Key takeaways
- Pontem has released Lumio's layer 2 devnet on Solana, enabling developers to deploy applications across multiple blockchains without code changes.
- Lumio framework supports Move VM, Solana VM, and EVM, with plans to add WebAssembly and Cairo to reduce developer vendor lock-in.
- Pontem's Liquidswap AMM is launching a governance token to transfer protocol ownership and decision-making power to the community.
- Lumio uses fraud proofs similar to Ethereum rollups to provide trust-minimized asset bridging between chains, addressing security concerns with traditional bridges.
- Developers can now build once and deploy across Ethereum, Solana, Aptos, and other chains, enabling DeFi applications to access native assets on multiple networks.
One app, any chain: what Lumio does
Ashton Addison: For viewers who haven't seen our first interview, what are Pontem and Lumio?▶ 00:42
Alejo Pinto: Pontem is a product development studio focused on three core products. The first is a multichain wallet that supports Aptos and other Move ecosystem projects, as well as EVM chains and Solana. The second is an AMM built on Move: fully immutable and permissionless, covering constant product and concentrated liquidity. A governance token is launching very soon, which means ownership will be transferred to the community to set parameters, govern fees, and decide on grants and future development. The third is Lumio, which is still in development and in testnet. We recently released it onto a public devnet on Solana. It solves a problem developers have today: vendor lock-in based on which virtual machine or language you chose to build with. Lumio is a framework to deploy with any virtual machine onto major L2s and L1s. Right now we have a deployment on Optimism, and we recently released on Solana.
Ashton: Why is developer lock-in such a hindrance, and how did you arrive at this solution?▶ 03:47
Alejo: That's where we started. We built applications on Aptos with the Move virtual machine, and when we wanted to offer our users native ETH and SOL, we ran into the issue. Move is a great language, with improvements in performance and safety, but you're vendor locked in because the L1s with native assets like ETH, OP and SOL don't support Move. To deploy there, you'd have to migrate and maintain two code bases, which incurs a lot of tech debt. As we talked to more developers, we realized we weren't the only ones. Uniswap is deployed across BNB, Avalanche and a host of EVM-compatible L1s and L2s, but something like Raydium or Jupiter can only be deployed on one. If we can solve this issue for ourselves, we can solve it for a whole host of other people. So we built an extensible framework. We're starting with the MoveVM, the SVM and the EVM, but you can add others like WebAssembly or Cairo. Covering these three reaches a large portion of the market, and for new developers, choosing Move on Ethereum or EVM on Solana becomes a lot easier.
Ashton: Could an application like Uniswap use Lumio to launch on Solana?▶ 06:34
Alejo: That's exactly it. Technically it's deployed as a layer 2. Take Solana: Uniswap could offer its users native SOL, rather than bridged SOL through something like Wormhole, and access to the meme coins trading on Solana. They would deploy on Lumio, an L2 on Solana, and people can then roll up or bridge their assets over. It's as trust-minimized as possible. It's architected much like rollups on Ethereum, using fraud proofs, and we're also researching validity proofs. You can deploy the same EVM contract without code changes and offer your users native Solana assets.
If we can solve this issue for ourselves, we can solve it for a whole host of other people.
Bridges, rollups and the road from devnet
Ashton: Bridges have been a major source of hacks. Does a rollup avoid those risks?▶ 08:07
Alejo: Bridges are one of the central points of failure, and many users still don't feel comfortable using them. A rollup still has a bridge, but the whole settlement of the rollup is secured in a trust-minimized way. There are fail-safes: if the sequencer were compromised, people can still force withdraw their assets back to the L1. So the implementation is arguably safer than something cutting across different L1s. Rollups also give you more bandwidth and lower costs. Even though Solana is cheap relative to Ethereum, most transactions still cost a few pennies, and 10 to 20 cents during congestion or for more complex transactions. Making that orders of magnitude cheaper matters for high-frequency trading or on-chain gaming, like earning experience points while chopping wood in an MMORPG, with settlement and compatibility with L1 assets.
