Making Ethereum Great Again | Lei & Namik

Transcript

the vast majority of people's interactions with blockchains are via centralized exchanges trading coins right and I mean the the research is there right when you have to wait 10 seconds for something to happen you lose your attention uh if you have to wait even a second or two for something to happen your intentions are interrupted nothing said on zerox Research is a recommendation to buy or sell Securities or tokens this podcast is for informational purposes only and any views expressed by anyone on the show are solely our opinions not Financial advice bacho Ryan and our guests May hold positions in the company's funds or projects discussed what's up guys welcome back to another episode of ZX research today I'm joined

by leay and namuk from Mega E thanks for coming on guys hey thanks for having us um before we get started it might be helpful to for you guys to give a background on what Mega E is yeah um so I can give a quick intro so Mega E is what we call a real time blockchain um a realtime blockchain is a blockchain that is um super fast in terms of both throughput and latency um we're targeting a through 100K transactions per second or 10 gigas per second and a latency of less than 10 millisecond or a block time of less than 10 millisecond um yeah so magga e is buil as a um ethereum Layer Two and

because that allows us to do something which we call um no specialization so we hyper centralize the block producer and at any at any given point in time there's only one uh set of nodes producing blocks which which ensures Ultra minimum uh latency and I mean Ultra minimum block time and on the other hand there's uh we support F noes and replica noes which permissionless join the network to synchronize split sequencer um to get the latest blockchain State and finally we settle back on ethereum the layer one uh with proof um first initially optimistic and later ZK to ensure the correctness of the states I was just going to say we're excited about megi because we believe it

enables a new uh design space for applications by making the design choices we have we enable applications which are real time leveraging our unique uh optimizations to build applications which haven't existed before perfect um so you guys answered my next two questions but I guess let's talk about this term real time blockchain um it's the first time that I've well you guys are the first team that is using it can you expand a bit on um what exactly it is yeah um so I mean I mean the notion of real ESS uh of a computer system uh is not I mean it's not there in in the blockchain I mean it's not invented in blockchain space but rather it's mostly to to kind of describe an operating

system or a piece of software that just responds to user interactions with so much responsiveness and uh gives the user kind of a a a a a feeling that they're talking and they're they're using something that is like alive right so a comparison would be so I mean I draw like two examples one is what do you call batch processing so that's basically the way how computers work back in the days like before 1970s um back then computers are mostly used as a tool to help researchers and what researchers do is to just submit I mean prepare a whole lot a whole lot of jobs for the computer to process and the computer take it as a batch and probably do do the work overnight and next

morning the researcher will get up and uh get the computational results and uh but however nowadays the way we use at least VAP two applications I mean our our daily computers um is that you click you click your mouth or or you you I mean you type something on the keyboard and you expect immediate response on the screen and that's basically what real time really means in in terms of ux and uh yeah basically our our goal is to make the map Street I mean the the world computer as responsive as uh yeah as the yeah as as our daily computers yeah that makes sense nam I remember watching a video on Twitter and it was you

mentioned it that like how we don't when I click on something on my regular computer I expect it to go through immediately or else I get really upset and then on the blockchain we don't really care that much as of State um but that's what you guys are trying to fix that it's as responsible at the daily computer yeah I think in a lot of ways blockchains are not truly usable which is why we have very few developers in this space and we have very few real users in this space the vast majority of people's interactions with blockchains are via centralized exchanges trading coins right and I mean the the research is there right when you have to wait 10 seconds for something to happen you lose your attention right uh if you have to wait even a second or two for something to happen your

intentions are interrupted right with Mega E the plan is to say hey let us leverage the security of eum to maximize for performance and bring a new wave of experiences on chain and that's kind of the north star of Mega E lot for us that makes sense um Le you kind of brought this up node specialization um and anom you mentioned that you guys made some design choices this is kind of I guess the most controversial topic about Mega E is that you may you may you have some tradeoffs um what would you say those are and how are you combating potential malicious behavior or potential I guess downsides yep um so the major I mean the most

