Layer 1 Stage | Fireside Chat - Illia Polosukhin & Jordi Baylina at NEARCON 2023

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

1. 0:03

   well thank you everyone who came for the third day of Nan I hope you all energized and did not drink too much yesterday so I want to welcome Jordy here Jordy is a co-founder of polygon he's a true OG of kind of web 3 in ethereum space and uh um it's been kind of on Wednesday we announced partnership with polygon working on Z VM together and Jordy is one of the brains behind kind of the push for zero knowledge at polygon and I wanted to sit down here kind of very grateful for Jordy to come

2. 0:38

   and discuss why zero knowledge technology is so important how we got here and what's coming next but first want to give word to Jordy to introduce himself to and also tell us more how you got into zero knowledge why it excites you well I am I'm I'm an engineer I'm an engineer since I'm 12 years old that's my my my heart what it is um I start working in the crypto space in 2015 so that's a little bit long time

3. 1:08

   long time ago um and uh it was uh very early maybe after one two years and understanding that the technology that uh actually I was trying to use this technology in real applications and I realized that this technology doesn't scale I mean it scale if you Cal lights but if you really want to the cental lights it doesn't scale very well so that was my obsession at that point I mean I was at that point was time of uh State channels and very complex things that they didn't work and uh uh at the

4. 1:43

   same time I was very excited about the the the the zero knowledge especially for the Privacy stuff so I start to study a lot here I I drink a lot from the zash uh people I think that all the zero knowledge comes from from them we have a lot of them a lot of thanks to that to the to this uh to this that Network by the time and uh I I was building I built a couple of uh I mean I was building circum which is a a language for creating zero knowledge uh

5. 2:15

   circuits and snar JS which is a pro that works very good very good in the browser right now is one of the most used tools in the in the space on that and from there we realized that this was very important also for scalability because zero knowledge is not only about privacy it's mainly about scalability actually scalability is the application that makes zero knowledge uh grow that much

6. 2:38

   people care less about privacy unfortunately and uh sadly yeah sadly yeah and uh and uh from there I mean we created the first rollup uh the only payment roll up that was called Hermes and uh after we launch Hermes um well we were start working in the H Grail that was uh this case in this case was a scaling ethereum scaling the evbm uh so the idea was to to prove uh an ethereum transaction in this case that executes uh thing this was a little bit of Holy Grail two years ago because I mean even myself I couldn't believe

7. 3:19

   that this could be possible it was hard to believe it will work because of complexity so yeah so at that time we start uh remember we had like we spent a couple of months with bar white hat from the foundation uh working together trying to okay uh it's impossible but let's try it I mean it's let's try to see what are the points and how we would do each of the pieces and we're like solving all the pieces and at some point

8. 3:43

   says okay looks like more or less all the pieces are there some of them very complex way to do it but it's it's possible so from there we start um working hard it was a a work of two years at that point we joined uh forces with polygon we merge polygon here was very important I mean we get all the funding and all the we could grow the team and and actually put resources on that and two years later we were launching in mayet the polygon zkm and has been a long journey we learn a lot

9. 4:16

   but uh I think this is an important Milestone uh not only for polygon or for ethereum also for for the full uh blockchain ecosystem because actually we see that it's perfect ly uh scalable so scaling blockchains is definitively possible and this is the and this is the way to go this is the clear way to go so maybe like for those who are not like very familiar with how zero knowledge

10. 4:44

    works as well as you know how it actually transforms you know provable Computing and applied to blockchain maybe you can describe kind of like in basic model how that works yeah Zer knowledge is mainly is what we call the best way to understand because there is a privacy part but the best way to understand is what we call the provable computation so here is the the idea how it works imagine that I have a program any program any deterministic program

11. 5:08

    this can be veryify a set of uh transactions but can be anything I mean any you training model yeah anything ex anything machine that a program that takes one week to compute that okay so have an input you know that the program will uh uniquely generate an output because it's a deterministic program and I give you the result to you so you know the program you know the the program and

12. 5:31

    you have the result and you want to verify uh you okay you have the result but are you sure that this result is correct well the obvious way is okay you can verify that I mean you just take the same computer the make super computer you run the program again uh stay for a week and then you see the result is the same okay the computation is valid okay but there is a better way to do that and the the better way is okay I I create I have the program I of course of course I need to spend one week running the

13. 6:01

    program but besides doing the computation I also generate uh um computation proof a z a z proof okay so it's a very small cryptographic like a kind of a signature okay as a cryptographic part and then I give you the result I give you the signature or this this proof this computational proof and the cool thing is that you don't need to run the program again you can verify that the computation is valid in very few milliseconds and the cool thing is that

