How Parallelized State Execution Unlocks Transactional Throughput at Scale | Logan Jastremski

Transcript

but the narrative in 2017 was really along the lines that ethereum particularly was going to bank the unbanked I paid ,000 for Unos swap transfer during defi summer and at that point in time I was like all right this is not the future of France if you are an Empire listener hopefully you've played around on chain and if you have done that you know that transferring assets across different chains is a pain to put it nicely that is why we are incredibly excited to have the Wormhole found as a partner of the Empire podcast stewards of the Wormhole protocol supporting over 30 different blockchains and six different run times stay tuned later in the show we have a cool thing that you can claim which is a wormhole nft just for Empire listeners

this episode is brought to you by PayPal here at Empire you know that Santi and I are really into real world use cases and always on the lookout for the next killer blockchain app we're excited to share that PayPal has arrived in crypto to unveil a way to seamlessly connect Fiat to digital currencies later in the show you will find out how you can use PUSD to check out at millions of online stores what's up everyone before we jump into the episode little plug for digital assets Summit coming up in London March 18th to 20th tickets are pacing so far ahead of schedule that we had to decrease the discount codes so instead of Empire 20 it is now Empire 10 head over to the website digital asset Summit Das London March 18th to 20th use code

Empire 10 and get 10% off your ticket see you in London all right everyone welcome back to another special episode of Empire today we've got Logan justy is the Americanized version of yky here from frictionless Capital Logan welcome to the show man thank you so much guys uh really looking forward to talking and excited to be on the other side of the mic yeah yeah pum for this okay so other side of the mic that's a good place to start actually when I first met you uh you were running around like a chicken with your your head cut off basically with this huge backpack that I think you'd spent like $10,000 uh you taking your savings from from your time at Tesla and you're like I believe the world works this way and

everybody believes something different everybody believes that the evm is the end all Beall and we're going to scale with l2s and I believe that they're wrong and I'm going to go create a podcast that and and a fund uh on the back of this thesis and so I think s and I both want to get your your thesis on Celestia and move and salana and you've been ridiculously right about a lot of these things um but actually it would be helpful as a starting place to just hear what is Logan's mental model and framework for like how you view what blockchains are supposed to do I mean it's a great starting point that time in New York was funny I I've gotten very efficient at uh recording podcasts on the go but I I think for me

I just wanted the industry to scale I I got involved in 2017 within the crypto markets during that bull market ultimately kept with it throughout the bare market and got really excited during defi summer as a lot of others did with kind of the idea of composability and these permissionless smart contracts but the narrative in 2017 was really along the lines that uh ethereum particularly was going to bank the unbanked and I paid ,000 for Unos swap transfer during defi summer and at that point in time I was like all right this is not the future of France like why are fees so expensive and really just started exploring different Avenues

and I think from my time at Tesla was really kind of beaten into me like think from for principles boil things down to their fundamental Primitives and build from there like question assumptions and all I saw throughout kind of my time in blockchain like even today it's really just kind of bag bias it's like hey my bags are better than your bags because I have a bigger bag or no one's really able to like articulate why these things actually work um like kind of from a engineering point of view and all I was really trying to do in my journey has been established kind of Base Primitives that can be comparative across the industry which ultimately led me kind of down the high throughput blockchain

rabbit hole and once I kind of went down that route it was hard for me to look another way and so I've just I mean I've been saying that same things for two years uh that high through put blockchains are going to bring Mass adoption to crypto and I still believe that today and ultimately we started frictionless Capital with the thesis that these blockchains are not 2X or 5x difference in performance it's quite literally 100x if not a thousand X difference so I'd be remiss not to say that every cycle we have uh the competing L1 trade um I'm hearing you say this time is different I have the benefit of having talked to you um over

better part of you know couple months uh maybe walk us through that Journey you know which when you say high throughput you know are you talking about salana Aptos say suie um what what are they able to achieve that ethereum has fallen short in their vision of scalability um and what are some of the more uh important breakthroughs that you've appreciated in this journey I ultimately realize that you're going to hit the limitations of these systems like I I I like to think of like all right you push each of these systems to 100% capacity what does that look like and then how do you further scale

from there and what I really realized was even if you use layer twos which I kind of view as like data compression you're compressing data down to to use the block space of the L1 more efficiently eventually even if you use 100% of the block space for the L1 for with layer twos you still at the end of the day are going to have to increase the amount of block space and this is ultimately ethereum's road map and what they're working towards with 4844 uh it's not a meaningful amount kind of compared to high throughput blockchains they're adding like 375 megabytes which in the grand scheme is not that much but meaningful amount for ethereum and then ultimately scaling with d sharting down the line which is a couple

years away and for me what I really realize for all these systems is you really have to scale the amount of data and this is kind of what is happening in the current narrative with like Celestia igen layer scaling data availability layers as the core bottleneck to all these blockchains and once I kind of realized this I ultimately just started going down the rabbit hole of what do these differentiated ecosystems look like and how are they actually to your point Santiago different than ethereum and so what I've come to appreciate is truly the fundamental differentiation between all these ecosystems is just how much is the hardware going to cost I think my

point of view on kind of like full nodes is the physics has to happen whether you that happens on the layer one that happens on the layer two if you want to do compute on layer that's fine if you want to do it integrated into L1 that's fine uh if you want to offload data availability to ig layer or Celestia that's fine as well but you have to do the physics you have to have high throughput blockchains and eventually you're going to have to have high computation in some form or fashion you can move those pieces around but you can't eliminate them and so when I started really going down and exploring this more and more I was kind of just

fascinated by how do these actually work from the product perspective and the engineering perspective and to me it just made a lot of sense to integrate them together uh you don't have the liquidity fragmentation you don't have the confusing ux that kind of layer tws have and the more that you can kind of simplify that user experience and developer experience it made a lot of sense but there's a couple key tradeoffs or not trade-offs but a couple Key Properties to me that are really different from kind of the ethereum killers of the past one that's being high throughput so you're doing a lot of data availability and then the second component of that being parallel

processing and happy to dive into both of those yeah maybe um paralyzation is something that has come to light and is very much a buzzy term right now so could you simplistically talk about what that actually means and who who have this feature and and which blockchains don't so the ethereum virtual machine was ultimately really Innovative at the time because it allowed kind of Turing completeness and the ability to build smart contracts on top of ethereum which was not possible really at the time with Bitcoin and during that process though

it was a relatively simple virtal machine it executes transactions sequentially just meaning one after another so if I send you Santiago $10 but uh my brother wants to send my mother $10 simultaneously even though those transactions have no relation to each other they have to be executed one after another but and that's one obviously fairly low uh it it's much harder to do that from it it's more simplistic from an engineering point of view but it's harder to scale throughput in that way because you're always uh doing transactions one after another Solano is ultimately the first blockchain to do

