Monad Co-Founder: Here's What I'm Focused On Until Mainnet

Transcript

Is that better for society? Is that better for um markets in general? Is that better for your 401k? You know, like these are all like the um philosophical economic debates that people have all the time and going on for years. >> What do you need to ensure a really really high throughput quality system for your end users? >> Welcome back to the rollup. We are your hosts. I'm Andy. >> I'm Robbie. And this is your home to learn from the biggest guests, the brightest minds at the tip of innovation.

>> Our goal is to help spread the collective knowledge of everyone in our community. We are breaking crypto out of its echo chamber and taking it to the real world. >> Sit back, relax, and enjoy a quick word from some of our awesome partners who make this show possible. >> CryptoUX broken AI can't trust it. Let Magic Newton set you free. It puts your onchain transactions on autopilot. does all the work for you 24/7 so you can go do what you want. The vision of Magic Labs, the creators of embedded wallets used and backed by the best. See for yourself. Simpler crypto UX, verifiable AI agents. Newton.xyz.

Introducing asset runes. The future of decentralized finance. Stable coins natively on Bitcoin. Hedge against volatility. Diversify your portfolio. All directly from your Bitcoin wallet, secured by ZK Technology. This is Asset Runes, available now exclusively on StarkNet, >> the infamous uh Rick and Morty uh PFP in the building today. James, it's good to have you in the city, man. It's been a long time coming. Uh thanks for coming out.

>> Yeah, thank you for having me. >> Yeah, man. there's been a lot of conversations uh that you've taken part in on CT over the last couple years really but also much more recently around throughput and um that's this has like been a a focus of our fund of my research of our podcast just trying to identify ways to scale this ecosystem and so this current uh idea that I think is shared amongst a lot of people that is throughput and UX are really bottlenecking the overall scale of the space and and you know I think we can kind of keep cracking at those two things we'll get to a a better place uh with more adoption etc. Obviously at um Monad you guys are trying to push the limits of throughput at the

infrastructure layer from the jump. So before we jump into Monad and all these kind of different debates and conversations, I would just like to get your perspective as to the state of highfrequency trading throughput uh and the infrastructure today and kind of how you're seeing this landscape currently amongst L1's, L2s, apps, uh and let's just start to get a foundation of what the state of high throughput is in the space today. I think kind of the genesis for me was the legalization of um sports betting. And I remember like around COVID times, pre-COVID, you know, you would go to like the New York train station and there'd be all these like posters, um DraftKings and different things and

and then CO hit, right? And um I took advantage of you know the work at home sort of thing to travel, spend time with family and my cousins and and um you know other relatives were in sports betting and you know like the stimulus check came out and all this sort of stuff. So I think um during co it kind of launched two sort of gambling realms. One was on crypto side and one the other was sports betting, right? And uh I just remember sitting like on the couch drinking a beer with like my cousins and stuff and you know my cousin would have like three different TVs the games would be going >> and have the phone open. We making bets making bets in the middle of the game

that sort of thing. And uh that's kind of like the first um use case that popped in my head for high throughput >> is I was thinking of like a ton of people sitting at home with their with their cell phones just like betting. >> Yeah. >> And um I kind of also viewed that it would just go to this like degenerate case where you would just be betting like >> is it are they going to get a first down? Are they going to you know all this sort of the pass going to complete? Like I I could just imagine just like >> real time. >> Yeah. Just like thumb just like Yeah. just like thumb pounding on the screen sort of betting and uh >> um that's kind of the case that yeah just popped into my my head of what could be on blockchain and then you

start like >> thinking through numbers how many people are sports betting how many people place bets how many bets would they want to place over some time period and like could the chain handle that and the answer is no you know there was no chain really that could handle that so if you think of and it's also kind of bursty like people bet all at once And then you get a kind of a lol where there's nothing going on and say you know you can't even really like take the averages. It's more sort of like handle those peaks. Um >> so like uh you know asynchronous execution is something that we have at Mona. It kind of came from that sort of like handling the peaks sort of um thing where you can burst faster than the chain can actually go and you can accumulate sort of things. So

>> um yeah that was really my inspiration. Um, I'm not a big sports gambler myself, but just watching other people do it and participating a little bit kind of inspired me of like what would you know what would this look like on chain? >> Yeah, that's actually ironically playing into this broader uh this broader idea from this guy good Alexander on Twitter. he's got this like post fiat concept and he's been writing a lot about AI doom and and and I think there's there's two kind of uh there's two pathways happening once is just the speculization or or the speculation and the financialization of a lot of things at once and then at the same time just like the world kind of going towards a more like degen I call it like degenation. Um

but but you know sports betting is kind of in in in the middle of this. But the core idea I is that people are going to want to interact uh globally with people whether on on a chain or or not at a very very high rate and the state of things in the onchain world is just barely getting there now. >> Yeah. Um you mentioned AI as well. You know AI agents is another sort of thing that would require high throughput. Um people have been talking about that for decades where like >> agents are interacting with each other on behalf of some human. >> Yeah.