important design decision that we took was to um centralize the sequencer or centralize the blog producer um yeah and uh I mean by extension um we are we are calling that no specialization because we we not only do this kind of I mean purposeful design for the sequencer but we also do it for for say for the food notes for replica notes and for the pro notes um so this desci sitution really I mean I think it Roes back in a post a blog post written by metallic um which basically kind of describes an idea that all performant uh blockchain would eventually require centralized block

production and uh the trade-off here or the benefit it provides to us is that because there's only one computer that needs to pump out blocks there is no like cross internet communication that needs to happen um to produce the very next block after one has been produced so imagine like I think a comparison is a traditional layer one let's say ethereum the main net and uh you don't really know where uh the next block I mean where the next block is going to be produced and who's going to be the the the block Builder or who's going to be the proposer of next block and the the moment after you have produced your block you have to send the block to the person and the person has to um apply the block on their local state so that

they can produce their block on top of it and this process of communicating and the replicating the states takes time um there's a lower bound on that which is speed of light right so by kind of collocating um the the I mean by collocating the machines that produce the blocks you can minimize the latency and we and Bic what we do is to push this architecture to the extreme which is to just I mean ram them into just one single machine and let the single machine produce blocks and the the resulting low inter block time which or or block time and the resting high throughput is the major benefit and um I mean the trade-off on the other hand is yes we are essentialized blck producer

but what we're what we're doing to combat this is by rotate the sequencer so instead of requiring a cohort of nodes to be always online and always participate in a consensus algorithm in a critical path to produce every single blog what we do is that we set up a a bunch of a lot of like candidate um blog producers and we rotate this role of the active block producer among these candidates um so we achieve both um liveness because if one sequencer is down we can just rotate it out and we retain all the benefits of this centralized block production that makes sense and I think

this is um kind of what a lot of chains are trying to do is they' we've realized that consensus is the bottleneck um and we're just either collocating or single centralized sequencer um one thing I wanted to ask is so in general with l2s um the Northstar has been to decentralize the sequencer we see this with I guess slower l2s um but this is obviously something that you guys aren't interested in do you think you can get the performance that you guys want or let's say like 10,000 TPS which is lower than you guys with a decentralized sequencer um I would first say that we are I don't think it's true that we're are not interested in decentralizing the

sequenc and uh I mean ultimately we see decentralization as as a means to an end right what users care about our our like guaranteed correctness of execution our our sensorship tolerance very strong sensorship tolerance to be honest and U um but decentralization in the form of like injecting a consensus ptical into the critical path of block production is just one way to do it and I mean this sequence of rotation which I just mentioned is I would say another way to also decentralize sequen but not in the traditional definition of it um but in terms of so I I guess your your question was can we achieve the same um performance if we have a consensus protocol in the critical path I mean the

consensus protocol would be participated by at least tens if not hundreds of notes then my answer would be very hard because if you want it's the it's kind of the same uh it's kind of the same argument that I make right because if you want computers across the world across the internet to reach consensus on what is the previous Buck um it takes at least several run trips across the world and that's almost 500 600 milliseconds and uh there's no chance that you can do um a block time that is lower than that if you still require this consensus to happen before you produce the next block it's just a lot of physics so but I mean suut a probably a different story um but still very hard

yeah just to kind of highlight there um I think for us fundamentally We Believe like settlement da and execution are all different layers that need to be optimized in different ways and we are building the exec layer so yes other nuances that one can take sure other uh one can mid curve this architecture to say hey you know what we can actually still reach a very strong level of you know performance and be you know with some consensus in the sequence architecture but it doesn't serve our purposes uh to the greatest extent right so in that regard we are quite opinionated and why will while we will have a rotational sequence to architecture we're not particularly interested in building less performant

