Is there room for more than one general-purpose blockchain in the world?

Clearly, the answer is yes.

But there will also likely be only one major winner. While the topic is currently hotly debated, I believe that the endgame for the crypto industry is already written, in an almost deterministic fashion, into the choices that were made in designing the current contenders.

I believe the final equilibrium for the crypto industry in the financial domain will form around a super large-scale, efficient, neutral asset platform serving as the “Internet of Money”, a term popularized by, amongst others, Andreas Antonoupoulos. Crucially, this one platform will have to be decentralized. The winner has already been decided, and it is the one that best provides decentralization, scalability, and security, without compromise. We will examine why Ethereum has already won by design, although Solana will put up a worthy fight.

The Internet of Money, built on a blockchain protocol, will be the future infrastructure of global finance, directly and indirectly serving billions of people, with an asset scale of at least several tens of trillions of dollars. A large variety of assets will be issued on the Internet of Money and because these assets are on the blockchain, they naturally possess "programmable" attributes, enabling efficient handling around the clock: transfers, trades, mortgage, bundling, unbundling, issuing derivatives based on underlying assets, and so on.

Why Does the Blockchain Have Value?

Why does the blockchain have value? This is a question all crypto investors have asked. The recognized answer in the crypto industry is: because of decentralization. I believe this answer is correct. However, when we talk about "decentralization," what exactly are we discussing?

In my view, "decentralization" is a means, and the goal is "trustlessness."

So, what is trustlessness? Let's first discuss what trust is. When you trust someone, you give them the "power" to harm you while holding a positive expectation that they will not harm you.

A fine example of trust in the financial system was when people initially stored their gold in vaults, which issued a deposit receipt, promising to return the gold whenever you presented the receipt. A depositor essentially had to trust the vault, which now had the ability not to return your gold, but they felt it would be fine, assuming the vault would return it. As we all know, vaults realized it was unlikely all depositors would withdraw their gold at the same time, so they lent out a portion of the gold to earn interest. Eventually, this developed into the "fractional reserve system." The vaults became banks, which then repeatedly faced bank runs. In 1971, the promise of dollar-to-gold conversion was broken, the "deposit receipt" was directly invalidated, "US dollars" became unanchored "dollars," and we entered an era of unbridled fiat currency issuance, moving into the credit money era dominated by fiat currencies.

What then is trustlessness? Trustlessness means you do not need to give others the power to harm you. "Trustless service" means you can obtain services without giving the service provider the power to harm you. Blockchain provides trustless services. In the blockchain world, as long as you control your private keys, no one can take or freeze your BTC or ETH; as long as you pay the blockchain miner's fee, you can send coins to any address. Trustless services are especially suitable for the financial domain, including services such as issuing assets (BTC, ETH) according to pre-agreed rules and handling assets in various ways, such as transfers, trades, and mortgages, among others. The blockchain is the basis for the Internet of Money, because it is ruled by code and not law.

The blockchain that builds the Internet of Money must be: (A) sufficiently decentralized; (B) able to provide enough throughput. These two points must be met simultaneously, without exception. While Solana and other L1s will put up a worthy fight, Ethereum is the only contender in this race.

Why must this base infrastructure be sufficiently decentralized? Recalling our previous discussion, the attribute of decentralization provides trustless services, and trustless services are the foundation of the Internet of Money. Why is trust, or rather "trustlessness," so important?

What would Satoshi Do?

Blockchain expert BitGulu notes that if the Bitcoin blockchain were not decentralized but ran on a centralized server:

• Satoshi Nakamoto would have to open accounts for every user on the Bitcoin network, and during account opening, Satoshi Nakamoto would have to review each user's documents, proof of home address, etc. Satoshi Nakamoto would ask you where your BTC came from. Please provide proof of funds!
• Satoshi Nakamoto would have to apply for operating licenses from governments.
• Satoshi Nakamoto would have to report various suspicious transactions to governments.
• Satoshi Nakamoto would have to provide tax-related information to governments.
• Satoshi Nakamoto would have to accept government directives, freeze BTC as and when required, and sometimes even transfer the frozen BTC to designated accounts.

