Latest / Crypto Coin Show / Polygon’s AggLayer and Type 1 Prover changes EVERYTHING in Web3! | Blockchain Interviews
Transcript
- 0:00I'm Ashton Addison from the Crypto Coin Show and today on Blockchain Interviews with Brendan
- 0:05Farmer, co founder of Polygon. Brendan, Welcome to the show and thank
- 0:08you for taking the time today. Yeah, thanks session, thanks for having
- 0:12me. You're very welcome. Excited to dive into the latest updates with Polygon
- 0:17and what Polygon is doing for the blockchain industry overall. I feel like twenty
- 0:20twenty four so far is kicked off to an amazing start, with the further
- 0:25increase adoption with the Bitcoin etf the having potentially more incredible updates for all the
- 0:32layer twos that are coming out, and I know that there are some issues
- 0:36that need to be solved to really unify blockchain as one holistic ecosystem, and
- 0:41I think Polygon has a part to play with that. So let's dive into
- 0:44the latest updates that Polygon and the team I've been working on. I'm really
- 0:48excited to hear the details. Maybe you can talk a little bit about first,
- 0:53I've seen this egg layer update, and you know these are some big
- 0:57This is a big project, so we might spend some time on this,
- 1:00but I would love to just kick it off to hear what you and your
- 1:04team have been working on this whole time, and then we can dive into
- 1:07all the details. Yeah. Sure, So the aggregation layer, the eye
- 1:11layer, is a big, big one for us. We've really been thinking
- 1:15about these sort of the underlying problems for a long time. And just to
- 1:21take a step back, if you think about how we think about blockchain scaling
- 1:25right now, the debate is sort of stuck between the monolithic side and the
- 1:30modular side. And so the monolithic side says, look, when we talk
- 1:34about scaling and crypto, we're really talking about scaling access to liquidity, access
- 1:38to shared state, maintaining composibility. If we create a bunch of chains that
- 1:44can't access liquidity where it's sort of available and useful, then we haven't really
- 1:51scaled in a meaningful sense. And so the modular side, on the other
- 1:56hand, says, actually, the way to scale is like, we'll never
- 2:02be able to bring Internet scale crypto with a single chain, Like we'll always
- 2:08run into bottlenecks, there'll always be contention for sure resources, and so fundamentally
- 2:14we need to look at a multi chain approach to scaling. And so I
- 2:20think that both sides get something right and something wrong. I think the monolithic
- 2:24side is correct that really what we're trying to do is scale access to liquidity.
- 2:29Like if liquidity is fragmented, that creates, you know, bad user
- 2:35experience, It creates complexity, but it also creates cost for users because they
- 2:38encounter slippage, their execution is worse. And then on the other side,
- 2:43on the modular side, I think that they're correct that we need horizontal scaling.
- 2:47We need to be able to add block space so that we don't run
- 2:51into this situation where we have like these intractable limits to scaling, but at
- 2:57the same time, like we need some way of taking fragmented block space and
- 3:02unifying it so that users can access shared liquidity insured state. And so fundamentally,
- 3:07that's what we've been thinking about with the aggregation layer, is how do
- 3:10we take a modular, multi train ecosystem and unify liquidity and state across those
- 3:17trends. Yeah, I appreciate that explanation, and I feel like the sweet
- 3:22spot is somewhere in between, right, because we're seeing chains get really big
- 3:28and running into issues, but we're also seeing so many different blockchains coming out
- 3:31that could probably be working better together. To unify their liquidity. I feel
- 3:37like that's very important if we're going to be getting to the next stage of
- 3:39adoption where bigo it want us to come in. Yeah, so for us,
- 3:45I don't know if you've ever seen the meme like thesis, antithesis,
- 3:49synthesis, and so for us, like one way to look at it is
- 3:52that the original thesis from Bitcoin and ethereum, and then for blockchains like Solana
- 3:58is like the thesis is monolithic and then the modular thesis arose in response to
- 4:03the limitations of the monolithic thesis. And so for us, like the synthesis,
- 4:08the way forward is this aggregated thesis where we have a multi chain ecosystem,
- 4:13but we have the mechanism to unify state and liquidity across those chains.
- 4:17So maybe you could talk about the vision. How did you guys plan all
- 4:21this out, because this is a huge initiative to try and unify blockchain together.
