[REDACTED] 2024 | Fast Finality, DA, and Why Modularity Is Surging

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

1. 0:00

   hello hello all right H it's good to have both of you guys here all right so let's just like have a conversation right like because you guys are both doing da and a bunch of other things slightly different approaches let's just jump right in uh first Alton how do you do approach Your da and stuff like that and then we'll go to you and how you do it kind of brief and then we can kind of discuss yeah so we when we started doing da um we we were a team at near so folks

2. 0:24

   are here um and and the first approach we took is like there's a lot of block space around um and near as perfectly suitable for this is with sharding um and and da is a thing that like pretty much uh near sold for in 2020 um so we went back and said hey let's just build something build something on near um and that's how we started um for for for building the it's it's it's scalable it's cheap and everything you need so um I know onok has a has a different stance

3. 0:49

   here but um we have um we we believe in a world where you can just use blocks from everywhere so that's the that's kind of and there's a lot of enough l1s to do that um that was that was the that was the approach we did awesome so you're the L you use the current L1 all right let's hear your approach yeah so so I'm coming from a different uh perspective right like so I started off in crypto starting polygon started with building a plasma one of the first

4. 1:15

   offchain scaling solutions for ethereum then we went on to do this rollup thing with zkm and stuff like that and around 2020 we realized that this rollup Centric road map was going to become pretty huge and um so where we we came from that was like if we were going to this more rollup Centric multi-chain sort of a uh future um you know like of course ethereum was on those that execution sharding road map which near eventually implemented but uh we were coming from more from the ethereum side

5. 1:48

   uh which is saying okay this's is uh rollup Centric road map and if you're going to look at the rollup Centric road map you will see that uh the fundamental requirements are of course you need lots of data availability and also you know things like interoperability between these two uh between rollups in general and we thought that let's start with the hardest problem first let's tackle something called Data lity sampling um

6. 2:13

   which ethereum has on its uh dang sharding Road what is called dang sharding in general um and we started with that and then now we are making uh Way Forward after building scalable da we are moving on to things like interop um and Sh security and so on okay so you're doing and and you built your own L1 right that's so what let's go a little deeper in that so why did you do that approach is it to help follow up

7. 2:36

   with the ZK and all that so you just kind of um so if you kind of look at uh um data availability is um and sampling the implementation itself right like it of of course requires this the base layer requires this whole cryptoeconomic security guarantee like it cannot be um defined by cryptographic uh proofs in General so you really need some sort of cryptoeconomic uh game there and we thought if we had our own L1 it's just

8. 3:07

   just better to do um um you know all sorts of combinations in that sense can I ask a question so is do you still think da is an important problem to solve I know like we've we bunch of problems like we have a lot of l1s that popped up I know there's a bunch of folks that raise more money doing da like what's what's your take on that one I think da is just the starting point right like I mean it it cannot not be just the only uh feature of uh the

9. 3:33

   solution so if you kind of look at what our road map uh is it's that is just the first step right like what do you really like if you look at the landscape of rollups today right like this it's just too fragmented for example each major rollup stack is doing its own interrupt thing you know optimism doing a super chain thing you know like ZK sing is elastic doing elastic chain and so on right like so everyone runs sequences

10. 3:58

    for example C sequencers there is there are different protocols for decentralizing the sequencer different protocols for interoperating between these ecosystems so it's of course you know we can we are of course in the middle of this whole modular sort of uh Revolution but the reality is that um when you kind of look at the other extreme end of the Spectrum which is the more monolithic integrated sort of

11. 4:20

    spectrum everything looks like neat and tidy everything working together this side looks very messy at the moment right like so so you it's it's kind of hard to kind of um uh ensure that we have these more heterogeneous rollups all work together and so you really need to be thinking end to end vertical stack not only just one layer okay I I I see what you're saying okay so here's my question for you so that we got a little technical there which is good I want to see like people

12. 4:51

    listening how do we go to like I'm at a place how do I explain da to like a non because I do know like maybe the people that you're focused on are going to be more Tech heavy but how do you just describe da and kind of like explain like I'm five we'll bring it down a little bit and then we'll go back tech again I think yeah I mean uh see datail is a is of course a very technical Concept in general uh most users won't

