Reaching Escape Velocity Architecting Next Gen Web3 Apps v3

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

1. 0:08

   all right well welcome all the survivors this uh this became from a really nice venue to like a sauna I think that's uh the transformation we've uh experienced here so the you're saying the discussion is heated up yeah I mean we need something spicy here clearly okay good we can bring that so I I figured we'll ask some questions I don't know if you want to start or I can uh go for it okay well so I think the thing that the the oh

2. 0:42

   there we go all right um I want to start with where does this all end up right five years from now if chain abstraction makes a difference what's the difference that it makes you know what what what what have we changed in the world if we succeed at this stuff what kind of if is that yeah I think I mean it's it's it's a really good question because from my perspective there's been

3. 1:07

   maybe three phases of blockchain so far it started with Bitcoin right kind of introducing the idea and everybody kind of was excited about it was trying to see what they can do around it then you had ethereum which introduced kind of this programmability environment and that got you know a lot more people in there was more ideas most of them didn't work but you know some things actually made progress and then we have this

4. 1:39

   like explosion of infrastructure that happened afterwards and it was kind of driven by the fact that people tried to build things in 17 and they couldn't right none of none of the actual applications worked and so I think kind of what chain abstraction is doing is bringing this next phase right that infrastructure kind of going away on the background and to give you like the maybe a way to contrast this right when internet was

5. 2:10

   developing right we also had like set of phases right you had arpanet and kind of you know Academia where you know there's some Geeks like sending packets to each other you had like represent that you had early internet right where you know you had like you can send an email to each other and was cool and you know there's early early web pages and then there's like massive build out of infrastructure that happened afterwards

6. 2:35

   where people were like okay everything will be on internet and you know it didn't really work yet uh for various reasons and then it took some years to actually build proper applications slide with a picture of the sock puppet from from uh uh uh pets.com that died in one of the internet bubbles looks astonishing like like a board ape you put those next to each other and it's so clear that that

7. 2:59

   was same pattern anyway yeah yeah yeah I mean then the yeah the do were pretty much me meme coins right of the of the past so and so I think like we are in this transformation this is what we I think driving with chain obstructs like we're transforming away from like okay everything has a doom everything is a meme coin that you know you need to buy in early to hey these are actual useful

8. 3:22

   applications that coming out of this you know few Will Survive most will be started afterwards and uh I I think we'll come to a world where you know same as you don't think about like the fiber that's you know probably serving Wi-Fi right now and like if the Wi-Fi network is Cisco or something else right we're going to get to that stage as well and you know there's still a lot of things like security and recoverability all

9. 3:47

   kinds of stuff that needs to be like polished and solved but that's kind of the world I'm looking forward the application world the use their own internet world there's sort of there's a phase I'm curious whether it'll line with this of you know um uh crypto launched and got the ball rolling with pure Financial stuff I think a big part of that reason is because those could be on chain and interfacing to the dirty real world is

10. 4:12

    uncomfortable and difficult and challenging and requires the kycs and the and the you know disciplined execution and that sort of stuff the the the to me the chain abstraction where it's very user focused it's oh we actually have to solve problems us users have that's really important there's sort of at the same time this real world asset thing coming in which is you know painful stretching its legs out of pure digital environment into actually

11. 4:38

    interfacing to the systems we have in trafi that people actually actually want to see do you see any connection between you know uh um chain abstraction and real world assets or they just different threads that will eventually stitch together um you know where how does that show up in your world yeah I feel it's going to kind of be parallel threads where you know the chain abstraction is

12. 5:00

    the user interface the the way you access the assets and you know the assets will come and be on different places we'll have you know the Omni assets we'll have the I mean like if you think of it usdc and usdt are an omni asset right it's available in multiple chains there's a way to like natively transport it between and uh uh we'll see more of those especially for your world

13. 5:25

    assets it makes sense to have you know them available on multiple and then we the other side we'll have S chain abstraction more of the aggregated liquidity where you don't actually need to think about which chain the asset is issued you're just going to get it and uh you know there's set of guarantees around that so I think like it's kind of two sides of yeah like real world asset is more of a supply and then chain abstraction is more of a demand side

14. 5:50

    interesting one of the things you you mentioned something that the that um I started to see a little bit of people confused about with respect to chain abstraction where you know it's not hiding the details it's abstracting the details example I had this morning or the you know the analogy this morning of travel sites where you know I just want to fly into Brussels some people will

