Rollups, Staking, and Infrastructure w Amrit Kumar from Altlayer

Captions as YouTube publishes them, usually machine-generated, so product and people names can be misheard.

1. 0:00

   and then again what you don't want is uh that you go to a and you have to select 15 you have this drop down list of 15 chains or 20 chains out there and you have to select one of those chains oh did I choose the right chain did I choose the right RPC at other so I think a chain abstraction will become even more important especially in this world of 10,000 rups nothing we say here is investment advice legal advice life

2. 0:23

   advice or representative of the views of the organization we are part of or affiliated with see chain abstract.com for more information hey everybody I'm Alon uh this is chain abstracted uh obviously in your podcast uh we have om today uh nice to have you om thank you very much for inviting me great to see you again yeah likewise so I'd love to hear a bit more about your background I know you have been in the

3. 0:46

   space for a while uh and love to love to hear sort of how you guys came together to do um to do allir yeah I've grown grown out say quite old in the last six seven years I suppose um so my journey started actually I was doing my PhD in computer science in Paris not Paris but in France and this was one of my last year of my PhD and this was around the time when um researchers were taking Bitcoin seriously up until that point it was more like hey it's a developer thing you know people are investing in this and

4. 1:19

   that's it there was nothing of academic interest right but that was changing around that time um you a lot of people were publishing papers around Bitcoin I think one of the first on was uh published by um um some of the Israeli researchers as well so so there was this kind of appetite in the academic Market that hey this is an interesting topic of research and we should look into it and so um unfortunately I was uh my PhD was

5. 1:45

   funded through a certain kind of project and so I didn't have much Liberty to actually do something else but I knew that once I'm done with this my PhD I will go and do something in the blockchain space so um at that time by the way this was around 2017 2016 2017 so uh this was the time when iOS were happening left right and center and Singapore was kind of the center either either it was Singapore or Switzerland right there were the two places where

6. 2:08

   most of the people were setting up the entities and there was a lot of hype around these two countries right so I said maybe I should go to Singapore and see something uh to work on on crypto um unfortunately even though Ico was at it speak I decided being an academic I decided to join uh n us which is the National University of Singapore to work as an academic researcher and when I joined there my advisor said look you're too late to the party it's

7. 2:34

   there's nothing left anymore U at least nothing interesting left anymore to do maybe you should do machine learning and so he kind of forced me to do machine learning for a bit to be honest with you it didn't click with me um one because I came from a security background and I felt like blockchain had the right elements into it it has it had applied cryptography it had security and network components which is not really what machine learning had um so but I did I

8. 2:59

   did drill for I think about a month or two on machine learning and my adviser was like he was observing me and I think he realized that the guys were passionate this this in this work um maybe I should have you know stayed in that domain know now that Ai and everything is eating up um would not have been a bad choice anyway so I I I my adviser kind of gave up at some point and said look um me and my some of his

9. 3:23

   phsd students had published a paper called sharding around sharding this was I'm sure probably you know them so this guy I'm ly Lou he was um the founder of kyber network um so he had published a paper he was a PhD student of my advisor so he had published a paper on Shing and then my adviser said I'm thinking of kind of commercializing this paper in some way um and so he said would you like to join the team and said yeah sure

10. 3:48

    so this is what ended up becoming zil eventually and this is where I also met my partner yaoi so yaoi myself and there were other folks who became the co-founders of Zila there was another guy called Shinu who now works uh with Babylon you may have met him in one of those circuits I suppose um so yeah the these kind of the same gang uh we were all associated with the same lab we had either done PhD or had done we're doing

11. 4:12

    PhD at the same in the same lab in same University so we started this company Z um um Shinu and yaoi left uh around I think after the launch of meet so yaoi went on to join par um and I stayed for a little bit longer and I left in 2021 one again to join ads with yai to build what today is called vlay so that's kind of a quick journey of what what I've done in the past and how did so I guess like you guys how did altl the idea of outl came

12. 4:44

    along yes so um I guess Yi already had this idea of what he wanted to build um so I got attracted with alll mostly because yeah we had this idea of um what we initially call flash layer or emal rollups and the idea was that hey if you like not everyone needs a blockchain for like ever right so for example you may there's certain use case and scenarios where you need a block infrastructure but only for a short period of time so imagine for example you're playing games right uh games have a very short

13. 5:16

    lifespan so you play for maybe a month two months and after that kind of the excitement around that game kind of dies away and so imagine if you're building a chain let's say a cosmo style chain for your for this game after maybe 3 months this chain will die right because there's no demand for that chain so the idea was to kind of have something that is disposable in nature where you would run up you would spin up infrastructure

14. 5:40

    you would host your chain run that chain on the infrastructure when you feel like you're done with it the game is not inand anymore you can basically shut that down and we realize that this is exactly how rollups work by the way right if you think about it right if optimistic rollups or even Z rollups you know in a similar way so you kind of run things off chain and then then every now and then you commit State on chain right on ethereal and you also post

15. 6:04

    transaction data and basically if you wanted you could at that time you could cut down the system but you could say you know what I have committed the state of the system and that's it so um in a way this disposable or female rollup is basically a shorter version of rollups in sensus right um so this this kind of one idea that kind of attracted me to to alll the second thing was that he felt like hey optimism and arbitrum they're great in terms of kind