parallel processing uh it enables you to really one do more transactions per second because you can instead of doing all transactions sequentially if they're not overlapping or non-contentious pieces of State you can do them in parallel one that massively increases the amount of throughput or trans actions per second that you can do on the network and secondly what it allows you to do is to isolate fees to specific applications you can think of like more granular fees than like what exists in uh say the ethereum virtual machine uh in the ethereum world you have say The Bard yach Club mint happening uh a bunch

of people are trying to get access to that state but I want to do a Unis swap transfer and because everybody wants to get access to The Bard at yach Club mint uh it is making my Unis swap transfer very expensive uh everything on ethereum spikes even layer 2os uh because they use or most of ethereum most layer twos because they use the ethereum virture machine what salana ultimately enabled and why it had some downtime was it started with parallel processing but soon after the fact figured out it need to localize these fees to specific applications because they're using more resources and you only increment that specific application for slightly higher fees um and ultimately what this enables

is multiple applications in like a single ecosystem uh it's one a much better user experience and two it's much easier for engineers just on like resource pricing um it allows multiple applications to live in a single ecosystem which is like a massive breakthrough yeah I want to stop there because before I continue down this optimistic and like like paralyzation there's two different kinds my understanding like why um did ethereum in its initial State didn't consider par parallelization was there some sort of research breakr or technical upgrade that was possible I I no I mean parallel

processing has been like kind of like low hanging fruit or like best practices I think since like 2010 ultimately Engineers figured out that it's going to be harder to add more transistors to a chip uh so every year you see Apple come out with its new silicon it goes from 12 nanometers to eight to four and two and eventually you're going to hit the limits of physics where you literally cannot add any more transistors to a single chip and so the way that they uh increase compute capacity it is add more chips or more cores um so they've Rewritten and this has been kind of the best practices for the last 10 years or so to for

software to take advantage of multiple cores and you spread out the workload on these additional cores and it wasn't like some unique thing per se that salana came up with I think it was just perhaps Anatoli kind of having a background in like high performance Computing uh at Qualcomm that allowed him to like identify this uh nowu does this apos does this monads big thing is we're going to paral lies the ethereum Birch machine um but again this is some of the limitations with even l2s because l2s do not really they haven't innovated on the virtual machine design and so they're kind of taking the worst parts of ethereum virtual machine and bringing that to layer twos and so I mean when

arbitrum was doing uh their token drop you saw fee spikes to I think like $510 uh base you've seen their fee Spike uh so like I'm very curious how these L2 are going to adapt over time that have adopted the ethereum virtu machine because they're going to run into the same limitations that the ethereum did yeah we talk about this idea of um could ethereum Implement some of these new like um technical features the the component that we constantly in the term that I think we hear is backwards compatibility like how is is it possible for ethereum to upgrade to and become a parallelize like parallelize the evm you

talk monad um are you aware of ethereum at some point throughout the last you know so many years thinking about this and and and maybe it's not just possible to do it I I think there has been a EIP that allows developers to specify what trans what state they're going to touch within the ethereum ecosystem but and that's like a great Next Step but I think there's issues around being able to do the localized fees it's really parallel processing enables two big things which are ultimately today nice to have but in the future going to be mandatory that is unlocking additional

throughput by going from sequential processing to parallel processing and then unlocking uh Global fee markets to localized fee markets both of those to me are Paramount for really future scale I don't know if it's ethereum from your point Santiago if it's going to have some backwards compatibility or not uh if it I I think honestly I mean if it were super easy monad wouldn't probably exist uh monad has really started up because some of the limitations of ethereum keeping some of that backwards compatibility with ethereum but making it high throughput and paralyzed Logan if I had to try to

summarize this conversation up until this point let me try to do it just to kind of get get my my mind around your thesis basically it's uh you you believe blockchain should be built to scale but fees are uh preventing that fees are driven by two factors block space and state contention uh on on the Block space side ethereum blocks are very very small 4844 increases the size and then DK sharding will increase the size a little more but when you compare it to something like salana or something else it's still you know maybe one 100th of the size uh so that's block space and then State contention um is uh that's where like this paral idea of parallelization comes in is that kind of

like your so that's when you basically say all right we need a new solution I know you've been really excited about salana now you're starting to get excited about other things is that kind of a fair Assumption of your thesis today yeah I mean at the end of the day there's a massive difference if fees are $10 versus $5 versus 50 cents versus 5 cents versus uh a thousandth of a scent like the types of applications that you can make are fundamentally different and allow different levels of scale and I I think the unfortunate thing about kind of the ethereum ecosystem thus far is any application that has started to achieve some levels of product Market fit kind of has anti- Network effects where the

network and the application actually get worse the larger scale it becomes and so all I've really why started looking at alter alternative systems was because I thought crypto was relatively Niche uh we're still I mean if you look at daily active addresses which may not be the M spectric but is some proxy to activity on chain these are generally measured in hundreds of thousands of like daily active addresses not hundreds of millions uh and when you compare that to web two it's just a fraction and I think if we really want the industry to go to mass adoption scale is the number one thing that has been limiting us and engineers and I think once you kind of

go from this dialup of ethereum world to broadband and then fiber optics you can build new applications like YouTube uh to the equivalent of web 2 that you could build on high throughput blockchains and I think that is why we're so excited about what is happening uh in these high throughput ecosystems really salana taking off uh the newer chains coming as well but there's really fundamentally new applications that are unlocked that are not possible on these low through putut chains because of the limitations that you described um you you mentioned something that I think is is a key property I think which is this idea that for the most part every blockchain we've seen has struggled from anti- Network effects

right the more users that want to do a particular activity at any given point in time you're ultimately competing for Block space that has a spike in fees and that degrades the experience um you mentioned you know Network effects and how that could be possible what are the big unlocks that Ena that uh is it like local fee markets and who actually has that because I think we talk about like this idea of localizing fees um but my understanding is even salana like it's still very much in development and hasn't been fully implemented so can you just talk walk us through kind of the what network effects look like

in in which kind of blockchain can we will be able to see that and and what are the like breakthroughs that enable that I I do think it's a combination of being able to have more granular fees over time so like resource pricing of the network at the end of the day this is a shared network and there's not an infinite amount of resources and so the more granular that you can be in that resource pricing the better I mean just because ethereum was first and kind of the early adopter for this I think they kind of took a basic point of view but today because of that you have these kind of global resource pricing which just makes it hard for any one if one

application hits scale on ethereum or an evm L2 you you start to suffer suffer these anti- Network effects and I so to your question it really is like one being able to granular fees to kind of be more granular on resource pricing over time and then increasing the block space and I think both of those kind of have to be done in Tandem and this is um maybe it would get down to the debate of like how big of node should be and like the differences between Celestia and salana but I think in terms of answering your question on like Network effects I really think it's

just like the more state so like more people care about things that are going on in blockchains the more valuable that state becomes so it's really like where are the users going and where are the engineers going which is kind of cliche but I think these new blockchains and why I've always been more excited about high throughput chains than kind of the ethereum or low throughput chains of the world was because scale is a requirement um unless you want to stay Niche and so to me ethereum kind of optimize for the incorrect thing with keeping Hardware relatively low and focusing on like profitability before hitting scale where these high throughput blockchains have