>> So that's another thing that can just like multiply the amount of volume of transactions um beyond what even humans can do themselves. So >> yeah, I mean I think that um you know we've seen this idea around throughput being being a bottleneck and I have throughput and UX is kind of my core parameter. Do you think that that uh throughput is something that is solved uh you know in in the span of you know this this next couple years right? Is is it a longer term thing like like maybe you could just walk us through what how you see this evolution playing out with uh some of the earlier stage you know you you had the Ethereums you had the L L1s L2s um monads coming out soon now you know we're going to get a 100x

improvement of throughput um but you know where do we go in the next couple years uh is that going to be you know kind of like where we end up >> well I think Like every action has some value, some economic value or societal value. And you know the the the more frequent something is, the less likely that individual thing sort of has a lot of value. Um you know you there's not a billion dollar transfers happening all day long, right? There's like a few here and there or whatever. So, but there's a lot of like low dollar transfers, $3, $5. Somebody buys an ice cream cone to McDonald's, it's like $3 or something.

So, >> I think the there's going to be some equilibrium by which the cost the security of those transactions like meets the value and that's kind of like where things will stabilize. So, will everything be on chain? No, because it's just not like economically worth it to put it on chain. You don't need that security. Do I need to put like every time I turn my light switches on and off like onchain? I know some people will say like, "Oh, the electric company we should we should record everyone's electric usage or something, you know, on chain, but that's that's not really like worth it." So, there is going to be some sort of equilibrium. And I think the you're going to have Yeah, it's just basic economics, supply and demand of blockchain space with with what people

want to do and and whether they value that security. If they value that security uh then then they'll pay for it and that the supply of blocks chain space will grow to meet that demand. >> A lot of that demand in the last couple months has came from uh you know you hyperl's done a great job of doing no KYC purp whatever you want to call that is a different story but I think there are clear pockets of that demand. Now, it might not be light switches. It might not be, you know, every uh text message, but but but there's going to be this kind of, as you say, this e economic equilibrium for where you as a person will want to have some sort of

guarantees in your, you know, daily life, specifically around your your your financial life. >> Um, now coming to some of these areas of demand, recently crypto Twitter has discovered clubs. Um, I put a tweet out saying that they were uh one of the most important new DeFi innovations since unis swap got absolutely meme to hell for that one. >> Um, when I when I say new, I mean that the infrastructure is getting there to actually support them. >> Um, and and and I think that's particularly interesting. So, I know that you guys have quite a few clubs that are being built on Monad. I'm curious what you think is the ideal base layer for a high performant um you know

exchange like a central limit order book is is is is what you guys are doing complimentary to something like a single specialized chain like hyperlquid where do you see the the ideal setting for something like a club? >> Yeah, it's all it's everything is tradeoffs, right? So you're trading one thing for another. Um like what are you optimizing for, right? I think typically when people talk about clubs, they're talking about like market makers. Like you need liquidity, what are the market makers going to come with the liquidity? Like what's ideal for the market maker? Um, you know, being able to uh have tight spreads, um what's what's good for a market maker to be able to do that? So, you have some, you know, people say, okay, we

prioritize cancels over fills or, you know, there's all these different sorts of ways of >> of trying to do that. Um and then you have just like in in your traditional exchange like CME you have an FPGA which is just timestamping um all the incoming orders and then on the backend system they get sorted and cancels orders whatever they get sorted and executed that way. So it's all sort of like time purely time based you know um I've tweeted uh you know like L2s should just sell data center space right like you you know so like if that's what you're kind of optimizing for >> what does that mean >> well like exchanges make a lot of money just from selling data data center space

they make money by connectivity right um >> and so if you have a place that is latency sensitive and people want to trade there, they'll pay to to get closer to that that thing and they'll pay to have better connectivity. They'll pay for all different sorts of things, right? And um it's always a balance like you know at HFT firm do we have 10 lines connecting or do we have 20 you know like is >> are we actually saturating those 10 lines or are we are those 10 lines good enough for bursty sort of situations that it's not worth paying for the 20 >> lines because the exchange is going to charge you for all that. And so you got this like fixed cost

>> of your sort of monthly stuff and then you're trading revenue and you kind of like do the numbers accounting numbers and you say like is it worth it sort of thing and >> you know if we added two more lines would we make more money? So there's all these sort of games you play in HFT to like optimize this stuff. >> Um you know uh there's always sort of inconsistencies within the exchanges uh uh internal technology stacks and you're trying to optimize for that and and play these sort of latency games. But um you know it just really goes to like what you're optimizing for. So is there ways to get liquidity without like the the best systems? Of course. There's really shitty >> centralized exchanges. You know crypto exchanges

>> have been known to be terrible compared to >> um you know traditional market sort of >> exchanges. And there's still a lot of exchanges in the world that are using >> 20-year-old software that's not that good. And you know there's a lot of variance. Yeah. And that's really frustrating when you're trying to like why did my order why didn't like why did my cancel come before the order? Well, they actually have all this like priority inversion in their system that like it's not >> so you know you can still have a ton of demand. You can still have a ton of liquidity with with a with a relatively poorly designed system. There's other ways sort of incentivizing um you know certain behavior. CME will give for a new product we'll give um you know credits right and so