16. 6:28

    of build building a scalable chain but at that time this was uh I would say 2021 and at that time optimism arbitrum were doing about 50k transactions a day so kind a very early stage for our B optimism but it felt like if you have optimism like this you could basically Fork it to have your own chain and then we have this idea of app specific rollups instead of app specific chains

17. 6:52

    like what cost more on poar hat and we we were quite bullish on that um and so what we ended up doing was we end up ended up building our own stack uh we didn't have a name for it but basically it was a stack that allowed you to have either persistent rollups so rollups like optimism orbitum that would stay there forever and then you also had a stack that would allow you to spin up these emerals or disposable

18. 7:15

    rollups um unfortunately we had to kill um the persistent rollup idea very soon afterwards when optimism released op stack because we feel like very difficult to compete with them um but it kind of solidified our notion or at least our thesis that hey app rollups are going to be here and op stack launching op stack was kind of clear uh message that hey we in the right direction so yeah out say the main

19. 7:38

    thesis for all play was appops and appops would be a better alternate to app chains that was kind of the main thesis okay so I guess coming from Paul K do it's you talked a little bit about right like that that um and I know that you guys do work in in in the iary ecosystem like we do um how how did that come along I guess the the sort of the rollups are different obviously you know it's been driven by op and arbitr but there's this like a shared security idea I know polygon did it like cosmos had

20. 8:11

    some sort of experimental ideas around it um how did you guys come come come up with that yeah I would say it was it was mostly a result of a demand that was coming from outline so um we had this idea of rollups and so when op stack Arbiter morbit polygon cdk got released we kind of ventured into this Ras side right so um people would come to us because the thing was op stack even though it was released it was not super mature right so for example in the first

21. 8:37

    version you could not uh put your own G custom gas token so if you had your own token you could not use that as a gas token for that rollup um you couldn't use let's say alternative DS that came much later on right so a lot of clients were coming to us and they said hey we want to use op stack we want to use arbitra mbit uh but we would like to make these customizations they didn't have the ability internal ability to do so and because we had built buil a

22. 8:59

    similar stack in the past we had all the experience to be able to tweak the stack to be able to fit whatever customers that the clients were looking for so we kind of kind of moved away from this hey let's go build a stack to let's support existing stacks and this ended up becoming Ras so we ended up becoming the RAS provider where we would CL come to us we would run and manage rollups for those

23. 9:22

    clients and so we were we were working with I think about 10 15 clients at the time and this was I would say end of 2022 and a lot of clients came to us and they said you know what this is good because we don't have to worry about managing any nodes it's it's all simple for us now but where do I get my token used right um that was a common feedback and the second feedback was they feel like maybe this is not the model that will stay there forever the RAS model where an external vendor will

24. 9:51

    basically lock you in for for a while right basically every Vendor by the way has a similar model where a client comes to them and the RAS vendor will ask them to sign an agreement and you're normally logged in for one or two years depending on what kind of agreement you have with them and some of these clients felt like this is not a good model because what if in the future um they would they could hire for example they wanted to reduce the cost so they could hire devops themselves and they could for example

25. 10:16

    bring all of this in house and they wanted this assurance that they had this ability at least to move away from a Ras vendor to either in-house or to be Community run in some way and so there was this common feedback that hey how can we de a system where we don't have huge dependency on a Ras vendor and so um this is where because of these two feedbacks we feel like they has to be some mechanism where one the rollup has some sort of decentralized

26. 10:41

    components where one this allows community members to get involved but also then token gets involved um and so initially felt like maybe what we could do is we could have some of Rec sequencing framework where people could launch let's say disly sequen Network and they could have let's say their token as a gap token right so this is something that they could have done and so um we we I would say the we were fiddling with

27. 11:06

    this idea of staking for to build a decent Network even before Ian Le kind of launched anything concrete in that s but then when IG got got launched we feel like okay now that IG is there we could basically use ig's reaking uh mechanism to boot stab those networks quite easily and yet you'd still be able to use the token that I guess the r the DAP has for or the drup has of you know

28. 11:32

    to be able to St and secure that Network so I would say the motivation came from the client demand and then you feel like I la was a tool that was a available for us to build that in so we kind of married them together and that's kind of what became uh restate rups so the idea that we bring reaing to rups in different shape and forms interesting so the idea of I guess shared sequen or Le any any offchain

29. 11:56

    work I mean this is quite offchain work you you grab security you bond that crypto conv to do something and then those ideas came along while just uh while um I there was pretty much launching um I have I have a question so like how do you maybe let's Deep dive into the restake rollups how would you sort of differentiate the characteristics of like a normal op stack rollup that's you know let's say um op main chain with one of the restake

30. 12:28

    rollups that you guys have on Main net today so what are the differences and um how does it matter to the developer and the user also that would so in in terms of tech difference there's no Tech difference so basically what I mean by this is we don't modify the op stack code base so you could have a restate rollup um or or for example if you take a restate rollup and if you compare with

31. 12:50

    an existing uh let's say rollup that's that does not have reaking at all the only difference is you have something on the side that is attached to this rer but it is not required in that sense so we are not modifying or in any in any way whatsoever the Ling Stacks or the roller puts exist on its own so for example tomorrow if we shut down the reaking component the roll would continue to work the way it was working previously so there's no Tech there's

