Improvement is an Understatement

On a technical level the validators of Ethereum work to reach consensus on the correct state of the network, and although not a simple task, there is at least a well defined set of operations to run, check, and agree upon. But move up a few layers, beyond the world of bits, and you find a much messier, more human, level of consensus that needs to be settled. The world of Ethereum Improvement Proposals (EIPs) is the arena in which any changes to the blockchain must prove themselves not mathematically, but socially.

Among the EIPs — which map roughly to the Internet Protocol Standards that define the internet — there are proposals that cover low level decisions effecting the execution of code on the EVM, standards for sending data between dApps, and suggested APIs for implementing interactions between end users.

Core EIPs tend to be more involved, and can require the network to change the lowest levels. Decisions here tend to be hard fought as enshirning more functionality directly into into the system comes with some tradeoffs (see this popular travel blog for details)

The less strictly enforced ERCs sit at the opposite end of the spectrum. As they don’t technically require any approval, an ERC can be used by anyone as soon as it is public. While Core proposals focus on technicalities of blockchain infrastructure, ERCs are an attempt to codify an idea into a composable interface; an API to interact with the blockchain state.

ERCs define the functions which end users experience when interacting with the blockchain day to day, and a select few have gone beyond mere technical documentation, or utilities in apps. From here we will look at the ERCs which captured the most fundamental use cases to not only establish themselves as standards, but to spawn fully fledged industries.

‘Tokenizing’ things like voting power in a group, or rewards for playing a game can be useful to standardize comparisons between participants. The issue is that the use cases for a token to represent voting share, and a token to count how much of an in game asset you can spend are very different.

In the ERC-20 spec, essential things like total supply, balance of a user, and how transfers should be carried out are defined in their simplest terms. This lets any smart contract developer inherit the standard, then customise the base functions to their needs. So although the voting power in your group may not be transferable, the in game asset will be. This gives app developers something to work with, since they can build sets of functions that make predictable calls to the base functions, regardless of what internal logic the specific token has decided to add. The biggest example of an app capatilizing on this composability is Uniswap.

Uniswap has built a $3 billion business around facilitating swaps of tokens between users in a permissionless way. Multiple exchanges and trading platforms have popped up to support other actions made possible by tokens like lending, borrowing, or staking. Combining all ERC-20 tokens on the market gives ~$300b in value… not bad for a few lines of code written in 2015.

Perhaps the most extreme example of an ERC producing unpredictable upside is the, at times notorious, ERC-721. Here the concept of non-fungibility was incorporated into an interface similar to ERC-20, allowing each token to represent a unique item, and in many cases, an associated image or metadata.

One major difference between the fallout of this token and ERC-20 is that as well as marketplaces doing billions in trading volume, these tokens have generated everything from ticket platforms to IP rights for movies. The subtle difference of assigning each token associated data allowed artists to produce works with predefined scarcity, or incentivise ownership by acting as ‘token gating’ to online communities.

Some of the more inventive use cases include minting an onchain SVG to redeem a matching watch, or playing chess against an onchain engine and minting the result. And with standards like ERC-6551 giving NFTs the ability to control assets, the next wave could be exciting for a lot more than just price speculation.

Until now the ERCs have mostly been examples of how a widely accepted idea can generate an industry, but the next highlights the risk of having application standards rather than protocol level enforcement. ERC-2981 was not the first (or last) attempt to put NFT royalties onchain, and the debate around the philosophy of recurring revenue heated up in 2022. Whether artists could survive long term without an ongoing revenue stream, or if it was excessive to expect income from a single artwork for all time divided artists and marketplaces alike.

Royalty payments were generally determined by the marketplace, with Opensea making a push for the NFT industry to empower creators to be able to enforce royalties on their collections across all marketplaces. Their rival, Blur, managed to hijack volume through clever incentives for both creators and buyers, including disencentivizing cross selling assets on multiple marketplaces.

In the end ERC-2981 could never ‘solve’ onchain royalties; as with many ERCs it was a standard interface, not a compulsory part of the protocol, and to this day there is no strict enforcement of royalties for NFT creators.

The final industry defining ERC that we will look at is, by some way, the most ambitious. It does not just aim to implement a new token standard, but to emulate the entire transaction and account structure of Ethereum.

ERC-4337 defines an application layer infrastructure in an attempt to replace conventional EOA accounts with smart contract accounts (SCA). From bundler services which send the SCA transactions onchain, to paymasters which can allow dapps to compensate gas spend to drive usage, the industry generated by ERC-4337 is wide reaching.

Companies such as Pimlico, Candide, and Stackup are building this infrastructure and given the billion dollar valuations of other early providers of ERC tooling like Uniswap, the coming years could be an exciting time for both these companies, and SCAs in general. With complimentary ERCs like ERC-6900 (Modular Accounts) users can have much finer control over the actions their wallet can take, and with ERC-7212 (Precompiled for secp256r1 Curve Support) even increased security by signing with a passkey rather than a seed phrase account.

While all types of EIP are capable of producing major shifts in the ecosystem, the most colourful are the ERCs. This is not to say that Core changes can not produce creative new features. Things like the addition of CREATE2 unlocked lots of fun use cases where smart contracts could be predictably deployed to a set address. But the fact that they require such coordination to happen makes them fell a little less magical.

The optionality of throwing a novel design pattern out into the world and having it capture the attention of the network and be implemented at scale is something very important for the still relatively early crypto industry. That said, power laws apply to many aspects of life, and protocol standards are certainly one of them. Of the thousands of proposals which have been made, a small fraction are accepted, and even fewer can claim to have single handedly spawned billion dollar industries. So before you go out and create the next big token standard, go check out the wonderful list of all EIPs so far to see what you’re up against. Good Luck!