15. 6:13

    care whether it's you know United versus delta some people will care first class versus coach but once they make that decision they push the button right a year ago you you know a year and a half ago you mostly didn't care what bank you were doing but wow suddenly in May a lot of people cared about Silicon Valley Bank because because you know suddenly it was no longer a counterparty I could ignore and so there's this it's not about hiding the details it's about once

16. 6:36

    I make my decision then I can just use it in in a uniform fashion I'm curious whether you see that show up in the applications you're talking about it whether you're focused more on the hiding stuff or I think I mean there's definitely people want to like kind of maybe the way to explain it there's two sides right there's one side which is kind of assd isure developers who do care about the counterparty risks more because at the end of building a business one way or another and they care about this they

17. 7:08

    want to choose the you know the best network but then they don't want to limit themselves to just the kind of user base of that ecosystem of that Network yeah and so that's where the other side of this is users who want to you know be able to participate in the economy in the Assets in the applications and and I think users even I mean with the banks the same right like don't fully understand all the repercussions and you know don't like I don't think anybody's like sitting there and like fully reading in

18. 7:40

    what if DIC actually does and you know what is the like all the corner cases Etc right and so we kind of need to get to the same level where maybe you have like you know some insurance on top or whatever but it's kind of on the developer and asset is side to really provide you with um uh the experience right and the thing is like if you issued an asset and that chain for example you know stopped and crashed and burned well it doesn't matter like if you if as a user I chose a more secure

19. 8:14

    chain I still needed to access that asset and it crashed and burned at the end anyway so like this is example like if you bought you know USD on ethereum it was still going to zero right or whatever 0.1 somehow it's still not zero right so so I think yeah like this is kind of the like at the end it's going to be you know set of product decisions as well where same as you know as you

20. 8:38

    said travel websites kind of surface information about okay well you know maybe you should not buy like some random airline that you know showed up yesterday or maybe looking at you spirit and Ryan yeah or you you know you show maybe like how many uh you know planes have crashed of that Airline when you're choosing what to buy right so the the travel sites did in fact add a selector you could say and I don't want to fly on these Boeing Airlines exactly yeah so so I think like we'll

21. 9:08

    we'll see more of that where like the the the product will provide kind of additional like I mean I actually think all of the information about assets all of the information about applications should be on chain and like registering the asset should happen on chain where you publish whatever the information distribution Etc all of that should be and then every wallet every exchange every app can pull that information

22. 9:32

    attach it to your token and when you're buying some asset you know you're going into Bon and says like well you know this is inial distribution this is what happened and then you're going into like some mem coin launched five minutes ago and shows like hey this asset exists five minutes it has initial distribution of 99% in one address and uh you know and like you can make a decision maybe this is like the stuff you into and then

23. 9:55

    you go or maybe you go and do something else right so I think we need that level of experience and so it's it's abstracting the complexity of like going and figuring out like in the Explorer right maybe there's a chart somewhere but actually bringing the relevant information in relevant context I I've seen you know so chain abstraction has this spectrum of people working on different parts right we we got the the

24. 10:19

    folks that want to do it in the wallet the account abstraction is a part of it there's people talking intense Etc um it is one of the things that I'm excited about there is you see some applications where they're really trying to deliver a good application but they have to do it right now offchain right you know and and uh uh Barry was up here from skip that have their you know IBC fund and that's see you know take my token from

25. 10:44

    there get it over there do whatever steps you need and then you have to it presents you with all the signatures you have to sign in order to make it happen but at least that plan and that route and all that could do for me and so it's sort of half offchain half onchain with a big part of the computation off offchain and you know there's sort of this this how much can we Migrate onchain how much how much do we end up

26. 11:06

    having to rely on third parties offchain my goal of get it all onchain right is exactly that you talked about these phases of you know web 3 there was some sense the zeroth phase which was smart contracts before blockchains right to me you know what we've always been about is the behavior in the smart contracts the you know software that's enforcing the terms of relationship and you could do that on website I mean eBay PayPal Veno

27. 11:31

    they all do smart contracts they just clearly have a trusted guy in the middle right and some combination of that and getting those smart contracts up onto a blockchain which is much higher reliability and so you don't have that that trusted intermediary or that vulnerability in the middle uh would be great and I think that that that chain abstraction is currently straddling you

28. 11:53

    know those those offchain versus onchain things for the so you know a you know where does that show up on your radar and you know yeah I think the there's definitely a spectrum there and I think per use case there's a very different things you need to do um I think for a lot of the trading use cases and financial use cases in general I do think the intent type architecture is just so much simpler and it allows to