- 4:28What was the vision and how did you start building this aggregated layer in
- 4:31solving all of the different problems without making it too complex. Yeah, sure
- 4:36so our I guess the main design principle with the ag layer is simplicity and
- 4:44minimalism. Like, we want to build a protocol that can support Internet scale
- 4:48cryptoidoption, and I think in order to do that, we need to build
- 4:53something that's minimal and that's not opinionated and can accommodate a range of chains running
- 4:59on different executions environments, using different tokens, using different mechanisms for sequencing,
- 5:04and so for us, like the I guess the foundation of this vision is
- 5:10ZK tech. And so the question was like, okay, we know that
- 5:15right now, for two chains to interoperate, they need to like two L
- 5:18twos. You need to go from your initial L two to ethereum and then
- 5:25to your destination chain. So if I want to move tokens from chain A
- 5:29to chain B, I need to withdraw from chain A. I need to
- 5:32wait until the proof of my withdrawal hits Etheroreum and that block on Etheroreum is
- 5:38finalized, So that's like twenty five minutes, and then I need to deposit
- 5:42into chain B. So that's another you know, however, twelve minutes of
- 5:46latency. And so the the concept was like, how do we create low
- 5:53latency interactions between chains in a way that's safe and trustless. And it turns
- 5:59out that we do that by using zk tech, So we can let Chaine
- 6:03A and Chain B interoperate however they want to. They can run full notes
- 6:09for each other, they can use a shared sequencer, they can rely on
- 6:12builders or relays. Like the coordination mechanism is sort of out of scope for
- 6:15us. But what we do guarantee with the AG layer is that if chain
- 6:20B receives a message from Chaine A at lower latency than ethereum, Let's say
- 6:25like chain B receives a message from CHAINEA and two seconds later wants to start
- 6:30building blocks on that message that they received from CHANEA. Then Chain B can
- 6:34commit to the sort of state of Chain A that produced that message, and
- 6:42we can use zk tech to ensure that chain B will never be finalized and
- 6:46accepted by ethereum unless that block that Chain B is relying on from Chaine A
- 6:56is valid and is also finalized to ethereum. And so this approach like generalizes
- 7:00to a bunch of different cases, but fundamentally it's about like enabling chains to
- 7:05interoperate at really low latency while still maintaining safety. Like you know that the
- 7:11destination chain the chain bill will never be rugged by anything the chain A does,
- 7:16and that opens up the design space and enables them to coordinate in ways
- 7:20they currently can mm hmm. Okay, it's a lot of digest but I
- 7:28know that your team's focusing on the minimalism and simplicity, which is good,
- 7:31and I think people when they see how it works from a DeFi perspective,
- 7:38if you're just somebody who's going to make a swap or make a move of
- 7:42some crypto coins to another blockchain, they won't have to understand all of this.
- 7:46They'll just be able to have a seamless experience. I would hope definitely
- 7:48not. You will not have to know anything about ZK tech or what the
- 7:51chains are doing on the back end. But the sort of interface that we
- 7:57enable is what's what we're calling like bridge and call, So you can if
- 8:01you have funds on one chain, you can bridge them and call some other
- 8:05contract in like one interaction. And so we think that you know, if
- 8:11you have like you said, if you have USDC on one chain and you
- 8:16want to buy ether on the chain that has the deepest liquidity, then you'll
- 8:18be able to do that in a really really simple and user friendly way.
- 8:22That's very cool. And now, in terms of unifying the liquidity on these
- 8:28different chains, how exactly will that work? Will it be to say,
- 8:31there's twenty EVM chains that all have a million of liquidity that's not communicating with
- 8:37each other with each other, and that sort of moves into a single pool
- 8:41that's worth twenty million. That lowers the costs and the slippage and et cetera.
- 8:46Yeah, so I think this is a really good question. So I
- 8:50think that there are two ingredients for unifying liquidity. There first is bung ability.
- 8:56So right now, if you think about how third party bridge is work,
- 9:00like, if I want to bridge from polygon to arbitrum via wormhole,
- 9:05then what I'm doing is I'm locking my eth and then I'm minting wormhole eth
- 9:11on arbitrum. And the problem is is that wormhole eth is not fungible with
- 9:16regular eth. I need some way, I need some liquidity provider to get
- 9:18me from my wrapped synthetic version of an asset to the native version of an
- 9:24asset. And so providing fungibility in the polygon ecosystem is something that's really important.