>> you know there were there were a lot of times a jump we ran a uh trading strategy that lost money >> but we were net positive because CME was basically paying us to make the market right so there's all these sorts of like >> a rebate >> yeah there's rebates right rebates is a big deal for especially for new products um >> and then over time as the market gets more efficient and gets more activity that those rebates they solely kind of like take those away But um yeah, I think there's a lot of different things that go into whether a market is successful. Do you need to be like CME to be successful? No, you don't because there aren't very many CMEs in the world. There's like URX. There's a couple other exchanges that are like at

that level where they're like FPGA is timestamping stuff and you're down to single digit nano sort of thing. >> Um you know, you got people cutting cable lengths trying to make them the same length. like you can you can take it to that extreme but there's plenty of markets that are not that >> that that technology advance that do that do fine and have tons of trading activity. From your past experience, what what are the uh parts of a really quality trading system? Like what do you need to ensure a really really high throughput quality system for your end users >> on the exchange side? You mean like Yeah. I mean it's >> what I'm trying to understand is basically like like what makes uh what makes a certain system far like far

better than a different system? Like what are the parts that matter the most? because then I'm trying to like understand how you guys have optimized your base layer architecture to optimize for those parts that matter versus other ways. >> Well, I think a lot of it just comes down to um the higher sort of like UX like having you know streaming sort of data um any of that sort of stuff is necessary to even build an automated trading strategy to begin with. Um but you know it's that there's differences obviously between a decentralized system like Monad where you don't well in the current case you do know where the next leader is going

to be but it could be you know it could be all over the world somewhere. >> Yeah. >> Um and uh a centralized system that's like a centralized exchange or L2 or something the single sequencer. So, you know, is it easier for a market maker in that case to compete on latency? It is because they can just say, "Okay, I know that server is in >> right in that location, >> this location in Singapore. I'm going to put my server next to it. I'm going to find the data center that has the best connectivity and put my server next to it."

>> And so, yeah, that makes that makes life easier from that perspective. you know, if you had a decentralized chain where you had a significant stake in Europe, significant stake in like Northeast, US sort of thing, >> um then you then you have to put two servers if you want to kind of race those sort of games or you start getting into the >> um um you know, sort of secondary services like me or that sort of thing where you're paying basically to >> to get in front of other people. Um but yeah, I just again everything's kind of just tradeoffs and um you know in in a decentralized system we're not try we're intentionally trying to be

globally distributed. >> Okay. So so you put out a post uh arguing about how L2s are kind of fighting over pre-MPS and uh really trying to get block times as low as possible. Um 100 milliseconds 10 one and then kind of going all the way back to the data center argument. And then with with your with your uh thought around uh basically single sequencers being like a centralized system. So of of of course if you're uh trading on those venues and you're very close to the where that uh sequencer is with with a highly performant data center you'll have an advantage over every everyone else in a more kind of uh decentralized uh L1 setup. you've got this uh globally distributed set of validators which

forces basically more more competition or higher higher threshold to actually be able to be closer to uh like the core uh stake like where geographically that is >> um it when you introduce sequencer collocation into this from an L2 perspective you start to get a little bit closer to where uh monad will be but >> you still trade off the pure sense the the pure like amount of stake and validators in exchange for still faster performance. Um I I think that it will be better to have these high performance applications on something like that than on a L1 with a a a need to have redundancy baked into the

uh consensus. Um obviously you don't feel that way. Um, but when we did the when this derby race was going on on Twitter, which was kind of funny, it was all these different L2s and Salana and Monad test nets and a couple others and like obviously the L2s that were just a single sequencer and just extremely uh juiced up were like first and second. Um, so I guess it's going to come down to like what the users actually care about or you know what trade-offs you're willing to accept as a builder. Um, so I guess the question then is what you guys are essentially trading off is like is the shared state of this globally decentralized L1 that is worth it more that the the composability that you get is worth more and having to achieve that

consensus is worth more than having just a single sequencer colloccated or three sequencers, four sequencers across the world in its own kind of silo or isolated environment with higher throughput or higher performance as a result of the setup. And so that's basically the decision that you guys have made and you're you're going to bring that to the market. >> You know, first of all, Monet isn't designed for a club. Like it's not >> it's a general purpose blockchain similar to Ethereum or Salana. So we didn't say >> let's build a blockchain that's a club.

Otherwise, we would have built hyperlquid, right? Yeah. We could have built hyperlquid instead, right? So we didn't we chose not to build hyperlquid. We chose to build something else. So, um, >> yeah, I think the the there's there's two we talk about like UX like user experience for as a as somebody as a retail trader, let's say. >> Yeah. >> Like you care, you know, you click the button and you want to see the order filled right? >> Immediately. >> Yeah. So, you know, what's what's human perceptible sort of thing for that? You know, I think with the some of the changes I outlined in that one Twitter post, it was response to the derby.