14. 6:32

    no matter how long is this program no matter how big it can be one week of computation 10 years of computation can be the computation of a full blockchain I mean it can be whatever computation you want that the verification it takes very few millisecond just to verify and this thing that's quite new I mean this uh suen part the zero knowledge the sucin computation part this was uh until

15. 7:01

    2010 or so was not even possible so the research level was at at level but this is what allows blockchains to scale because what what what's the big limitation of the current of the current blockchains of current layer ones is that I mean the consensus is expensive and if you want to reach consensus that means that all the all the notes in the network they need to I mean when somebody propose a note all of them in order to accept this consensus they need to execute that no they need to take the

16. 7:33

    transactions and and check the the the signatures check that they have enough balance if it's a uh VM uh machine I mean they need to execute the the code or the smart contracts the transactions and um if you want and these notes if you want a really decentralized networks these notes cannot be huge notes I mean well you can put big notes and you scale to five machines yeah but but I mean if

17. 7:58

    you want to if if you but this again decentralization I mean if you want to have relatively small notes and you want to scale this is limited so the number of transactions that each note can execute is is is limited so the cool thing here is that the zero knowledge allows you to okay maybe somebody uh gets transactions somebody uh creates um this zero knowledge and then in the consensus layer they don't need the nodes they don't need to re execute all these transaction they just need to verify this proof and

18. 8:31

    agree that this proof is valid and this proof is I tell you it's very few milliseconds actually you can do it in a smart contract um in in in ethereum in near I think there you also implement the pings thing so it's uh and so this is in in the you you you can do that and this is what really allows us to to to to to scale blockchains actually there are two pieces one is this execution

19. 8:56

    this this this technology this ZK uh technology and the other of course is the data availability data availability okay this is cool but you need to put that the data needs to be in the uh on to verify the proof you need the to see the data what the transactions went in and you need everybody else to be able to actually reconstruct the state exactly so thata needs to be available because I could I mean the idea is okay I could

20. 9:19

    get get do I mean get reach a consensus for some valid estate but if you don't know how to this estate was built nobody will be able to build the next state because okay it's a hash route okay but you need you need information yeah yeah you need the information there so an information requirements is that needs to be available so that anybody should be able to to have access to this information this is the data

21. 9:45

    availability and I mean this is looks like a simple problem a sing simple problem but actually it's one of the hardest uh uh problems uh to solve and by the way I mean near is is is is is is leading the space in this data availability near is one of the uh networks with the best shed data the sharding with that availability I mean this is good in in in in in that site so yeah so this is the yeah this is complimentary yeah to the kindes yeah and I think kind of maybe to to expand a

22. 10:22

    little bit on this the idea that any kind of computation anywhere can be proven with a cinct proof right verified by anybody else within milliseconds it also means that some of the things that right now like you can only do with consensus you can now kind of do in like you know on the personal server on the some set of servers and then just provide a proof and so this kind of else unlocks a lot more provable compute

23. 10:50

    which is bringing us to the kind of broader Vision right where people you know are in control uh and you know you like right now if you using any service right they can change their database like there's engineer who can like literally go and run an you know update query on your data and you will not be able to like even detect that right but here you have this probability and

24. 11:12

    that's kind of fundamentally changing both the blockchain space but also broader space and so kind of bringing this to to web assembly um so for us when we started um we actually banned the zero knowledge words in the in our office back when we had an office and we started and this was for a reason that in 2019 it didn't really work it didn't scale it was really slow uh it worked

25. 11:37

    like on smaller things like back then like it was hard to imagine evm would be provable and it would be too easy when we were designing near to say well actually this can be done with zero knowledge proof but we cannot implement it and so like the yes there is a design with zero knowledge proofs but we cannot implement it and so we ended up building the data availability but also shed execution and kind of uh the ex them

26. 12:02

    around that and people who've been at the talk yesterday for Bowen seen kind of the evolution of that but with zero knowledge now working right and especially kind of starting to uh scale to this more generic computation the being able to prove this zero knowledge kind of for web assembly and web assembly is a generic VM it's VM that runs on every device like every device you have right now there is a web asem running on multiple different applications it runs on edge on iot

27. 12:33

    devices it's a generic technology that has kind of very wide adoption being able to prove that will unlock kind of provable compute on the on edge it will provide you know provable compute on your laptop and it can provide provable compute for blockchains and in turn also allow to continue scaling the blockchains and continue scaling the um kind of what we can do with this technology yeah there is um to mention maybe two important applications in the blockchain because one thing is of

28. 13:08

    course the scaling blockchains themselves I mean building call it layer too but I mean the end is this set of data availability and uh the computational proof to to scale but this uh technology also plays an important role in thrust l briging i mean when I mean when you want to prove some a chain in another chain uh you I mean if the other chain are able to verify you can prove the full chain that the full state is in the other chain so this is important for this trustless breaching I I like more the word it's a little bit