29. 12:23

    aggregate liquidity and aggregate kind of uh access without needing to sign every single thing in the middle right like imagine you know I say hey I want to buy whatever a Bitcoin for salana usdc like I don't want to think about how is that's going to be you know executed right like is this going through like ethereum you know through sord chain through like IBC bouncing through osmosis into ethereum and then going through wbdc and then unwrapping like

30. 12:57

    this is not something I want to know you just poked at one of my problems with that I see in the chain abstraction in some of the chain abstraction rhetoric so much of it is focused on the swaps right intense is mostly focus and there's a lot of volume that's a huge part of crypto right now but it's a relatively small small part of the economy well but but I the way I like swaps are more of of a so like the way I I I I have this like what is blockchain really good at right it's it's a good

31. 13:29

    good registry for identity it's a good registry for ownership it's good for payments and it's good for marketplaces and swaps in a ways represents marketplac right you can swap you know a guitar for some something else right for a piano right and uh like that is a swap that is an intent I have a piano I want a guitar and then we somehow figure out how to settle it um and you can have everything between right like you have a compute Marketplace for gpus for AI you

32. 13:58

    can have and so that is I think like a lot of our life is in general marketplaces right I want to apply to a university right there is lots of universities there's lots of applicants it's a very inefficient Marketplace which is by the way judged based on how many applications are so there's a lot of things put out an intent which is okay I'll pay $5,000 for a four-year education and I don't care where well you don't say you don't care where I I'm saying I want to be able to pass this

33. 14:24

    test afterwards right and I want to have this knowledge so you need to specify what you actually want to receive right you want the outcome not the forear education is not an outcome that's just you know wasted time I want an outcome of this knowledge what I mean is I want to go to that school so so the key thing that I mean by that is so there's there's a lot of the stuff that you see in in like the cake stack and a lot of the the intense stuff which is it's all

34. 14:48

    about as said it's all about these swaps but anything where you're doing you know lock up collateral to get to get a stable Tok I don't want to lock up collateral with a random contract on some chain I don't know about I want to lock it into IST and mint IST right I I want I've got there I want to a lot of the economy and what I hope a lot of the new economy is is Long Live trustworthy

35. 15:11

    Arrangements rather than mere Anonymous transactions and so the question is you know the the the a a question for intense is how do they help us move to that and maybe the answer is they don't I mean there's plenty of stuff to do well I mean what you just describe it isn't I mean to be clear I'm not like a int here but I'm just specifically what you described I I have this dollars or whatever I have this ceral one is

36. 15:35

    literally an intent right fair enough so you have IST and you have aoll you know you have a depth uh collateral on on your account but like generally speaking I I agree intents are a extremely useful tool for a subset of use cases there is another subset of use cases where you do want to like a specific plan of execution that you can um run and this is where my point is again you don't

37. 15:59

    want to have this like multi-dimensional like signing process across you know every single intermediary you want a way to facilitate that and like route that and have somebody else to figure out all the grungy details of how to pay transaction fees and related transaction and which RPC to send it to and you know is that nodes are down and I need to reroute it Etc like and this is where I mean change signatures are kind of the tool for that

38. 16:22

    is you know it's able to say well I want to deposit that collateral and IST and here's a transaction action you guys figure out how to send it right it's not it this is this is actually not the intent this is literally you know a proper transaction but I as a user don't need to think about all intermediate steps there either okay so people may have heard about chain signatures but but a softball right there's there's

39. 16:46

    General chain abstraction there's chain signatures and there's account abstraction which you know balic uh uh um is very interested which has slightly touches on chain abstraction but it's not the same can you sort of clarify from your perspective the the the the difference the dichotomy in one minute so I think the the high level is this right the chain abstraction is a mindset to me like this is like where we

40. 17:10

    remove we trying to remove this complexity from the user account abstraction was more of a like how do we do it well we we don't want the user to have a this binding relationship to a public key on a blockchain where now they're not able to do anything else right you don't able to pass it to anyone you're not able to recover it on different device you're not able to you know set up um you know that your friends you not able to pass pass it to a custody and so that's kind of where

41. 17:42

    account obstruction come in is like Hey how do we make accounts to be more of a kind of programmable object that you're able to do a lot of things with the kind of so near build that as well from day one so work and so now the question was okay well now we have different account abstractions on different chains evm is doing its own account abstraction we have safe which is really good and

42. 18:06

    provides account obstruction but then it's like a whole separate op like like thing that like you need to use in a very different way right you need to open an app inside their app right and stuff like this so and then kind of what we again from a chain abstraction mindset is well how do we unify this how do we make it such that you can use kind of one let's say master account and then log in everywhere how do you able to you