focused on scale and then once you hit scale you can eventually get profitability Logan there are your initial conclusion to this was salana but now there's a lot of other folks on the market so there's Celestia approaching this from the data availability perspective there's monad who's kind of rebuilding the virtual machine there are are Aptos and suie there's SE obviously what has your been your like evolution of uh V1 is okay salana solves a lot of this stuff excited about salana V2 is is what I don't know if there's a V2 and and that's the kind of hard part all these blockchains are now kind of writing the software to take advantages of

increasing bandwidth and increasing kind of parallel processing so when you add more bandwidth with to the network or more block space the network you don't have to rewrite that every time that happens if you go from 100 100 megabytes to 200 megabytes the network should automatically know hey I can ingest more data and send more data out uh same with uh kind of adding more cores to the network and all these new blockchains are kind of slight variations of what salana is kind of like first stated of hey we're going to increase throughput and we're going to increase computation but these are not I mean 10x improvements over salana uh going from

ethereum to a salana enable parallel processing and doing high throughput was quite literally 100a and like performance design these newer networks are slight optimizations in either how they do parallel processing how they come to a consensus um and more optimizations than like massive leaps in terms of of this is a network is fundamentally that much better maybe this would be a good time to go into the two different types of parallelization right deterministic and optimistic and you know there I think as I understand it those are probably the biggest distinctions between salana sui monad apto say like

um so yeah in in these different networks you can do parallel processing differently um salana andu both do deterministic parallel processing all that means is you have to as an engineer you have to state which contracts you're going to interact with um ahead of time and it's kind of annoying from an engineering perspective because it puts more work on the engineer to do this but ultimately what the result is is you know ahead of time if to uh if you're going to be conflicting and transactions uh the conflicting transactions are super important to know because that's when you need to know hey uh Jason you

really want this m trade versus me you're going to actually bid more than me and because of that you're going to have to pay slightly more to get access to that um and by having the engineers state which contracts they're interacting with you're going to know ahead of time if you're going to be able to have these be processed in parallel or if you're going to have that state contention um s or Aptos say and monad all do optimistic parallel execution they kind of push the problem from the engineer down to uh the virt Machine level where it's a little bit more engineering

friendly because they don't have to State those those dependencies but what they'll do generally is run all the transactions like they would have been run in parallel and try to so some of them will run some of them will not run obviously there's going to be some conflicting transactions and for those that do conflict they have to rerun those again um so there's slight differen and one the performance and two kind of the user experience I think the more interesting thing is with optimistic parallel execution if you're not always knowing whether you're going to have state contention or not there may be some variable and fees um that may be a

little bit more onerous on the users than say the approach thatu and salana are taking which is deterministic when I what I just round out this when I first heard this distinction I felt that deterministic was a superior design long term while you are burning Engineers I think just mechanically I I just feel like that's more elegant solution and one that would allow you to really discriminate conflicting transactions and focus on that and optimize fees optimistic parallel execution will only ever be as good as deterministic right so deterministic

like even I mean it's nice to have optimistic because Engineers don't have to think about it and it's handled by software deterministic is really nice because that's the best possible performance you have because you know whether things are going to conflict or not the things that don't conflict you can just do in parallel and those are fairly trivial but optimistic the best case you can get the performance to is deterministic so yes deterministic is nice uh from the performance standpoint and user standpoint because you'll always know how much fees are but it's a little bit more onerous on Engineers we could keep going down other features of blockchains you know consent

well ask go ahead go ahead no no I was GNA transition but I want to maybe you go first I have one more question on parallelization Logan which is what do you think about folks who are bringing parallelization to the ether rollup ecosystem like folks like um you know maybe yeah like NE what Neil and is doing at eclipse and folks like that you're so bottlenecks even if you I think I mean funny enough I think they're the best l2s uh because you think the best l2s are the are the rollups that are basically leveraging something like the SPM you're saying because they do localize fee markets uh you have to localize fees because

today even coinbase's fantastic has been a great Champion for the industry they don't have localized fees on base so if frch comes roaring back and they start scaling Frac is going to be the single uh driver of fees in the entirety of Bas and all other applications built on Bas will suffer and that's true for optimism that's true for arbitr that's true for polygon all the ones that have copy imposta the ethereum virtual machine as it stands today will suffer from those issues so you have the same anti- network effects that you had on ethereum you have to be able to localize fees and this is why eclipse and movement have

really I think leapfrogged them despite you don't have kind of the backwards compatibility with ethereum and Engineers have to rewrite their smart contracts which is really in my mind the only downside which it is a big downside just to be honest because you have to rewrite your app like that takes quite a bit of time um but just moving the virtual machine to a ethereum of the world you're still limited by the block space so because there's not that much block space on ethereum uh eventually you're going to hit that capacity limit and then fees will Spike and so this is why generally ethereum is viewed as the settlement layer instead of the data

availability layer because you're going to both need to scale throughput and you're going to need to scale computation which um these newer virtual machines do with parallel processing but again you can't obate the physics like if you want uh you need High throughput e if they don't want to do that at the Bas layer that's fine uh arbitrum Nova will do it igen ler will do it Celestial will do it but that throughput has to happen you can't do low throughput and scale same for the virtal machine you're going to need High compute if you want to integrate that totally fine if you want to put that on a layer two also fine but the physics has to happen you

can't get around the physics of scaling these things um I remember seeing a great chart that you have and maybe we'll link to it um but just give us a sense of perspective of of of on on a relative like scale like ethereum some of the l2s like where do you start reaching that end point and and limitation and how does that compare to a nextg blockchain and then salana in its current state salana with fire dancer just to give people a um a rough sense because every time I talk to an L2 or or a project it's like well I don't get the sense that many understand what that limitation will be yeah uh but I think you've done a lot of work on that on kind of really spreading that comp if

you well was it this one Santi this one's a little bit old but yes so what I would love to do is create like a API that ultimately could pull some of these things but this is directionally kind of correct um so ethereum today does 80 kilobyte blocks uh there is a, kilobytes and one megabyte there is a th000 megabytes and one gigabyte and there is a th000 gigabytes in one terabyte uh just for everybody's reference so every time you go from one megabyte to or one kilobyte to one megabyte it's a 10000x uh it's a little bit hard to

understand these things because the numbers are just vastly different but ethereum today is relatively small uh what 4844 is going to do is is ADD 375 megabytes per uh per block but the block times are every 12 seconds ultimately DK sharding which will ship in 2025 or 2026 will be 1.3 megabytes per second as you see here selesia they're today I believe doing eight megabyte blocks every 12 seconds so these numbers kind of change igen layers targeting 15 uh megabytes per second and salana today is doing 40 megabyte blocks every 400 milliseconds so the difference again between ethereum