- 9:30And the unified bridge and the AG layer allows us to ensure that L
- 9:35one native assets and L two native assets are fungible across chains, so I
- 9:39won't have to think, Okay, I have eth on the zkavm chain and
- 9:45I want to move eth on the pos Chin. I'm not minting zk evm
- 9:50eth. I'm just getting eth on the destination chain. And so this is
- 9:52a really really powerful thing for unifying liquidity. To your question, specifically,
- 9:58what will what will the world look like? Will it be a single super
- 10:03deep pool of liquidity or will it be a bunch of pools that that sort
- 10:09of automatically rebalance and and and you know, have arbitrage opportunities between them,
- 10:16so they effectively have the same execution. I think that that is an open
- 10:22question. You can imagine a world in which there's a single pool that has
- 10:26all the liquidity, and the problem is is that, you know it like
- 10:33getting best execution on that pool and getting priority becomes really expensive, and so
- 10:39normal users might be priced started using that pool. But one thing that you
- 10:43can imagine is those users would sort of batch their transactions and on low cost
- 10:50chains, and then that batched transaction would be executed on the pool with deepest
- 10:54liquidity, and so you know, the transaction costs would be amortized across all
- 11:00the users that are patching. I think that there are like similarities between that
- 11:03model and the model that you describe where you just have a bunch of different
- 11:07pools that are all automatically rebalancing and where orders are sort of being routed between
- 11:13them. I think that both are uniquely enabled by the AG layer and by
- 11:18sort of the vision that we're building, and I think it'll just be interesting
- 11:22to see what dominates. Yeah, it's very interesting, And I like that
- 11:26example that you gave with bridging through the wormhole. I feel like that is
- 11:31a setback right now where I don't think you can't you can't really tell from
- 11:35the front facing perspective, but you may not necessarily have the exact same ether
- 11:41on Polygon that you would on a different blockchain. And same goes for stable
- 11:46coins as well. You know a stable coin USDT or C that's on a
- 11:50certain blockchain, it's not the exact same coin on a different blockchain, and
- 11:52that's why the liquidity can't be combined, because it's not the same asset.
- 12:01Yeah. Yeah, and I'm curious also on the ZK perspective. You know,
- 12:07a lot of people just know zero knowledge proof says something that adds speed
- 12:11privacy, being able to confirm things with zero knowledge. Are there extra benefits
- 12:18to introducing this ZK tech into the agg layer or maybe you can touch on
- 12:22that a little bit more. Yeah, sure, So ZK tech is sort
- 12:26of the it's like the fundamental ingredient of the AGG layer, because we need
- 12:31a guarantee that, like, we need to be able to prove to ethereum
- 12:35that we have this giant multi chain ecosystem with thousands or tens of thousands,
- 12:39and we need to be able to prove to ethereum that actually all of these
- 12:43chains are sort of like operating on consistent views of the same state, so
- 12:50you don't have like chain A telling chain B one thing and sort of double
- 12:54spending on chain C. And so being able to do what we call proof
- 12:58aggregation is a really really important thing and has like a sort of a few
- 13:05different purposes in the ag lare. It provides, you know, a guarantee
- 13:09that everything that happens across all chains is valid. It provides a guarantee that
- 13:15chains are relying on consistent views of other chained states, and it also provides
- 13:20the safety guarantee where we know that if a single chain is compromised, that
- 13:26attackers cannot drain the entire ecosystem of funds because we're able to enforce that no
- 13:33chain can sort of withdraw more than the tokens that are deposited in that in
- 13:41that chain, even if that chain's prover is unsound, and so it's it's
- 13:46a really important ingredient in this overall vision, definitely. And with some of
- 13:54the EVM chains right now, you know, there's so many that are coming
- 13:58out, but from what I understand, some of them have some kind of
- 14:01ZK technology that they're looking to integrate, some of them not as much.