29. 13:40

    old but this double pegging just to distinguish uh the bridge of multi6 I mean you know this this work around I mean it's real bridging of blockchains this is what allows you this a trust l i mean you don't need to trust anybody it's just the math and and and then you can I mean you can prove one chain to another chain and one chain to another chain chain so this allows us to connect different chain so uh I mean connecting for

30. 14:05

    example near with ethereum is is is per perfectly for possible with this uh technology and I think I mean that's an important note because indeed right now the bridges are either a multi sake of some parties or what we've had at near and some other blockchains have is a light line verification right where you just verify the block headers but you don't verify that compute was correct and so there is you know an economic

31. 14:30

    attack that's possible where if the value of the bridge is higher than let's say 30% or kind of uh you can start actually doing some interesting things now with your knowledge if you have a full client proof right the proof of all the computation there's no way to actually violate the rules of the blockchain with whatever am out of stake like even if they have 100% of stake they cannot actually violate it right

32. 14:56

    and you need to change the circuit on the other side deci side first to do that and so that is kind of the benefit but it also means that now the security of all the blockchains is getting kind of a line I call kind of chain obstruction because you now have you know the security of all the blockchains is getting more and more unified and there's a lot of like you know details how this is done but now it's less about that and more about the applications and

33. 15:22

    kind of things on top that people benefiting from yeah I mean I think that having the wasm uh ZK wasm I mean this should allow to proof the full near uh chain wherever so this has a a bunch of applications another application that I think is important to mention is for like clients uh like clients I mean right now if you want if you have a mobile phone and you want to verify the chain you either trust some server somewhere or uh you you can have a I mean you have the last block

34. 16:01

    and then you have a proof that the last block is the the one that's correct and this is and and you can't verify that block in few milliseconds right now a light client they need to synchronize a full chain or at least from certain point so this zik wasm is also important I mean uh also for verifying the the the full chain in a in a mobile phone so that's why I mean I so excited on this on this uh for this technology I think that this is for for near is a a step

35. 16:35

    super important step uh for for for for for scaling and we'll see that uh this will change um the space I mean this technology will change the the the space of uh uh how the how the how the layer ones are I mean we were maybe May in a place where there were like a lot of layer ones uh any application any DB or any application was thinking in creating their own layer one and well you know the difficulty that means uh creating

36. 17:10

    layer ones so I believe that's going to be a world where with much less uh layer ones and much more layer twos with these uh um execution environments and of course wasm is is is an important piece uh for these layer TOS yeah I think in general it's it's the boundary between what is layer one and Layer Two is also kind of becoming less clear right because uh consensus I mean when when I when when you talk at least in my in my mindset when when I

37. 17:42

    hear layer one uh I'm I always thinking in the consensus layer uh probably I think it's a better definition that layer one because layer one is more yeah so layer zero and layer zero yeah um so I think yeah I agree that there will be more like even if there is more consensus layers because you can prove each one so like ethereum as well right polygon is working on type one prover which allows to compress the whole ethum chain as well so you will be

38. 18:11

    able to do the same thing with ethum and so you can put that ethereum proof on near you can Pro that ethereum Pro on polygon POS on on tender on like on C on different chains I mean in private chains I mean you you can you can prove the full chain in the and so similarly yeah you can run a private private chain or in our case private Shard and prove that and then like you don't need to reveal the data that and transactions that happening there because for example it's a private chart like for private

39. 18:41

    transactions but you can still verify to everyone that it's correctly executed and it can hold balances for example on public chain for the users on private so like there's new use cases that are kind of because of this unification of security that are opening up as well as just generally kind of expanding what's possible with this right you can have you know single server and have less

40. 19:04

    sensorship resistance you can have you know a like separate consensus Network to and have censorship resistance you can leverage other networks to kind of have Force transaction inclusion so you can have all those different options and uh indeed like kind of this technology really opens this up and and we just seen kind of the beginning of this exactly so I mean and kind of within this topic I think might be interesting

41. 19:30

    to to talk about the the value Network and the kind of the aggregate liquidity and aggregate you know user base right because uh kind of I mentioned this that you know we're really trying kind of a lot of uh work in the ecosystem went into kind of making the experience easier right making the interactions with blockchains more straightforward easy for Consumer and at the same time because of this kind of zero knowledge

42. 19:59

    you know full client proofs aligning uh UNIF unification of security what it means is now the value can smoothly move around right and so you guys kind of introduce this value layer uh nomenclature exactly that the Z I mean this this this the cool thing of this proofs is that once it's generated I mean you don't need to White uh uh it's already proven okay this is compared for

43. 20:24

    example with optimistic ways that I mean you need to wait for this challenging period and so on so if we are able to uh generate to create this proof fast enough that means that the communication between the chains uh become become fast and if you have and this is the idea of the idea of the polygon 2.0 the idea of polygon 2.0 the idea is to have like a different set of networks uh with