>> Yeah. like you can get that down to a couple hundred mill a couple hundred milliseconds, right? And so can a user really >> tell >> tell I don't think they can. So >> market makers could but again you're not optimizing for something like >> so the mark yeah so then we go to the market makers. So what can the users tell from the market makers? Well, okay. So, then we get to like is there is there a lot of liquidity, right? Like if I go to the book and there's no liquidity and you know, I remember I remember this from when I used to trade different coins. Um I haven't recently, but >> you know, you go to the book and you're like there's no liquidity here. Like there's like one token at this price, two tokens at that price, you know, like there's nothing there. Like it

>> it wouldn't take very many dollars for me to just like sweep multiple levels of this book, right? And that that that's not the way it should be. like there should be a lot of liquidity there. So, um and then it's the the spread the width of the width of the book and you know how much is it paying me to aggress to cross the to cross the book, right? And so, um sorry, how much am I paying to aggress? And the that's kind of like what users see. So, if you're doing if you're trading shitcoins, do you care about any of that? Probably not, right?

Like you're you're buying that coin, you're buying that meme coin hoping that it's going to 10x or 100x. >> Yeah. >> Do you care if you pay like a 3% spread versus a 1% spread? Like you just don't give a [ __ ] Like you're not >> probably don't care about the the you know 100 millisecond difference or 50 >> Yeah. Yeah. I mean you just don't you're so you're hoping for so much more gain than what you're what you're paying in fees that you just don't really care. >> And and people trade on the centralized exchanges and those have terrible sort of fees and spreads. They just don't care at all, right? Like, um, I can't believe that people even trade on these centralized exchanges. Like, it boggles my mind. Like, why don't you, you know, I understand there's a UX thing, custody, and other sorts of things that

make it nice for >> for the average sort of retail person, but as a trader, I can't imagine a real trader would ever trade on some of these exchanges. It's just terrible. And you >> especially if you go through the the basic sort of interface, which is just, you know, >> that's a scam. >> Yeah. It's a total scam, right? The fees are terrible. The fees are terrible. Spread's terrible. Everything's terrible. >> You know, there's people buying Bitcoin on Fidelity, you know, like the the spread is I think they charge like 1% or, you know, like this stuff is just terrible, but people do it. So, >> I think um where does it where is sort of the depth of book and the width and all that sort of come in play? It comes into sort of the majors, the the the the BTC, the ETH, these sort of things, the

futures, you know, where there's a there's more size. So that stuff matters more there than sort of the shitcoins. >> Yeah. >> And um memecoins or whatever. >> And you know so then that question is for the user is like well how do you get as we already discussed how do you get market makers to put liquidity there? Like how do you get them to put liquidity? How do you get them to quote tightly? And so yeah, if you you know, one way is to say um you know, we have the centralized server. You can put your HFT firm, you can put your server right next to the centralized server, >> and you will guarantee beat all the retail people who come in to aggress, you know, when they see the news tick,

right? And you're going to be able to cancel all your orders >> because you're so fast. And then you get into this game though with like you get a you get a a quant trading firm that comes in and they put their server next to it and now it's the fight between um the quant trading firm which has these sort of like aggressive strategies and and the other ones which are have passive strategies and they're fighting each other based on this latency game and that that game is a race to single-digit nanconds. That's where that leads. >> Yeah. Um, that's fine. You know, like Jump has made a lot of money off winning singledigit nancond races. You know, Jump is a very successful firm at trading at CME >> because we have all that technology, all that hardware.

>> Is that better for society? Is that better for um markets in general? Is that better for your 401k? you know, like these are all like the um philosophical economic debates that people have all the time and going on for years, right? So, >> yeah. >> Um so, so going back to the sort of decentralized system like Monad, yeah, can you can you create a club on Monad that reaches the level of um you know, liquidity and sort of markets that you want that people find it's valuable enough for them. you know if if it's a difference between a spread of 1% or 0.85% 85% like do people really care? Probably not. So I think there is

some threshold there. >> Yeah. >> But yeah, this is what >> you know to talks about this. A lot of people talk about this. >> He's been in the trenches. >> Yeah. There's a lot of people that talk about what are different ways we can accomplish this, right? right? >> Like I mean all these ideas just keep getting taken to their logical extremes >> like it just it's just going to keep happening you know like like okay you know we're just going to push everything to the limit of like being onchain um but bringing all the way to like the least amount of validators the most amount of speed and then all the way to the other side as well. Um, so I'd like to talk about some of the economic stuff as well, but one last uh thought around kind of like this latency stuff is I

think that people are underindexing on latency and overindexing on like uh block times or like overindexing on TPS and and these types of of metrics. I think that's that's been something that's been a little bit misconstrued. Perhaps you could like unpack and educate us as to why latency matters in these systems more than some of these other uh terms that are more popular per se or or just maybe like what the what the ramifications of of latency is in in these trading systems. I feel like we've covered it pretty well, but just want to get a firm understanding. Introducing Summerfi, one of DeFi's oldest and most battle tested platforms. Summerfy automates your exposure to DeFi's highest quality yield so you can sit back and enjoy this summer. Deposit

stable coins at USDC or USDT and let Summerfi route your funds into optimized low-risk lending strategies. With over a 100 million in instant liquidity and support for over 34 protocols across Ethereum, Arbitum, and more, you always stay in control. Choose from a wide range of curated strategies reviewed by independent risk experts all in just a few clicks. So why work hard for your yield when you can be lazy? Start earning now at summer.fi the home of the lazy summer protocol. Innovating finance global institutional and onchain. Deep liquidity to drive capital efficiency.