and these high throughput blockchains is massive like I don't think we would have started frictionless Capital if it was a 2X or a 3X difference it is quite literally 100x if not a THX uh salana fired dancer Target if you go to their page they've shown them doing 25 gigabit uh transactions per second on four cores and they have a plan to scale that to to a 100 gigabytes or gigabit uh so I I kind of think in long term these systems do end up closer resembling kind of like highfrequency trading where you're ingesting hundreds of gigabytes per second parsing that and then doing execution on it than what most people

believe is like the ZK um version of the world so when I look at this and assuming fire dancer is coming later this year I not totally talk about targeting break Point um which would be in September but I mean it is it puts it at a whole like order of order of magnitude and a half or so relative to even nextg blockchains people don't realize the difference it's massive yeah so two questions one is translate the megabytes per second to us speak you know um because I've heard

anatol say the the question would be like does fired Daner like can we get Visa can we get NASDAQ on the blockchain with fired answer um does that have like I want to give people like a practical sense of what these numbers are so fire dancer was demonstrating 600,000 transactions per second I think on 10 gigabit um and Visa currently at its peak I believe is doing 65,000 transactions per second uh I think it's more steady state around like 2 to 5,000 if I remember correctly so I mean there's a lot of Headroom in these systems I think I mean I think the biggest takeaway that people should get from this conversation is you can't get

around physics like if you want to move data availability layers to somewhere else fine if you want to move compute somewhere else totally fine but you still have to do it and if you know that then it's integrating or no coming to the decision of how do you want that user experience to be do you want to integrate them or do you want to modularize them I think the integrated approach is much cleaner I think it uh concentrates liquidity I think uh it removes the fragmentation you don't have uh arbitrium usdc and optimism usdc and base usdc and fundamentally it I I think it's just much better um but it's massive in

difference of performance I I started asking teams with this question which is why would you build on anywhere else like why not just build on salana if you understand this fire dancer if you believe that's going to happen why even bother looking at and building anywhere else that's at some point you know I so I mean we can get into like the differences between like Celestia and salana or like sui and salana there are some differences in tradeoffs and like how the approaches things but at the end of the day the high compute and high bandwidth are going to be requirements I I just don't think

Engineers fully appreciate this um and how large of a difference these blockchain systems actually are and I think it will really take one kind of breakout consumer application to kick off that flywheel but I I think it's happening I mean to me salana really reminds me of what happened in 2018 with ethereum where all the Bitcoin people were dancing on ethereum's grave at that point in time because they're like it's only used for ponies it's for icos uh it's never coming back and bitcoin's the Holy Grail and obviously the engineers that stuck around with it on that time I mean are now Legends uh the the AES the stonnie compound Robert like those guys

are ogs in the space because they stuck around and kept building and I think salana was in a similar position post FTX post all the downtime with the bugs that it had the people that stuck around believe in the technology and understand the difference in performance and now it's not the copy projects of ethereum so to speak they're really innovating on how do I build high performance blockchain um applications because you have a new Sandbox to play in and I think that's what we're excited about at frictionless Capital is these new zero to1 applications that are only possible on high throughput blockchains Logan if you're wrong about this thesis why why will you be

wrong I I I try to ask myself this I really have a hard time uh articulating the bullish side um of kind of an ethereum of the the world and I I really try because the the goal Santi recently well there's two buckets it's not just ethereum there's like in my mind there's kind of three buckets that you're laying out there's kind of ethereum which is still has 90% of the Mind share 90% of the developers 90% of the liquidity like we got to remember that so there's ethereum the the developer report I think had ethereum or Electric Capital developer report was like 3,500 or 4,000 engineers in the ethereum ecosystem that

were full-time I think salana was at uh 500 um so it's it's not like a 10x 20 maybe it's not 90% but let's go it's s to one Whatever any you get what I'm saying which is the the the overwhelming mind share and number of people are still in ethereum land so there's bucket number one is ethereum scaling with l2s whether it's ZK sync or polygon or arbitrum or optimism bucket number two is where the thesis has been planted which is salana bucket number three is what I'd call like new cycle new things like Celestia um Eclipse uh monad sui Aptos SE like that bucket as well which you know maybe some still use the evm but

like they're they're they're they're new experiments so like the ZK of like world view you're doing I view layer tws as data compression and new execution environments you've created a new execution environment you're still going to have state contention I mean most likely because multiple applications are going to live in that network uh So eventually people will bid the fees up to that economic value within that layer two uh so you generally have similar prices as what you would at the layer one but I kind of view them more as like data compression you're batching a certain amount of data and posting that

somewhere else which is totally fine you can do that but again because you're doing data compression you want to really post your data in my point of view on the blockchain that has the most data to compress like going back to that chart that frictionless Capital publish ethereum is a very low throughput blockchain uh and there are newer blockchains like salana that do much higher throughput if you want to use a layer two totally fine but you want to use layer twos or data compression on the chains that have the most data to compress you can't infinitely compress data um you can do like a ZK proof of a ZK proof like you can recursively do these things but at the end of the day

it's kind of like one app chain to another there's like real world latencies that you can't like abstract it gets very messy um it's like is my Unis swap uh pool on this Shard or this Shard and like trying to communicate across all of them is a little it's not super clean so yeah I'm very skeptical of that worldview yeah maybe go back to just a broader question of in your thesis you very opinionated you have been for a while what are the edge cases where can you be wrong um and what are perhaps the known unknowns as you think about the implementation of some like a full

version of a local fee Market or whatnot like I'm just kind of curious how you think where this Pie as well could break today generally the point of view has been modularity we need modularity for customization and for example if you cannot customize the technical stack enough in a salana s AP toos of the world I have to go create my own app chain or I have to go create my Layer Two because I want more control of that and I think it's where does more of the how you acrw like is it in that 80 to 90% that could fit on a integrated high throughput blockchain or is that value

captured in an L2 or or an app chain uh because you need that customization I think we're really in the early days of kind of the high throughput integrated blockchains and they will be more customizable over time so that application specific chains are less useful than they may have been in the past hey everyone wanted to give a big shout out to today's sponsor Wormhole Foundation stewards of the Wormhole protocol if you are like Santi and I and you play around on chain you know how bad the cross chain experience is today well Wormhole has set out to solve that powering cross chain transfers for over 200 different multi-chain teams

including some of the best like Unis Swap and circle so what does that mean for you the Empire listener this opens up a huge number of multi-chain use cases across defi nfts governance oracles and more by supporting over 30 different blockchains and six different run times including sui salana different eth l2s ethereum and a whole bunch more that means you have now the most powerful interoperability platform at your fingertips if you're a developer you'll be excited to hear that Wormhole provides an extensive Suite of tools and infrastructure so that you can securely build multi-chain appliations but don't just take our word for it obviously Wormhole protocol leads the industry in

alltime messages transferred with over 900 million cross-chain messages 900 million that is close to a billion and that's a big number of messages as a thank you Wormhole Foundation is dropping exclusive nfts that's right we got some exclusive nfts for Empire listeners hit the link in the description to claim your unique Wormhole nft today the days of not using crypto for really anything in the real world world are over it is time to start using crypto in everyday transactions whether that's shopping online or just buying a bagel on the street we're excited to tell you about py USD PayPal's entrance into web 3 PayPal is proud to share an open letter to the crypto community that outlines their

