart contracts and digital tokens to establish a decentralised and trust-free asset management system. The idea is that ownership of something can be transferred directly via the blockchain, through the transfer of specifically designed coins which are linked to a particular item. This allows for the creation of “cryptographically-activated” assets, such as a smartphone that can only be used by spending a particular token, or a car that can only be driven by the person who owns that token. Instead of transferring the ownership of the car, transferring the token associated with that car is sufficient to achieve the same result.
And the other is the concept of crypto-property. This is extremely interesting in that it allows for algorithmical entities, which are neither moral persons nor legal persons, to own currency or particular assets as if they were their own property. So, as opposed to standard property rights – which have been defined by the law and can therefore, in certain situations, also be taken away by the law – crypto-property rights are both defined and automatically enforced by code! This means that they cannot be seized, but that also, once they have been stolen, there is no possibility of recourse.
Returning to Ethereum, this essentially means that Distributed Autonomous Organizations have absolute sovereignty over their own resources, which cannot be seized by anyone unless this is specifically provided for by the code of these organizations. That brings us back to what Lawrence Lessig had already identified over 10 years ago: basically, that in cyberspace, code is law. I think we all understand that by now.
So the question is: if code is law, how can the law regulate the code so that it actually regulates our behaviors in a way that remains compliant with the law? This brings up some more fundamental questions: how do we want to regulate Distributed Autonomous Organizations? Should they be regulated in the same way as standard corporations or organizations, or do we need a distinct body of law that would better account for their specificities?
As I was researching these questions, most of the material I found was related to the question of the regulation of intelligent robots. This was surprising at first, but in fact it makes complete sense since they both share this commonality of being autonomous and self-sufficient.
This bring us to the third point, which is the issue of liability and responsibility. Let’s take the example of a Distributed Autonomous Organisation designed to send a copyrighted song to everyone who transfers the equivalent of $1.00. Here, the main challenge is to determine who is in charge of, and responsible for, this kind of activity?
It could be the creator of the Distributed Organization, but then we run into two problems. First, the creator might be difficult to identify if the distributed organization was created anonymously. Second, even if the creator could be identified, it would be possible that the creator would no longer have the power to control the organisation – which will continue to operate as long as there are sufficient funds for it to operate on its own.
Or, should the users be held vicariously liable for the services for which they’ve paid? This would only apply to the extent that they knew or had good reason to believe that the Decentralized Organization was doing something wrong (but, in this case, users might actually not be unaware that they are purchasing an infringing song).
Perhaps the Distributed Autonomous Organization itself should be held liable for its own actions. But then we encounter an ever bigger problem in terms of law enforcement. It is virtually impossible to recover damages or to obtain an injunction unless these measures have been specifically encoded into the contract/constitution of the organization.
So, we find ourselves in a state of legal limbo, as we cannot rely on traditional legal means to regulate the code of this technology. The question is: do we actually need to?
The supporters of Ethereum would argue that we don’t. In fact, if Bitcoin was designed as a decentralized alternative to counteract the corruption and inefficiency of the financial system, then Ethereum constitutes a decentralized alternative to the legal system as a whole! This refers to the somewhat anarchic idea of decentralized law, where everyone is free to implement their own rules within their own contracts, creating an interconnected system of rules interacting with each other in a reliably predictable way and not dependent on trust between parties.
Of course, the flipside is that Ethereum could potentially be taken over by big corporations, financial institutions, or even by the State, in an attempt to recreate the same economic system and political order that we have today – except that this time, it would be much more difficult to escape from that system. This could lead to the establishment of a totalitarian society that is (almost exclusively) regulated by self-enforcing contracts, which establish the rules that everyone must abide by, without any constitutional constraints.
I would like to conclude with a quote from Yochai Benkler, which says that there are, in fact, no perfect freedoms, just different sets of constraints. What we have to ask ourselves is whether we would rather live in a world where we are constrained by the rules of law – which are universal, but also more democratic, more flexible and not perfectly enforceable – or the rules dictated by code – which, once they have been agreed upon, will be automatically enforced without any possibility of recourse.
Transcribed from a conference held at the Berkman Center for Internet & Society on April 15, 2014. You can find the video of the conference here.