Little Learnings #1

ERC 8004: Reputation, Identity & Validation for AI Agents

This is the first in a new series (whimsically titled Little Learnings) of educational posts I will be releasing every Friday. The intention behind these are to break down some new interesting topic or development happening in the crypto space. God knows there’s enough happening, and it’s often complicated — so I’m here to try and simplify it a bit.

This post is sponsored by Open Campus

A very fitting sponsor for the first issue of this free newsletter, Open Campus is building the onchain financial layer for education.

They recently announced a partnership with the Government of Madhya Pradesh (a state in India) to bring 50 MILLION academic records onchain.

Every graduate in the state will have a verified digital credential on EDU Chain.

This is a pretty incredible achievement and will push forward the future of the role of blockchain technology in education.

Learn more here (check it out, it’s genuinely a very cool partnership).

---

ERC 8004: Reputation, Identity & Validation for AI Agents

This week we’re talking about ERC-8004. ERCs are basically just standards used by Ethereum. You may have heard of ERC-20s (the standard for fungible tokens), or ERC-721 (the standard for non-fungible tokens aka NFTs). They can be for just about anything, and ERC-8004 is a new standard being proposed for AI agents.

Here’s how it’s described in the official proposal:

This protocol proposes to use blockchains to discover, choose, and interact with agents across organizational boundaries without pre-existing trust, thus enabling open-ended agent economies.

Trust models are pluggable and tiered, with security proportional to value at risk, from low-stake tasks like ordering pizza to high-stake tasks like medical diagnosis.

The way it works is by creating three registries: an Identity Registry, a Reputation Registry, and a Validation Registry.

• The Identity Registry basically creates an NFT for every agent, giving them a unique onchain identifier that is censorship resistant and decentralized, theoretically allowing it to be implemented into any protocol.

• The Reputation Registry is an interface where a reputation score for agents can be created and determined. Basically, feedback can be given (both onchain and offchain) to determine how reputable an agent is.

• The Validation Registry basically allows third parties to check and verify an agent’s work (using ZK proofs or other methods).

Combining them all together, you get a way to have AI agents act and interact with other agents, and the real world, trustlessly.

It’s pretty magical. This is the kind of stuff that excites me about crypto. Not some new meme coin, not another ponzi, not speculating on what the price is gonna be tomorrow, and certainly not betting on a prediction market what words someone is going to say on live TV.

But using Ethereum, the world computer, to create a protocol that increases the potential of AI agents? In a way that doesn’t seem possible without crypto technology and blockchains?

Hell yeah. This combined with the x402 technology I wrote about a little while ago, makes me very excited for the future of AI agents interacting onchain.

Gonna keep these issues short and sweet, but if you want to learn more, this is a fantastic article that goes into much greater depth.

Thanks for reading, see you next week with another little learning!