Does Ethereum have a future?
Posted: Sun Oct 11, 2015 7:07 pm
Ethereum has passed through several phases in its long history, each time surpassing the expectations of many people in the altcoin community. At first, with little work to show for itself, the project issued a "crowdfunding" sale that raised millions of dollars and likely violated SEC securities regulations. Shortly thereafter, Ethereum's developers chose not to sell the bitcoins raised, wasting more than half of its funders' money. When launch finally occurred, the price of Ethereum rose and then fell, causing some to question whether the allure of Ethereum had faded. Now, with talk of implementing Ethereum as a bitcoin "sidechain" circulating, whether Ethereum can compete in such a rapidly changing world is in question.
To understand the arguments for and against Ethereum, the purpose of the network needs to be reviewed. Unlike bitcoin, which allows limited operations on its coins, Ethereum allows any possible computable operation to be performed, such as loops. That means that Ethereum can be thought of as a "world computer," a term that the project itself uses on its website. In theory, anyone who is willing to pay enough can execute any code they would like on the Ethereum network.
The ability to execute Turing-complete code leads to interesting possibilities, but for all the talk of the intelligent agents that can be programmed to use Ethereum, none have actually been implemented. In reality, the "world computer" runs at a speed equivalent to a system considered average fifteen years ago - and everyone in the world has to share that obsolete computer's power. The major limitation to Ethereum is that every node on the network executes every instruction, an incredibly inefficient architecture that significantly limits its usefulness. Another lesser limitation, which can be worked around today with automated trading scripts, is that users have to use an exchange to convert Ether to bitcoin.
One alternative to the current Ethereum network would be simply implementing Ethereum's features on bitcoin's blockchain, so that instead of "Ether," users spend bitcoin directly. But as doing so would require a huge amount of data to be appended to the blockchain (and therefore is incredibly expensive), a current proposal is to implement Ethereum as a "sidechain." In sidechains, coins are mined and change hands, and then one large transaction periodically settles the balances between the bitcoin blockchain and the sidechains. A sidechain implementation of Ethereum would combine the benefits of not polluting the bitcoin blockchain with Ethereum transactions and also allow people to be paid directly in bitcoin, rather than in an undesirable second currency.
There's only one problem with this plan: bitcoin "sidechains" don't exist, there is no implementation written for them, and it is unlikely that they will ever exist. Sidechains require a hard fork, and one that changes the bitcoin philosophy from a pure monetary system to having the potential to implement other applications. People like theymos who are opposed to even the most basic change (a capacity increase), are certainly going to actively block any attempt to expand the definition of bitcoin beyond money.
Since Ethereum can't be integrated with bitcoin, that leaves one system or the other to survive. Both systems currently have significant flaws, but Ethereum's problem (the system is too slow) is a technical problem, while bitcoin's problem (individuals who are opposed to any suggestion of change) is a people problem. The Ethereum technical problem can be resolved by a future version of the network requiring a limited number of nodes to execute code, checking to see whether they agree, and then trusting their results (as the reCaptcha system does). If only 5% of the network executes any given contract, for example, the cost of computing power declines by a factor of 20. Anyone can implement this code and as long as it were secure few would oppose it, whereas the bitcoin network is crippled by a small group of people who fully control all of the source repositories, news sites, and forums around it.
Two years ago, I dismissed the Ethereum network as one of those VC projects that had plenty of businessmen and money but no working code. Now, given the state of the bitcoin industry, I'm not so sure. While Ethereum will never be implemented as a sidechain of bitcoin, the technology will be far superior if a solution can be found to the universal execution problem. If such a solution is found and implemented while bitcoin continues to fall behind in a state of paralysis, I would be worried about whether bitcoin will be able to compete with such a capable system.
To understand the arguments for and against Ethereum, the purpose of the network needs to be reviewed. Unlike bitcoin, which allows limited operations on its coins, Ethereum allows any possible computable operation to be performed, such as loops. That means that Ethereum can be thought of as a "world computer," a term that the project itself uses on its website. In theory, anyone who is willing to pay enough can execute any code they would like on the Ethereum network.
The ability to execute Turing-complete code leads to interesting possibilities, but for all the talk of the intelligent agents that can be programmed to use Ethereum, none have actually been implemented. In reality, the "world computer" runs at a speed equivalent to a system considered average fifteen years ago - and everyone in the world has to share that obsolete computer's power. The major limitation to Ethereum is that every node on the network executes every instruction, an incredibly inefficient architecture that significantly limits its usefulness. Another lesser limitation, which can be worked around today with automated trading scripts, is that users have to use an exchange to convert Ether to bitcoin.
One alternative to the current Ethereum network would be simply implementing Ethereum's features on bitcoin's blockchain, so that instead of "Ether," users spend bitcoin directly. But as doing so would require a huge amount of data to be appended to the blockchain (and therefore is incredibly expensive), a current proposal is to implement Ethereum as a "sidechain." In sidechains, coins are mined and change hands, and then one large transaction periodically settles the balances between the bitcoin blockchain and the sidechains. A sidechain implementation of Ethereum would combine the benefits of not polluting the bitcoin blockchain with Ethereum transactions and also allow people to be paid directly in bitcoin, rather than in an undesirable second currency.
There's only one problem with this plan: bitcoin "sidechains" don't exist, there is no implementation written for them, and it is unlikely that they will ever exist. Sidechains require a hard fork, and one that changes the bitcoin philosophy from a pure monetary system to having the potential to implement other applications. People like theymos who are opposed to even the most basic change (a capacity increase), are certainly going to actively block any attempt to expand the definition of bitcoin beyond money.
Since Ethereum can't be integrated with bitcoin, that leaves one system or the other to survive. Both systems currently have significant flaws, but Ethereum's problem (the system is too slow) is a technical problem, while bitcoin's problem (individuals who are opposed to any suggestion of change) is a people problem. The Ethereum technical problem can be resolved by a future version of the network requiring a limited number of nodes to execute code, checking to see whether they agree, and then trusting their results (as the reCaptcha system does). If only 5% of the network executes any given contract, for example, the cost of computing power declines by a factor of 20. Anyone can implement this code and as long as it were secure few would oppose it, whereas the bitcoin network is crippled by a small group of people who fully control all of the source repositories, news sites, and forums around it.
Two years ago, I dismissed the Ethereum network as one of those VC projects that had plenty of businessmen and money but no working code. Now, given the state of the bitcoin industry, I'm not so sure. While Ethereum will never be implemented as a sidechain of bitcoin, the technology will be far superior if a solution can be found to the universal execution problem. If such a solution is found and implemented while bitcoin continues to fall behind in a state of paralysis, I would be worried about whether bitcoin will be able to compete with such a capable system.