32. 13:16

    Tech change but there's no Tech changes to the stack itself okay um and the reason why the way it will it works is because we take op stack and we take blocks that gets generated from let sequence or the valer that comes out from that and we pass it down to this D supplies Network and this Des network will then verify that the block looks okay if you were to reexecute those block you would get a state that would match the state that sror claims it to be and so it's kind of like an external network of to simplify this lot of let's

33. 13:47

    say full noes that are verifying the the the state transition function for the rup so there's we are not making any changes so so on Tech aspect is exactly the same thing the only difference is that have now an additional Network that is giving you extra security that you could benefit from in terms of benefit for let's say faster finality potentially for uh bridging potentially for let's say withdrawal if you want it

34. 14:11

    but it does not change the underlying stack okay Michael you have something to to jump into or well I didn't want to take away from the the strand of where we were going with the restake rollup but one thing I I actually think uh people would find interesting is that you're probably one of the few people that's done serious research on sharding and then also worked on the you know rollup stack and usually I guess if if you would ask someone how they would think about rollups in relation to shards they would

35. 14:44

    refer to one of ital's blog posts where he says that we're rolling up until ethereum is able to Shard almost to make it seem like rollups are a temporary solution and then the long-term game is to Shard I would love to hear your thoughts on just how a rollup compared to shards and you know you seem to be quite well versed in both I would say they're very similar in my opinion um especially on the execution component so when you know I guess ethereum or

36. 15:10

    vitalic and the team decided to kind of move the execution out of ethereum you basically created shards in some way but the only difference was that it was execution shot so basically a system that were executing transactions and committing State back to back to back to the main chain L1 uh the other difference I suppose was um and by the this is something that I think Kevin from was it Kevin I think someone from op stack I think did a very

37. 15:34

    good blog post or I think presentation on U how to understand rollups right and basically his explanation which is very simple where you say look ethereum is like an external system e doesn't care what happens in the ruup right it cares only when you want to bridge things out right only when you move if if you want to prove to ethereum that hey something has happened on the L1 on the L2 then

38. 15:56

    only you need something to prove to ethereum otherwise is okay so ethm sends you a bunch of transactions the input transactions and anyone can basically those transactions and claim that here's the output of those transactions this is kind of what in a way this is kind of what A Shard looks like in the end the only difference is that when you want to withdraw from the rollup to the base chain you

39. 16:19

    have to have some sort of mechanism to prove to e that hey this is the valid State and therefore this interaction is allowed to happen because in the end remember your assets first get logged on E right so when when you're bridging in your assets are logged on ethereum and so you are asking eum to let the assets go and so eum has to be convinced that the the the right things have been done on the L2 so that you know this could happen so in a way for my in my in my opinion they are very very similar um

40. 16:45

    and you could see that happening as well like for example in super chain World actually something that I guess optimism trying to build where the idea is you'll have something Central in the middle and then you would have all these chains coming out of them and this Central thing would allow all these chain to interrupt or talk to each other right which is also kind of very similar to I guess what Cosmos has uh you know with all the IBC and everything so I would say they are very very

41. 17:07

    close um and so in my opinion execution sh execution sharing is what rollups are in my opinion yeah we we have this I mean I I worked at near for two years like in the the one of the core teams that was one of the first thing that we're like ah this looks like one Shard that needs to communicate with other um other shards uh obviously it's it's asynchronous World model which I mean everything is right uh what do you think about that sort of discussion I've I've heard that

42. 17:38

    quite often in ECC this year where people are like oh you should be all synchronous when you're talking with different rollups or we should be all in saying there's a huge debate between um I think one of the guys at polygon and with Ben Fish was was espresso so and you have any thoughts on that one yeah look Asing is already something that we can do already today right it's not a it's not not a problem you know through bridging and other things you can also

43. 18:01

    already do async composability kind of it's already there synchronous I think it's is is ideal but I'm not sure if something that people actually need in real world right um and it's also complicated to build and people of course building Atomic you know inclusion Atomic execution across across domains I do agree that this there's a in order to in order for that to happen there you have to have some Central entity or some Central set of notes that see everything in some way and that's

44. 18:31

    kind of where you need something something of shared sequence something something of that kind without that I feel it's very hard to achieve U you know synchronous uh composability of some sort and if you could see that with Justin Drake example as well and he he has been pushing it through the base kind of configuration right where you say okay you have base sequences and they see everything and if you have a shared base sequencer then I can I think you can allow cross to interrup ability

45. 18:56

    to happen in some sort of synchronous way um but I would say at least for now e synchronous compos is already there you can actually use it through all sorts of bridges uh synchron is something that would need some sort of a shared model some sort of shared node that can that has visibility across different domains and you you mentioned a base rollups and this is a something like

46. 19:21

    what are base rollups for folks that have been learned about before and like is there going to be a demand for them from the RO rollup side not from etherium side because what you're saying is Justin Drake coming in and say this is actually the best world that I want to see because that's that's usually what uh researchers allowed to do but is is there like a is it practical from from the roll out perspective so a needed B so just to just to give um

47. 19:45

    everyone some some context so B rollup is simply a rollup that uses ethereum valuators as the sequen okay so instead of having a an outside sequence of for example in case of optimism optimism runs or optimism Foundation runs a sequence of that takes transactions from users and orders them and then you know gets exer right um instead of having this external party what you want is the validator on ethereum to become one of a SQL and they could do that because in