or infrastructure that makes sense and the full nodes as far as I understand are used there to verify that the last good state is good right um the full note so okay so um the food node will reexecute the transactions yes and uh they're they're going to verify that the state they hold is correct uh we also have replica nodes who do not verify transactions but merely serve as like horizontally scal um like read replicas um yeah and also Prov noes which not only verifies the transactions themselves but also proves to the world that hey the transaction the state snapshot is correct fantastic um I guess one more question about your

architecture is you talk about memory computation in memory computation and how you translate smart contracts into machine code on the Fly can you explain those to me like um I don't know that I've never done this before just so I understand um yeah of course so these are actually two separate um optimizations um in memory computation is a kind of a trend that happened um in the data center World um about 10 to 15 years ago so the whole story is um yeah so the whole story is that um yeah most applications require uh accessing and actually very frequently reading or or writing to some State uh to some storage

some database right and the inherent limitation here is that if you write um if you store your data on a kind of a persistent medium which like which for example um your SSD your your your hard drive um the latency or the time you have to wait your software have to wait before the data is fetched or the data is written is actually much longer than if you store those kind of data in the main memory of your computer the ram so the idea is basically that uh now that Ram is cheap enough uh last time I checked it's about1 uh dollar per gigabyte uh in the DDR for server memory so now that the ram is cheap enough you

can actually afford to store all your hot data in the main memory to benefit to benefit from all the latency improvements actually multiple magn multiple a magnitudes of of improvement um by Yeah by staring us the main memory so um yeah it's actually kind of a simple idea right you just put all your data in the main memory um but the execution is actually what I think is pretty interesting here because as I mentioned um if you store it in a persistent medium you suffer from high latency but if you store it in the main in the main memory yes you you benefit from low latency but the memory I mean Auto data in the memory is lost the moment you unplug or someone unplugs

your SE your sequencer so ensuring um the data persistency even if you are storing the data in the main memory um while running the sequencer is a major technal Challenge and actually um our other uh Tech co-founder elom um he and his lab studed in memory computation very extensively uh during his years of his PhD at Stanford so yeah it's a very interesting problem but we have also a very interesting solution um on the other hand I think you also mentioned the translation from from uh evm B codes from Smart contracts to machine code so the the story here is a kind of a technique called compilation which is

again a kind of a a technique from the greater Computing world not just specific to blockchain and uh the idea is that um yeah people can only write uh uh scalably write uh programs in human readable programs so usually uh people write their programs in say C C++ go rust whatever and a tool called a compiler that translates these kind of human readable strings human readable languages to the language that your computer understands because your computer on the other hand is only good at like doing very mechanical stuff like adhering I mean reading and parsing uh programs that aderes to very strict and simple syntaxes so doing this conversion

is the job of the compiler and the in evm too there is a compiler I mean there's a translator because the evm programs are written in what we call EVN bite codes but your computer does not really understand at least your Hardware does not really understand the EVN bite codes so you have to translate it from the bite codes to what we call the machine code which your computer understands and U in traditional evm execution models um there are two I mean Major Performance sufferings in this translation first is that evm um the VM architecture is a much much simpler architecture than your your modern Intel arm AMD CPUs so um basically if you run if you kind of

emulate um an evm on your Modern Hardware you are kind of not using much of the hardware ACC acceleration Hardware capabilities of your CPU hardware and this is lot in the translation and second this translation itself takes a lot of time uh in the CR path because literally you are when you execute a smart contract when you execute a transaction you are literally grabbing the contract from the database do the translation live before you can even round the contract on your Hardware so what we're doing is basically first we're doing this translation ahead of time we're moving it out of the critical path so that it does not really cut into the time of your like of your block

building and second because we do the translation ahead of time we can afford to do much more sophisticated stuff to maximally preserve the hardware capabilities of your mod CPU so that the performance loss of emulating a very simplified evm BM is minimized fantastic I um I want to ask a question about inmemory computation and then I think we should move on to um application layer but in terms of inmemory computation so you said that if a computer turns off and the ram is lost what happens if let's say the sequencer turns off um the nodes surrounding the sequencer have when is the last like good State saved is it per block or is

