Founder of Cyber Capital: Don't just rely on L2 for Ethereum expansion

It is a dangerous idea to rely solely on L2 to achieve expansion.

Written by: Justin Bons, Founder, Cyber Capital

Compilation: Luffy, Foresight News

Treating L2 as the only blockchain scaling solution is a dangerous idea, with poor user experience, trust decay, and centralization. These will inevitably lead to failure and users will eventually move to other scalable blockchains.

From Arbitrum to Optimism, all major L2s are centralized. The reason for this is that L2 does not inherit the security of L1, and even after they transition to a decentralized orderer, L2 only have a fraction of the security of L1 because they still rely on a separate consensus layer.

Ironically, the solution lies in reshaping the decentralized consensus of the orderer. In the end it was just a futile effort. In the PoS mechanism, security = the amount of pledged funds, instead of distributing funds to hundreds of L2, it is better to give one L1.

L2 doesn't solve the problem, it just shifts the problem to a weaker model. Since smart contracts still need to manage keys for upgrades, they are managed through DAO. It is true that L2 can give away their own administrative keys, but this is not the case.

**This is why L2 sequencers and managing keys end up facing the same challenges as L1. **

L2, unlike most major L1s, is not scaled to address these issues, and even if the orderer cannot steal user funds, it can censor and frontrun transactions, which is unacceptable.

L2 breaks down the whole point of using cryptocurrency. First, L2 smart contract management keys can steal user funds because it can change smart contract rules. These keys are now primarily managed by trusted multisigs, including Optimism and Arbitrum, the biggest players in the L2 space.

That's not even the biggest problem with L2, the biggest problem comes from user experience, which is almost unresolved in a competitive market environment.

To demonstrate this with a simple example, a user transfers money between L1 and L2:

The operation of two users transferring money in L1 is very simple, just scan the QR code of the other user and click send. Operations in L2 networks are a bit more complicated, as users need to know what L2 their friends are on and how to bridge between the two.

Even more troublesome: users also need to check that these particular L2s are secure and decentralized. The crypto market is a free market and there will always be custodial and centralized L2s. Even today, most L2s keep smart contracts managing keys and run centralized orderers.

Also, L2s are not fully compatible with each other, which makes the L2 experience even worse. We're still waiting for more L2, which is certainly a user experience nightmare. The free market keeps competing L2s from merging on interoperability, nor should they, and it would be irresponsible to make a fully decentralized L2 interoperate seamlessly with a managed L2.

When trust models differ, users must be given more choices. However, it is this user choice that affects the user experience.

Users will naturally turn to finding a custodian in order to simplify the process. Bitcoin's Lightning Network is a good example. In fact, most Lightning users are now using hosted solutions, which is exactly what I predicted back in 2015. It's a total user experience nightmare, and we shouldn't expect too much from the average user.

I predict that if Ethereum continues like this, it will face the same fate as Bitcoin. Ethereum's original roadmap also included implementing sharding (horizontal scaling), but this path was too difficult for Ethereum. But blockchains such as NEAR, EGLD, XTZ, TON, etc. have proven that performing sharding is feasible. This means we can scale on-chain without sacrificing decentralization or pushing most users to custodial solutions.

The Ethereum community should not ignore this fact, it could move to implement sharding. The negative effects of sharding on composability, interoperability, and user experience are overstated: especially compared to L2, which is much worse on these fronts.

Sharding is a more competitive solution that will defeat the modular scaling theory as it will provide users with all the benefits of L1 chains without any of the inherent costs that come with L2. This can be done with Enshrined L2 or Sharded L1.

A new question arises, why doesn't Ethereum do this?

Revealing the elephant in the room: L2 is orders of magnitude more funded than L1; L2 tokens are worth billions of dollars while L1 development is only in the millions…

This is a huge bias against L2 scaling, even to the point of arbitrarily limiting L1 capacity and ignoring L1 scaling techniques. Bitcoin is an example, which arbitrarily limits the block size and goes against the original vision of Bitcoin.

Ethereum is repeating history. Its block size limit (gas limit) used to be determined by miners, and is now set by clients. Ethereum’s roadmap to remove execution sharding is a violation of their social contract.

Incentives for L2 are grossly misaligned. Major L2 companies such as Arbitrum have directly acquired major clients such as Prysm. History is repeating itself, and this is exactly what has happened to companies such as the Bitcoin ecosystem Blockstream and Chaincodelabs.

L2 is becoming the biggest corrupt force in the industry. They benefit from not being able to scale L1 in the short term, and developers turn into multi-millionaires through L2 tokens with the help of VCs. Given enough time, all systems with such perverted incentives will tend to corrupt. Blockchain is no exception as it can still be controlled at the center.

History is repeating itself, and this is where the real tragedy of mankind lies. In the long run, Ethereum will be as unscalable as Bitcoin.

What we need is funds that favor Ethereum L1 over rent-seeking L2 funds. This can be achieved by putting part of the block rewards into the treasury, and through on-chain governance, funds are allocated to solutions that can promote the long-term development of the protocol, thereby finding a source of funds for L1 expansion. This is not a new idea, cryptocurrencies like DASH, DCR, and XTZ have been in the works for many years before DAOs became popular.

However, even with all the flaws of on-chain governance, it's still a great option. In that sense, it is like a democracy, flawed, inefficient, corrupt and prone to mob rule. However, it is still the best form of government we have.

**I have nothing against L2 solutions, I absolutely believe they have their own niche use cases. However, I disagree with limiting L1 capacity in favor of L2 scaling, and the market should be left to decide which is best. **

Capacity expansion can occur in both L1 and L2. Supporting L2 scaling by limiting L1 is a user choice. The purpose of this restriction is to "force" the user to use L2. In effect, this will cause the user to move to the scalable L1.

I was a supporter of Bitcoin from 2013 to 2016, but after Ethereum came out, I started to support it and participated in mining with a few machines in 2015. It's very sad to see history repeating itself in this way. Ethereum is a huge improvement over Bitcoin, but it can also be replaced.

**I still hope that Ethereum will go back to its original sharding roadmap, or pursue Enshrined Rollup instead. **However, this will wipe out billions of dollars worth of L2 tokens and VC investment, it remains to be seen whether Ethereum leadership will do so.

I really believe we can provide the world with a lot of useful cryptocurrencies. So this critique comes from the depths of optimism that we can solve the scaling triangle problem. For the really good future, hopefully the Ethereum community will take this as constructive criticism.