Interview | Web3’s next fronteer is P2P transactions: Yellow

Alexis Sirkia, Captain of Yellow Network, explains how P2P transactions can fix the scalability problem that continues to plague blockchains.

Blockchain adoption has come a long way, with both governments and institutions recognizing the technology’s potential. However, core issues, including speed, scalability, and decentralization, persist. Network scalability can scarcely keep up with the demands for computing power.

In an interview with crypto.news, Alexis Sirkia, Captain of Yellow Network, outlined his vision for a missing layer in Web3 infrastructure. According to Sirka, Web3 needs a trustless peer-to-peer communication system that reduces blockchain overhead without compromising security.

crypto.news: What are the benefits of Web3 and DeFi compared to traditional finance, applications, and centralized servers?

Alexis Sirkia: The main benefit I see is this idea of trustlessness, which is a system where you don’t need to rely on human intermediaries. That’s what decentralized architectures like Bitcoin and Ethereum introduced. They allow you to build applications that automate processes previously controlled by centralized entities.

With Web3, you can have autonomous entities, meaning smart contracts or DAOs, controlling business logic, not people. That’s powerful. Think of companies that don’t depend on human oversight. They run autonomously, efficiently, and without risks of corruption, insider trading, or biased decision-making.

That’s the dream: someone can build the next unicorn startup, launch it via smart contracts, and it runs itself. Like Uniswap, once deployed, the business just runs.

So Web3 is about liberating people from tasks computers can do better. Compliance, trust, supervision, these don’t have to be human responsibilities anymore.

CN: What are some examples of real-world use cases where this trustlessness would be an advantage?

AS: Take Amazon, for example. It already provides a kind of trustless experience. A buyer purchases from a seller, and Amazon ensures the transaction happens smoothly — it acts as an intermediary.

But with Web3, that role can be replaced by a smart contract. You don’t need a centralized authority — you can create systems where autonomous logic ensures that if something goes wrong, there are clear, pre-coded rules that handle it.

The promise of Web3 is that we don’t need to rely on people anymore to guarantee these outcomes. These companies can run themselves without human input. That’s what makes it so powerful.

It’s not about removing human decision-making — it’s about removing dependency on humans for trust, enforcement, and operation. Computers aren’t corruptible. If I have to choose between people and math, I’ll choose math.

Of course, there are trade-offs. Smart contracts aren’t as flexible as humans. For example, if there’s a strike and goods are delayed — is that the seller’s fault? A smart contract wouldn’t necessarily know how to handle that nuance.

Humans can adapt to new or unexpected situations; smart contracts follow rules. So right now, they can’t fully replace the gray areas that exist in the real world.

CN: From your perspective, where are we now on the journey toward fully autonomous Web3 applications?

AS: I’d say we’re at a similar stage to AI in the year 2000.

The concept of AI had been around since the 1960s, but by 2000, the infrastructure wasn’t quite ready. The tools were there, but the computing power and platforms weren’t widely available.

Web3 is in a similar place now.

The term “Web3” itself was only coined in 2014 by Gavin Wood. It’s only been about a decade, and we’re just now reaching the point where the infrastructure is solidifying.

What we’ve done with Yellow is part of this missing infrastructure. And now, with Bitcoin, Ethereum, and Yellow, we finally have a nearly complete foundation to build the next generation of autonomous, decentralized applications.

CN: Can you go into more depth about what was missing in the infrastructure — and what exactly you’re building with Yellow?

AS: Sure. With Bitcoin and Ethereum, we had the beginning of trustless infrastructure — systems that removed the need for intermediaries in transactions involving three or more parties. For example, in a financial transaction, there’s the sender, the receiver, and a third-party authority like a bank or a clearinghouse. Ethereum replaced that third party with a smart contract — essentially a programmable judge that guarantees what’s been coded will execute as agreed.

But what was still missing from the Web3 stack was trustless peer-to-peer communication — direct, cryptographically secure interactions between two parties without relying on centralized servers or full blockchain verification for every step. That’s the layer we built with Yellow. It’s what enables two people or two systems to do business together without needing a third party to constantly supervise every message or transaction.

CN: Why is that peer-to-peer layer important?

AS: In the real world, most business happens directly between parties. You go to a store, pick up some items, and only at the end do you pay — that’s when the system settles. The tax authority, for example, doesn’t care about every apple or orange you pick up, just the final bill and taxes owed.

Web3, up until now, was missing that dynamic. Instead, we had 30,000 validators verifying every single step — every action and handshake — when really, that level of oversight is unnecessary for most interactions.

So what we created with Yellow is similar to Lightning Network for Bitcoin — but for Ethereum and any smart contract chain. These are called state channels. Two parties open a channel between them, agree on a set of transactions, and cryptographically sign each one off-chain.

You and I could agree to trade a bitcoin for $100,000, then another trade — say, Ethereum for $4,000 — and each step is signed. Only when we want to settle do we involve the blockchain. If one party tries to back out, the smart contract adjudicates based on the signed proofs and penalizes the dishonest actor. Both parties have collateral in the contract, so there’s a meaningful risk of cheating.

