EigenLayer has recently released the whitepaper for its protocol token, EIGEN, introducing several novel and complex concepts such as Intersubjective, Work Token, Token Forking, slashing-by-forking, and Intersubjective staking. These terms quickly became the focus of community discussions.
The original intention behind the design of the EigenLayer protocol was to generalize the use of Ethereum nodes, allowing them to execute other business logics to increase revenue and provide additional income for users participating in ETH consensus. However, while increasing the utility of nodes brings benefits, it also introduces risks. These risks include observable objective risks, as well as uncertainties that exist in the subjective and objective realms. Although such uncertainties cannot be completely guaranteed through cryptography and mathematics, they are based on “social consensus.” This is what is referred to as “Intersubjective,” which can also be translated as “social consensus.”
Because it is based on “social consensus,” the protocol token of EigenLayer also needs to be forkable, which requires the design of a dual-token isolation model and the initiation of challenges for forking, among other logical arrangements. It is more like Ethereum implementing a complete set of network logics, encompassing both (social) consensus and nodes, and the ability to fork, but it is not a blockchain.
In addition, I also recall some interesting anecdotes related to Work Token and Weak subjectivity.
Work Tokens have been discussed since around 2018, and they gained more attention thanks to Kyle Samani of Multicoin Capital. In simple terms, under the Work Token model, service providers can only obtain the right to perform network work by staking the native tokens of the network. In other words, nodes need to provide tokens and services to receive rewards. With the popularization of on-chain liquidity and DeFi protocols in recent years, the tokens that nodes need to stake can be contributed through third parties, eliminating the need for nodes to use a large amount of their own assets. This is also the model adopted by EigenLayer: users provide ETH, the Liquid Restaking Protocol provides liquidity, node operators provide hardware, and AVS provides business logic.
For more information on Work Tokens and other token models, I recommend starting with this article from Multicoin: https://multicoin.capital/2018/02/13/new-models-utility-tokens/
First, let’s define two key terms: Objective and Subjective. Errors caused by blockchain and decentralized networks can be classified into four types based on their nature:
Objective errors: These errors are based on data and cryptography and can be explicitly verified, such as the execution process of the Ethereum Virtual Machine (EVM).
Intersubjective errors: These errors involve social consensus among groups. When certain behaviors or judgments go beyond this consensus, they are considered Intersubjective errors.
Unobservable errors: These errors are only known to the victim and cannot be observed by others.
Subjective errors: These errors are entirely based on individual experiences and perspectives, resulting in a lack of consensus.
EigenLayer believes that unobservable errors and subjective errors are difficult to correct, and therefore proposes using ETH to avoid objective errors and EIGEN to avoid Intersubjective errors.
The concept of Intersubjective can be seen as a state between “objective” and “subjective.” The term is derived from “Inter” (meaning “between” or “mutual,” similar to “interactive” or “internet”) and Subjective. Therefore, Intersubjective refers to a subjective state formed among individuals, which is a consensus formed within a society through interaction.
For example, in the financial market, the statement “1 BTC = 1 USD” is generally not accepted. This consensus error can be classified as an Intersubjective error. Thus, if we were to explain Intersubjective, we can understand it as “social consensus,” which is the collective acceptance of certain ideas or facts within a group.
Although there are subtle differences between “social consensus” and Intersubjective in academic and professional discussions, with Intersubjective focusing more on describing the sharing process of subjective experiences and knowledge among individuals, while “social consensus” focuses more on the outcomes of collective decision-making and actions.
To briefly review the EigenLayer protocol, users can deposit ETH into the Liquid Restaking Protocol, and these protocols will stake the ETH to run Ethereum validation nodes. These validation nodes will also run various middleware businesses such as oracles, cross-chain bridges, and data availability services to provide services to end applications.
For AVS, it can also be divided into two types: objective and Intersubjective. Objective AVS is based on cryptography and mathematics and can be quantified and verified explicitly. In the EigenLayer design, these AVS can rely on ETH staked through Restaking as security guarantees. On the other hand, Intersubjective AVS, such as oracles, can only rely on social consensus among nodes since off-chain data cannot be verified on-chain. Sufficiently recognized data by multiple nodes is considered trustworthy.
In summary, Restaking ETH will serve as the work token for objective AVS in the EigenLayer protocol, while EIGEN will serve as the work token for Intersubjective AVS.
Token Forking is also a novel concept. The ability of blockchains to fork usually refers to the open-source code or network itself, rather than ERC-20 tokens, which are theoretically not forkable and do not possess this capability. This is because tokens, as smart contracts, rely entirely on the objective attributes of the EVM.
However, EigenLayer believes that token forking is a safeguarding measure within their framework, even if it is a once-in-a-century event. If the number of malicious attackers in the EigenLayer network exceeds half, regular users can fork the tokens in this way, allowing all users and AVS to choose the corresponding tokens based on their own circumstances. This is referred to as “slashing-by-forking,” and it is derived from an article written by Vitalik Buterin nine years ago.
To support this forking capability, a significant amount of logic needs to be added. For example, if the tokens are forkable, can EIGEN still be used as collateral for lending protocols? Therefore, they designed a dual-token isolation model where EIGEN cannot be forked, but another token called bEIGEN can be forked. They also designed the process of initiating forks, as well as compensation logic, and more.
In fact, Ethereum’s PoS consensus has long had a concept called “Weak subjectivity” (also invented by Vitalik Buterin), which is another concept that lies between “objective” and “subjective.” And only PoS blockchains have the attribute of Weak subjectivity.
For PoW networks, the cost of computational power competition is significant, making the longest chain the safest, making it entirely “objective.” For PoS networks, the cost of block production is negligible, and the cost of attack is low, so nodes that have just joined the network need to retrieve social information to find this “Weak subjectivity” and objectively participate in the PoS process. Therefore, there are some “subjective” factors before entering the correct network.
However, for nodes that have entered the correct Ethereum network and participate in consensus, all consensus processes and EVM operations are objective, guaranteed by cryptography and mathematics. For example, the determinism of EVM input and output, and the rules of double signing that would result in slashing, are all explicit rules.
This article is authorized for reprint by ChainFeeds.