commitment their road map and their goals in the digital currency space PUSD seeks to transform how you interact with your digital assets available today you can send your crypto to your PayPal account swap it for py USD and then use it to check out at millions of stores PayPal invites you all the Empire listeners to be a part of this journey hit the link in the description of today's episode to read PayPal's open letter to the community it gives you a really good sense of what their vision is take the next step by signing up for a PayPal account today the future of crypto payments starts with PayPal Logan how do you make bets then at frictionless because you're

overwhelmingly excited about salana but as a fund you have to outperform right so I'm assuming you can't just buy a bunch of salana and you also probably can't just invest in the salana ecosystem so when something like suie or SE or manad or something comes across your guys's desk like how do you think about investing in that I mean at frictionless we most people will bucket us as a salon of fund I mean just because we are one of the few funds that were was publicly bullish salana and obviously we're extremely excited about the salana ecosystem but how I like to classify ourselves is as a thesis driven fun like our thesis is that you need high throughput and you need

paralyzation Sano was the first to kind of enact that uh but that thesis also expands to other high throughput blockchains because some of them are doing interesting things um that are different than say a salana and so S is doing interesting things with move um apos is doing this parallel processing optimistically which maybe Engineers are like hey this deterministic parallel processing is really sucks like I really like how APS is doing that um there are differences between them they are M more minute as I said but we're very open-minded to the fact that look there are a couple hundred thousand people

that use blockchains today it's not couple hundred million and we need to be open-minded from a kind of fun perspective to where users are going to go um and where Engineers like building I do think move is a very interesting programming language um that was specifically designed for blockchains that could potentially be the most dominant programming language uh in the future uh so maybe that's s maybe that's apos salana can also ultimately compile move uh down to l lvm so salana is really program agnostic um but there's different trade-offs

yeah um we've talked a lot about um engineering components and physics of these blockchains anything else that is before we transition more into the as Jana was kind of going to anything else that is worthy talking at the physics level at the kind of you know distinctions between these high throughput l1s that think is noteworthy I mean I appreciate we haven't talked about so many other things like cons you know how they reach consensus and a lot of other things but I'm curious if there would be anything else you'd call out I used to think consensus was more important now I think it's less of an importance I mean obviously you need to agree on what state transitions happened but I think generally they kind of are

all in the same vein um I I think people do not understand though generally the different levels of performance that these systems are compared to the low throughput blockchains even Celestia today is not very high throughput um it's still catching up to these high throughput blockchains like salana um and it has a long way to go to do that um so I think the narratives are just narratives which is fine like we're moving in the correct direction like blockchains do need to transition from dialup to broadband and after that they'll have to go from broadband to fiber optic but people do not understand the state of how far advanc these high throughput blockchains are compared to a celesia or

in ethereum of the world yeah on this point I just want to get your takes on a few things and the criticisms that we hear about salana are pretty dated and and the chain has come a long way in a short period of time you know I think uh the maybe the more general question is we hear it's not sufficiently decentralized it's been shut down you know like we see these memes SP SPF turning off the chain turning it back on um what would you say to those criticisms and others in terms of the salana specifically and then maybe get your take on some of the other ones High through boot chains so again I I kind of view this very much from an engineering lens in two buckets one is how many

replicas of the state exists in ethereum people often misquote the amount of people that are St 32 eth that is not really representative of how many copies of The Ledger exists last time I looked ethereum generally bounces between 4,000 full nodes and 6,000 uh salana today has about 3,000 full nod so anywhere from like say 75% to 50% of like full node decentralization I'm very curious though to see where the market ultimately ends up determining how many nodes is sufficient for decentralization uh s has 100 apos I think is around that

100ish as well say is around 39ish um so will the market say hey that's fine or will these need to um further decentralize because Celestia uses tendermint tendermint has an upper bound of how many full nodes you can apply to the network that's 200 so I think Celestia doesn't have very many full nodes either um so salana is much more decentralized than a Celestia it's much more decentralized than these other High throughput blockchains um and I I think that will probably continue I think what people ultimately will care about more is that realtime censorship resistance the Nakamoto coefficient so the number of independent parties that

are needed to collude to stop the network from achieving consensus and salana again here is relatively decentralized uh I think it's like 22 uh independent validators apos is 18 I thinku was somewhere around there as well at like 16 or uh 14 but that realtime censorship resistance just means like no single party can stop you from getting an me trade like that is very important for Traders and Defi and this is a fundamental limitation of the current uh Layer Two landscape because you have one sequencer and one party that

theoretically could censor you if they don't like you obviously you have that escape hatch uh generally but you're paying so much money and you kind of have this like um the lay time that you would actually go to the layer one and you probably miss your opportunity anyways so that real time censorship resistance piece I think is probably what people are going to care about longer term after you get sufficient centralization which TBD what the market decides there yeah the the criticism around uh just quickly on on the cost of running these validators is something that we keep hearing on on Salon like it has a higher Hardware requirement um of course there are three different cost buckets

as we think about like longterm okay you know you can certainly process more transactions you have storage costs you have uh you know Broadband cost and you have the hardware costs can you just quickly run us through kind of how you think those three components will evolve over time for something like salana and maybe just other chains I think once you transition from low throughput to high throughput the biggest cost ultimately will be the bandwidth cost so how much it actually costs to propagate these networks uh propagate the blocks to all the nodes in the network when you go from today like kilobytes ultimately to gigabytes you're sending a lot of data uh like

in I think sana's docks in their or may have even been in their white paper the once they're doing one gigabit per second uh consistently for a year it's four pedabytes of data so 4,000 terabytes uh that is a lot of data um so there are going to be like real world Hardware costs associated with that but I think the biggest cost will ultimately be the bandwidth cost in these networks what we'll probably see happen long term is that node operators move to locations that have the cheapest cost of bandwidth um because it will be a large part of trying to reduce their overall cost to maximize The Profit that they're making

from the network the difference is in these high throughput systems and what like you'll ultimately boil down the conversation to in these decentralizations debates is how much should the hardware costs um ethereum obviously there's incredibly smart people some of the brightest Minds in in the industry I think honestly where they went to stray was optimizing for low Hardware requirements so being able to run on a Raspberry Pi being able to run on low internet connections if you have to do these things if you have to eventually have hire compute uh it's fine to move that to a layer two but you're still like you're centralized by that single sequencer at that point um and so you're always in my opinion bottleneck by the weakest link and so I