13. 5:13

    even need to understand that much it's mostly the rollup devs who want to understand it but one of the side effects uh or Main benefits is that if you want to uh especially with data sampling if you want to verify um the chain uh yourself without running this full node for example um that's that's pretty much the easiest way and if you combine it with things like execution proofs like ZK execution proofs also then the chain becomes user verifiable for example um when you think

14. 5:43

    about is like blockchains are about promises that you keep so validators keep promises right they they basically say this is going to be the ordering of transactions that I will be committing to execute um this is sort of similar uh so you make a promise that this is going to be the transaction ordering of an L2 um l1's do the same thing right so you come in you want to get that promise and

15. 6:08

    make sure that that promise is correct so I think that's very important for blockchains and the space as a whole like promises to be kept and I guess that's where cryptoeconomic security comes in uh that's why ZK comes in um yeah so we're quite excited about that like the promise keeping promise and we can talk more about that too definitely and now I've got my source from smart devs questions right because I was like yo I got to reach out to the community let's talk to devs I I'm talking to

16. 6:34

    highly technical stuff so let's go with cross chain data verification great segue um what approaches are being developed to handle cross chain data availability verification in Bridge protocols and how do these improve interoperability start with ZK again y ZK no so see um essentially if you kind of look at this whole rollup Centric road map or the way that we've kind of been looking at it right like it's you

17. 6:58

    can say data is a scale problem I would say it's also a interoperability problem right like is really like if you like either you build one big fat chain that can house all the world's transactions or you want to get into these multi-chain multi rollup sort of a future and how do you make this all work together if you are in this multi-chain thing right like and so also kind of U

18. 7:21

    when we say that users can verify this it's not only for users right like other chains need to verify other chains like how does that happen like in general before rollups it was all light everyone had to build a light client uh to each other for example like in the cosmos world for example and so here essentially if you want to move out of the notion that um um all the world's transactions will not be on one big

19. 7:48

    super blockchain for example right like it'll be very similar to the inter where apps on the internet don't really live on one big superc computer they live on different servers um uh scaled independently and so all of this if needs to work together um interoperability is a key goal right like how do does one rollup trust the other how does one chain trust the other to do transactions if you have a 100

20. 8:14

    million transactions how can you Shard them between these chains so one one um implementation is of course execution sharding but uh what we believe in is the more scalable role of centric path so we have a I guess different approach so um we love ZK uh it's it's definitely going to be the thing that in the future but to be more um sort of practical uh we use cryptoeconomic security and I know a is working on the same thing as well there's a lot of tokens lying around so there's near

21. 8:45

    there's cardono there's all these to tokens there's even arbitrum the AL have like billion dollars of tokens so how do you take them and use them to make promises so you basically uh take those tokens and use them to do cross-chain verification um instead of doing ZK how do you attest to them with with the with the tokens with the security that you get um we work with Ail on this one too where obviously ZK is great ZK VMS are

22. 9:10

    coming up but how do you do this with optimistic um rollups um it's a good way to use uh cryptoeconomic security to um attest to different blockchains and say hey this is going to be the state um so yeah the two solutions can be combined to build something better and I guess that's that's that's something that we we talk with the quite a lot um yeah so I've been thinking on this problem a lot right like I mean uh

23. 9:35

    essentially uh what I've I'm thinking right now is that both these approaches are very important like the cryptoeconomic security and the cryptographic part because in general um not all chains will be you know like ZK verified um immediately so there will be a lot of optimistic chains and there's also this distinction between change with very High liquidity and chains which are new maybe with low liquidity and so how I think of this is like there

24. 10:05

    will always be chains that are you know like big like high liquidity chains then there always be a long tail uh maybe non-financial use cases or you know things like that and really if you kind of look at the cryptoeconomic security uh thing you also have to pay rent so it makes sense and even from a liquidity provider perspective I think what is eventually going to happen is that from a liquidity provider perspective these migrate to high liquidity chains but for the longtail you'll need a ZK sort of

25. 10:35

    component because liquidity providers or paying cryptoeconomic security is expensive for this long tail Okay cool so let's like talk a little bit because you um so like how does like data storage cost impact the overhall tokenomics of a blockchain right and I don't know if you it's probably wouldn't be for you because so it's probably wouldn't be for you but for you and then um and What mechanisms can help maintain sustainable like fee