43. 18:31

    know go to any chain and not think about how to send transactions how to pay for fees what feed token it's used Etc and so that's where kind of chain signatur is coming from is it's a technology that allows near accounts to send transactions from another chains and not worry about all of the details of routing payments Etc who's running those MPCs so it's starting in a more uh kind

44. 18:54

    of permissioned way because uh of some of the limitations but the idea is over time we want to add both near stakers and igen lery stakers uh to kind of become validators of that as we can scale the number of notes so now I'm U now I'll I'll talk a tiny bit about orchestration right I talked about orchestration before and that's you know from our perspective or rather our focus

45. 19:17

    on our emphasis is uh you know is programmability of this async activity across multiple chains and that addresses you know both chain abstraction or a big chunk of it of of you you know you've got this connection but now how do I do a use case that spans these things and can roll into it everything from CCP to to socket to router to ig layer all you know and all these different connectivity things to be able to program interesting

46. 19:41

    cross-chain use cases I'm curious you know a you've seen us talk about that and we've been at multiple events you've seen some of the agoric stack but I've not actually gotten a lot of your reaction or feedback or thoughts about it I'm curious if you have thoughts about it that you you know want to share with you know 50 of your closest friends you're like putting me on the spots here

47. 20:02

    oh I can yeah there I am didn't mean to do that yeah yeah no I I I like that the yeah he did ask for spice yeah no I think as as I said I think for different use cases we need different set of tools right so if you do want to execute something and like again if it's a long lived relationship you do need a very different setup like I I mean a Nora is very much saying like everything can be done with intent I don't really subscribe to that either uh and so I do

48. 20:37

    think you know putting a lot of this relationships on chain allowing to have this traceability on this relationships is extremely important and uh allows us to really move forward on on like a bunch of new use cases I think the the questions I'm mostly thinking is how do we actually break out of the like some of the conversations we we have been having right for past like year now but

49. 21:02

    actually how do we get more people to build applications on this like what what is the what is the level of uh because maybe like for context right when we started chain abstraction kind of like trying to get this mindset going and when you go to you know some of the Venture Capital they were like okay this is this is not going to happen right uh like the best answer I've got okay this is five years out right right and they were like you know nobody cares about this right now and so to me that's kind of bigger question is

50. 21:33

    like we have all those pieces they're really all interesting on its own right the challenge is like how do we give examples like a very specific example like for somebody to build an app right that's why like swaps are just like so easy because everybody understands them lending is like good example of like or or borrowing stable coin swaps this swops the swaps version of I have asset over here and I want asset over there

51. 21:57

    and those are on different EOS systems that's the swaps things but it's nightmarishly hard without all of the the stuff that we're building so exactly yeah but I think like we do need some more examples which go beyond that and I think this is where I mean you guys been pushing a lot right and we kind of thinking about this as well uh that's why like I you know I don't know if you saw my talk but I was using examples of like subscriptions right this is something that like long lived relation

52. 22:22

    long relationship yeah that you need to facilitate where this is something that been impossible to do in web 3 for like pretty much forever and it's something that like people want because nobody wants to pay a you know a mill ment every time and sign on the thing I just want to pay like $5 a month and just access a stuff right and so I think this is where we need more of those examples and I think like I mean you guys doing a

53. 22:46

    lot of a good job there um I say my model of fixing problems is usually ship code so we are shipping our orchestration core it's in test net now and it should go out on Main net next week so and then you know calp so we'll demo later and they'll be deploying on top of that so so you know in terms of moving things forward you know you guys are shipping we're shipping you know I think that everybody's shipping not everybody's shiing you're you get a

54. 23:10

    chain and you get a chain uh but but to me we're we're we're at the Tipping Point where suddenly these things are actually possible that we've been talking about for for a year U I agree yeah I think that's that's a transformation that's happening hopefully this summer and so by September we should all have of production use cases and uh instead of talking about it people can just use it we can wrap up and you know all just

55. 23:35

    enjoy a beer that's right that's right the other thing so um at the beginning of the year I went to an investor event this was you know the house of web 3 or something and asked so what what do you know about chain abstraction and about onethird of the the investors there had heard of it it's like okay yeah sort of interesting maybe whatever ever since then ever since February our first chain abstraction day that we did uh every event I'll go and ask a have you heard

56. 23:58

    of chain abstraction not only have they usually heard of it now but they can play back to me a pretty reasonable you know thing about what it is use your Focus use you know stuff on any chain here's what I think of it um multiple funds I was a actually we were both at an event uh uh a couple weeks ago where multiple VCS reported yes they're getting chain abstraction proposal after chain

