Ethereum is going through a crisis right now due to problems related to The DAO, which was the smart contract platform’s most notable project up to this point. While Ethereum supporters tout it as a more advanced, more powerful alternative to Bitcoin, those added features also may cause issues in terms of the reliability and security of the applications built on top of the blockchain.
Although not an Ethereum basher by any means, 21 CEO Balaji Srinivasan recently shared some of his concerns with Ethereum at an event hosted by Boost VC. When initially talking about Ethereum in response to a question from the audience, Srinivasan noted that he respects Ethereum Founder Vitalik Buterin and plays around with the software a bit. He even added that 21 will likely support the Bitcoin alternative at some point in the future.
1. Ethereum Has a Much Broader Attack Surface
Now that real projects have been built on top of Ethereum, more hackers are taking a closer look at the platform. In a way, The DAO was a more than $100 million bug bounty built to test the reliability of the kinds of smart contracts that can be written on top of Ethereum. Ethereum has seen an increase as a percentage of Bitcoin’s overall market cap this year, which means more people are now paying attention to the project.
During his recent talk, Srinivasan noted the number of people looking to find vulnerabilities in Ethereum will grow as the market cap rises. He stated:
“Ethereum is close as a percentage [of Bitcoin’s market cap], and if it ever did become really close, I think you would see a level of Internet trollery and DDOSing that you’ve never seen before.”
Srinivasan added, “Ethereum has a larger attack surface, and Bitcoin has been more stress-tested.”
In addition to possible issues with Ethereum’s base protocol layer, there is also the issue of smart contracts being extremely difficult to write properly -- as was evidenced by the recent fiasco with The DAO. It may be true that Ethereum allows developers to create more complex smart contracts than what’s possible with Bitcoin, but it’s also important to realize these smart contracts are extremely difficult to verify as safe -- as Cornell Professor Emin Gün Sirer recently pointed out in a blog post.
2. Complications with a Transition to Proof-of-Stake
Ethereum developers have also been planning a move from proof-of-work to proof-of-stake for quite some time. In a proof-of-stake system, hashing power is replaced by holding tokens (in this case ether) for the mining process. During his recent talk, Srinivasan referred to proof-of-stake as “a major, open technical question.”
Srinivasan then explained one of the issues he sees with a proof-of-stake system:
“With proof-of-work, you have a fundamental measure of how much computation went into the length of the chain, whereas with proof-of-stake, you only know how many votes were taken. You could imagine a situation where people do an attack called ‘grinding’ where you have a hydra-like thing with lots of competing Ethereum blockchains. And you have to figure out which one is actually the canonical one to follow, and then you’ve added trust into the system -- as opposed to the simple heuristic of just picking the longest chain with the most proof-of-work.”
“Vitalik says he’s got an answer for this, and he’s a smart guy, so I’ve an open mind on this. But it’s an unsolved problem in my mind,” Srinivasan added.
3. Ethereum is Attempting to Solve Difficult Problems
When comparisons are made between Bitcoin and Ethereum, one point that is almost always brought up is that Bitcoin tries to keep things simple. While Ethereum is attempting to be Bitcoin 2.0 and “decentralize all the things,” Bitcoin developers are focused on making sure their particular network is able to maintain a digital bearer asset.
Srinivasan touched on the difficulties associated with expanding on Satoshi Nakamoto’s ideas and applying that sort of game theory to other aspects of society:
“It took Satoshi-level game theory to figure out a way to do a decentralized wire transfer. That’s what Bitcoin is -- a decentralized wire transfer . . . If it took that to get a decentralized wire transfer, how much more challenging is it to decentralize corporate charter?”
Srinivasan then explained that Andreessen Horowitz, where he is a board partner, knows all of their limited partners, and they tend to be pretty conservative institutions. “They all know each other. We all eat food together or whatever,” he noted. “Still, our charter runs to thousands of pages with all kinds of weird edge cases and stuff.”
In the case of The DAO, those known institutions are replaced by anonymous individuals who don’t necessarily need to protect any sort of reputation attached to the real world.
In terms of when these sorts of problems may eventually be figured out, Srinivasan stated, “Decentralizing corporate law, I think, is going to be like at least ten or twenty years . . . When you start doing game theory not on two people but on 2,000 people, that’s an unsolved problem.”
Smart contracts may be to cryptocurrency what autonomous robots were to computers. Doable eventually, but the breakthroughs took decades.— Balaji S. Srinivasan (@balajis) June 23, 2016
Srinivasan clarified that Ethereum is about much more than just The DAO, but he also added:
“[The DAO] gets to a really core thing about Ethereum, which is that the programs -- their difficulty may not be in the implementation but they may be in the game theory and the human dynamics around them. So I think it will take a lot of iteration for Ethereum to become useful beyond simply the fact it has 17-second block times and a commitment to on-chain scaling.”
Of course, it’s possible that Ethereum will eventually find solutions to the various issues that it’s facing right now. After all, the Ethereum blockchain hasn’t even been live for an entire year. Much like Bitcoin, Ethereum is still very much an experiment -- a much younger experiment.