A deep understanding of the significance of Intersubjective staking in EigenLayer: collective subjectivity, tyranny of the majority, and forkable tokens

Author: @Web3Mario

Introduction: During the May Day holiday, Eigenlayer released its Eigen Token white paper. Strictly speaking, this is not a traditional economic white paper aimed at introducing incentive models and values, but rather presents a brand new business system—Intersubjective staking based on Eigen Token. After reading the full text of the white paper (without delving into the appendix) and considering the interpretations from predecessors, I have some personal thoughts and understandings that I would like to share and discuss with everyone. To conclude, I believe the significance of Intersubjective staking lies in its proposal of a consensus system based on a forkable ERC20 token model, which can be used for decision-making on certain "group subjectivity" issues while avoiding the tyranny of the majority.

What is "Group Subjectivity"

Correctly understanding Intersubjective is a prerequisite for grasping the significance of this system. There seems to be no unified conclusion on how to translate this term on the Chinese internet. After reading Professor Pan Zhixiong's article, I found it quite agreeable that the concept of "social consensus" can be well used to understand its meaning. However, I feel that using "group subjectivity" to refer to this concept seems to align more closely with the literal translation and understanding. Therefore, in the following text, I choose to use "group subjectivity" to refer to Intersubjective.

What exactly is "group subjectivity"? In the context of EigenLayer, it refers to a broad consensus among all active observer groups within a system regarding the correctness of the execution results of a certain matter. Thus, this matter is said to have intersubjective characteristics, or group subjectivity. We know that one of EigenLayer's core values lies in decoupling the consensus layer from the execution layer, focusing on the construction and maintenance of the former, thereby service-ifying consensus and reducing the development costs of Web3 applications, fully tapping into the potential market demand. In the narrative of the white paper, EigenLayer seems to position itself as a decentralized digital public platform that can execute digital tasks for third parties. Therefore, it is necessary to analyze the boundaries of its services, clarifying what types of digital tasks can be "trusted" to be executed by it. In the context of Web3, "trusted" usually means that a system is designed with cryptography or economic models to prevent errors in the execution of digital tasks. Thus, the first step is to classify the possible execution errors of digital tasks. EigenLayer categorizes the execution errors of digital tasks into three main types:

Objectively Attributable Errors: This type of error refers to execution errors of digital tasks that can be proven through a set of objectively existing evidence (usually referring to on-chain data, or data with DA) without relying on trust in any specific entity. For example, if a node in Ethereum signs two conflicting blocks, this error can be cryptographically proven. Similar cases include the fraud proof process in OP Rollup, which can determine errors by re-executing a set of disputed data in the on-chain execution environment and comparing the results.

Group Subjectivity Attributable Errors: This type of error refers to execution errors where all participants in a system have a consistent subjective judgment standard regarding the execution results of a certain digital task. This type of error can be further divided into two categories:

• Errors that can be identified at any time by backtracking past data. For example, in a price oracle, if the spot price of BTC on Binance at 00:00:00 UTC on May 8, 2024, is $1, this error can be identified at any time afterward.
• Errors that can only be observed in real-time, such as malicious censorship. For instance, if a transaction is maliciously refused execution by a group of nodes for an extended period.

Unattributable Errors: This type of error refers to execution errors that have not yet established a definite consistent judgment standard among the group, such as judging whether Paris is the most beautiful city.

Intersubjective Staking aims to effectively address digital tasks with group subjectivity characteristics, which means it can respond to execution errors of digital tasks that are attributable to group subjectivity. It can also be seen as an extension of on-chain systems.

The Tyranny of the Majority in Current Solutions

The so-called tyranny of the majority is a political term that refers to the situation where a vast majority in a parliament unites to forcibly pass policies that infringe on the rights of the minority. After clarifying EigenLayer's goals, let’s look at the types of current solutions to such problems. According to EigenLayer's summary, there are two main types:

1. Punishment Mechanisms: These mechanisms typically use cryptoeconomics to deter malicious behavior by punishing the staked funds of malicious nodes. Staking slashing is one such method. However, this approach can lead to a troublesome situation. Imagine if an honest node submits proof of wrongdoing, but most nodes in the system decide to collude and commit wrongdoing, they can choose to ignore the proof and even retaliate against the honest node.