48. 20:09

    the end all your transaction have to go to eum anyway and that happens through Val so all you're saying is the validators are sitting on ethereum they of course can go and process transactions on ethereum the way e transac get handled but also they have the right to now order transactions coming from the L2 as well that's kind of what we sequencing you know in a simple explanation is right now one of the Bas main challenges is um I guess

49. 20:34

    one of the benefits of this Bas seening is the livess property that e gives you right e is extremely like have you have a lot of redundency you have a lot of De even if one of the vars are down someone else can pick up your your block and kind order transactions this is something that we have seen very difficult to do in a centralized world so for example the arbitr was down a couple of times in the past the sequence was down a couple of times in the past I think optimism probably too but I'm not

50. 20:58

    sure uh so we have seen that you know even bigger chains some of the most established chains out there in the L2 World they have had sequencer is issues in the past and so you want a more decentralized network uh to provide you that liess guotes and that I think that's one of the benefits of of um of eum and then again I guess Justin wants to do ethereum alignment so that also gives you some alignment with it you know ethereum as well um I think the

51. 21:23

    mean Downs is of course latency component right because um if you look at a very simple transaction today that goes inum um it takes 12 seconds to get not even finalized but to get into a block and then you have to wait for a certain number of epochs for that block to be to be uh finalized and that can take up to 12 to 13 minutes right in practice and so uh lency is of course a major component and that's kind of what people that's kind of why people are

52. 21:48

    proposing this base pre-confirmation now so the idea is of course you can't wait for the full uh you know guarantees from ethereum so but what you can have is the sequences of is they can use something like IG again uh or some some sort of a bond mechanism where they would say look here's my bond and I can guarantee you that I can take your transactions and at some point in the future when my turn comes in I will include your

53. 22:11

    transactions in the block um and um again all of these things are still being built but the claim is that you could have you know subsec latency um for pre-confirmation not for the finality but the pre-confirmation that have you cany uh I it's hard to say how and by the way all sorts of complexities here for example when uh you have to one you have to pay something extra to this valid to give you that pre-confirmation

54. 22:36

    and then you have to make sure that hey if I'm paying you something then I need a guarantee that my transaction will indeed get confirmed or will get a pre-confirmation right eventually so all of these things are still being built are there caveats around it but I believe that latency at least on the preon side will can be handled so I think it is doable it is manageable to bring it down to around 1 second or below one second in terms of getting this Readiness so I so far I know one

55. 23:02

    team that is building this uh called Spire Labs so um again these are ethereal Foundation researchers I think and um and by the way the guy who's building this is is also was the guy who propose this idea of um what is called pre-confirmation Gateway the notion is that now you could have many rollups coming in and these rollups need based PE confirmations and you can't do that with the traditional RPC so you need some kind of a fancy RPC that takes your transaction takes a tip and takes then

56. 23:32

    pre-confirmation from from varas and all that complexity so that's why you need some sort of fancy pre-con confirmation RPC but I would say it's complex it's being built it will take at least six months for us to see uh what sort of real latency you see in the market um to add to that shared sequences are slow compared to anything like centralized sequencer today right so if you look at

57. 23:56

    uh any of the Shar sequences today if if you just use share sequencer and if you want to wait for the final finality coming from then most likely your load your rollup transaction will have three to four kind of at least latency which is not ideal right we have to see whether you could bring it down with some sort of preon confirmation but is yet to be seen I I would say I'm optimistic simply because I like the fancy idea of B Bas spe confirmation it

58. 24:20

    looks it looks nice uh it's complicated sty but we have to see how how the latest will look like and precom is a is a Marketplace so that would be a pretty much you can you can outrun people right so that's that's exactly yeah exactly yeah you you can think about mechanisms where you can actually yeah it's the same same thing as like M because you can one roll up you can do something you

59. 24:44

    can you can put the put the G you know the tip put it up and then you do something else before everybody does something else I mean yeah it comes the same comes the same problem of of um I guess B will probably use some of the tools that was developed by flashboard guides and and some yeah yeah very very or like a hedge fund or like I'm just waiting for the transactions to come in I could they come in I'm just going to do something at the same time by the way I know this is going to get

60. 25:16

    prec conf and take that one um yeah I mean it's it's the it's I think getting that out is going to be very difficult anyway yeah yeah yeah yeah I agree cool but um this is the this is helpful um so what are you guys shipping at outl today what is the what is the focus for you guys like um in in terms of I know you guys are a a bit you you guys are good at shipping things but what are the how does the how

61. 25:41

    does the next sort of two to maybe maybe a year because crypto it's two years is a bit of a hard question but how is the how does the next year look like for you guys so of course we are we heavily into the AG ecosystem we are heavily into the r side so far we are we support about 30 odd clients across all major stacks and and there's a new stack coming up every single day by the way um I was talking to someone the other day and they're building um a parallel VM stack so par

62. 26:06

    VM stack uh there's a team building Sol VM stack so if you wanted to have an svm chain um that's coming up very soon SPM stack is coming coming very soon so uh and then now that bar is coming up a lot of people will want to build something on top of barain as well so I think there's a lot of activity still happening in the broader rollup or execution space as well um move um is coming up so we would be busy on the raft side for sure to to on board some of the guys who want to have access to those new execution engines or

63. 26:38

    or new new run timeses um on the second side we of course are having on the IG layer side as well so the reaking more generally speaking um we so far have shipped Mark which is the fast finality AVS uh we have five or live there we plan to extend that to some of the other avas that we wanted to build more one is called vital which is basically a decentralized challenger Network we were not building it quite aggressively