it a longer do you get what I'm trying say yeah I I see your point yeah it is per block and I think it's a very interesting question because if the data is in the memory and uh if it turns off then the I mean naturally one would think that the most reason data would be lost and uh we have to roll back to some like reason like snapshots so this is actually where this uh e- memory competition is non tribal to achieve because um when you produce a block it's kind of a contract to with with the other noes surrounding you right it's like I will never uh roll back the block so you have to make sure that the data once leave once it leaves the machine it's never lost even in the most bad

disaster which is like a to Total power loss and uh yeah we can guarantee that we can guarantee that um the data is never lost once um aign block has has left uh the signature fantastic okay um let's actually do L2 competition because I want to hear you guys this thought do you think that the current state of l2s is good enough um and do you think that they can compete with monolithic chains which we've seen kind of dominate this cycle in terms of Mind share yeah I think there's a couple reasons why uh existing l2s have not been able to be competitive with monolithic trains like salana we can touch on those afterwards but I want to answer the first question which is do we

think that the current layer twos are good enough and I mean obviously we don't think that's the case right that's why we're building Mega for us there are fundamentally different design choices one can make by being a layer two understanding the role of execution and hyperfocusing on execution we don't think that those have been adequately optimized SL maximized right I think there's a lot of great High perf players that are coming to Market and those will be good I'm sure I don't necessarily have any insight into their timelines and what they're doing but we do believe that we will see the next wave of truly High performant block space exist only as a

heterogeneous architecture right no blockchain that has consensus will be able to be as performant as a blockchain that has no consensus right I I may have fumbled my words there but tldr consensus slower no consens is faster Mega E no consensus and we just think that's like a basic law of physics fantastic um I guess we've seen other l2s try and scale horizontally um I think arbitrum and optimism they've both gone with l3s and other l2s um yeah I I would say base is scaling vertically and there's clearly demand to scale vertically and you guys are clearly scaling vertically

um is there any chance that you're going to do also some form of L freeze or do you just not care I assume you just don't care but just want to ask it so I think here we can actually answer the the question about like why these monol monolithic trains have been so much more successful than um let's say like l2s as a you know basic blanket statement and I think it's it's exactly because we have been a bit intellectually lazy when it comes to Vertical scaling and we've said let's just have a kajillion l2s right and let's create un believable levels of fragmentation and uh breaking the composability of crypto right like why are people in crypto there's generally two uh sets of individuals one it will

be someone from an emerging market that understands the power of not being effectively locked into a social system social economic system right so myself I'm aeran Cy Le is Chinese right like we come from like the global South right but there's a very very large portion of the population that sees the immense value and atomic composability right it's an environment everyone can come and play in and that is for many people the true pmf of crypto and that is something that existing L2 architectures completely break right because when you have a stack where you have some horizontal product what you end up doing is you end up have saying hey you know what to move liquidity between these

different chains in the most trustless manner has crazy uh withdrawal periods right and then you have like a literal domino effect risk so I think that's one of the main reasons why we haven't seen success of let's say layer twos compared to layer ones though I think one exception could be something like a hyper liquid which is an app chain and that is because they have a very particular use case right like per Dees and that can be doesn't need to have uh insane composability of the rest of the ecosystem so yeah I think like for us Stacks are cool horizontal scaling is like interesting and all as well but like we are vertical scaling first right like Mega E with our design choices will

be able to handle more computation than any other L1 and L2 combined it would make absolutely zero sense for us to focus on building a stack as opposed to creating a really very very rich ecosystem of applications that makes sense I have um one more question the evm why did you guys choose to I mean obviously you want to build an L2 um but why the evm this is like something that people have started asking over the past four months and I get you guys on it I'll give like the left curve take and then L can give the right curve take uh we're quite complimentary in that regard uh I mean I think a lot of this alt VM stuff is very overrated right uh the evm has the