57. 24:22

    abstraction proposal so it really you know it is very different from where it was even at the beginning of the year are what what's sort of the most interesting thing you've seen in terms of signal around that well I mean I think that event we were both at that was kind of a COR so it was like in San Francisco right you know small event and then two VCS were saying like oh yeah I had three pitches today of chain abstraction and the other VC is like

58. 24:45

    yeah I also had two pictures today of chain abstraction and we're like high five that's right that's right our we curious not done but now we can go do coding yeah yeah so yeah I think that the I think that this being a really good example of kind of building out the the movement and really showcasing what you can do and now yeah like now we need to ship applications that's really the

59. 25:08

    yeah that's why no no now they need to ship app yeah they need to ship applications and uh yeah I think I think overall um I mean the other signal is just like I mean there's people who are just like starting to build something on you know our Tex Tech on other Tex TXS who kind of just thinking that way already right so the mindset is changing and I think that's the that's the

60. 25:30

    important part I think the piece that I still uh think not there is the the way people evaluate ecosystems and the way people think about what success looks like because I do still get like oh but what's a value of cruel those questions are like what's a tvl why is tvl in near so low right it's like well if you assume chain abstraction what does tvl near even mean right is is application that uses chain abstraction that has assets across 50 chains is that tvl on

61. 26:02

    near is that tvl on those chain like those questions toal value accessible is what I want I mean the reality is is like how much money developers made at the end that's what that's what that like if you think of like any platforms at the end it was how much success did people on top of the platform had right and everything else is just like a kind of a proxy so do you measure that I mean do you have a is there an ability to track that yet not

62. 26:27

    really because I mean outside of just like basic fees that swaps are making and and some Landing Protocols are making it's really hard to capture like how much but that that's kind of what like I'm thinking like what is the to Total Market of the applications building on top of again not even near but like the whole ecosystem the whole web 3 and like I we actually bringing value like that's that's a part that I'm

63. 26:52

    mostly like I had this like long tweet thread because I got kind of annoyed um and so like there's been so much wealth creation in crypto but the value creation have definitely not get caught up with that right and and again if you look at like outside world the way it's judging crypto is based on value creation not based on total market cap of all coins right and so kind of we kind of need to fix that and and chain

64. 27:17

    abstraction to me is you know the primary narrative that is about how do we deliver value to users I mean that that's exactly yeah how do we' deliver value but also again it's a mindset change from tvl and you know token launches Etc on any specific chain which I think drove this very weird kind of economic incentives and fact that like you can fund every single chain and like copy the same stuff and it's going to make

65. 27:42

    ton of money but actually starting to think about like how do we create value because again this is like we have users on boarding they're using the Stuff Etc but I think we are over time oh we're over time all right all okay so in that case that means we have like 30 seconds for the how does a I relate to CH chain abstraction not to blockchain but to chain abstraction so there's two things one is uh something for example mbase and now it's called BTE wallet does which is you

66. 28:16

    can literally describe what you want right like I want like the example I like because it's a little bit it's very complicated to execute which is I want to build an index of top 10 tokens that were issued in the last three months and buy $100 worth of it uhuh uhhuh right and so what they do they pretty much have agents that parse this types of transaction they figure out how to execute it and then they use chain

67. 28:41

    signatures to execute it across all the chains where this assets will be right so that's kind of high level like if you think you have all of this complexity now across many many chains well natural language is really good way to kind of uh navigate in this multi-dimensional space um and you want some user interface Etc like it's going to be a mix mixed experience but this is a really good example of like something

68. 29:04

    that if you try to do manually like even if you have chain signatures you still like you know you're going to spend ton of time figuring out how to like build that index you know how much to buy Etc so I think that's like AI is kind of as an interface is really good and again we see this in web 2 we'll see it in web 3 there's a on the other side obviously you know we we moving in a world where we have ai agents paying each other

69. 29:28

    transacting as apis Etc and like AI agents don't really care uh like they want to kind of transact across all chains and be able to find the best you know again it's a Marketplace in some way like what's the best execution across like force specific intent that agent has to be done and so again this is where a lot of this tooling comes in to really facilitate that and uh not need for each agent to have like balances everywhere figure out execution all of the stuff which like is is

70. 29:58

    complicated and should be done by a professional good good I you know people had that question for us chain abstraction gives us a channel to be able to consolidate what we'll bring out to web two developers as they start moving into this space so it's it's sort of critical for all these different uh uh initiatives so thank you all for listening thank you for this conversation next time we should tell people what we're planning to do between

71. 30:19

    aor and and and near so sounds good but not till next time tune in for next time he