The Stablecoin Surge is Coming. Can Blockchains Keep Up? | Money Rails 2025 | Polygon Labs

Transcript

Thank you also to the panelists in the previous panel. Was an amazing amazing panel and a perfect segue for what I'm going to talk about. I am Gregori Roshu. I'm the co and founder of Pi Squared and we build the fastest decentralized network for instant payments for humans and AI. We heard in the previous panel about the surge of the Genius Act and the stable coins and we also know that the AI agents are also all about payments and autonomous activities and and microp payments. All of this will put an immense pressure on the web 3 technology.

They will demand millions of transactions per second. The question is are blockchains ready for this? At pi squared we believe they are not. We believe that something else is needed. On this slide from A16Z we have different transaction volumes for different money rails around. And what is fascinating to see is that crypto stable coins actually are second and more than than uh than credit card transactions and even the adjusted transaction volume is higher than PayPal

much higher 10x higher all this is telling us that we are going to have a lot of transaction volume pressure on blockchains very soon. And this is not everything. We also have the AI agentic future which is already here. In this picture going from top to bottom, we see how large institutions like Open AI partner with Stripe in order to incorporate instant checkouts in Chad GPD where your AI agent can pay directly for things for

you on your behalf. Similarly, we see Microsoft, oh sorry, Google, sorry, Google introduced the AP2 protocol for a similar reason so that agents can communicate with each other and and and pay each other. These are protocols done and proposed by private institutions and they are private protocols. However, going bottom up now, we see also open protocols like the X42 or its cousin H42 which incorporate web3 powered payments in the HTTP protocol where the agents and humans

alike can pay through websites for everything they want to using crypto. Again, all these will put a lot of demand on transactions per second. The two together, the Genius Act plus all the agentic payments will overload blockchains massively. Even one of them in isolation will already put more pressure on blockchain that they can sustain today. both together will make blockchains essentially unusable the way they are today.

Because of that, at Pi Squared, we decided to challenge the elephant in the room, the blockchain itself. And specifically the big problem that very few tend to agree with is that blockchains impose a total order on all the transactions. Look at these marbles and the funnel, right? Imagine that those are payments. They have nothing to do with each other. I pay you, you pay me. An agent pays another agent. Why the heck should we sequence them? Why should we enforce a total order first place? Let's not even

worry about blocks yet because blocks are just an optimization. The big problem appears before we even talk about blocks. The fact that we want to sequence all the independent transactions, embarrassingly independent transactions, we force ourselves to sequence them. This is a very very serious issue that almost nobody wants to address frontally. We believe that that is the problem that is what stays in the way of scalability in web 3 and blockchains.

And unfortunately we see a certain phenomenon where many people identify web 3 and crypto with blockchains. How else without blockchains we can talk about crypto and web 3? No, blockchains are just one mechanism, one way to implement web 3 and crypto and give us everything we want. Decentralization, peer-to-peer, payment, security and so on. You can have all those without the total order imposed by blockchains and enter facet our protocol but pi squared.

which eliminates the scalability limit from the blockchains. We joke but not really internally and externally that the blockchain trlemma which says that you cannot have decentralization scalability and security at the same time. We joke that that the blockchain dilemma is a dilemma of blockchains not a trilema of web 3 or crypto. In fact, our protocol facet has an amazing speed because it is embarrassingly parallel is that it operates at the speed of the internet. All you need is to send a message and receive confirmation which is on average 100 milliseconds on the

globe. That's the finality time. Each transaction is finalized, validated, settled in parallel and independently of all the others. So you can have unlimited TPS with our existing devet. By the way, go to fast set.xyz to play with the network if you want. With our existing network, we already achieved 250,000 TPS without any optimizing it. So we are going to optimize further by mainet. We will have 1 million TPS. But that's not even loadable.

I think what is bad is to aim as a new blockchain especially at 100,000 TPS. say well we are going to have 100,000 TPS that's ridiculous that's too little for the future that is already here you need uncapped TPS and we can have it these days scalability comes from the massive parallelism there is no sequencing no blocking nobody blocks nobody else everybody is settled finalized independently of anybody else and finally the cost is ridiculously low The actual cost for each transaction is hundreds of a cent because you don't have to spend time on stupid unnecessary

consensus to ensure strong consistency which you don't need anyway first place. So metaphorically this is the difference between blockchains and facet. Imagine that gravity is the internet, the bandwidth of the internet, right? So to the left is what blockchains do. They inhibit gravity. They inhibit the internet by forcing a total order unnecessarily on all these beautiful marbles, the transactions, they don't need to wait for each other.

No matter how smart you try to be to sequence in the smartest possible way, to group in the smartest possible way, in the end it will be slow because of the total order. There's nothing you can do because everybody has to communicate to everybody else in the network to agree that this is the total order. And that's wrong because you don't need a total order first place for payments. And we can get into that if you're interested, but that gets technical. What happens in fast set to the right is that you just let all the marbles fall under their own gravity on our under weight at the maximum speed of the internet. So you get literally verifiable settlement at the speed of the internet and we should forget all this nonsense with

sequencing transactions unnecessarily. And with that, I'd like to thank