26. 10:58

    structures and stuff like that can you dive into that just a briefly yeah so see data availability is um not really uh you know like to be very Frank it's not data storage uh really of course there are data storage U uh implications on what you can store but really the guarantee of data availability is is to these are all commitments right like I mean when when users send transactions when to rollups or rollups send transactions to the deer you're making like a Time Stam

27. 11:27

    commitment that I sent this ordered transaction data to you and here has this commitment that this you know like won't be rolled back sort of you know like that's really the guarantee and we keep the data around for a while right now the network is new so we don't need to keep that um Purge that data or prune the data but in general of course there will come a time when we will prune the

28. 11:49

    data but there always be archival data methods to kind of uh do it it's not the same problem like how we structure it is our chain doesn't doesn't at the moment combine execution and data availability and so it doesn't lead to things like State blood which is a data storage sort of a problem with traditional blockchains uh and maybe you can answer this I know well from from the architecture how does I guess if you have only da how do you think about like

29. 12:20

    the Avail usage in the ecosystem so um at the moment uh of course you know like it looks like da only layer but there are components that we have been working on uh there's something called Nexus which we've been working on which is you can think of this as a more of like a proof verification execution environment so it doesn't do General smart contracts but can verify proofs from you know like ZK rollups uh we also we just tied up with

30. 12:50

    symbiotic uh to you know kind of bootstrap something that we can use for crypto Economic Security for example uh so things like that right like so at the moment looks like we like we only provide data availability but it's I mean the intention was never to do just that it's much more I joked about it last time when we talked it's like everything we built has been going in the same direction so like the last time like symbiotic we work with I so it is

31. 13:16

    it has been going on the same direction which is kind of funny but then we we were're very friendly so that's the um I I play in the unification cup that we're sponsoring football so that's that's quite funny to see like how um um how this evolved over time so in in both your guys as at this is individual where's one area you think people should be paying more attention to specifically that they aren't that you're like that's the spot that like a lot more attention should be but people aren't looking

32. 13:39

    there because they're looking here yeah I can I can respond to this one so I think uh IG space abss are very interesting um because we've been building blockchains for I don't know like last seven eight years and it's all about keeping promises so it has stay blow it has all these problems that you keep on the blockchain so we believe in a future at nle to use these tokens from anywhere um to be able to build uh build

33. 14:01

    the economy of promises so what do you do in world of promises you run oracles so you provide data to the blockchain that's number one you do zkt LS which is uh which is a technology that allows you to take any data from outside provide it to a blockchain so almost like Empower these apps outside of the blockchain to make them more usable because apps are the thing that of the of the future so that's how we think about it and and

34. 14:24

    what we call this is enough protocol so it is contracts everywhere like near arbitrum anywhere um you can stake your tokens we use those cryptoeconomic security and plus could be E2 honestly um and then we can use that um those the the promise provided by these assets to do applications outside the blockchain offchain um so we're quite excited about that uh that's kind of the and it uses

35. 14:47

    da and fast finality at the back end right so um dude that's cool I feel like you just dropped some like Alpha there that that was really cool man uh what do you think where are you kind of looking then I also want to follow up a little bit on your project specifically yeah yeah yeah so I think um when we began ail back in 2020 that was that was the time the rollup Centric road map for ethereum was just shaping up in general uh and we've been spend we spent the

36. 15:09

    last 3 4 years uh building out this hard component called datability sampling uh which we finally launched in July this year uh but over the course of building this we also saw a lot of stuff that was happening uh in the modular space in general right like we see a lot of fragmentation as I already mentioned right like I mean different uh people running let's say Ras infra sequencers

37. 15:33

    you know like prer networks uh interrup providers for example so all of this is a lot um so like I see I kind of think of this as these two extremes of this monolithic integrated side of spectrum and the modular side of spectrum and I think modular has become a little bit too unopened as well um and so I think the answer is really in the happy medium uh you know somewhere in the middle path

38. 15:59

    like where we kind of integrate a little bit more um so that uh users don't really need to kind of think about um you know like especially rollup devs don't really need to think about maybe like doing their own sequencing for example running their own infra so we are on our upcoming road map we are thinking much more in that um area and essentially all of this is also kind of in service of you know what I've been kind of seeing in the last couple of years is that in general not only crypto world the entire um you know like world

39. 16:33

    is moving towards a more fast mobile first sort of ux like if you kind of look at Solana uh what's happening there in general like even if you look at it's dismiss of this as mem coin Mania the infrastructure is you know like doing all all of this today it's used for mem coins today tomorrow it can be used for anything really right like I mean but if you kind of look there's a very strong emphasis on mobile first US