richest uh public good public goods that have been contributed and poured into it and the evm has not been optimized into its final state right so why would we kill the network effects why would we kill the mo for anything other than just like a nice narrative right like the reason to my understanding that people love slana is not necessarily because of the SPM it's because of all this other properties the spbm is actually has has a lot of constraints so the whole feces here is like let's focus on uh effectively working what we have and bringing it to it ultimate conclusion and just continually optimizing but they

will give a better answer yeah I think it's along the same line of thoughts um so evm has been there for I mean the longest amount of time in I mean in gring complete smart contract programs uh platforms and uh the the amount of tools and expertise accumulated in the developer Community is too big to to ignore really um like like the static analysis to the fs um all the security tools all the libraries uh so that you don't have to really risk writing uh uh I mean incorrect smart contracts right these are all there and uh it would be too destructive if we even dare to to kind of abandon this the reach asset and also I I just took my lesson from from browser Wars right um I I personally I

hate JavaScript but somehow JavaScript lived um until today and it's striving and uh I mean all those big companies they are what they do is not really invent a new like browser in browser programming language every five years but instead they accumulate 20 plus years of I mean of work into like optimizing the engine behind JavaScript which is uh the only uh VB programming uh platform uh nowadays so yeah I just took my lesson here uh even if you like you probably you can uh make different trade-offs make design decisions if you are given the flexibility of spawning an entirely new execution environment

usually it's better to just go with the mainstream and just like pull the work from all the minds here to just kind of co-optimize the same execution environment what they said which is really really important is like the tradeoff between um the new vm's supposed guarantees right be that like security guarantees and the lack of third- party tooling and libraries to also solve that same problem right I think security is a great example how do you build the most secure application well you know I think a really big selling point for like move is that it's like very secure right but it won't have that same uh Rich third party

contributions to be able to build a really brilliant application day one that the evm would have but that's it what's up guys um welcome back to ZX research today I'm joined by Antonio from from dydx thanks for coming on Antonio hey thanks so much for having me um I guess it's been kind of a big um past few months for you since you rejoined dydx as CEO um how has that experience been it's been great yeah I mean I feel like I and the company are more revitalized now we went through some pretty big structural changes for the company we're operating much more as a startup um shipping pretty fast we released dydx Unlimited couple months ago and development continues Full Steam

after that so yeah it's been really good fantastic and um with you joining have you shifted the vision of dydx a little bit I'd say some but the vision of dydx has always been pretty longterm and pretty Broad and even when I was away from the CEO role I was still obviously really involved in the company overall and still executing towards that vision and yeah it hasn't changed too too much overall we really just want to build powerful products that give everyone the opportunity um for financial upside and have has the um priorities for the team shifted a little bit as defi has evolved it has some so we've been full steam ahead on dydx unlimited and then

that shipped about a month or two ago um now the biggest things that we're working on are really just structal improvements to the product overall making it a lot more performant a lot more reliable and making sure that that dydx is the best decentralized exchange to trade derivatives on perfect um could you run over dydx unlimited for a little bit Yeah so dydx unlimited was one of the biggest product updates that we've ever made to dydx and the biggest headline features of that are instant Market listings and then the mega Volt and these two are really synergistic the mega Volt is a protocol wide Vault that allows automatic liquidity provision and

it makes being a market maker really simple so you just go one click deposit into the megga Vol and then those funds are used to start providing liquidity automatically on basically all markets on dydx fast forward to today there's already $70 million locked up in The megga Volt last I checked um and it's just a really great simple easy to usee product where users can get yield it's also great for the protocol because it provides a lot more liquidity and it especially provides liquidity in longtail markets that's where we've seen automated market makers be really successful and it pairs very well with the second feature I mentioned of instant Market listings this is really cool because with instant Market listings anyone can come and list

