A major announcement just happened in Ethereum researcher channels.
Even those tapped into the crypto industry seldom follow these channels, never mind those outside of it. We can think of these channels as where long-term plans get codified into the blockchain.
This is where discussions and debates take place regarding the establishment of new standards for Ethereum.
These conversations included things like allowing new tokens to trade on Ethereum (ERC-20s), making new assets such as NFTs (ERC-721 and ERC-1155), and, some years ago, even discussions around the creation of staking.
These conversations eventually led to the altcoin boom in 2017, the NFT frenzy in 2021, and the rise of staking yield on the Ethereum blockchain.
It’s even where smart contract standards like those that led to DeFi primitives – think onchain swaps and borrowing – originated… and the creation of onchain vaults that can act similarly to hedge funds, minus the middlemen.
All this to say, if a standard gets the green light in the researcher channels, it often leads to a gold rush of new investment opportunities.
It’s where new asset classes in crypto are born.
And the most recent dialogue is centered on one specific use case. One that will fuel a topic we’ve hit on several times before…
The Web3 agentic economy.
Developers who want to make a change to Ethereum can submit what’s called an Ethereum Improvement Proposal (EIP) to the community.
Anybody can submit an EIP. It’s not a centralized body or private organization.
The same goes for who can comment on the EIP. The original idea can be submitted, and if it gains traction, it moves into the draft stage.
This is where an idea gets a formal EIP number. The proposal we’re talking about today has just moved into the draft stage and into what’s called the ERC bucket.
ERC stands for Ethereum Request for Comment. It’s mostly for ideas that create and define different types of tokens, like altcoins and NFTs.
Essentially, this distinguishes the proposal as an “idea of merit”. And while there are literally thousands of ERCs that have hit the draft stage since Ethereum began a decade ago, we’re bringing this particular one up for three reasons.
The first is that the authors are well-known. Two of them are from the Ethereum Foundation itself, a group of core developers that research improvements for the network. The third is the artificial intelligence lead at the biggest Ethereum-based wallet, MetaMask.
These are well-connected and established experts in the industry.
The second reason we bring up this proposal is that the community is taking note.
Within hours, various industry insiders were chiming in with their thoughts on the matter. This included author researchers at the Ethereum Foundation, research leads from major layer-two chains, and even developers from some top infrastructure projects.
So, it has grabbed the attention of all the biggest components of the ecosystem.
Even the debates within the discussion boards show an unusual level of excitement.
We typically don’t see this type of response on the heels of a proposal. That tells us this specific proposal has significant interest behind it.
So what is this proposal, and what will it do?
That leads us to the third reason we’re bringing it up to you today.
The proposal is formally called ERC-8004 Trustless Agents.
It was created on August 13 by MetaMask AI Lead Marco De Rossi, as well as Davide Crapis and Jordan Ellis from the Ethereum Foundation.
The summary of the proposal reads:
This ERC extends the Agent-to-Agent (A2A) Protocol with a trust layer that allows participants to discover, choose, and interact with agents across organizational boundaries without pre-existing trust.
Cutting through the jargon, this means the new standard will create a system for agents to interact with one another autonomously. Think of it as one agent considering a transaction with another agent.
How can one agent trust another that they are trustworthy and won’t conduct anything malicious? Or that they will follow through on their intention?
That’s where the next part of the summary comes in…
[ERC-8004] introduces three lightweight, onchain registries – Identity, Reputation, and Validation – and leaves application-specific logic to offchain components.
The new standard proposes establishing a registry system as a way to check historical records in terms of how trustworthy an agent can be. And in doing so, it allows AI agent creators to move things like compute offchain.
This is crucial…
Compute is very costly when it comes to AI, making it unsuitable to sit on the blockchain itself.
It would be like trying to run a fully self-driving vehicle on a car built in the 1980s.
It would require a lot of custom solutions that would be prone to failure and would hinder the wide-scale adoption of self-driving vehicles by the masses.