40. 16:59

    because the younger audience 15 to 25 age group is is on mobile right like I mean not everyone uses a laptop desktop browser extension that sort of a world I think the Solana guys have been able to focus their ecosystem on this more younger audience mobile first with hdks and stuff like that and I think uh the rest of the ecosystem need I mean I'm not saying they're not active on the mobile side but um I think we need to do

41. 17:25

    much more focus on getting to fast experience is mobile first experiences and all of these infrastructure buildings should be in service of that I just kind of did a tweet yesterday saying you know like uh if you look at app something like Apple pay Apple pay is you know like tap to pay very easy um fast mobile um experience but the reality is that's there's a lot of infrastructure powering that which the user is not um seeing uh I think we we have to slowly move there and I think

42. 17:57

    the technology is there I mean or it will be there in a year for example so those are I think areas to focus on I have a question so commitments are very technical terms if we take that as a promise in in the world right and the is a thing that is very interesting in from um from a blockchain perspective do you see any applications of Das outside of just blockchain infra and you think like

43. 18:20

    could be useful for mobile future or something that could enable us to do more than just blockchains um um So within blockchains of course you know like we are experimenting with this whole put the ZK execution proof on the light client sort of a thing and you know like either users can verify on their mobile wallet for example and so on or other blockchains can verify for example but

44. 18:43

    outside of blockchains also see really what we are what we ended up building is also like a Time stamping uh server really um in general and so I think we have we've been talking to a few projects outside of the blockchain who are looking to use it for various other uh requirement not necessarily for running a blockchain for time stamping reasons uh even you know like um uh one particular messaging uh protocol wants

45. 19:08

    to use it for uh different use cases for example um and so I think there is utility in having this whole scalable shared chain that people can you know use as a common timestamping um server in that sense right like so we've been seeing those sort of use cases as well dude I'm digging the like time stamping concept and all that stuff and I also think mobile oh yeah totally speaking of but

46. 19:33

    yeah totally thinking of all that that's like really cool and the mobile first thing I agree like that should be like the beginning and then it can go onto a desktop it's kind of hard to ship that so okay so what is a no let's do I like this one so what advice would you give a developer navigating like how to start building or start working with what what what you all are each doing like you're targeting a Dev you're talking to a Dev

47. 19:56

    what would you say where to start where to go how to like what are some like things you would advice you'd give them well yeah first one um definitely don't build build a new blockchain so that's that's one um because it's it's been just been there's been a lot of different blockchains that have like almost became zombies right like that even though there's incentives for that um number two is well take some of the things that in the world that we we make

48. 20:21

    promises and try to build some tech around it um like I gave an example of ZK TLS I think uh one of the teams from the iice ecosystem is building it to verify Uber ratings on chain so that you can build really cool applications around that um and that's what like for example there's a lot of use cases in the real world like sweatcoin on on near things like could be used to do um to do pretty much verifiable data that allows

49. 20:46

    you to do really cool stuff um so look around there I think apps are apps have been undervalued in a sense uh because the yeah the the incentives were in there um I think after after poly market and pumped up fun I think people are going to get more excited about apps as a chain and also the Technologies here so like aail Nexus for example um uh with near fast final layer which is the

50. 21:11

    um NFL that we built you can actually do it any any rollup right you can connect any rollup so infrastructure is coming I think need to think about like how do we use these promises that we make in the blockchains usable yeah I think uh see in general the the industry in general has um I would say of course including us uh we've we' kind of sort of failed to educate the people outside of our very cute crypto circles that you know like what blockchain really is right like so

51. 21:42

    we have like this one team building our ecosystem which is doing this whole uh gaming play right like it basically takes in data from some of the most popular games and then builds tournaments for the longtail really like if you kind of see you know like gaming tournaments are for the top one person and this is like gaming tournaments for the long tail and it takes in data from these um big

52. 22:04

    Publishers like you know walve and uh those sort of people um and actually um some something like ZK TLS for example is really cool for them because they're able to take that data and put it into use trustly and the when they started they wanted to kind of do escros for let's say community like for the tournament um uh uh like pool for example doing that in traditional fintech is really really hard doing an escrow in traditional fintech is really