2. Committee Mechanisms: These mechanisms usually establish a fixed committee of nodes that, in the event of a dispute, will approve the accuracy of the proof of wrongdoing. However, whether the committee is trustworthy becomes a major issue. If the committee nodes collude to commit wrongdoing, the system effectively collapses.

Both of these solutions evidently encounter the problem of the tyranny of the majority. This highlights the difficulty of solving such issues. Although there is a consensus on the accuracy of execution results, the lack of objective verification capabilities forces a shift from trust in cryptography or mathematics to trust in people. However, when the majority chooses to act maliciously, current solutions appear powerless.

Avoiding the Tyranny of the Majority through Forkable Work Tokens

So how does EigenLayer solve this problem? The answer lies in designing a forkable on-chain work token and leveraging the social consensus capability brought by staking this work token to handle group subjectivity digital tasks and avoid the tyranny of the majority.

What exactly is the so-called social consensus capability brought by forkability, and how does it avoid the tyranny of the majority? First, EigenLayer points out that its inspiration comes from the study of ETH PoS consensus. It believes that Ethereum's security comes from two aspects:

• Cryptoeconomic Security: By requiring block-producing nodes to stake funds and designing a punishment mechanism for malicious behavior, it ensures that the economic cost of wrongdoing exceeds potential gains, thus deterring malicious actions.
• Social Consensus: When malicious actions lead to a chain fork, due to a consistent judgment standard regarding the correctness of execution results, any well-meaning or honest user can choose the fork they believe to be correct based on their subjective observations of different fork execution results. Thus, even if malicious nodes control the majority of staked funds, the tyranny of the majority will be accompanied by users abandoning the malicious fork, causing the value of the forked chain to gradually surpass that of the malicious chain. For example, most CEXs will choose the correct but less staked supporting fork and abandon the incorrect but heavily staked malicious chain. As social consensus becomes widespread, the value of the malicious chain will gradually disappear, and the forked chain will re-emerge as the "orthodox fork."

We know that the essence of blockchain is to reach a consensus on the order of a set of transactions within a trustless distributed system. Ethereum has designed a serial execution environment (EVM) based on this foundation, so when transactions are consistent, the EVM will achieve a consistent execution result. EigenLayer believes that the evaluation of execution results for such transactions is mostly objectively attributable, but there are also cases of group subjectivity attribution. Specifically, this refers to the evaluation of chain liveness. In Ethereum's PoS consensus mechanism, there is a special Inactivity Leak mode. When more than 1/3 of the nodes cannot produce blocks correctly due to some unknown circumstances, the cryptoeconomic security of PoS will be compromised. An extreme example would be if a war causes the internet in one region to be completely disconnected from another region. In this case, Ethereum will experience a fork. When the consensus mechanism detects this situation, it will enter Inactivity Leak mode, where new blocks will not be rewarded with inflation, and inactive nodes will gradually be slashed until the staked funds of active nodes exceed 2/3 again. This will allow the two forked chains to gradually regain their respective cryptoeconomic security.

After this, regarding which chain will become the so-called "orthodox fork," it can only rely on users to actively choose based on their own judgment standards. This process is what is referred to as "social consensus." As users actively choose, the value accumulated by the two forks will shift until, under the competition of cryptoeconomic security, one fork achieves a clear victory. This process can be seen as security granted by social consensus.

Summarizing this phenomenon, EigenLayer believes that Ethereum relies on social consensus to identify and resolve group subjectivity errors related to chain consistency, known as chain liveness attacks. The core of this social consensus capability originates from forkability. When a divergence occurs, rather than hoping to immediately judge which side is at fault, it relies on subsequent users voting with their feet to resolve the divergence through social consensus. This avoids the protocol from suffering the tyranny of the majority, as a minority of honest nodes will not be immediately slashed due to collusion, allowing for a comeback capability. In addressing group subjectivity issues, this approach demonstrates its value.

Therefore, following this reasoning, EigenLayer references and upgrades a consensus model from an on-chain gambling protocol called Augar, proposing a forkable on-chain work token named EIGEN. An Intersubjective staking mechanism is designed around EIGEN to resolve execution consensus for group subjectivity digital tasks. When there is a divergence in execution results, conflicts are resolved through the forking of EIGEN and relying on social consensus within subsequent time windows. The specific technology is not particularly complex and has been introduced in some articles, so I will not elaborate further here. I believe that understanding the above relationships will help grasp the significance and value of Eigen intersubjective staking.