64. 27:03

    simply because um none of the stacks out there had that ability natively so for example op stack did not have fraud proofs and they I think they recently shipped fraud proofs and they got have revered it back because of the bug so we can build that decentralized challenge netwk because that assumes that there's some sort of working fraud proof system right uh we couldn't do that for arbitrum orbit either because arbitrum had some sort of

65. 27:27

    a know it had Rod proofs but it had this white listed kind of framework right now they're building it right they're building something called bold which allows for permissionless validation of orbit chains on Orbit on chains so the idea would be you could extend um you know ABS you can extend I lay to these bold or the Challenger Network where if you want to become a validator for

66. 27:50

    arbitrum or for op you have to come in stake and then or bring let's say you know St or any of the restake assets and then if you can prove or challenge if your challenge successful then you know you get some rewards and other PE person on the other hand will get slashed so so we definitely want to expand the the AVS component uh two different rollup Stacks by building some of these other Eves that we had initially planned but we couldn't build it because it was not there not possible to build it so that's on the IG side uh

67. 28:22

    then we also also want to expand um our reaking Beyond ethereum as well and that's kind of what's happening with Babylon um so Babylon is about to get launched I suppose in a few uh weeks time I suppose so that's something that we are we are getting ready for um and the idea is again uh once babyon comes in you can come and stake your Bitcoin natively on bitcoin and then be able to secure a chain like Babylon and then the idea is can you

68. 28:47

    expand that to other chains in parallel so through again reaking so um you you secure Babylon but you could also potentially know secure an ABS that supports let's a Bitcoin rollup for example so it could be a mark which is a fast finality layer for that rollup or it could be a decent challenging n that rollup so that's something that we are doing we also are aggressively working

69. 29:09

    on making it easier for people to deploy ABS this something that is still not pretty yet but the idea is to kind of develop tools for abs builders in particular um hopefully in a few weeks time we' be able to announce something more properly on that so I would say yeah keeping ourselves busy on the raft side getting support for all the newer Stacks coming up um kind of extending our ABS uh exposure

70. 29:32

    to different kind of things that we have initially planed build and then expand into the Bitcoin ecosystem okay cool I I just want to add in there um maybe to distill a little bit of this up to the surface of what your general VC would be will be kind of looking at or asking questions on broadly relating to as the podcast is called chain abstraction um in in kind of what you just laid out for us what does chain

71. 30:03

    abstraction mean to alt layer and I guess how in this next year as you build to this do you expect the crypto space to be changing with um the these Stacks coming live and with these new infrastructure opportunities uh uh you know uh becoming functional well J St was never more important than that today right because previously like if you go back I don't know a year ago two years

72. 30:26

    ago before op St before the stacks you had maybe five l1s that's that right or maybe 10 l1s 10 major l1s now you have 10 major l1s and and then you have 10,000 rollups or thousand rollups um not everyone will not each one of them will survive but at least you will have 10 big ones coming up so you basically have expanded the scope of all these domains to much much larger and then again what you don't want

73. 30:50

    is uh that you go to a and you have to select 15 you have this drop- down list of 15 chains or 20 chains out there and then to select one of those chains oh did I choose the right chain did I choose the right RPC and things like that so I think chain abstraction will become even more important especially in this world of 10,000 rollups um sadly I would say defa hasn't quite picked up in

74. 31:13

    this in this cycle so much and so a lot of the chains are being built today they are kind of self sustained like selfcontained in some in the sense that uh people are building games that do not necessarily need to compose with other chains that much people people are building let a social fire applications that again do not need to compose with other chain but I think once Dy really kicks in um this These Chains will need a lot more of this chain abstraction

75. 31:37

    tool that you know that are being built today or um you know that will have to be developed especially in the context of you know these thousands of 10,000s of rups and again across many run times right so it's not just about evm anymore previously it was just evm now you have evm mov VM Solana VM and waram and other other run times are coming up so I think change of structure has even more

76. 31:58

    important than than it was before yeah I mean for me train abstraction is also reaking and I mean because you don't have a chain right so you just use verific I mean you just use restake assets to do something offchain um what are the challenges you see in the market today I mean uh We've chatted this quite you know like what are the what are the things that should happen for restating to take off uh not only within ethereum ecosystem but everywhere

77. 32:31

    else okay I think one of the challenges that we're seeing at least from an Evas Builder and I'm I'm not sure if you agree with me about this but my feeling is one major challenge is there's a lot of Supply there's less demand for it what I mean by this is demand not in the sense that like they're not you know Builders but I mean demand in terms of assets right say so for example you have I don't know how many billions of eum assets available for eaking but as an AVS Builder how much do

78. 32:57

    you actually need in Practical terms um and so there's there's there's that that I guess demand and supply problem in s right um we still in the early stages of evss so we haven't really seen a major EVS come out and say look I'm we have the ability or we have the need to actually absorb I don't know billion of assets for example that that that's something that is still have to be seen and and develop maybe other model would be uh you know in a in a restricted setting you could say I only need this much I don't need more