53. 22:37

    hard like and they um you know like cancel their plans to do that because escros are actually standard in smart contracts for example right like it's much easier for them to scale to a global Audience by using smart contract infra for this so I think what I'm trying to say is like from an app builder perspective like I think most most Builders should be looking at what problems are there out in the world that

54. 23:01

    they're trying to solve and how blockchains can help them right like I mean things like escros pretty cool things like tapping into a global audience pretty cool for example um even there's a lot of things right like um uh asset collateralization right like if you kind of see something like a a it looks like we are you know kind of doing this digital native um crypto tokens here you know like we swapping them put into WS for example but just for a moment think these are stocks or

55. 23:33

    you know like let's say gold for example or anything that can be tokenized in the real world and you suddenly see that if you have like some sort of app that you can let's say buy gold or buy some sort of asset which is not available uh in in a liquid Market uh so today if like India for example um gold purchases is very uh prevalent right like and you can buy gold on these apps

56. 23:59

    um uh but the gold really stuck there in the app right you can't do anything with the gold you cannot uh collateralize it to take a loan against it for example and so these are all examples of these siloed systems being built in the real world and every system like that is purpose built for a particular use case blockchains are this very composable structure where you know like once you kind of tokenize one asset they they can

57. 24:24

    get access to this whole you know like set of building blocks so I think it's um so one of course use case fited the use case but also we should also be doing a much better job of you know like explaining what blockchains are really good for got I can I can also say one one more which is one of our our C talks about it service Agreements are very important right so you you work with Amazon you work with Google you work

58. 24:48

    with all these folks um and they they say that they're going to provide a service but usually when they don't um they there's no disincentive for you not to do it um so that's why like crypto could be very useful for and that's the future of of of of like the future that we envision that's awesome all right so we're we're we're coming to the end here but here's my question so does like the development of AI

59. 25:14

    affect anything that you guys do and if so how like right we just had a big presentation today about like AI agents they're coming right and there's G to be a wave of them where does this fit in or does it not right like that could be a possibility too ZK yeah yeah yeah I mean see uh of course AI is a very exciting development uh in general um I'm still looking at that whole AI agent thing still trying to figure out what will come out of it I mean you can of course

60. 25:44

    think about lots of a agents interacting with each other doing a lot of microtransactions micro decision making those those sorts of things and uh for sure like if if there is an economy that emerges with that AI agent thing they need some rails right like I mean I mean it's not going to be you know like they're not going to kyc open a bank account and do stuff like that it's going to be crypto for sure right like I mean um in general it's also like uh

61. 26:09

    it's AI is not only AI agents right like it's also you know like um pattern matching for example and so you have a lot of problems in blockchain that are not C all on chain right like you have things like intent matching for example and so on that can be solved by this you can have like um a lot of wallets getting smarter uh because most wallets are very dumb today right like how can you um

62. 26:34

    make them uh understand what the user intent is for example right like so so there's all sorts of things that can apply um so I'm cautiously optimistic for sure yeah and when you say wallet's getting smarter I really like that because I'm like I'm always shouting from the rooftops like near like wallet structure and account structure I'm just like that is like so cool um what do you think there's a point though AI is getting smarter but how do do you ensure

63. 26:58

    that AI doesn't rug you right and that's the that's very cool with with cryptoeconomic security is let's let's just do a swap um but instead of near it hallucinated and did a Bitcoin swap right like you should desent Des incentivize that behavior so for that um again promises are very important so you need to make AI to make promises about you using the the assets that we've already built which is a at this point I think it's coming to three trillion Market but it was a two trillion Market

64. 27:27

    uh was a month go um so that's the that's quite exciting for us uh that we're looking into and I already posted about this which is one example is this kind of a AI Federated learning ABS um that there's a lot of these things that you can build in AI that is using a cryptoeconomic security that we have built in this industry uh to do something that can that can't rug you that's awesome and and it's exciting that like now you're thinking about it

65. 27:51

    and as we're ending here I I just can't wait to talk to you both again in six months from now and see where you're thinking about how AI Folds in are these agents yeah you know as they take off were you folded in so thank you so much and with like we're out of time but do you have like one last thing you want to say or anything or we're feeling pretty good that was pretty great see you in the next panel because every time every every conference I have I have something with under rocks at this point it's it's been a tradition so I'm excited for that

66. 28:15

    cool thank you so much and uh let's have a great day guys cool thank you