basically any Market that they want to trade on dydx so listing is not up to me not up to governance it's just really simple again you can go to dydx just click once and the market is live immediately these two things work together because whenever you list a market instantly on dydx it requires a deposit into the megga volt for that market so there's immediate liquidity on that market and you can start trading right away this really is important towards our vision of allowing users to trade basically anything on dydx that's really what we're building towards all right let's talk a bit about Mega Mafia um so there's a lot of performance chains out there and I feel like the big thing now is getting as many apps on

your chain as possible and making people want to actually use your chain um do you guys want to run over what Mega Mafia is and then maybe we can dive a bit deeper into how it works out sure I'll talk about the inspiration for the Mega Mafia um every us everyone who's listening has come to the same conclusion we've made some Extreme Design choices right we've looked at these trade-offs and we said let's take them to a logical conclusion and the the promise Behind these decisions is like crazy performance well if we didn't have any applications which were leveraging this crazy performance well what on Earth are we doing here right so a year ago we kind of came to this conclusion we came to this another conclusion which is like okay you know the Venture Market is a bit weird right I come from Venture

background as well where people aren't willing to take a risk on individuals right unless they have some sort of pedigree and already established and that's where we saw our Edge come in we have this new blockchain it's contrarian we have a new design and this design could conceptually bear new applications let's do our best to find the smartest people we know make bets right asymmetric bets on people who Market does not price well right and you'll see that of a lot of the mega mafia they aren't your pedigree you know 10 years at Google or like have been the top crypto companies they're young Scrappy hungry people that have great ideas and are willing to basically give it their

all to build applications which didn't exist before so we found them we put them in a room together and we just basically cooked so the first situation of that was in Berlin we basically flew 14 teams to Berlin we all c-li together for a month work hard play a lot less hard and effectively like what came out of that was 10 great applications right uh top tier teams have been able to raise a lot of venture financing from Top funds and building applications which are only possible on Mega E right and that could be defi applications which leverage uh 10 millisecond block times uh unbelievably high fruit put uh coo meaning real-time price oracles and

price feeds right but it can also mean non defi applications right like things that bring amazing emotional value by having such abundant amount of space that you can now put things on chain that you couldn't have put on before because data is useless right examples of that could be games a fully onchain Minecraft it can be a dating application which leverages your knowledge graph to give you recommendations on who you should go meet it can be uh game shows right like trivia games we've been able to incubate SL accelerate now north of like 17 18 teams which are all building really unique applications and they're all young hungry Founders which are like

the Next Generation we think of etherium fantastic um when you say incubate and accelerate how much support are you guys offering these teams um sure so it's really interesting because I think that the mega Mafia was more so a function of a pre-main net Mega Eve which will not be the future forever it's not for long honestly and it was done with the idea being like okay I'm actually putting my entire weight behind you right A lot of these ideas for context Mega takes no equity in any of these teams right for us to our win condition is if this project becomes a billion dollar business on Mega because of our design choices Mega if is

validated so first we take nothing how how Hands-On are we unbelievably right Le is a really busy guy he has weekly of all of the engineering teams I'm not nearly as busy as lay I also have weeklys of all of the teams right we ideate from product to uh you know economics to the uh go to market strategy to the user behaviors to the tokenomics to the fundraising all of it we do hand in hand fantastic um last question is what are well you mentioned a few of these n but what are some apps that are uniquely enabled by Mega's performance and architecture uh one example can be

so let's start with defi right I think that the The Grail of the Holy Grail of defi has always been like the central liit order book it's a bit boring in some ways because everyone talks about it all the time we have multiple teams working on Club architectures both on um both lending as well as like traditional like spot and Cur architectures right so that's pretty interesting The Lending side we haven't seen a lot of people explore lending for clubs uh something that's being done uh on we have some really interesting uh trading architectures right so another one of the teams which is being built uh you know part of the mega Mafia is effectively saying hey there are all kinds of really interesting trading strategies which are now possible with