Now maximize yields on institutionalgrade liquid staking with M. Unlock trillion dollars of capital via F Bitcoin. All of this secured by Mantle's treasury of billions in stables and ETH capital pools. This entire ecosystem is coordinated by a vast network of seasoned VCs that are aiming to create the next generation of banking and financial technology. This is mantle. >> Yeah. I mean people talk about performance. Performance is like a general word. There's different ways of measuring performance. One is throughput. Um throughput is you know like how many transactions you can do

per second on average or something, right? Um, you could have a system that doesn't that that produces one block every hour, >> Ethereum, >> but that block has like, you know, trillions of transactions in it. Yeah. And it so by definition that has that chain would have the highest throughput. You know, there's this old sort of joke, >> you know, you could load up a semi-truck of hard drives, >> drive it across the United States, and it'd still be faster than like sending that data over fiber from one coast to the other, right? and um you know it has very high latency but it has extremely high throughput. I don't know if that's true anymore but it was years ago when people used to run this mental exercise.

>> Um so that's so that's your TPS. So that that to me goes back to what I mentioned at the beginning was like sports gambling like you just >> you have a 10,000 people who want to place a bet right now right they're not on individual basis they're not concerned about latency >> they just want to place a bet right now. So you need that sort of throughput on latency. Um, you know, I don't think again for retail traders, it doesn't matter that much, right? It's a UX, you know, you don't need to execute the trade faster than the person can actually click the mouse.

Like that's just silly, >> right? Or faster than your brain can process the information, >> right? Right. Yeah. Um, >> but so in the latency uh world, there's a concept called jitters, right? where where there's like a discrepancy sometimes in the performance of a given distributed network based on based on the I don't know based on something is like could you explain that idea of like jitters and like how how monad is set to basically uh I guess reduce the amount of of variance in transaction processing time >> well the transaction processing time for monitor right now is basically fixed at 500 millisecond that's what I said in that >> yeah yeah >> that post So even though the system can run much faster, it's, you know, it's

it's 500 milliseconds to do one round of communication. >> Yes. >> Um, so I I'm not sure are you talking about like I'm in New York and the and the leader is bouncing around the world. Is that what you're talking about? >> Yeah. And then like on a given like let's say on a given day there is there's different spikes in terms of how long it takes for information to be propagated from different like validators to each other or from from different nodes. And so like the goal is to try to reduce the variance to be able to give validators and users like a very clean average of like okay here's what you can expect.

>> Yeah. So we're not really trying um right now it is consistent just because of that minimum thing that we have. But yeah, >> you know, if if if you're sitting in New York and 30% of the stake is in New York for Monad, then 30% of the time you're going to have the best possible connectivity to the leader. >> Yeah. >> But you have, you know, you have a block time like it's not the time that a leader receives transactions and then builds a block and sends it is not instantaneous, right? Like it's there is a window over which is receiving. And then people try to play timing games like I want to wait till the last possible millisecond to get my transaction in because I've given I've got more information

>> at that time. And so, um, but yeah, there's, you know, I think I think a general sort of assumption is like if I'm in New York and you're in New York and the next leader happens to be in Singapore, I guess we're equally matched, right? Like, >> um, you're going to send your cancel, I'm going to send my aggressive order and, you know, whatever happens happens. But um >> you know, this is this is, uh, in that sense, it's fair. Now, if I'm sitting in New York and you're sitting in Singapore and um you know, like Tony gives examples, right? like okay the information comes out of Singapore you know well then you have advantage because you're going to get your thing in but you know things are not in in in um

in traditional markets things are ordered by time like CME timestamps orders and cancels as they come in and and you know they're fair in that sense in that time sense is blockchain doesn't really work that way because you have block builders and they accept bribes and other sorts of things and so you know there's other ways of sort of dealing with this latency. Anyways, >> all right. So, I'd like to kind of pull out of the technical side and move towards some of the economics stuff that uh is particularly interesting to the current meta per se, but also to the state of how to think about uh the economics for users and uh validators and developers as well of applications.

And so, um one thing that that's obviously been a concept of valuing these kind of L1s has been REV, right? real real economic value and and a lot of that is actually generated by me going to validators. Um and there's a bit of like a incentive misalignment issue between the validators of like let's say Salana and like Jupiter the team or like pumpf fun right like you know there's a there's a there's a lot of leakage of value from these applications on their base layer chain to to the validators. The the counterargument is obviously that um well you if you weren't on the chain you wouldn't have any value.