- 14:05When these new EVM chains move into the AGG layer, does that make them
- 14:11all sort of have a unified equal level of ZK technology implemented into the chain
- 14:16or how does that work? Yeah? Sure, So that's a good question,
- 14:20and it's a timely question because we just announced our type one prover,
- 14:24which sort of you know, achieves what you're describing. And so the way
- 14:28that I would put it is, in order to participate in this ZK powered
- 14:33ecosystem that runs on the AGG layer, chains must include ZK proofs like we
- 14:41need a guarantee from each chain that everything that happens on that chain is valid.
- 14:46And so if we think about like what the current state of the world
- 14:50is, we have all of these VM chains, like optimistic roll ups,
- 14:56like polyon pos, like ALTA ones use the EVM, and there's a ton
- 15:01of liquidity and state and huge user bases that are locked in these EVM chains,
- 15:09and they can't interoperate in the way that we're describing with the A layer
- 15:13because they don't support zk tech. And so what we released last week was
- 15:20the Type one zkbmprover and this is a really really cool piece of technology because
- 15:26it allows us to take any existing EVM chain, whether it's an optimistic roll
- 15:30up, whether it's a side chain like the polygon POS, or whether it's
- 15:33an altil one, and we can instantly start generating proofs for that chain,
- 15:37and so that will allow those chains to onboard into the EGG layer. And
- 15:41so this is like a really really important component of the overall vision, which
- 15:46is to take all of this fragmented liquidity in state and these user bases that
- 15:52are currently dispersed among a bunch of VM chains and unify them together so that
- 15:58we can deliver better experience for users. Hmm, okay, that's great to
- 16:02hear. And with those chains implementing the ZK technology, how easy is it
- 16:07for all of this to be done? Do they just have to flip a
- 16:08switch? Do they have to rebuild the spaceship? No? No, no,
- 16:14So there are no client changes that are required, like nothing changes for
- 16:18the users or validators or full notes. We just turn on a prover and
- 16:23we just start generating proofs for that chain. Obviously, if they want to,
- 16:27if they're like an L two and they want to migrate their L one
- 16:32native assets to be able to have fungibility with L one native assets on the
- 16:37agg there, they would have to do a bridge migration. But if they
- 16:41don't want to do that, they can just immediately start generating proofs. Amazing,
- 16:45And so does this sort of came about from the polygon's growth from the
- 16:56POS chain moving into the ZK EVM chain, and you guys are going to
- 17:00be running both of those chains. But this technology does it affect primarily the
- 17:06ZA the ZK e v M chain, So it actually affects POS because the
- 17:11POS chain is not using ZK technology and there's a ton of liquidity. There
- 17:15are a ton of users. I think it's one of the most actively used
- 17:18chains in the ecosystem, and so taking everything that exists on POS and onboarding
- 17:23it into the A layer so that all of that, all the assets that
- 17:29are on POS all the users can seamlessly move between POS and ZKVM. That
- 17:34was sort of the genesis of the teche one prover and the reason why we
- 17:38started working on it. That's awesome. And now in terms of the transaction
- 17:44costs, zk's supposed to help, you know, lower the cost to as
- 17:48well. Some the cost varies from what I've personally experienced in different layer twos
- 17:56on ethereum, you know, going from a dollar to ten cents to a
- 18:00few cents. How will this affect the costs of the different chains? Will
- 18:04that be improved even further? Yeah? Sure, So. I think it's
- 18:08really important to note that the cost forrail twos is generally dominated by the cost
- 18:15to post data to ethereum. So if you're a roll up, every transaction
- 18:18you have to post some amount of data to ethereum and this has a cost.