44. 20:53

    different maybe different token with different governance some of them more centralized sequencer or the more decentralized sequencer but they share uh single security rules I mean is you do what whatever you want but the rules that manage the the main Bridge or if you want the the the liquidity the liquidity pool you cannot do what you want okay so you need to follow the yeah

45. 21:17

    you cannot like print tokens exactly so that's and and here the the the rules uh the rules are set in the in the in this uh ZK and if you uh I mean if you agree on these rules on on this ZK then the idea is that the transfers the value transfers between the different chains can happen really really fast and this is the little bit the the idea of the of the system so that uh having different networks but single single liquidity but

46. 21:49

    then you can you can prove you can transfer the ownership through this zero through this zero knowledge and this allows us to to scale in a parallel way here we are talking always the scalability goes in two directions one thing is um I mean scaling a single chain this is zero knowledge is important but then if you are able to connect the chains in in in with with fast transactions with this proof then

47. 22:14

    you have like it's not having a processor that's fast it's having like many processors so you have many processors and those processors are fast here it's a new space uh I mean here we open to a new a new a new era in in the in the in in the blockchain space yeah so I mean near actually was designed exactly like this right just without zero knowledge lots of parallel processors trying to interconnect them as fast as possible so you can actually move value between them and then

48. 22:44

    unifying The Experience so users don't need to care about that and so I think now with your knowledge working with polygon there's kind of an opportunity to to broaden what that means right not just for near ecosystem but also polygon ethereum and kind of even broader than that and I think for you know you always kind of I mean you obviously personally been in ethereum EX system very deeply

49. 23:07

    polygon is obviously very uh close in this I think like I'm curious how you think about this kind of ethereum alignment and kind of generally like how do you think this evolving uh from from the more like you know lenss of everybody working together well this is a I think that this I mean this technology allows you to connect this Network so uh I feeling that the dis solation maybe probably right now there is a lot of uh layer ones and they are quite isolated to one each other and and

50. 23:40

    I mean they have their own system their own Community is different difficult to transfer but this I mean these systems allows this connections allow this so here is and at the end I mean we're sharing the same values more or less all of us we come from Bitcoin I mean so it's the the idea of decentralization at some point so uh I think that this uh technology is an opportunity to uh to work together work together again I mean to to to work on that uh of course each

51. 24:11

    one comes from each network is background but uh joining and having like a single like having a single internet having a single chain if you want a set of chains that are constellation of chains that that that connects one each other I think that this dream uh um it's possible thank you to this technology we are very early in the stage I mean this technology is quite new I mean but I think this is the clearly this is the the the way to go yeah it's like connecting you know we have right now South America internet

52. 24:43

    North America internet European internet and we actually starting to connect them and being able to these highways I mean this at the moment we have highway so we have airlines that fly from one play place to the other so this of course will transform the the world or transform the world of uh block CH this Cas so what are you most excited in the next year and kind of the space well I think that this seeing this this this

53. 25:05

    networks uh working together working together connecting being singho scaling for me I mean I'm me I'm in the space for for the people to use it uh as an engineer for me makes no sense to build things that that that's that useless so uh my dream is to to start seeing applications start seeing many applications I think that the the possibilities are huge and we haven't we have not started I mean the technology uh is not ready yet at least

54. 25:35

    for most of the applications that needs to come so uh I hope that in the coming uh months and years uh we start seeing this expansion and this uh I mean especially in the number of users and users that take profit of this uh technology that's what I'm more excited about yeah agree I mean that's you know our Focus have been on like how do we bring more consumer applications which

55. 25:59

    can benefit from this technology which can leverage it and at the same time you know continue investing and innovating on the underlying layer such that the there's more and more scale but also you know versatility of the uh infrastructure and functionality and so I think this partnership allows us to do a lot of cool things it kind of allows to scale near blockchain itself it allows to connect to other blockchains

56. 26:25

    it allows to also push things in more wider range of provable compute across uh kind of the whole not just web 3 but uh broader technology space at the same time it allows to kind of start to work together in in a more I would say kind of friendly way in in such sense that we're actually not you know trying to build our own uh kind of small tribes which is which always felt weird to me when we say like I mean especially we we

57. 26:54

    frame everything as open web and it's really weird when when people are trying to close and like okay this is our ecosystem right it's like well it's kind of we have a lot of indeed like shared value we have a lot of uh things that we're working together on uh we've been working with a lot of other blockchains from kind of early on at least talking and making sure that uh the ideas are shared the kind of um the technology is you know moving along for everyone and

58. 27:19

    we are able to actually innovate here and so yeah I'm really excited I uh really appreciate you coming here and uh I hope this was a very informative exciting and we'll see a lot more of uh kind of innovation coming from this technology stack thank you