>> Mhm. right? You know, I mean, pumped out a fun went from blast, I think, or was it blast or some L2, they moved to Salana, it was like boom. Um, so I'm just curious how you think about applications on monad and perhaps if this leads into like the future of will apps on L1's have consistently have some sort of application specific sequencing or some sort of ordering logic for their own application that they can then either return that value to themselves or to their users or just kind of how you're thinking about the that economic incentive game for apps on Monad and just L1's in general. Yeah. I mean, if an app doesn't value the chain, it won't be there, right? It'll be somewhere

else. So, um, >> but it's more like symbiotic than that. It's like you help each other. >> Yeah. You you have to make this sort of decision. Um, and it's somewhat some game theory decision as well sort of thing, right? And >> like uh I I didn't I wasn't aware of the pumpf fun thing. I I thought they were on base at some point, but was that was that before or after? Man, >> I'm not really sure, but >> they moved from some app, it was some L2 and they went to Salana and it just went boom. >> I think first even in a even in a fixed fee sort of space like AWS is not the cheapest cloud to host your stuff, right? So why do people >> still stay with AW AWS? Why don't they go for the absolute cheapest? Um what

are they getting from that? You know >> um are they getting stability? Are they getting uptime? Are they getting something else? So um there's there's that aspect >> uh at a at a sort of even a fixed fee level you have that sort of thing >> but application specific sequencing absolutely I think that's a big part of it um >> that'll solve some of these um latency gains that you see as well right if if the app says a block is invalid because the cancel there was a cancel after an aggressive order then then the then validator is incentivized to play along with that sort of game otherwise the

validator loses that app business altogether as well. So, >> um, I almost kind of see this in in traditional, uh, in in the trad world. It's just these are negotiations, right? People get you get some people in the room and they negotiate over like how this is all going to work together. >> Yeah. But it just happens on chain. >> But yeah, but you don't have that. You don't have that on chain. So, you kind of have these games. So then maybe the question is like is there a economic point in which an application would have more like would have the negotiating strength to come to monad and be like hey guys like we're driving this much value.

>> Yeah I think >> like like you know how do you think about that? >> I think chain should have sort of generalized frameworks by which >> um these contracts and intents can be specified and and executed and enforced. >> Yeah. And so that's something I've been thinking a lot about recently. Um, >> tell us about it. Okay. >> Yeah. I don't want to say too much because it's not built out yet, but you know, it's like what we're saying like >> uh an app has a specific sort of requirement. >> A validator desires the business of that app or the chain divider is the business of the app. I mean, the validator could always just say, I'm not going to include your transactions at all.

Anybody who interacts with this contract, screw you. Go talk to the next guy. Yeah. You know, but if if there's some balance there by which the app gets what it wants in terms of its specific sequencing and inclusion and other sorts of rules and and the validator gets what it wants, which is the fee revenue >> from people using that app, >> then they can find that sort of equilibrium and how do you like >> how do you express this and execute this sort of contract? It's not really a contract, but this sort of agreement uh >> programmatically, right? and and building a framework that allows you to do that I think is like the next generation of of this stuff. And so I've been thinking a lot about that sort of recently how to build something like

that. It's an economic uh game theoretical like scenario, but it also comes down to like off-chain stuff, right? Like bey, you know, stuff that is like not just purely economic, but at the end of the day, like there's this populist movement right now, obviously, right? Infert apps, you know, >> infra is dead. Infras infra is not cool anymore, James, right? Apps are the only cool thing. >> Well, I I I think that's just >> my head took too long. Infra is dead. >> Sorry. I think people are burned out on infra just because a lot of infra hasn't been innovative >> and it's just like oh we got another one oh we got another one and and Infra is just >> they've also been uh I've said in some

of my tweets like one trick ponies right like >> especially going back to like 2017 2018 2019 era especially you know it was just like >> these guys came up with this one little improvement to a consensus algo and that you know somehow led to an entire new chain Yeah. >> And why? It's not really their fault. It's the people it's it's the people that go along with it and throw their money at that thing's fault, right? Like >> if you could if you could convince somebody that, you know, uh this new sho string for shoes is awesome and people are >> launch a whole new collection.

>> Yeah. Launch a whole new company like Nike and people give you billions of dollars for that little shoestring invent, you know, like why wouldn't you, right? So, um it's that sort of um you know, FOMO or I think a lot of it just comes from people not really understanding technology and like what this thing really does. Um like I said about that Derby game for example, it's it's something that an average person can kind of understand, >> but that's not really a good >> No, but you look at it and it makes my look bad. It makes fun. I look bad, but it's it's not uh well the one there was one that failed and there was one that like had like terrible latency, but like if that if that is your level of

thinking, you probably shouldn't be investing in technology at all, right? Like you need to go a lot deeper than >> a horse taking steps across the >> I mean, it's good marketing. >> It's good marketing for them. Yeah. Um and you know it did it did competition is good and it did generate some discussions within the team and and you know we talked about some of these well that thing I told you that vote pacing mechanism right which slows the chain down intentionally. >> Um but uh infra is not people are people are burned out on infra.