want to 10 m second block times and proportionately there's really cool aror opportunities between that blockchain between C exchanges and between other blockchains what if we were able to take all of these trades all of this yield and give it back to users so there's some really interesting things going on in the form of stable coins there um on the non-directly defi design space uh I think games are really cool right everyone loves the idea of games but the issue is there are none in crypto none of they're successful and I think it's a function of like usability right if I move my character on my laptop right and it takes 12 seconds for that to be represented like that's an horrible ux right yeah on Mega that's no longer a

problem and we think that's going to be an unbelievable um unlock I I think like you'll you saw a note announcement of a mega Mafia just a week ago that is building effectively autonomous world engines on Mega E right and it's the whole idea there will be like you can do all this crazy stuff with pure crypto composability and have seamless iux we have some other really cool stuff coming up as well around live streaming uh really fun games uh couple interesting ponzi's which are intellectually stimulating we have a wide range and to anyone who's listening right like the best way to stay in touch with the mega Mafia is really just zerox

Mega Mafia we will more or less retire to program after mainnet because you know we don't want to be working so closely some particular teams um but we're really really excited of the caliber of uh Founders who've been able to inspire to join us fantastic um yeah guys make sure to check out um mega Mafia somehow like on zerox Mega Mafia and on the I guess it it used to be on your website I it's still on your website we have a new website coming very soon by the time this pod is live you should check out our website mega.com or mega. systems both rout to the same

domain it looks pretty sick very good um one big thing that you guys did over the past few months is I guess over the past month uh is the Echo race um what was the process like are you happy you guys did it and I guess why did you guys decide to do it sure uh you know I think we are I wouldn't I don't know we want to say old school but like definitely quite sentimental for the days where like crypto walk the talk right and like what do I kind of mean there when I go on crypto and I had like you know a little a little bit of money and I had equal opportunity and equal access to projects that I had belief in

and I had conviction in and I was able to like do okay for myself right and kind of have a step in crypto and start growing and like meeting people and like I was able to pay for my first ever for like real IRL event right because I had made a little bit of money in crypto and I was able to go there I mean now if you aren't like an angel investor in this previous bare Market of like the Primo like really good deal flow you you didn't make money right because what ends up happening is a lot of these guys end up looking at um users as exra liquidity right that's not really how we see things right I think on our end um we didn't need to raise more money right

we were able to raise a decent amount amount of money to build our Ambitions what we really needed to do was make sure that we can get uh everyone to have equal terms right so we're very excited to have been able to use Echo to effectively bring in 3,200 investors to participate in mega at the exact same terms as vitalic and Joe and dragonfly and Kobe and all of these other guys and for us it wasn't really a question of like oh well we're diluting our company right we there is Lally no point of doing this if it is not to enable like an entire new generation of Builders and Founders and users to kind of accelerate and come on to

crypto very good I think it's a very vant Valiant um effort and goal and I I don't know I thought it was very cool when I saw it um I guess to wrap it up a bit what are some mistake that you guys have made uh with Mega E since the launch and what is something that you definitely would have done differently well I I I don't think we have made too many mistakes on the technical side um yeah well all all the I mean all the all the initial optimizations that we envisioned um in our very first right paper um in memory computation um paralyzation with uh provable that is probably free of cascading

boards um compilation uh a new peer-to-peer Network design yeah they they all work well actually so turned out that every single one of the uh optimization that we conceived worked well I think we're a really Scrappy team right on the non-technical side at least and we sometimes were just too quick right like I think we really took the idea of like you know move fast break things and it would could be a bit like a bit confusing right at some points you were saying millisecond block times other places we say one millisecond block times uh I mean in actuality we can go to submillisecond block times that's the