I think being able to have kind of these various costs if you want to really increase if if you're only increasing cost by 10x but you're increasing performance by a THX to me that's like worthwhile tradeoff yeah definitely I think that's where I've certainly it took me a while to really appreciate that what makes salana I think Superior is you overcome the initial idea that it will cost you 3,000 bucks instead of 30 to Pony up initially to run but the economics will be far greater if you believe that there is orders of magnitude increase in

demand for for Block space and the economics for running a validator will be superior I think but this is where I want to get your opinion because you you do see the argument of one I think uh like we should salana Implement something like e559 mechanism or like the margin's pretty high I think today but like the current state of the fee markets and Sal is not perfect I think I've heard totally say this I think most people would agree with that I am curious like like fees are extremely low and you need to believe that demand like is is far far greater than it is today to be like probably as economically profitable

than in ethereum etherum is just a really expensive luxury chain that is not I hear you like necessarily viable long term but in its current state I'm just curious how you think about the economics of these chains and the sustainability of it and if at some point we're going to have to see an up an increase in fees in salana or just a a better pricing mechanism of these resources to make because if you don't have if the validator you're ultimately competing for mind share validators in a system and if they can if they can be more profitable you know going on another chain then that's bit of an issue right so I'm just curious I mean how I kind of think about ethereum today is it's like I think the Jeff Bezos saying is like

your margin is my opportunity and I when I think about ethereum they've optimized for like their margins and they have crazy margins on block space if you think blockchain block space is that product that blockchains are selling then ethereum is killing like they're selling crazy expensive block space uh but that being said they have not optimized for scale and so this is where the high throughput blockchains have ultimately taken the other side and said we will optimize for scale first and then if we hit scale then the economics will start to make sense because you're doing much more transactions but they're at a smaller amount so today I think Solana's uh inflation and other

blockchains or high throughput blockchains is around 7% six to 8% somewhere in that time frame for salana specifically the goal is to get inflation to down to 1.5% a year but to offset the validator cost really there's three main components one is just increased network activity I think we put put a Blog about this uh post FTX on friction list website um running these numbers salana had to increase like Network traffic like 11x or something to get to profitability that's just like purely off uh Network traffic so you can either increase Network traffic graic so more transactions on the network increased priority fees where you want

to pay more to access a state uh because you you want you want access to that state before other people and then the last one uh is me and these like hotpots uh and like redistribution of meev so I think everybody likes to criticize high throughput blockchains today because they are inflationary but they just really haven't had in my point of view the amount of time in the market that ethereum has and people like like to uh forget like I mean SLO only came out in 2022 uh apos andu launched last year so like these are relatively new ecosystems like monat's not even

live celesia just started so I think it will just take some time for these networks to mature and to actually hit scale but the key difference is they can actually hit scale um where unless the community really wants to change the hardware requirements for ethereum I think it's going to hard be much harder for them to actually hit scale yeah absolutely um maybe a last question before we transition into just some of the topics or applications that you're really excited like only possible in salana or some other ecosystem um like what do you think is the biggest challenge

or problem of salana today and I pick on salana because it's the farthest ahead and I think it's got a ton of attention and you followed it quite a bit um I mean Anna toally recently has been talking about so you can think of a blockchain is like scaling the rights like if you and what I mean by this is if I post to Instagram that's one right to a database but the reads like people that are viewing this content you could have a 100 people Swipe by that Instagram post and 100 people view it so you have one write but 100 reads and you can kind of think of the rpcs as the reads for these networks um just making sure that rpcs are good also scaling

it's kind of like once you go from a low throughput blockchain to a high throughput blockchain you also have to optimi optimize everything else in the stack to be high throughput uh the oracles uh uh obviously you have to scale the computation the rpcs like it's not just one thing it's the entire stack that has to go from kind of this old architecture to a new architecture and I think salana particularly is just kind of hitting that um their stride now um maybe downsides I don't know it obviously if the network goes offline that's horrible they've been doing a very good job about optimizing the network for uptime fire dancer will add additional redundancies the client side

um but it's it's hard for me to articulate and that's why like I'm very bullish but I try to talk with a lot of people to tell me where I'm wrong because you're always going to be some degree wrong you just try to be less over time this is why I've try to open source my thinking and just being more vocal of course it fall on for years but just want to make sure that I'm not missing anything and I from a all kinds of different players in the ecosystem Builders investors I think most people still very underweight this AA that you know it's still very much an me like sorry evm Centric World um so that probably changes so so maybe transitioning go ahead I think this is where like the podcast is uh extreme

Alpha I mean going back to your point Jason like when I first met you I was literally starting my podcast uh and trying to get people to talk to me like yourself in New York uh but that has really been one of the best sounding boards for me is just talking with extremely smart people and seeing where their views differ and how the industry is going to scale most VCS I've learned don't read white papers let alone talk to the engineers Logan are there any folks who you've had on the podcast who you've been remarkably impressed by like any Founders any investors um who kind of blown you away I mean I I think sham is very smart from aanl uh him and I were dming uh talking about

different things that they're working on Igan da I think I'm interested to see what sham pulls up to me he always can kind of pull a rabbit out of that um obviously Anatoli has been really steadfast in his vision I I really admire the founders and the investors that don't waver I mean you you should change your mind if you find better information that's like a part in my mind of like being either a good investor or good Builder like if someone's able to articulate things to you that are much level uh much deeper than you have really put thought to it and they can invalidate your thesis it's kind of immature in my mind to continue to hold that point of view Anatoli has obviously done an amazing job at like

being pretty steadfast and he's never really wavered in terms of like hey I want to synchronize State at the uh as fast as possible Sam blacksher uh the founder and CTO ofu uh very smart I did a podcast with him I believe the first uh podcast with him in Palo Alto when they're launching um Sam's level of depth on the engineering I was just very surprised by which I shouldn't be but uh very smart had great conversations with Avery um generally I would say those people um now I'm interested in that Engineers though actually building applications because to me as much as I love this stuff and like love the infrastructure

talks they're kind of like to an end like it's dumb to me that we're having this conversation in 2024 when I don't know how long the industry has been around and we still haven't hit scale and we're talking about infrastructure because it doesn't scale like I want applications that actually hit Mass adoption and that's what I'm excited about at frictionless yeah I mean let's go there what are the kind of buckets of applications that you're you're most excited about and I think as Santi was asking before like what are the applications that get uniquely unlocked by this whole last hour of infrastructure conversation that we've had I mean the cliche one is like order books um I mean amms were really bed from the limitations of having a blockchain that can only do 12

transactions per second I think it's been interesting how kind of say Unis swap has continued to evolve uh it's starting to look more and more like an order book um but it's still not an order book on chain I think there is obviously always always going to be place for amms in the world they help bootstrap liquidity kind of get that zero to one moment flywheel for tokens to get started but for real price Discovery for larger trades I think order books just make sense um and that is uniquely enabled by a high throughput blockchain when you're going from 12 transactions per second to 500 to a th000 transactions per second and even

more um and that I would say that like broadly is one very interesting um kind of Point that's only really possible on a high throughput blockchain per se um what else I'm very interested like higher frequency def applications you can kind of think of um me as like Arbitrage between time that has passed in the world and the longer the block times are and like the less you're updating your world view the larger the me events happen so like uh flash Bots has made [ __ ] ton of money