>> It's not that infra innovation is dead. There's going to be a ton more infra innovation. A ton more. >> Yeah. But people are burned out because there's just been so many um very very small delta improvements. >> Yeah, marginal improvements. >> Very very marginal improvements that raise tons of money and get dumped on retail and all that sort of stuff. And >> you know what's that that George Bush quote? Shame me once, shame me tw you know people just get like >> Yeah. Yeah. >> I can only take this abuse so many times sort of. Yeah. So many times sort of thing. And so >> I don't think infra is dead. But I think people are just burn out on Infra and and it's just because of the the past has not been efficient at all.

>> So when you guys are leading up to mainet here um how are you feeling about the application ecosystem for monad um if we're thinking about this from like an L1 perspective right there's a lot of uh similar DeFi applications you got your your your dexes your lending etc. Um, are you guys are you seeing uh the technology that you've built allow for new ideas that are being built on chain or kind of, you know, do do you do you have any cool ideas of what you'd want to build on the chain? And also, I mean, this is like three questions in in in one here, but there's a new trend of like teams wanting to just dog food their own tech and being like, look, you know, I'm just going to build the thing

that I know that my tech is really good for, right? So, let's just say like I'm I'm a DA layer. I'm just going to build execution, build a really high performance chain or I'm a, you know, I'm a I'm an app. I'm just gonna build a chain for that and just go, right? Just kind of like uh flip-flopping the way people think of just internally dog fooding their own chain. Curious if you guys are doing anything specifically for the chain from like the internal team like the lab side or just how how you're feeling about the application e ecosystem more broadly. >> First of all, I'm a lab so I'm not you know I'm not building I'm not a BD guy.

So >> yeah, of course. Um, but you know that you have your you have your sort of like blue chip apps that already exist on other chains. You people kind of expect those to be there. So you know you you >> you get them to sort of build on your chain as well. Not build but deploy on your chain as well. And >> there's just a ton of um BD that happens you know like we have USDC and launching and we have a lot of other sort of like dra um blue chip sort of DeFi stuff there. Um, but is that going to bring somebody from like Ethereum over to Monad? You know, like they expect it to it's a necessary but not sufficient condition, right?

>> And so you have to have something that's like sort of um makes people energetic and you know there are teams that are building >> um new stuff. We have we do have club teams. Um um we have some games. We have some other stuff that's sort of innovative social media, social financialization stuff which um >> um but I think uh sports betting we got we got that as well. >> Uh I think one thing that I kind of learned when I was at Jump, you know, building for Salana is just like developer support. And I've always said from the very beginning of Mona that like one of the most important things is for the the the team that's working on the core technology to

um work with the app teams to help that make sure their stuff is optimized and you know running performance. their expertise is presumably in the thing that they're the area they're building, you know, but so they might be experts in uh trading or they might be experts in >> um sports betting or whatever, but they're not maybe experts in like how can I optimize this app to run as quickly as fast as possible on a blockchain. And so I think that's like that collaboration needs to be there. >> Um frankly, we haven't done a great job of that as much as well. I'm usually harder on myself than than other people are hard on me. But like so, you know, we're working more recently trying to >> get more involved with the teams, make sure and you know, there's a lot of

stuff that's onchain but also offchain, right? A lot of these apps have off-chain components. Um, you know, data uh statistics automated processes, they have off-chain components that sit there as well. and those need to be performant and how do those interact with the chain, how what's the RPC, all that sort of stuff. So, um that has all been a learning lesson for me. I always focused on the chain and not on the >> integration piece and that that really hurt us when we first launched test net because our RPC >> was very had terrible performance and other sorts of things and it might might not still be even that great. We're still working on it, but >> um we've started rolling out, you know, things where allow >> much uh higher in performance sort of

integration and that from the app teams. We're getting a lot of like good reviews like they're very happy with that sort of work and so >> um yeah, it's something that I always knew from the beginning that I wanted to do and had to do and then we just didn't get time to do it. >> Yeah. >> And now we're trying to play catch-up on that side of things. I think a lot of the infra block explorers, RPCs, indexers, these type of things, they need to catch up to the uh high throughput chains actually. >> Yeah, it's it's both. It's it's and the the old sort of methods that they use some will pull like the RPC for the block and download the entire block like that just doesn't work in a high performance chain. uh you need a streaming sort of like

>> streaming set of events coming from another process to your to your explorer and indexing and so yeah there is a >> um there's a lot of just technical sort of backend problems that need to be solved. >> We're we're going to wrap this one up by playing James reacts to his own tweets. Is that all right with you? >> Sure. >> Most things are only interesting because they're unfamiliar. Reationships. Try to find someone that you still find interesting after they are familiar. >> Interesting. Yeah. What do you want me to say? I mean, it's true, right? Like dating, you've you've you've been dating, you know, you go and meet somebody and like they're different than the last boyfriend or girlfriend you had, and so >> you find them really interesting and then 3 months later, 6 months later, you're like, "This person's boring as

[ __ ] so I'm getting out of here." And you go on to the next one, right? So, >> so find things that remain interesting over time and stick with them. >> Yeah. I think you just need to distinguish the difference between something that's actually interesting and something that's just unfamiliar and new and not >> and people often confuse those two concepts >> and they're interpreting unfamiliar as interesting and that's just >> are there good examples of automated slashing that were truly from malicious behavior as opposed to incompetence such as accidentally running two nodes with the same key. >> Did you did you find anything in there? >> No, I don't think I really found anything interesting. So we were thinking about this, right? Like uh a node starts behaving in Byzantine behavior, >> right?