Obviously, a single server could not run the Bitcoin network. So why a decentralized network? Because decentralization is an "army" that prevents providing the blockchain network with a form of "sovereign independence," thereby providing the Internet of Money with neutral, independent, predictable security services.

So, how much decentralization is enough? Everyone's judgment is different, and this threshold is dynamically changing, related to the severity of the external environment.  Dozens of consensus nodes are definitely not enough to build the Internet of Money; a few hundred may not be enough; a few thousand nodes may start to make people feel at ease. The degree of decentralization, in addition to the number of consensus nodes, is also very related to the nature of the nodes themselves. For example, if the hardware requirements for nodes must be data center-level, then even with a few thousand nodes, this "army" is still fragile because the privacy of nodes is almost nonexistent, and "soldiers" cannot conduct guerrilla warfare. Thus, the Ethereum community believes that it is very important for ordinary people's computers to be able to run consensus nodes, the crucial basis for Ethereum's decentralization.

Throughput is Equally Important

The blockchain building the Internet of Money must not only be sufficiently decentralized but also able to provide enough throughput. However, before the proposal of second-layer technology in Ethereum (English: Layer2, hereinafter referred to as L2), the crypto industry once popularized the "impossible trilemma" theory. This theory posits that it is impossible to simultaneously achieve scalability, decentralization, and security, with the best being two out of three. Obviously, security cannot be compromised, so one must choose between scalability (i.e., high throughput) and a high degree of decentralization. As a result, many blockchains compromised on decentralization to achieve high performance, such a compromise has already disqualified them from the race to build the Internet of Money.

Today’s L2 technology solves the problem posed by the impossible trilemma. What defines an L2 is simple: whether the L2 system can ultimately achieve the "trustless" level of L1 (Layer1, i.e., the underlying blockchain) in design. L2 is an extension of L1, forming the entire blockchain internal ecosystem together with L1. If it loses the most important "trustless" attribute after extension, then such an L2 system is not part of the blockchain ecosystem and cannot provide independent space for building the Internet of Money. Otherwise, logically speaking, centralized exchanges could also claim to be L2, because after you deposit (rename as bridge) to a centralized exchange, you can also transfer and trade.

Leaving aside those "pseudo-L2" systems that claim to be L2, among the real L2 technologies, the most important branch is Rollup technology. The working principle of Rollup technology is to compress a large batch of transactions into one Rollup transaction and upload it to the L1 blockchain. There are currently two types of Rollup technology: Optimistic Rollup and ZK Rollup, both of which break the so-called "impossible trilemma" in their own ways. Optimistic Rollup outsources the verification work that Ethereum nodes need to complete, allowing anyone to challenge the state after an Optimistic Rollup transaction on Ethereum within a specific period (typically 7 days). The challenge mechanism can be designed to reward successful challengers, encouraging active public supervision and challenges against any errors. In ZK Rollup, cryptographic zero-knowledge proofs ensure the correctness of the state after ZK Rollup, and zero-knowledge proof technology also allows Ethereum nodes to quickly verify a large batch of transactions compressed together with very little computational resources.

"L1+L2" is a potent combination.

Ethereum's future will be a combination of "L1 blockchain + L2 system equivalent to L1's trustlessness" (hereinafter referred to as "L1+L2"), especially after ZK Rollup solves the technology for general-purpose smart contract platforms. Such a combination not only maintains the current decentralization level of Ethereum but also provides high throughput services, making it the best choice to carry tens of trillions of dollars of the Internet of Money.

L2Beat (L2Beat.com) provides an overview of the various stages of maturity and "trustlessness." This website comprehensively presents the maturity of various L2 projects (including "real L2" and "pseudo L2").