because the block times in ethereum are really slow um and there's a lot of Arbitrage because a lot of things happen within that block time and so when you are able to reduce block times to a second to 400 milliseconds to 200 milliseconds to 100 milliseconds less happens in the world and so you're more kind of trading on noise than anything else um and you generally kind of reduce me between each block there's a lot of kind of like small examples like that I think one thing that I'm more excited about though outside of defi is just these decentralized physical infrastructure networks they're very I like them because you can reason about

them um you can kind of extrapolate them where like a hive mapper or helium of the world you really don't need that many people kind of setting up the hot spots or Dy driving around the world where you can see this like actually compound in a pretty meaningful way I think being able to go into the world and physically do things is a little bit more challenging than kind of the what I call the digital Commodities of uh dend which is like bandwidth compute and storage um I like that just because it's more passive but I think both of those are uniquely enabled by high throughput blockchains and they're much harder to do uh on through

blockchains one of the things we haven't talked about is this coordination I think ultimately as I think about blockchains are just could be really interesting instruments to improve coordination uh through better incentives uh helium is a good example actually passed on the seed because I felt that this was had been tried before and without like crypto incentives and then here we are the risk I see there of course is that Telos come in and really make their business difficult um um how do you think about when adoms meet bits and some of these kind of applications um you know people talk about as another example real world assets bringing a lot of supply and inventory from The Real World into a digitally native context

and that has a lot of a whole host of issues that I don't think are impossible to overcome but I'm curious to get your thoughts on on on those applications because in some ways dpen makes that crosses that Chasm if you will I think the Adams part is harder to coordinate but that is what generally the token is trying to incentivize people to do and I I I'm curious with hi mapper with helium if they can reach a sufficient level of scale I think they're going to have a pretty insane mode because it's just hard to coordinate people to do things in the real world um it's if you set up a hot spot for helium in Miami versus

New York those are fundament Al different places and it's hard for generally most people to set up in hotpots in both locations that could be a moat because it's hard to replicate I think on the flip side the digital Commodities of dend the bandwidth the compute the storage networks those probably can scale much harder um at least in the short term because in general you have kind of one-click deployment on your computer saying if I have additional Hardware capacity here rented out uh I'll earn five $10 a day whatever it may be and I don't really think about it I didn't really have to do that much in the physical world I already have those resources at my disposal and I'm just letting other

people use them so I think from that standpoint on the short term those are likely to scale much harder I'm curious though how defensible those will be as they continue to scale because they are kind of easier to bootstrap are there a certain type of application or projects that you will not invest in oh there's a lot I mean I I I generally there is this really cool Mark andreon post from a long time on his blog post that kind of articulates from a venture perspective how you should think about kind of product team and market and from that blog post he uniquely talked about how you at least

from a venture side you should focus most exclusively on the market and how large the market is going to become instead of the product and team which I think if you talk with most VCS they'll say uh it's the team the team the team which obviously the founding team is very important they need to have the skill sets but a good entrepreneur building a good product in a small Market is still going to have a small outcome and so for us I think we uniquely focus a lot on how big markets are going to be become and focusing on areas within there instead of like only focus on the the product or only focusing on the team the market matters

quite substantially even a mediocre team mediocre product where the market is compounding at like 50% year year-over-year to be honest most of crypto uh you can have a lot of mistakes be forgiven because the Market's growing so quickly a lot of the dollars have been historically allocated to infrastructure the market just has a clear conviction that it could be a billion dollar like a I'll probably take any infrastructure bet with a good team um sub 50 million because I think I see a cleaner path to a 10 20x than um you know something like a front end I believe that will change largely a lot of the things we've discussed here because you finally have the possibility to scale these things as

you think about where you allocate your dollars in some sort of ratio you know infrastructure middleware front facing app Sofi deep in whatever you want to call them how do you think about that tradeoff and today if you had $100 million to deploy or $100 it's really the same calcul I think I I think the current high throughput blockchains are probably going to still be the winners in like five or 10 years time frame I have a very hard time seeing how they get displ um I think generally infrastructure bets just for like say doing a salana in Europe or Asia markets or like doing application like l1s to

our specific geographies probably not going to do well and I think we truly are at the precipice of transitioning from infrastructure investing to application investing and there's going to be a lot of investors that are caught off sides because they have not done their homework and understanding how big of a difference these high throughput chains are from the low throughput blockchains and where applications what type of applications are uniquely enabled on High throughput blockchains um I think a lot of people are going to be C off sides we're going to be focused predominantly on like the application stack but generally kind of agnostic if a awesome Market exists and great entrepreneurs are going after it yeah I mean I think

you guys are big investors in backpack uh which is uh one of the top projects in salana um maybe that would be a good example of walk is through your thesis there uh because I'm curious and trying to get out value acrel right because I see no reason why salana can't be as big of ethereum from a market cap perspective and then just grow from there spicy take it could surpass it I I don't think it is candidly I mean that is more of an assumption not necessarily that ethereum goes down much it say that salana is just meaningfully you know from a technical architectural standpoint like I just I just don't think there's anyways none of this financial advice um but I'm curious know I'm curious like

maybe backpack is an example like how do you think about the relative trade between investing in backpack you have a liquidity you're betting on a team there's more risk there like like specific to to them versus just investing in the L1 token of like just buying Soul that's a good question I mean particularly with backpack we were very impressed by their ability I think when we kind of went down the exchange Rabbit Hole what we ultimately found it's very hard to obtain licenses uh so just literally getting to the starting line and being able to operate in key jurisdictions around the world is very hard um it is not easy to get those licenses and when we are looking at just how large kind of

crypto adoption is today it's still relatively small um maybe a couple 100 million people hold crypto one of the estimates I saw was half a billion people hold crypto to me that seemed optimistic but there's really maybe few hundred million people 10 million people that have ever interacted with blockchains at least once and so our thesis there was really will blockchain adoption continue to grow and if so will you need kind of a global player to help one with kind of the more centralized trading but more importantly the Fiat on-ramps and off-ramps into these ecosystems and I think coinbase has obviously done a f fabulous job with

the US market holding down the forth there they've been a little bit slower on the international markets um binance obviously is a massive Gilla internationally but perhaps cut some Corners with the licensing to be able to um um cutting some corners and just in terms of how they approached different markets and now are getting their licenses revoked and some jurisdictions and so we really thought if backpack can become a large kind of player in the ecosystem get a lot of market share I think there is a huge opportunity for just the crypto ecosystem to grow and

having a kind of compliant exchange from day one that's playing by the rules uh facilitating the on and off ramps between kind of the crypto rails and traditional um banking providers I think is going to be extremely valuable yeah well look I think this been a fascinating discussion we could go hours and hours on decomposing more specific elements of of these blockchains I think we're probably should do an episode later on this year um I would love to just maybe do a quick set of questions um sure if you're okay with that ran I know if you have any other questions around just more stuff