>> And like what what is the likely cause of that? >> Yeah. Just like you [ __ ] up or like you're trying to actually do something wrong, >> right? Or the team that wrote the software, which in this case happens to be Category Labs, [ __ ] up, right? And so we're sitting there thinking, okay, we're going to implement slashing and like maybe this maybe this node this validator does something wrong. >> Yeah. >> And we're going to take money away from the validator company or the delegators that delegate to that >> and then we come back later and we find out it was really a [ __ ] up on category lab software developer made a mistake >> and now what do you do? Right? So, you know, we've had these issues in trading and banking and other stuff where we made errors and we had customers and

>> and we screwed up and you know, I there was several times I remember Goldman Goldman literally wrote them a check, right? It literally mailed them a check. >> Yeah. >> And is that what we're going to do here? So, you know, um I know >> uh I've seen threads with Tuli about Salana and automate. you know, Sauna doesn't have automated slashing if I'm if I know correctly and um you know, relies on some sort of social. But if you have automated like >> what is the likely case that something what is likely cause of something going wrong? Well, somebody misconfigured something, right? That there's been that case where people have spun up two validators with the same private key um or there's a code bug or they're actually trying to do something malicious. The malicious thing is far

less likely than the other two, right? And so does automated slashing really make sense in that regime? >> That's interesting. Another interesting one here from you is uh around the NYSC, the IDGAF chain, the I don't give a [ __ ] chain, single sequencer, IDGF, ponzomics, copy paste code, misleading tech, sketchy MMD deals, >> multi-stig with two friends, centralized RWA, DM to angel. >> A little bit of satire here. >> Yeah. I mean, if you don't give a [ __ ] then you don't give a [ __ ] Why why bother with anything, right? Like You know, there's there's this just be honest. Like, >> yeah, >> I don't give a [ __ ] >> Just just Yeah, just >> you're here to play with ponzies, >> but you can't

>> come to my chain and play with ponzies and we'll ponzi together. >> So, you're saying just be honest. >> Just Yeah. Just I don't give a [ __ ] Like, why not? Like all these games to pretend like you're something other than you're not. Yeah. >> And who are you fooling? You're not fooling the smart people in the room. No, >> the smart people who are probably uh sociopaths will still invest with you because they know that the dumb people in the room are going to believe you, right? A lot of them. >> You wrote that those with good credit are not spending 2 to 3% between consumer and merchant. All my cards are 2% cash rebate.

>> The cost is one and one and a half% and you get a lot of benefits from that. This was in reference to how crypto has not yet won in consumer to business payments. seems like a very easy obvious use case like let's just plug in stable coins, right? But there seems to be just down there's uh there's drawbacks towards blockchain based payments when it comes to the benefits that you get from like a credit card. >> A problem you have in crypto sort of marketing and thought leadership is just sort of disingenuous behavior, right? >> And you're like, "Oh, we could replace this with a stable coin payment and save 2 or 3%." And that's just not true. like you're just lying right now. Do you >> are you lying because

>> uh you're dumb and you don't understand how credit cards and all that stuff really works or are you lying because you own you have some investment and >> I guess you didn't have investment until circle until now but you maybe now you do but you know like why why are you lying? Um, so you know there's there's a ton of services with credit cards. The the net between me and a merchant is like 1%. I always go for the 2% cash back except for like hotel rewards, that sort of thing. And >> and so the net is probably 1% and I'm getting value from that 1%. >> You know, if I were to switch to stable coin payments, it would be for like the McDonald's ice cream cone that we talked about.

>> Yeah. >> I I would do a stable coin payment for that because there's no real economic value there. Like if if they give me a bad if if the ice cream tastes rotten >> or it falls on the ground or whatever, like I don't give a [ __ ] It's $3, right? But am I going to am I going to switch to a stable coin payment >> for, you know, a new $3,000 TV? I'm not because I want to be able to call American Express >> and say Costco just screwed me over on this TV or Best Buy screwed me over on this TV and they refuse to take it back >> and I'm going to get that refund. So you're buying sort of insurance, you know, that's socialized amongst other credit card users.

>> Correct. >> So and so you know to pretend that you can swap one apples to oranges comparison is what I should actually say >> super popular in crypto. Everyone loves to compare apples to oranges >> and act like oh this you know this thing's better than the other one. Dude, you're not the same thing. You know, you're not even comparing fairly. So >> um yeah, that was that tweet. >> James, I appreciate your time today. It's been a fun conversation. Technical, economic, fun stuff. So, I I appreciate you coming out.