actual goal but not having consistency around language I think we also shipped our website a bit too fast I think a lot of people were like yo you raised all this money why do you have an AI mascot which is a totally valid point I think that's a very valid criticism uh I will say it wasn't fully an AI mascot we had a little bit of design as well but I think we were we were just really really excited to hit the ground not only running but basically sprinting and that kind of not backfired but kind of made things a bit more um chaotic I think also sometimes we can be a bit overly opinionated on Twitter right so we can

get into unnecessary fights or just like we can be a bit bantery I we think it's fun but understandably it can end a bit agitated so I would say those are the two too I think we've gotten a lot better about our mistakes I think we're really excited about uh effectively building this ecosystem out we I think we've been able to do that so far so well I think the tech is coming quite soon so that's going to be really exciting for Mega to kind of face its next Evolution all right fantastic um last question who is your biggest competitor in your mind and how do you see the L2 landscape shape up over the next year sure I can go first um I I I only see salana really um for a couple

reasons right I think salana represents a like I think we have a similar goal so one of the reasons why I think they're our largest competitors is because we have similar goals to enable Like This Global State machine right we have very different approaches towards it right salana is monolithic and its Most B simplest core while we are a heterogeneous Network by by Nature right and obviously we think that our architecture makes the limits much much higher and is more much more scalable but I think that when I see the our competition for talent for Builders for eyeballs I think that's all salana I think salana has the same value

propositions as us and it has the same uh North Stars as us in terms of like what's the kind of experience they want for their users um in terms of how you see the layer 2 landscape emerge I think there was a a pretty good argument SL conversation around this I think between Jesse and anatol on Twitter right do we see a million l2s or do we see a few I I personally lean more towards a few at least in the short term right in the next few years I don't really see more than a couple l2s existing uh as a function of like you know these are countries these are civil like civilizations these are users these are like ecosystems and if you can build vertically right you can your population

can can get so much larger and then you have massive Network effects of that um and yeah I mean I think that's my main my main opinion let's see a world where all of these things get like sent to their absolute limits then we'll see more l2s emerge uh and I think we also need to see a lot more success of the you know Atomic interoperability uh for these kinds of other architectures to really be valid they what do you think yeah I think yeah in my mind our our big enemy is actually a kind of an aify view of what a blockchain is and what a layer to how a layer to should be built right um say embedding because a trend that I see among among many layer TOS is that they are kind of becoming more and more

similar actually like I think you mentioned that many layer TOS are trying to decentralize by reintroducing consensus and I think that's kind of a walk back of um of I mean the the their two idea which is to to kind of be a architecture that's starking different from a from a layer one architecture so I think a kind of an aify view of how Layer Two should be like is actually a major is actually a major enemy to us because um yeah and but also I think an opportunity for us because basically the goal for us is to show that hey there is a different way to build a blockchain and it will be equally secure equally um useful but much more performant um yeah

so I think that's probably one of the enemies and yeah I I would agree with namik that um yeah our yeah I would view Solana as a team that has a very similar mindset in terms of the trade-off between performance and the and the other aspects of blockchain right and uh well I I would say that we go slightly more extreme than them uh still um but yeah yeah we I really appreciate how product focused salana is and it's like it's like a corporate principle in absolutely everything they do still a breakpoint compared to um like I don't an e conference they're fundamentally very different right and you see that in the way they look at uh

productizing the road map perfect um is there anything that my fellow listeners should look forward to um that you guys want to bring up I guess like maybe over the next few months when is main net when is test net when are we launching the product is my question um so I guess like what you can what we can say is this right uh we are really really really happy with the private test right uh we're talking about multiple gas in terms of performance of single single credit performance we are moving very very rapidly towards making that into a public test from a public test to a main net let's hope this public test n works right knock on wood

Mas um out side of you know the technical road map I think we are really really just excited about getting as many people that are value ad into Mega Community as possible right we are really excited to kind of bring more people into the fold and we're really forward we look forward to having hopefully some of the listeners here join the mega Community as well all right fantastic um guys thanks for coming on this was a lot of fun likewise thanks for having us m [Music]