- 18:22And so generally all of the transaction costs that you see for L twos
- 18:27are based on this one data availability cost, and so I think that the
- 18:34way that I would look at it is there are other costs that exist on
- 18:40l twos, and so one really obvious one is that optimistic rollups have this
- 18:44downside where if you want to withdraw from an optimistic roll up via the native
- 18:49bridge, you need to lock your funds for seven days to sort of wait
- 18:56for the duration of what they call the challenge period. So it allows anyone
- 19:00to submit a fraud proof if the operator for the optimistic rollup is misbehaving,
- 19:03and you know, you can ensure safety. But the problem is is that
- 19:07this seven day withdrawal window imposes a real cost on users. No one wants
- 19:12to lock up their highly volatile crypto assets for seven days and take on duration
- 19:18risk, and so what they do is they rely on third party bridges that
- 19:22impose fees on users. And so these fees and aggregate are actually like it's
- 19:29a lot of money, so it's millions of dollars. And so the nice
- 19:33thing about the Type one prover is that we can convert all of these optimistic
- 19:37roll ups into ZK rollups and the proving costs, so we're effectively taking this
- 19:42duration risk and these fees that are paid to third party bridges and we're replacing
- 19:48it with proving cost. So instead of forcing users to pay for bridges,
- 19:52we just force them to pay or to cover the cost of generating zero knowledge
- 19:57proofs for the chain and so they can have instant with role. And it
- 20:00turns out that the proven cost is actually really really low. So right now
- 20:04we're seeing proven costs on for generating proofs for main net ethereum blocks as like
- 20:10two to three tenths of ascent, and in the next few weeks we're going
- 20:14to release upgrade, so this will go down to like between one and two
- 20:17tenths of ascent, and then in the sort of more medium term we expect
- 20:22a thirty to fifty x reduction further and so yeah, it'll go to like
- 20:30I guess like four thousandth of ascent and so this like like in terms of
- 20:34cost, this is like you know, orders of magnitude less than the capital
- 20:41and efficiency costs that are imposed by optimistic roll ups and so so it doesn't
- 20:48solve the data availability costs, but it does solve these other sort of roll
- 20:52up costs that are still born by users. Yeah. Wow, that's incredible
- 20:56to hear. I can't imagine not how to pay a dollar or two.
- 21:02This gets me so excited and that's crazy. So your team you have this
- 21:08is a huge update obviously, but then you're going to be working towards even
- 21:12further upgrades that you're talking about that will make it fifty times cheaper and faster
- 21:19in the near future. Yeah. Yeah, And I will say, like
- 21:25from the transaction cost perspective, this model allows us to end this sort of
- 21:30the ag layer and the multi train model allows us to launch what are called
- 21:34filidiums, so they're l twos, but they don't require you to post data
- 21:40to Ethereum and so for these like the transaction costs can be super super low.
- 21:45And so yeah, I think in the next year or two we'll see
- 21:49a pretty huge reduction and transaction piece very cool. So looking towards the very
- 21:56near future here is they're going to be you know, what are your team
- 22:02sort of focused on in the next month or two and ensuring that you know,
- 22:04this all rolls smoothly and everyone's getting on board and really getting the realization
- 22:11of these benefits for everyone that's using DeFi on all apps in EBM chains.
- 22:15Yeah, sure. So the ad layer is going to be released in a
- 22:18few different stages. So happening this month will be V one, which is
- 22:22effectively a unified bridge, and so this will provide that fund ability property that
- 22:29I was talking about earlier where you can move assets across chains, but it
- 22:33won't have the low latency property yet. So low latency will happen in a
- 22:41few months. But yeah, so in the new term it's about providing pung
- 22:47ability, about getting chains using a unified bridge, and then in sort of
- 22:52the medium term, we'll be releasing further upgrades which will enable the low latency
- 22:57mode for cross chain composed building and interaction that I was describing it awesome.
- 23:03I'm looking forward to that. I would love to have the team back on,
- 23:04you know, as the phases develop and we can see just the benefits
- 23:10people reaping the benefits in these applications. Yeah, for sure. And what
- 23:17is the best way for people to follow along, well, first of all
- 23:18to learn more about ag layer and the Type one and for these updates as
- 23:23they're rolling out, to sort of understand better and capitalize on these improvements that
- 23:30your team's making into blockchain. Yeah, sure, so you can follow Xerox
- 23:34Polygon on Twitter, you can we have discords or Polygon developers and then an
- 23:42overall Polygon discord, and so we run AMA's and we're pretty responsive to to
- 23:48questions and feedback. And yeah, I think that's that pretty much covers it.
- 23:56Twitter and discord sounds great and I can leave those social links in the
- 24:00show notes below as well. Appreciate your insights into this. Congratulations on making
- 24:04such an amazing improvement to blockchain. I'm excited to reap the benefits myself and
- 24:11looking forward to the coming months as this develops further on. Thank you so
- 24:15much Brendan for your time today and all the best with everything Pobbly on moving
- 24:18forward and let's follow up in the near future. Yeah. Thanks, Ash,
- 24:22really appreciate it and thanks for having me