CN: So what’s the actual benefit in practice?

AS: It means we can do trustless business without involving 30,000 computers every time. I don’t need to trust you because the system guarantees that if anything goes wrong, the smart contract will enforce the rules. And the best part? It’s incredibly fast.

We’re trading at the speed of light — peer-to-peer, no global consensus required. It’s like having a local blockchain just between two parties, and it’s far more scalable than anything happening fully on-chain.

That’s the thing. Blockchains are fundamentally limited in processing power. They’re not built to be high-performance computers. In fact, the entire Solana network today has less computational capacity than a 1984 Intel processor, or a DEC VAX system from the 1970s. I’m not exaggerating — those machines had one million instructions per second. That’s more than Solana currently delivers.

So it’s not that blockchain is a bad system — it’s just not meant to run an entire app. You need it for consensus, settlement, and arbitration — but the heavy lifting should happen off-chain. With Yellow, we’re saying: keep the blockchain where it makes sense, and shift the rest to peer-to-peer layers that are faster and more scalable.

CN: And technically, where is the information stored? What if one party disputes a trade — where does the smart contract get the data to decide who’s right?

AS: Good question. Every transaction in the state channel is signed cryptographically by both parties and contains a reference to the previous one, much like a mini blockchain. So each new transaction forms a chain that builds on the last — and the newest one always wins.

If there’s a dispute, the smart contract simply checks which party has the most recent signed proof. That version is considered valid. The logic and rules for resolution are already coded into the smart contract in advance, so there’s no need for oracles or external inputs in most cases.

CN: So, as long as you have the latest signed proof, you can resolve any dispute?

AS: Exactly. And because both parties signed it, there’s no ambiguity. The contract doesn’t care about claims or arguments — it just looks at the math. That’s what makes it trustless. And because you’re not involving the broader blockchain until the final settlement, the entire process is lightning fast and doesn’t bog down the network.

In real-world business, you don’t ask a court to validate every handshake — you only involve a third party if something goes wrong or at the end for settlement. We’re replicating that with cryptographic proofs and smart contracts. It’s a shift from “trusting the network” to trusting a minimal, verifiable interaction layer — which is what Yellow provides.

CN: What do you think is the most immediate or compelling use case for this peer-to-peer system you’ve built?

AS: Initially, we built this for trading — that was the obvious one. We built a broker called NADAs and a protocol called NLAX using all of this infrastructure. But what surprised us is how many other use cases emerged once we opened the Yellow SDK. Developers are doing things we never imagined. At the ETH Prague hackathon, for example, there were over 800 participants, and two of the finalists were using Yellow. That’s incredible validation.

We’ve seen payments apps, trading apps, games — everything from Tinder-style trading interfaces to multiplayer Tetris, and even a real-time Snake game between two players. One developer built a swipe-based trading UI in just a weekend. Another used Yellow to power an instant payment system at a rave in Ukraine — real people using cryptographic payments live, in a chaotic offline setting.

We even saw this tech used at a real-world rave in Ukraine. A payments app built on Yellow allowed people to instantly pay for drinks and items at the event. There was no waiting for confirmations — it was just fast, cryptographic payments. Because there was no dispute, there was no need to settle on-chain immediately. That’s exactly how real business works.

We’re launching the mainnet in two months. But even before the launch, we already have real developers building real businesses on it. That’s been amazing to watch. At Korea Blockchain Week, we saw someone demoing an app they built on Yellow just weeks earlier at a hackathon in Europe. That was surreal — the ecosystem is already moving ahead of us.

It reminds me of when Vitalik was first launching Ethereum. Back then, people didn’t really understand what it was for. I actually got in during the presale — bought 20,000 ETH at 27 cents. I remember Vitalik spending two hours explaining the vision. It felt revolutionary. This feels the same. People don’t fully get it yet, but the ones who do are building fast.

CN: Are there any trends that people are now overlooking, perspectives you think are missing from the current conversation?

AS: Yeah. When Bitcoin started, barely anyone understood what it really meant. I remember back in 2013 — even earlier, really — we were organizing Bitcoin meetups, building some of the first crypto companies. I had a banker tell me that I should be careful because this was illegal. That was the perception: that it was dangerous, criminal, fringe.

People didn’t get it. Almost everyone was skeptical. But those of us who had a technical background — I worked as a software engineer at a space center — and also understood the history of money, we saw it differently. And there weren’t many of us. It was a small handful of people who understood both worlds: the tech and the economics. That’s who made the biggest bets early — and they’re the ones who’ve done incredibly well since.

Then Ethereum came along. And again, most Bitcoin people didn’t get it. There were maximalists — obsessed with sound money, but blind to the broader concept of trustlessness. They couldn’t see that the same principles could apply not just to money, but to contracts, businesses, and apps. It was a huge mental leap.

Now, we’re in that same place again. Most people don’t understand that the concept of Ethereum — programmable trust — can be extended one layer deeper into peer-to-peer systems, where two people can do business directly, without any trust, without relying on a global consensus for every step.