Even in its libertarian variety
Blockchain technology is one of the hottest things in the economy of today, and perhaps none of the existing blockchains promises more than Ethereum, especially as far as libertarians are concerned (e.g. this article by Jeffrey Tucker). However, the recent split of the Ethereum community and blockchain into two highlights the dangers of mixing real-world technologies with technological utopianism, even in its libertarian variety.
However, before addressing Ethereum in more detail, let us start with a brief introduction to blockchains. A blockchain is essentially a distributed ledger (database) maintained and constantly updated by a decentralized network of computers, each of which is programmed to reach a consensus at every iteration, that is to keep the same version of the database. The iterations of updating are called ‘blocks’, thus the word ‘blockchain’. Mining computers are set to select the longest chain at every block, i.e. the chain that contains the most blocks.
Currently, the predominant way blockchain consensus is achieved is through the proof-of-work (PoW) mechanism. Under this arrangement, a subset of the nodes in the network — the miners — use computer power to find through a random search process a string of numbers (nonce) that loosely matches to the one assigned to the block by the relevant algorithm. Another algorithm then verifies that that string of numbers is correct by applying a special function to the inputs of the block concatenated with that string. Mining computers have to use random search because the function and inputs are chosen such that it is prohibitively difficult to use any other approach. The computer that first finds the correct hash mines the block and receives a reward for it in the blockchain-specific cryptocurrency.
The power of the blockchain technology comes from the fact that unless a malicious actor or group of actors control more than 50% of mining power, the record of transactions on the blockchain will tend to reflect the will of the transacting actors without the need for centralized control. That is because if miners controlling less than 50% choose to change some past blocks, the remaining miners will tend to mine blocks more frequently, and the chain they mine will be the longest. Hence, the two key features of blockchains: relative decentralization and immutability. It is around those features that the ideological struggle erupted in the Ethereum community.
Ethereum and the DAO Hard Fork
Ethereum is a blockchain that took the technology a big step further than the first widely successful predecessor — Bitcoin. Where the latter only allows to transfer amounts of its native cryptocurrency, Ethereum allows to execute directly on the blockchain smart contracts of various kinds: essentially, any contract where implementation can be verified by a computer. The examples of such contracts with applications already being developed for them include renting objects secured by smart locks (Slock.it), paying for music (Ujo Music), ride-sharing (Arcade City), idle computer power sharing (Golem), prediction markets (Augur) among countless others.
However, the first big smart contract the Ethereum ecocystem has produced turned out to be a costly failure. That contract involved the creation of a decentralized autonomous organization called The DAO. It was supposed to be an organization governed by code that would allow investors who contributed their ether coins in exchange for DAO tokens to vote on funding various Ethereum start-ups. In a very short period of time, the DAO attracted $150 million worth of ether.
The DAO set-up allowed investors who disagreed with the majority decisions to withdraw their funds through executing a split, i.e. moving their ether into a child DAO. However, unbeknownst to the community, the split function together with associated functions was vulnerable to a to a recursion attack. When withdrawing ether into a child DAO it was possible to move one’s DAO tokens back onto the balance of the original DAO before it has been updated, thus allowing repeated withdrawals of ether with the same small number of tokens.
The issue had been revealed several days before the attack actually happened but it was too late to stop it. On June 17, the attacker(s) managed to drain 3.6 million ether ($60 million at the then current rate). This created a difficult problem for the Ethereum community that it first tried to resolve through the least contentious methods possible but failed. One of the attempts involved a fascinating cyber-battle against the hacker(s) in which a group of developers aided by some DAO token holders managed to outcode the hacker(s) and drain all the remaining funds from the DAO into a safe account. The only viable approach left to regain control of the stolen funds was to execute a hard fork.
A hard fork consists in a modification of the code governing the blockchain that miners and other nodes must either accept or reject in which latter case they cease to be part of the forked blockchain (hence the ‘hard’ part). A vocal minority of Ethereum users stood vehemently against it emphasizing decentralization and immutability.