and but I guess like uh tell me where I'm wrong too like where where do you think I know uh I mean I'm always open to it I I think I want to learn just as much as like I hold my opinion strongly like I could be wrong but I've done a lot of homework in this and I mean as I've said like I've been trying to disprove this thesis for the last two years uh since we fir met first met uh Yano like my thesis has not really changed because I haven't found new information to like discredit the point of view that I hold but I always look for it so I'm very open-minded I just want the industry to

actually hit meaningful levels of scale and I've been saddened by that hasn't been the case yet I don't think you're wrong Logan I think what's going to probably happen is uh some of the problems that ethereum encountered Salon is going to encounter and is is going to have to deal with things like uh governance things like maybe there's going to have to be its own version of one 559 one yeah 1559 on salana like things like that feel pretty inevitable and if salana uh ethereum was already so hardened as a community that the tech stack became very hardened and if the mindset of folks in salana also get very

hardened they're not going to be able to when someone like let's say suie builds something that's five times more efficient and five times easier to build on I'm not saying that about suie but I'm making that up but uh salana needs to be able to adapt to that so that that's the only that would be like the why you could be wrong I think if anything we've seen that Maxis kind of do not win long term and like and and that's why like our fund like I say we're a thesis driven fund like we we have had the kind of idea that parallel processing is going to be a requirement because you have localized fees and ultimately you'll have to continue to scale throughput um and other than that we're kind of open-minded to how we get

there I just want scale I I want the industry to be more impactful than what it is because I truly think this technology is the one of the coolest things that's happening in the world and I'm saddened by the fact that it's only in the hands of few people if we really get back to like banking the unbanked which was the original Vision that got me in like I would be super excited yeah I would agree there and both of what you guys said I mean ultimately I think it will become more apparent of The Have and Have Nots because as you finally start getting applications hopefully uh that are nonp speculative per se just more like you know referral networks or just like putting a lot of

these type of programs on chain and really battle testing uh not purely from a security standpoint but also from a scalability standpoint I think we will become more and more clear for everyone in the ecosystem all the stakeholders users and also Builders of where they what's a viable you know chain to build on and so I I think ultimately like we've been in the state for like L1 loyalty which is going to be largely irrelevant when you on board billions of people they will not care about you know your bias towards a particular L1 or your bags they'll go where they can do stuff faster better cheaper and it's already starting to

manifest itself in user activity and you look at Jupiter you look at tensor I mean it's just you're seeing that and I guess I've often thought about well you kind of saw that with binance smart chain like at one point everyone went there and traded in pancake Swap and it sort of like was for I'll be briefly more popular amm than Unis swap but then that kind of went away and I think it just like I think it is different like the salana and security the trust like you know that the caliber of the team and so in many ways uh you know like binance Marin was an implementation of tender Min um which you know but um I think yeah I generally think users don't

care about security or decentralization they care about and those things do matter it's just like how much they matter and then like there's a once you get sufficient decentralization hi from dragonfly actually had a great like blog post on this which I think it was like called sufficient decentralization um where once you get to a certain point you really get like diminishing uh like returns if you continue to decentralize and I I think to your point Santi new users they don't care about these different ecosystems they want the one that they can use and that's why it got me really excited when I saw just even the nft stuff going on

in salana kind of last bowl Market is because because those were the people that were price out of doing things on ethereum like they wanted to do probably nfts on ethereum they just couldn't because gas was so expensive and I think the more that we can get people just trying things on chain and participating that's what gets me excited um I think the people that have allegiance to ethereum or the people that made ethereum the people that have allegiance to ethereum they often got rich by ethereum and they continue to Champion that the new users have not made money off ethereum and are not as much um as bias yeah absolutely maybe a parting question just one to which is uh what is in your journey in crypto what is

something that has surprised you the most [Music] um the cliche thing is all the frauds I think I mean but maybe more a personal one I think one of the sayings is like you can make friends or you can make money like do you want to be right or do you want to make money um I think I've had to learn especially because I've not really changed my point of view um just on high throughput blockchains really since like I started my blockchain journey it was very hard for me to be comfortable with kind of taking the opposite side of what majority of people think um it's not

exactly fun like if you go to a party and you're like hey I'm not ethereum for XYZ and reasons and here are all the reasons why I think you're wrong you're not very popular at parties uh and I think for me what I've really come to appreciate is just like um it's okay to not be the most popular person at the end of the day my job is to make LP's money and to be okay kind of in your skin and voicing kind of different points of view in the market it's not always fun but I think that advances the space and moves us forward and kind of breaks us out of that that Echo chamber yeah that's well said uh I mean we are crypto is still very much

non-consensus so we're here talking about like different factions within a very much non-consensus trade of of our generation of our lifetime but no I hear you uh it could be very uh it could be very uh difficult at times when you get the criticism um but ultimately you're in the pursuit of Truth and being right true investing is um to me it's finding the truth and then manifesting the Investments through truth you have to find it before other people and to me it's like the ultimate puzzle which I like from my over over engineering brain nice uh Logan uh a true pleasure uh I'd be uh obviously I'm excited about you guys are building uh potentially we'll invest in the font so you know

there is bias here in these questions uh but I I I do remember the first conversations that I had with you we spoke one day and then very quickly the following day I spoke with you because I found you to be really deep in the tech have a very very fine appreciation to the point that you're able to talk to Founders like Anatoli and the founders of all these other blockchains I very quickly reach my limitation um within a matter of seconds you are able to carry a conversation with them and it's very apparent people should go check out your podcast because I think that is a public good that has helped me personally understand um a lot of these projects that perhaps I was an investor in or a lot of them I I am an investor in but I haven't had that level of appreciation

uh and and I think uh so definitely would encourage people to go listen to that uh if you want to get deep in the weeds of the technical stuff because ultimately I do think that the best tech will win even though we constantly go back to these VHS betamax like marketing um is important but Tech here is is like things when they break in crypto it's not fun and you can't really fake it maybe you've gone away with it over the last 12 years or six years because we haven't had real consistent stress tests but now what's really exciting is we're probably going to see these multiple type of applications that it will become very apparent which chains have that Tech and which actually don't so anyways um thank you for coming on and sharing

your your your thought process your views and yeah really exciting where you guys go thank you and thank you both to you Jason and Santiago uh I've I mean the reason why I started my podcast uh was to get my foot in the door and I was always the nerdy person that did not get invited to any of the parties and I was the one that was just watching your guys podcasts and your Twitter's uh Twitter accounts and so for me to be here today I'm just very appreciative uh you guys are a great force for the industry and I appreciate each of you uh contributing to push the space forward appreciate you Logan congrats on everything man thank

you hey everyone Thanks for tuning in to today's episode don't forget to claim your free Wormhole nft exclusive to Empire listeners hit the link in the description of today's episode and fill out the form to claim your unique Wormhole nft today hey everyone thank you so much for watching today's episode really hope you enjoyed it we wanted to take a second to just remind you about our upcoming digital asset Summit in London March 18th to 20th Santi and I got your seats are limited if you heard it earlier in the podcast there's a little competition running at blockworks to see who can drive the most number of tickets so when you register for the digital asset Summit make sure you use our code see you in [Music]

London