Skip to content

The CCS Interview

Supra's Jon Jones: Scale the layer one, don't split it into modules

Jon Jones, co-founder of Supra, explains why the project has gone from an oracle to a vertically integrated, multi-VM layer one. He also covers its bridging plans, the Project Blastoff 2.0 DeFi campaign and an AI trading agent in the works. The conversation was recorded in February 2025, shortly after Supra's mainnet launch, and what follows reflects what Jones said at the time.

Interview by Ashton Addison Recorded February 2025 5 min read

Key takeaways

  • Supra is a high-throughput layer one blockchain with vertically integrated native oracles, automation, cross-chain communication, and randomness services built into the protocol.
  • Vertical integration provides shared security, shared liquidity, and composability by having validators run consensus on both transactional and oracle data rather than relying on external protocols.
  • Native oracle services eliminate the security weakness of depending on external third-party protocols, which become the weakest link in decentralized applications.
  • AI agents will likely conduct most future DeFi transactions and blockchain interactions, requiring efficient vertically integrated stacks optimized for agent experience beyond just developer experience.
  • Supra builds its layer one and protocols modularly, allowing components like consensus algorithms to be swapped out if better alternatives are proven.

One chain, native services

Ashton Addison: What is Supra, and what makes it different?▶ 00:46

Jon Jones: Supra is a high-throughput layer one, but what we're bringing to the market is vertical integration. That means the layer one comes with native oracles. Oracles bring external data on chain, for example price feeds that power DeFi. We also offer automation, cross-chain communication such as bridging, and randomness. That's good for consumers, who get information fast and cheap, and for developers, who can build applications on these integrated services. If you want to build a DEX or a perpetuals protocol, you need a high-throughput layer one. You also need an external oracle and an external automation protocol. Automation is essentially an "if this, then that" system, and in Web3 it isn't something layer ones have built in. We combine all of it in one place. We also see a lot of future DeFi and on-chain transactions being done by bots, or agents. They will use the most efficient routing for cross-chain communication, and we believe a vertically integrated stack is the most efficient for agents too.

Ashton: Why does it matter that oracles are native to the chain rather than supplied by third parties?▶ 05:36

Jon: A computer should have a ledger, and it should be able to speak to the outside world. From our experience with oracles and getting off-chain data on chain, we realized quickly that you need both the ledger and secure communication of data. Ethereum can use external protocols for oracles, bridges and other services, but that's much more inefficient and slow. You also no longer have the security of the underlying layer one. L1 consensus rarely fails. But as soon as you introduce an outside oracle or a bridge, your application is only as strong as the weakest protocol, and you add complexity and extra security assumptions. With a vertically integrated L1, the validators run consensus on transactional data and on oracle data. We call that shared security. The three pillars are shared security, shared liquidity and composability.

Ashton: Is vertical integration the opposite of the modular approach? What are the trade-offs?▶ 09:27

Jon: Most software is built in a modular way, and we build our layer one that way too, so we can swap components in and out. For example, we have Moonshot, a high-throughput, low-latency consensus algorithm. We also have another called Sailfish that could be a contender, and if it proves itself we can swap it in. But in crypto today, "modular" means separate protocols, each with its own validators and nodes, its own token and its own security assumptions. We find that very inefficient and less secure. Some people argue that multiple protocols are more decentralized. But the layer one itself should be the maximum security, and it should be distributed and decentralized. If you scale the layer one and expand its validators and nodes, you get maximum shared security. We think you should scale the layer one and provide oracle services under one hood.

As soon as you introduce an outside oracle or a bridge, your application is only as strong as the weakest protocol.

MultiVM and moving assets across chains

Ashton: Supra runs on the Move VM, but you also talk about MultiVM. What does that mean for developers?▶ 12:51

Jon: We've been watching Move since the early Libra days at Facebook. Our researchers dug deeper and found that Move is a more secure and intuitive language. It's much harder to introduce bugs in Move than in Solidity. It's a language that's easier to not mess up. It was also created for Web3, with assets as first-class citizens. It's similar to Rust, so we thought it could onboard hundreds of thousands of developers. But you want to give developers multiple languages they may already know. Many applications are already built in the EVM world, and others spent years learning to build on Solana. So optionality matters. A multi-VM structure lets them build on different virtual machines on one shared-security layer one. We launched mainnet with Move, and EVM is already on devnet. Solana and one more virtual machine will be live later this year. That gives us four virtual machines and programming languages under one shared-security layer one.

Ashton: Is there a bridge for moving assets between chains?▶ 18:11

Jon: A bridge launched on Supra just last week, and more are coming shortly. Supra is also building its own native bridge. We'll be able to bridge assets from EVM to Supra, and from other Move chains to Supra, and back and forth. Some bridges are already live. The native ones should be live very soon, within the next month.

Blastoff, DeFi and the vertical integration narrative

Ashton: More than 500,000 KYC'd participants joined Project Blastoff before mainnet. What's next for it?▶ 22:32

Jon: We were lucky to be one of the first, maybe the very first, to do this Learn-to-Earn point system. Point systems aren't as hot these days. But being early gave us time to educate our community, which was quite special. We had to stretch it out because we had to wait for the token generation event and mainnet. It turned out very successful: 550,000 real people completed KYC and participated. We're in a different market now, so Project Blastoff 2.0 will be different. It's about going from learning to implementation. It will be aligned with our DeFi campaign, guiding people to participate in staking, swapping and more. We want to uplift Supra's DeFi ecosystem and bring value to the projects and developers launching on Supra, while rewarding users.

Ashton: The modular thesis was dominant in 2024. How do you get the case for vertical integration across?▶ 27:09

Jon: No one is really reading white papers anymore, so we've done it through education, social media, X and our community. The modular thesis was so strong in 2024, so it was an uphill battle. But most technology trends toward vertical integration, and we had this idea from day one. Solana is a great example of some vertical integration, with its focus on user experience, high throughput and low latency. Consumers like it. I believe that in 2025 we'll see protocols like Supra and Aptos trending toward vertical integration. Aptos is deploying its own randomness, and I'm sure the next thing you'll hear is that it deploys its own bridge. I think 2025 could be the year vertical integration becomes a bigger thesis and rivals the modular camp.

We think you should scale the layer one and provide oracle services under one hood.

AI agents that trade for you

Ashton: How might AI agents fit into Supra?▶ 29:48

Jon: I tend to think too far ahead about what's going to happen, so I have to ground myself. We haven't put much AI content on our socials yet, but we will shortly, and we'll focus on what we can do today. We're working closely with an indexing product, and there will be an announcement soon. We're going to implement an AI agent that crafts trading strategies using historical data and our oracle network. Users will be able to talk to an agent or bot, which crafts a strategy for them. Then they can hit a button and actually implement that strategy. You can only do that with a combination of vertically integrated services. You need the layer one, the oracles, automation and historical index data. For version one, I'd like to see it interact with Supra, maybe Ethereum and Solana. In the end, we want it to be multi-chain, interacting with almost every DEX on every chain.

This interview has been edited for length and clarity. Recorded February 2025.

Jon Jones
Jon JonesJon Jones is co-founder and Chief Business Officer of Supra, a vertically integrated layer 1 blockchain. He studied political economy at UC Berkeley and Chinese studies at National Taiwan University. Before Supra he founded the consultancy Blockcamp.io, served as president of the Taiwan chapter of… Full profile →
Share this interviewXLinkedIn
▶ Watching the interview