79. 33:28

    than this and then there was a Mech if there were a mechanism where you could of limit that uh the demand for you then it' be helpful because in the end at some point when you borrow because in the end with through restake you're borrowing security from some other place right and borrowing comes out a cost you have to pay some interest rate or some apy to to whoever is lending you this some people could argue hey you know what uh eum is already giving them n wield of I don't know 3% or whatever and

80. 33:53

    so even if you offer them 0.1% they would be happy my thinking is that's not how things work in Practical yeah in principle yes you know an Ideal World things will work like that but in Practical terms people expect something something more than that right and so um I would say there has to be some mechanism where you could specify how much reaking collateral act orate security actually need for eavs

81. 34:16

    and then you can rrap it off wrap it up over time so you could say look I'm just bootstrapping a network I need I don't know 10 million or 10 100 million of eon Economic Security and then over time my CH has become much more powerful much more strong much more economically viable and so I need much more there was there should be a way in which you could kind of modify that the

82. 34:33

    second thing again going back to this female rule of idea where you could say not everything you rent like in in the real world when you rent you don't rent it forever like when you're renting an apartment you rent for a short period of time so is there a model where you could restate you could borrow security but in know on a for a term basis you could say I'm going borrow this security for a certain while and then after that I don't need this anymore this also potentially would need some kind of new kind of applications to come up new kind

83. 34:59

    of ABS to come up that would borrow Security on demand U so that's I I would say one thing the second thing is I think we have to build again interesting in the end ABS is could have been at once in some way right could have been their own chains at some point um so and some of the that we're seeing of course some of the new things have come up for example um let's say faster fin didn't exist

84. 35:22

    before right um Co processor didn't exist before but any way to me personally seems like many of them are quite similar in some in some there's nothing like makes you think this going to change the needle dramatically so I think I think we need to come up with something that changes things drity something that enables new kind of applications which I like in theory yes it looks like they might but that hasn't quite happened and we have to kind of

85. 35:48

    reflect why why we have not been able to create new kind of applications with all this infrastructure that's something I think I keep reflecting every day when hopefully someone will come up the ideas of this uh what's your take on I guess I talks talks about verifiable cloud like is this something you see you know web 2 to web 3 has always been tried I think it's always been it's been a not a it's not been a tech game but something else right it's been more incentive game but like is there something you see where like oh

86. 36:20

    you know if this was an ABS in the real world that would have made sense or where it's just still early there that you haven't seen anything yet I say the reason why EVS makes sense is because as I said very fi is the ability to open it to other people I think that's what EVs or any other dis networks allows you to do so for example if you compare like back in the day in 2016 2017 people were building hey we going to replace AWS and that that's that's not going to happen

87. 36:43

    if you're looking purely from a cost perspective right most people in the real world care about cost right so can you be cheaper than AWS some people say yeah we are cheaper than AWS but most of the time it's because of token incentives like you know many of the D5 Pro are not actually making money they they if you look at their revenue that they generate and if you remove uh the

88. 37:06

    expense uh that you know that they the money that they spending in terms of token incentives there's very little profit coming out of this so I would say maybe it's not the lowering the cost uh loading the cost is not the right angle if I think that hey I'm going to beat AWS by providing a better cheaper thing that's not going to happen if it has to be something that like like Bitcoin right the idea is that it's not necessarily cheaper that for acting yeah I mean it's cheaper in the sense that

89. 37:35

    like if you're trans transferring less a billion dollars um you know internet is going to be much more costlier if you do it with traditional rails while with Bitcoin is going to be much cheaper but you know in you can't simply transfer let's say $100 worth of assets through Bitcoin Network today right it's it's way more expensive and so there is certain kind of things that make sense with crypto or blockchain that does not make sense like

90. 37:57

    would makes sense elsewhere I think this is where crypto shines I would say for verifiability it makes sense but it's not in context I think if you think like hey I'm going to run my Docker image on on AWS and now you could run this on AWS on on abs and here's the benefit I don't think that CS it in my opinion there has to be something more more tangible than that kind of to to dig in a little bit on the the token side of things you'd said before that back in 2022 there were

91. 38:28

    some problems on the functionality of some of these rollup tokens how do you expect with restak rollups how do you expect the the place of a token for uh you know a spun up rollup to change are we ending the era of of rollup tokens or are we going to have to find new ways to to create incentives around them I think roll up tokens are going to exist there forever I don't think that's going to

92. 38:51

    change uh one because in the end most of these rollups are underneath there's an app right most likely so if you you're either building a game you're either building def chain def app or something and then you need an infrastructure like a rup to be able to put that put that app on that chain so I think and if you're building in crypto there has to be token in my opinion like otherwise you would rather just be web to or build something that's um that's on the site

93. 39:17

    and we have seen that even centralized entities like you know binance has a token right you could argue you know yes it Powers BNB chain but before the BNB chain existed even then it was a there was a token right so I would say token has not going anywhere they are they're going to stay here uh no matter how much CC tries tries to kill them uh they're not going anywhere but I think what has changed is previously people were thinking hey the only thing that you could do with token was to have a

94. 39:41

    governance a valueless governance token uh and that hasn't played very well for most people right um and therefore many of the I guess in 2022 a lot of people were coming to us and saying we want to have our own gas token our own token app token as a gas token I would say that that demand has slowed down um quite considerably simply because they realize that I think it took a while for these people to realize

95. 40:07

    that hey if I'm going to use my gas token one it creates friction right so people who want to come in on chain they have to somehow Bridge something to pay my gas true is even if people are able to do that or willing to do that um and you know go with that pain the the demand that would come from uh you know having my token as a gas to is not going to be large because in the end it's a L2