The hard fork took place on 20 July and was initially a resounding success. The ether belonging to the original DAO investors was moved to a refund smart contract that started distributing it to them. However, despite the fact that the majority of miners and users have supported the hard fork, the minority stuck to the original chain dubbed ‘Ethereum Classic’. This ultimately led to the emergence of a cryptocurrency (ETC) competing with ether, and arguably created more uncertainty about Ethereum’s future. There was then a massive price speculation motivated by the presence of these two competing cryptocurrencies that at some point attracted hundreds of millions of dollars from people hoping that ETC would be able to dethrone ETH.
Ultimately, the chain with the much better fundamentals (ETH) withstood the attack (as basic economic considerations would have predicted) but all these events seem to have done damage to the reputation of Ethereum as a whole. At the time of writing the combined prices of ETH and ETC add up to around $13 (and at the height of the speculative attack, it was around $12), while shortly after the hard fork and before the emergence of ETC as a tradable cryptocurrency, ETH price reached $14.5 on July 23 and would probably have reached a far higher level by now. More importantly than prices, the existence of two blockchains with completely the same underlying technology and similar names may well be confusing would-be investors, especially those new to the blockchain space. There were already some unconfirmed reports on Reddit of people mistaken one Ethereum-based cryptocurrency for another, for example.
Why Absolute Immutability and Decentralization are Utopian Ideas
The intransigence of the anti-fork members of the Ethereum community, thus, seems to have inflicted a significant damage to the Ethereum project. The main question is whether there were really valid reasons for doing so.
The main reasons stated against the hard fork come down to immutability- and decentralization-maximalism. Many anti-HFs in the discussions on the Ethereum subreddit went so far as to claim that the hacker who originally drained the DAO was not guilty of any wrongdoing because she simply used a feature implicit in the DAO smart contract (in fact, the ETC community has just allowed the DAO hacker(s) to collect her/their ETCs on the ETC blockchain). The intervention in the form of the hard fork, from this perspective, amounted to a third-party confiscation of a properly-earned ether, and the pro-HF part of the community acted like a state.
In other words, this reasoning takes immutability and decentralization to the absolute limit, ignoring even the essence of what contracts are, which most people understand intuitively, and decent undergraduate law students can easily explain. Contracts are not reducible to their text, their text is merely evidence for the intentions of the contracting parties; evidence that can be overridden in some cases, if it is manifestly clear that what is written in the contract could not have been its subject-matter. And in the DAO case, we are not even talking about what was directly mentioned in the contract (the relevant code) but the unnoticed logical implication of one of its provisions (the possibility of recursion).
By analogy, imagine a contract of sale that by mistake would result in the object of sale being left for some time in an unguarded location where people would assume that the objects left there have been abandoned. If the object is then taken by the buyer before payment was made, few people would conclude that everything went as parties had envisaged when striking the deal.
Ethereum’s Hard Fork and a Free Society’s Potential
Fortunately, the success of blockchain technology (including in thwarting some forms of onerous government interference) does not hinge on absolute immutability and decentralization. Rather, the capacity of the relevant blockchain community to quickly and decisively handle those situations where smart contracts have misfired and caused a lot of damage is a necessary condition for the technology’s adoption outside of the circles of strong believers.
More surprisingly, rather than putting a quasi-statist stain on the blockchain ecosystem, the efficient way the Ethereum community cooperatively handled the DAO situation is a powerful illustration of how efficient non-state complex dispute resolution can be, not a betrayal of the libertarian spirit behind blockchains.
Just imagine how long and cumbersome the process would have been if state courts and law enforcement had been involved. Imagine what it would take to explain the blockchain, the details of the DAO smart contract and how it failed to prosecutors, judges, jury perhaps. In addition, a perhaps bigger hussle would have arisen from the jurisdiction problem. Ethereum’s blochain does not exist in any particular single country, it is instead distributed among a multitude of computers worldwide.
What is more impressive is that the community trusted the group of developers who implemented the hard fork even though a lot of money was at stake and they did not have any government-issued license or other official authority. Their reputation as development leaders, as well as the understanding that they were stakeholders in the reputation of their blockchain was sufficient for most community members. In the process, an enormous amount of cooperation (mostly unpaid) took place, even though the success was highly uncertain. Far from showing their supposed true Hobbesian colors, lots of individuals managed to solve a very complicated social problem, involving rule interpretation and enforcement without any prior practice. One has to hope that in the future history books of the free society, the DAO crisis and its resolution will be cited as one of the first major steps to its creation.