Running a reputation score helps streamline what we care about most – the trustworthiness of an agent.
It also aligns the incentives of the developer to maintain that trust over time, so their hard work doesn’t go to waste, forcing them to start back at zero.
It’s an elegant check and balance system.
The authors then go on to try to explain the solution in everyday terms…
Trust models are pluggable and tiered, with security proportional to value at risk – from low-stakes tasks like ordering pizza to high-stakes tasks like medical diagnosis. Developers can choose from three trust models: reputation-based systems using client feedback, stake-secured inference validation (crypto-economics), and attestation for agents running in TEEs (crypto-verifiability).
The proposal begins to paint a picture of a future where highly tuned, task-specific agents will be coming onchain. It also paves the way for AI models running on highly specialized software to begin running tasks in a 24/7/365 marketplace.
One that allows for easy integrations across virtually any application.
Permissionless technology is pretty crucial to understand here. What it means is that with just a wallet, anybody can interact with the application. There is no specific API that needs to be set up, partnership agreements, or intermediary technology that must enter the mix.
It’s a plug-and-play mentality. And AI agents will be able to tap into a growing ecosystem of solutions as they see fit.
Davide Crapis has been vocal on allowing the Ethereum Network to unleash what it calls GigaGas.
Without getting too into the weeds, “gas” is what’s used to pay for transactions on Ethereum. These fees go to validators for processing transactions.
And what he’s alluding to is an interesting concept called Jevons’ paradox.
William Stanley Jevons was a 19th-century economist who accurately predicted that the steam engine’s efficient use of coal would increase demand.
Jevons’ paradox suggests that more efficient use of a commodity or resource breeds increased demand and consumption of said resource, because increased efficiency makes these resources cheaper to use.
In cases of boosted efficiency, we’d probably expect the resource price to trend lower, but higher demand continues driving prices higher.
Simply put, increased efficiency leads to lower costs per unit but increased demand and consumption. We’ve seen it with coal, with fuel-efficient cars, with energy-efficient appliances, computational resources, and now… Ethereum’s “gas” fees.
Crapis is suggesting that AI Agents will be the biggest consumer of network fees and likely drive the price of these fees even higher, in the same way the steam engine did for coal.
He’s looking ahead to what AI Agents are about to Ethereum… And the network needs to make gas incredibly cheap because there will be so many agent-to-agent transactions. That’s what’s behind “GigaGas.” It’s a way to ready the network for a significantly higher volume of transactions.
That’s because most people’s expectations of how big a change AI agents represent to the market are too low. These agents will automate financial transactions, paying for services, and helping you with everyday tasks.
It’s no longer about users clicking a button a couple of times a week or month to make a transaction… We’re talking about dozens of agents conducting transactions throughout the day.
It’s an order of magnitude shift in blockchain usage.
The technology is fast approaching. New AI solutions will be introduced at a pace that traditional solutions will find hard to keep pace with. Iteration and innovation will accelerate.
We can also see how even the most modern-day solutions to AI can begin to make their way onchain.
And at a time when many individuals are beginning to worry about how centralized entities might be attempting to form public bias, these trustless systems will act as a check on that concern.
That’s what public blockchains are all about at the end of the day.
It’s an unbiased, single source of truth. One that is global.
Verified and reputable Trustless Agents running on these networks will impact our everyday lives in ways we can’t fathom today.
And the best part…
It’s coming sooner than most realize. ERC-8004 is gaining positive feedback. There might be some adjustments and modifications made, but we should expect to hear that a new standard is coming to cryptocurrencies.
In its wake, we’ll see the birth of a new asset class enter the market, creating a whirlwind of opportunities. And before you know it, you’ll have your own army of agents paying your car payment, shopping for insurance deals, monitoring the weather, rebalancing your portfolio, and so much more.
The future is coming onchain.
Your Pulse on Crypto,
Ben Lilly
Senior Crypto Analyst, The Bleeding Edge
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