96. 40:30

    it's not a theum so your gas fee has to be much much lower compared to Bas chain so the demand that comes from just gas is not going to be enough in anyway whatsoever the second thing is in the end you have to because it's a rup you have to go and you know pay some fee to ethereum for a DA or settlement right and so at some point either you have to upfront that eth cost on you from your own pocket and say you know what I'm not

97. 40:55

    going to convert my gas token that I've get to convert to e to to pay so you either have to affront yourself or you have to sell your your token that you got from all the users at the sequenc level they have to sell them for ether and then sell that you know put that on etherum so what are you really gaining from all of this right not not much so I think there was a demand people because people feel like I could use my token for gas I

98. 41:19

    think the eventual consensus is that that's not the best idea out there and so people now yeah sorry go ahead in nl1 right you pay this it it goes to the validators directly right so there's like the you you're not creating a self pressure exactly on your token but in L2 you're actually creating a sell pressure on your token that's one of the things I was just thinking about is like why top 50 is

99. 41:42

    L1 you can easily come to you can easily come to a conclusion we like oh well there's no like there's no self pressure there's validators they hold it already you know they they buy to get a state price so it's any any just just a random thought there because let's say optimis I'm not sure how much money let's say millions of dollars a day right or month month whatever right now if if it was in their token you it's guaranteed that

100. 42:06

     you're selling that tokens or somehow you're you're exchanging that token for to pay for the feed and so it doesn't make sense so yeah I mean so so so coming back to the point like these D developers they now come to us and they do not want the gas token what they want is some sort of way in which they can get the token used this is where kind I guess reaking comes to play because if let's say if you have a Mark of f fin lay you know you could actually get your token used as a staking token and to be

101. 42:32

     able to secure that Network and this does not create any sell pressure because becomes like a proper staking token so it creates a sync for your token and no no self pressure from that side yeah the only self pressure is got when people get the rewards yeah but that's that's yeah that's an L1 yeah in an L1 people will do the same thing too right so so you you can't avoid it anyway I mean you can you can complicate

102. 42:56

     you know things where you could say oh you know you lock for a year you get more remot things like that no scar that though the past uh for for quite a long time so I mean people are doing things like that but otherwise like I mean you can't expect people to kind of keep holding on to onto this um and this is also by the way where where I come recing comes into play right where you could say hey when like the whole idea

103. 43:20

     was one of the ideas for I or reaking was that hey if your if your token doesn't do well for whatever reason if it creates kind of a circle of death right know the spiral of death it keeps selling and then you're giving more reward it keeps selling again and then it kind of goes nowhere and your token goes to zero your protocol goes to security go goes to zero because the token value is zero right and so in a way them also helps because um you know you you avoid that kind of scenario the

104. 43:43

     the spiraling down of of of you token to zero at least initially or for some period of time yeah then everyone's crying again yeah um do you have a I I have another question Alton if if I can um I I I guess now you know a lot of people you had referenced this before about Defi and how it hasn't really kind of moved too much what types of uh daps are you really excited about for this cycle or do you think are going to come onto the scene this cycle and and kind of you know find product Market fit and and take the crypto You by storm

105. 44:23

     do any stand out from what's being built now or are we still in kind of the experimental Discovery phase I would say we are still in the experimental phase um I would say there two things that I'm really excited about um some of them are probably being built but we still have to see what how exactly they work and how exactly they look like one is I think entertainment um some people call it social F or whatever but the idea is if you think about it right in a normal

106. 44:48

     world if you use your let's say backing application right you maybe use it once or twice maybe five times a day that's it to pay for coffee to pay for lunch or whatever maybe five times a day and that's it um one that you actually use for a long time is entertainment like you you you watch YouTube for hours you watch you sit on Twitter for like hours right on a daily basis and so I think something

107. 45:11

     that has some sort of entertainment element into it would be would be much needed that's that's one angle that I think really excited about and then people like even if you have let's say just 10 users let's say and people often talk about it right how many users do you have on your chain oh I have only 10 or you have only 20 so even if you have only 20 user on chain and if all these users are addicted in some way or are spending a lot of time on that application because it's entertaining or

108. 45:35

     because it's it's engaging then you actually get a lot of lot of activity so I would say sometimes not necessarily worth it chasing I don't know 1 million users that's do one transactions every month compared to chasing finding let's say maybe 20 people who spend an hour every day on that application so we haven't seen that quite much yet where apps where people sit there for a while right I mean axi was there for a while but it didn't kind of pan out in a longer time frame but I think they still

109. 46:05

     get to see application where people sit play or do something on that on that application for for hours on a day that I think that's something I believe needs to be built that's a great honestly I I think that the fact the the idea that everything that we've used to date has been kind of a come andg go model you take what you need and then you move move on we are waiting for that thing

110. 46:29

     where it's like you're waking up and you're spending four and five hours at a time just locked in yeah because sounds like mein trading to me but that exists though that that exists right like so if you look at the applications that have been taken off the Bots did really well over like it's just because there was a lot of like tokens Printing and then you have pop off onun that did really well and I I

111. 46:55

     use like so I'm you know I like I like I like hanging out with like apps the other thing I used was like poly Market I was just said there's some things that you want to you know have fun on but you're not like spending that's why like the change like Solana did well the cycle because like everybody was just un pumped that fun um yeah so but like my feeling like people and the other problem I think I guess with we have