L2Beat judges the "trustlessness" of each L2, here "maturity," based on five risk factors. These five risk factors are (1) State Validation (verification of state validity), (2) Sequencer Failure, (3) Proposer Failure, (4) Exit Window (the window period for user escape), (5) Data Availability. For example, as shown in the figure below, only when all five risk factors are evaluated as green can a STAGE 2 rating be obtained. Currently, among all ZK Rollup projects, only one has achieved STAGE 2 rating, which is DeGate, as shown in the figure.

Why is it so difficult to technically achieve "L1 trustlessness equivalent" in L2? The core reason is that L2 systems are very complex, the more complex a system, the higher the difficulty of achieving secure operation, and the longer the construction time required for secure operation. Both Optimistic Rollups and ZK Rollups are new technologies, especially ZK Rollup's use of cutting-edge cryptography in the field of zero-knowledge proofs. In fact, the application of ZK Rollups is rapidly advancing the development of zero-knowledge proofs in the academic field. Among the L2 systems displayed on L2Beat, to my knowledge, the earliest to implement ZK Rollup, Loopring, has gone through at least 5 years from project initiation to now; DeGate, which achieved STAGE 2, took 3 years and underwent 5 rounds of "security audits" and a serious bug bounty program with Immunefi.

Recently, the blockchain industry has engaged in heated discussion about modular DA (Data Availability) layers, with some proposing to migrate DA services out of Ethereum to use other cheaper data services. If DA services are migrated out of Ethereum and Rollup systems can still maintain L1 level "trustlessness" in design, I fully support it. In fact, there are such schemes, and excellent teams are actively exploring and building in this area. However, recent discussions actually aim to abandon L1 level "trustlessness," downgrading the concept of L2 to "pseudo L2" for lower costs, which is unacceptable.

All financial application L2s aim to scale up and eventually become important members of the "L1+L2" system. Therefore, whether to abandon L1 level "trustlessness" from the start in design must be carefully considered. Abandoning "trustlessness" will severely hinder "pseudo L2" from scaling up. Currently, among the L2 projects running on L2Beat, the capital scale in value locked of "real L2" is more than 10 times that of "pseudo L2," indicating that the market cares about real trustlessness.

The Sun vs the Ether

There are many contenders for the race to be the Number 1 platform underlying the Internet of Money, amongst them Bitcoin, Ethereum, and Solana. For one, there is Bitcoin, which is the best known blockchain and the highest in market capitalization. Yet because it is not a general purpose blockchain, it is unlikely that it will be able to contain the many applications of the new Internet of Money.

More interesting is the challenged posed by this bull run’s major competitor, Solana. While the introduction of ZK compression could vastly improve throughput potentially, Solana has the problem of decentralization to contend with.  If Solana is the Sun of this bull run, because in the immortal words of Will Ferrell in the movie Zoolander it is “so hot right now”, one centralized ball of flame, it is also very likely to fail by that same logic. It is too centralized by design to withstand the eventual stress tests from all sides attacking it’s decentralization. This is not to say that it will not have a great run, and that there will be no value created in this dynamic ecosystem, before it’s eventual demise due to a lack of trustlessness, a “flaming out” if you will, in the albeit distant future.

Conversely, there is Ethereum. Historically, "ether" referred to a hypothetical invisible medium believed to permeate the universe and serve as a conductor of light waves. This name was not chosen without consideration. If the Ethereum seems relatively cool and detached by comparison, it is a feature not a bug. It encapsulates more things and more people, because it is more decentralized. This laissez faire attitude has led some to accuse it of being slow, but it has shown to be an all-encompassing church for all creeds, precisely because of how easy it is for the everyman to set up a node and be part of the ecosystem. Because Ethereum is both decentralized and high throughput by design, it has already won the race.

Note: In writing this essay, many ideas are indebted to the writings of BitGulu and Andreas Antonoupolos.