I caught up with Eliott, a rising young star in the blockchain space and newest recipient of a KEF Grant to further research and develop for Kaspa!
Chad: Hi Eliott, congratulations on the KEF grant! Can you briefly introduce yourself and tell us what you are up to with this work?
Eliott: Thank you! It’s a pleasure to be back working in crypto. I’m a Pure Mathematics Master’s graduate from ETH Zurich, where I first explored crypto through my thesis on oracles, supervised by Yonatan. I’m passionate about crypto because it allows me to tackle complex, theoretical mathematics while contributing to enhanced security and innovation in blockchain protocols. Applying my studies to crypto, specifically oracles, MEV, and markets, has been the first time I’ve felt I truly understand the purpose of my theoretical studies. I spent six months in banking in 2024, as I wanted to understand how banks worked and how they moved the economy. However, I quickly found the work unchallenging and misaligned with my values—I wasn’t proud of my contributions to society. So, I’ve been eager to return to working with Yonatan for some time now, as his philosophy very much aligns with my own.
*MEV resistance is a defence against “maximal extractable value” (MEV) attacks on crypto transactions. MEV is a hidden fee that can affect the price of transactions and lead to failed trades.
I’m currently working on oracles, MEV, and decentralized exchanges. These three topics are closely related and essential for security, exchanges, and fairness. Oracles are crucial for blockchains to function in real-world scenarios; for example, price oracles secure billions in TVL for DeFi platforms like Aave. However, during my thesis, I noticed academia lacks focus in this area, just as it does in MEV solutions. I want to tackle the challenge of designing better solutions to create more transparency and fairness in the underlying mechanisms of DeFi.
*Aave Pronounced (A-vey) Greek word meaning “Ghost.”
It is is an open-source, non-custodial protocol that allows people to lend and borrow cryptocurrencies through decentralized finance (DeFi). Essentially, it provides a peer-to-peer money market for cryptocurrencies, removing financial middlemen from the equation.
I will also start posting on Medium (@noclue122) once I consider my expertise sufficient to express myself without making too many mistakes. I’d like to share aspects of my work, and improve awareness about unresolved challenges, and how we are working towards solving them. This approach is highly motivated by reading Shai’s work, which got me into Kaspa in the first place. His Medium posts made me realize I could understand some mechanism behind crypto, which at the time seemed like an obstacle. That is what sparked my interest in the field.
Chad: Will your work be on Kaspa core network, or is it generalized deep work that any development team could implement?
ELIOTT: Technically, any team can implement this work once the research is published, as others can adapt and integrate the findings. My goal isn’t to gatekeep research but to advance academia, so I’m all for other teams using it to address existing challenges. However, as Sutton recently noted on X, multi-leader consensus is critical for robustoracles and MEV resistance, and we’re designing our oracle with this principle in mind.
Yonatan gave a great explanation during his talk, highlighting that no matter how good an oracle is, it often ends up centralized. A single consensus validator is relied upon to deliver the oracle price on-chain and could arbitrarily manipulate or censor it. Similarly, many attempted solutions for MEV front-running exist, but the final consensus leader can choose the transaction order, allowing them to censor transactions even in designs that prevent front-running (see Budish’s HFT arms race). Kaspa’s architecture makes the foundation of this work much easier and will certainly enable a superior design.
Chad: We Asked CHatGPT to explain Yonatan’s tweet on Oracles to an eighth grader.
The Problem:
Imagine you’re playing a game where you have to guess the price of candy from a store. You ask your friend (the oracle) to check the price online and tell you. But what if your friend lies or makes a mistake? That’s the oracle problem — you can’t be sure the information you’re getting is true.
Why Is This a Big Deal in Crypto?
In cryptocurrency, blockchains (like Ethereum) sometimes need real-world information, like the price of Bitcoin, to do things like process payments or trades. The problem is — who provides that information (the oracle)? If the oracle lies or is slow, people could lose money.
Two Scenarios:
1. The Easy-to-Catch Cheater (Good)
Suppose a company controls a bridge between two blockchains. If they cheat by changing the balance or lying, everyone can immediately see the lie and stop using the bridge. The cheater gets caught once and loses everything.
2. The Sneaky Cheater (Bad)
Now imagine the same bridge but with a central oracle providing data (like prices). If the oracle gives wrong prices (even by accident), it’s hard to prove if they did it on purpose or if it was just a bad internet connection. This makes it impossible to hold them accountable — meaning they could cheat slowly without getting caught.
The Solution (Maybe):
The person tweeting suggests using a group of miners (like people voting) to estimate the correct information — like a group of friends guessing the candy price instead of just one friend. This way, it’s less likely someone could cheat. The method would rely on majority honesty — the same way mining in Bitcoin works.
But to make this work, you’d need super-fast mining (many blocks per second), so the information stays fresh. That’s why PoW-DAG (like Kaspa) was mentioned — because it’s really fast and could help solve this problem.
Bottom Line:
• It’s easier to catch a company lying about money than to catch a company slowly feeding wrong information.
• Using many miners to agree on information (like the price of candy) might solve the problem — especially if the network is super fast.
• The person tweeting thinks fast PoW-DAG networks like Kaspa might help solve this big problem in crypto.
This is basically about truth in information and how hard it is to prove if someone is lying — especially when you rely on them for data.
ELIOTT: I think it’s crucial to help everyday users understand how oracles matter in crypto. ChatGPT did a good job breaking it down. The core challenge with oracles is figuring out the “real” price of an asset. But since no one knows the absolute “truth” (as it’s the whole point of the oracle in the first place), lies can be hard to spot—especially when your idea of the truth is influenced by those lies. That’s why penalizing false inputs is so tricky.
Take Aave, a DeFi lending platform. If someone can manipulate an oracle’s price feed—even for a moment—they could liquidate your loan unfairly. For instance, if Aave’s oracle only pulls data from MEXC and MEXC’s liquidity dries up (because not your keys not your coins), attackers can sell Kaspa heavily there, push the price down, and trigger liquidations—even if other exchanges still show a fairer price. This is how poor oracles create unfair outcomes and MEV opportunities.
When reading whitepapers of current oracle protocols, I find the proofs lacking. Claims are made like “participants will behave like this” without any mathematical modelling or proofs. This is a bit frustrating as my whole studies have been based around getting zeros when my proofs were lacking. It doesn’t sit right with me, billions are secured on these protocols that rely on assumptions, and unproved claims. I’d like to change that.
Chad: Kaspa solving real world problems is the main message for Kaspa onward. Utility and every day use adoption. Kaspa integrated into The IoT, etc. What are your thoughts on this?
ELIOTT: In crypto, there’s no truly great oracle solution yet—most rely on claims of security with no mathematical proofs, and make assumptions that won’t necessarily hold. For example, every current oracle solution assumes that there is at least 50% of participants that will be honest. And I explained in this X thread why this doesn’t necessarily hold. As long as this problem is set aside, the blockchain will rely on trust, that Chainlink nodes won’t one day have an incentive to lie, or a validator to censor the on-chain delivery. It’s a real world problem, and it isn’t talked about enough.
I’d like to design an oracle solution, MEV and trading designs that have the same level of beauty as Dagknight. This challenge will probably take a while given I have only been in crypto research for a year now, but I think the experience will be memorable.
MEV resistance is a defense against “maximal extractable value” (MEV) attacks on crypto transactions. MEV is a hidden fee that can affect the price of transactions and lead to failed trades.
X: @elimmea
LinkedIn: linkedin.com/in/eliott-mea