112. 47:18

     with crypto is like we have enough users like in the broader I guess peripheral right so you have a lot of people who hold tokens right that's for sure the question we are not able to convert them to to to use anything maybe I you know people use unisoft you use binance people some of the things but that's very infrequent you to build something where people as as Al was mentioning people sit there like people sit there hours

113. 47:41

     and hours on daily basis at least it gives them even if you have 10 users you get a lot of volume as simple as that that's Entertainment saying this is coins me coins are entertainment yes so funny like there's this one M qu called Nero which was like sister of do right and everybody's like ah it's so funny like you just sit there it's uh that's you know uh that's where things are and yeah this is also why I guess the last cyle

114. 48:13

     was all about entertainment honestly they put in a kind of like a sense right what I'm saying is that you could go beyond that as well like something where like like I I spend for example quite a bit of time on YouTube or you know you listen I don't know on Spotify things like that think that where you spend time yeah nfds were were big last cycle what else was there like nfts defi had a

115. 48:37

     had a interesting things it wasn't defi summer but like um a 2020 was kind of def5 silver yeah really yeah I don't see that yet I'm just curious what's going to happen like I when the there's also less money available um that that's very true the system that's very true like people like a lot of people VCS even say hey we investing in in in in consumer apps but you actually go and push them I think out of if you give them 10 infra and if you get them 10 apps I think they will invest in out of 10 they will invest

116. 49:10

     five infra and then one app right they'll do that for us inra Builders but it's also sad right because it won't work without all these app investment apps yeah and you know that that's a that's a great truth bomb honestly because you know EX exactly the logic of of what goes into that it's there's not a mathematical calculus for how a consumer app can become Mass adopted right it's a lot of oh you you found product Market fit through this community or you marketed it the right way or you you did something special a

117. 49:43

     lot of the infrastru stuff you can at least delineate some of the mathematical cost rewards uh uh you know yields Etc and it's a very safe investment to make uh compared to to this this consumer world that's a good point yeah it's a numbers game right like it's a numbers game like you need to you get in and you're like hey I want to do this one app and try and then try a growth metric and then if it hits right actually one of the good things of crypto is like the

118. 50:09

     fee is built in right so you're when you're doing it you can actually charge fee so you can actually make a lot of money like but the thing is like the world is weird right it's a numbers game and it's not it's very easy to build a map like like I built a telegram bot I was like a month long right versus what armit does what we do is like you need a lot of like thinking time a lot of deep Engineers which cryp has a lot of those

119. 50:32

     folks but they don't have a lot of these like hang there just thinker and like build stuff kind of folks so in that case and the investors are that way too right so then you get this like a lot of infrastructure not that much apps and people that do apps don't get funding so they they eventually try try try Nothing fits and then you know it's like ah let me let me leave here that's kind of the the setps you yeah way it's also an

120. 50:55

     opportunity by the way right and if you are a good good product guy right if if you have a good idea you you if you build an app and if it if it clicks you would be You' be the darling of crypto like everyone would like to have you right everyone would pay millions of dollars to come and deploy your thing on on on on their chain right so it's just nothing is clicking but at some point whoever clicks it I think he's going with the King he's going with the King

121. 51:20

     for sure another question though if that clicks are you launching your own chain or are you deploying an open if if I that Builder I would I would build my own chain for sure because I think that's what polar Market will do right definely like ask you stupid not to do it because if you think about it right if you're building on any chain whether it be base or op or arbitrum whatever right uh you might get funded right you you'll get some sort of Grant

122. 51:44

     or whatever from them um but in the end you like a normal person goes and builds on el2 because of distribution and liquidity and users because they feel like hey I may not have users and so I need optimism supports I need people who are using and transacting on optimism to give me that that that user base but if you already have built your distribution on your own like why do you

123. 52:07

     care you don't care about any of that you have your distribution of your own through that distribution you can make money as Al was saying you can basically bake a simple fee model and you can make money we have seen that to pump. fun you know they make lot of money so if you're distribution if you have if you have made your money at some point you would like to go there's no reason why you

124. 52:24

     would go on UHA yeah which becomes like a cycle that's why it's like I have an L1 or an L2 now we're built there I'm just going to launch my own L1 like it's just fragments which why chain abstraction makes sense so that's the that's where the things are but cool thanks for being with us we have like maybe two three minutes but uh do you have anything to share I guess about Al

125. 52:49

     there about what you guys doing or anything that you want to share with mainly like new YK Community but others as well so so of course we we are with you guys so um anyone building in the near ecosystem if you want to have a rollup if you want to use um you know near da or the tools that altern and team are building come and come and talk to us um you know we are very Hands-On

126. 53:11

     uh in many cases we We'll advise you on how what's the right approach to go with uh which stack you should go for um where the vibe you're looking for um you know you would find that so come and talk to us if you're building in the near ecosystem we're happy to support you in any shape or form in terms so kind of what you're building unfortunately I can't share something quite right today but keep an eye on our Twitter I think we'll have something in a couple of weeks time uh especially for the ABS and the icon

127. 53:37

     ecosystem awesome cool great thanks so much for the time we appreciate it and uh this was fun so hopefully we'll do it again sometime yeah thank you a thank you Michael good to see you