Ashton: This is a devnet release, so nothing is moving from Uniswap onto these chains yet?▶ 11:18
Alejo: What we have deployed now is a testnet on Optimism. The idea is to merge this implementation onto the Optimism sequencer once they move to a shared sequencer, which gives compatibility with other L2s in that ecosystem, like OP Mainnet and Base. Our goal is to deploy a collection of L2s, which we call the Lumio Collective. Each implementation is separate, just as Uniswap has separate deployments on BNB or Avalanche. So a Uniswap deployment on Solana would be an implementation on Lumio, which we call Lumio on SOL. That contributes to fragmentation, which isn't our focus, but I think intent-based cross-chain swapping will make the experience a lot better for users. Off-chain solvers, such as market makers, fill the swap, so the user doesn't have to switch networks or bridge, as long as the applications are deployed on those networks.
What Move brings to Liquid Swap
Ashton: What makes Liquid Swap stand out as an AMM?▶ 15:31
Alejo: Liquid Swap is the first core implementation of a constant product market maker in the Move language. Aptos was launching with a gap for this core infrastructure, so we built it and then added concentrated liquidity. Functionally it's similar to Uniswap or PancakeSwap, and most DeFi technology is relatively commoditized, so the differences come from small improvements. One is parallelism. With Move, you can run transactions in parallel as long as they don't collide on the accounts they target. That's a benefit of Rust-based virtual machines like Move and the Solana VM, and the EVM can't do it out of the box, so Move is marginally more efficient. Another is safety. Uniswap v4 is introducing hooks, which makes the AMM a platform for other applications, and the EVM carries risks like re-entrancy attacks or overflow errors. The Move VM has a native prover for formal verification. You can do it with Solidity too, but it's not built into the virtual machine, so you rely on third-party libraries. Move has a lot more guardrails to stop developers shooting themselves in the foot.
You do need a responsible transition to full automation, otherwise things can easily go off the rails.
Token governance and the push toward automation
Ashton: How does on-chain governance change what a product can be?▶ 19:50
Alejo: These protocols are becoming almost like little nation states, and their tokens are a tool for coordination. Uniswap is fundamentally no different from Robinhood or Coinbase in that it enables the exchange of assets, but those companies don't let users or third-party stakeholders change the product in real time. On a blockchain product, token holders can change the protocol's parameters in real time. Right now it's mostly humans making decisions, maybe writing bots and algorithms to implement them, but I think very soon we'll see more decision-making by AI, algorithms or even LLMs, with humans probably remaining the end owners. For now most parameters are still set by humans, and there are security councils and vetoes because governance attacks are still an issue. But I can envision a future, maybe not even too far from now, where most products are not run by humans. Setting risk parameters for a credit protocol is mostly human today, and you can imagine that being entirely automated, with token holders, humans or machines, operating the protocol.
Ashton: When the Liquid Swap governance token launches, what will holders be able to do first?▶ 29:14
Alejo: Initially you don't want to hand over full control. The safest approach is an on-chain mechanism that updates parameters but keeps veto power, a security council with a window to roll back what it sees as a governance attack. We'll start with certain parameters, such as fees for individual pools. Protocol upgrades could also be voted on by the community with tokens, but an upgrade to malicious code could withdraw all the funds, so you'd want the council to inspect the code and make sure it has been audited by reputable auditors first. You can also build in guardrails on fees, for example differing between a volatile meme coin and a stablecoin. Over time you could remove some of these guardrails, for instance if trusted auditors commit on-chain to a hash of the code they audited, though then the auditors become something like the security council. You do need a responsible transition to full automation, otherwise things can easily go off the rails.
Ashton: How do token holders benefit, and how is that different from owning shares in a company?▶ 35:03
Alejo: Take the fee switch for Uniswap, which is likely to be implemented soon. These protocols are very similar to existing corporations, except that they run fully on-chain and no one can shut them down. Even a state-level actor like China, the US government or a coalition like NATO couldn't shut down Uniswap. Right now the fee is zero, but the treasury could accrue fees, and token holders could vote on what it does, such as buying back the token or distributing cash flows, as a Coinbase would. People don't like the comparison to dividends, but I think it's fundamentally different. A protocol's revenue logic is coded, so you can inspect it yourself and you don't need to trust an intermediary to distribute it. A company could lie about how much it makes, and because it's not on-chain, you can't see the money. Removing that intermediary means you don't need a regulator to tell you whether something is safe.
This interview has been edited for length and clarity. Recorded July 2024.

