Zero Knowledge Proofs 101 / Daniel Lumi

Transcript

good morning guys glad to see so many people in the crowd let's begin another day and it will be full of conversations about zero knowledge at the first half of the day a small announcement we have live stream right now happening so if you have communities friends who are not here with us in Prague please encourage them to go to live.ithproc.com and they can watch and if they get a token so they can connect their wallet and have a question asked and then participate in quadratic quoting it's a bit complicated thing to get the token so if you if you need this token please reach me and I will explain how to get it so and thank you for

showing up so early we will now have a very cool conversation about the zero knowledge protocols for damage right for people who really do not understand how it works and give around the uploads to Daniel Lumi who is going to explain this topic to you thank you [Applause] good morning everyone how you doing awake yet I'm not um so first of all welcome to my speech on zero knowledge proofs 101 uh what is Zeke and why should you care

a little bit about me my name is Daniel Lumi I'm a research and consultant specializing in ZK l2's privacy endows uh today we'll be talking about um why should you care about ZK proofs uh what is a ZK proof how do they actually work in applications for ZK proofs and so why this talk right ZK is essentially a PhD level mathematics it's like a PhD on a PhD right and for many years it was a small select group of us math nerds that really love ZK and no one really paid attention and with the inventor of zkevm all of a sudden everyone's talking about zika is he get this CK that but no one really knows what it does so this talk is actually aimed for not just dummies but regular

non-mathematicians to understand ZK so maybe you start you go to one of my friends one of the cryptographers and they you ask them what is ZK and they essentially start talking about polynomial commitments or fry-based recursion right or maybe you get a little bit more into it and you start doing some research online you bump into aerial speech on Hassan's theorem he tends to forget that most people don't have phds in mathematics or maybe just follow some shitty YouTuber and he says you can make money on ZK please don't sue me big boy so why should you care about zks it's a new field of math that's very exciting come on

but you might say okay I like math but who studies it for fun or you might just say [ __ ] math I don't want to learn that zikas are very cool okay stay stay tuned hopefully I'll get you excited for them and just to get you a little bit of context for this vitalik recently said on uh Bank Bankers uh that he thinks DK snarks are as big of a deal as blockchain and I actually disagree with him it's a new field of mathematics I think it's much bigger than blockchain oh I disagreed with metallic what come at me buff bro so algebra in the beginning right a couple thousand years ago when we invented what's considered the modern form of algebra

um there were some Farmers that thought great I can Farm carrots a bit more efficiently or I can calculate the largest common denominator of a square to figure out what size the tile on the floor should be but thousands of years later algebra is literally everything we do right we couldn't have gone to space we couldn't have manufacturing we couldn't build buildings the way we do we would not have definitely blockchain or internet right and proofs actually have existed for a long time right you learned this in school but previously uh proves to prove any single thing you have to have this massive book of calculations just proving that that conjecture is true around 2016 to 2018 we invented the

first form of what's called an uninteractive generalized ZK proofs right it's a new field of math when were you last around when they invented a new field of math come on that's very exciting right and so today we're the farmer and we think great we can make ethereum a bit faster or great we can log into a website without showing a password but thousands of years from now they'll be talking about the moment that we first invented non-generalized sorry interactive generalized ZK proofs so what is actual ZK proof and so in today's world for us to show that anyone did a calculation or execution correctly essentially you need

to prove you need to redo the execution to check that they did it correctly right so now in those grocery receipt you need to list cucumber its price ice lemon it's price and then eventually you get to this total right in the Xeon Lodge Paradigm you don't need to show any of your execution right you can just see show the total and people know for a fact that you calculated it correctly right so ZK is a new um field of cryptographic proofs that essentially one party the prover can show to another party the verifier that they did calculations or execution correctly without actually showing them how they did it

so how do ZK proofs actually work and here's some examples the most common example is where's Waldo this French game you try to find Waldo in there and let's say that you found Waldo you have your friend there he also wants to find Waldo but you're not going to give them the answer he has to suffer on his own right so how do you actually show to him that you found Waldo try to find Waldo he's there it took me a really long time to find Waldo for this example I sat there for like an hour oops hold on I think I went too far yeah so how do you actually show to him that you found him right you can cut out the

picture of Waldo and then you can send it to your friend the verifier and he can check you found Waldo but I actually don't like this example it's the classical one it's kind of confusing uh people don't seem to really get it right so the more accurate uh example Sudoku and this is very accurate because that's actually what plunkish systems do underneath essentially Sudoku it's not quite the same but essentially it's the same construct right so again you have your friend and you you're both trying to complete this Sudoku puzzle you solved it you want to show to him that you solved it but [ __ ] that guy has to do it on his own right so how do you show it to him essentially you cut each round column you put those in a list maybe you Shuffle those numbers around a bit

you flip them upside down and then you that's what you actually send to the verifier is all the rows upside down right and the verifier can start flipping random rows and see okay he has one two three four five six seven eight nine and he didn't cheat right and the verified doesn't actually have to flip every single row um it's probabilistic mathematics if you randomly flip it the chance of them cheating is extremely low right and you might ask okay great that sounds good but how do you actually create a proof and that's something my girlfriend said she was sitting there scratching her head like Sudoku honestly it's a bit hard to explain without going deep into math right the way that you can think about this is

like it's hard to understand how computers and circuit boards work with zeros and ones but essentially you have zeros and ones those go through logic gates those logic gates on a circuit and Magic picture appears on your computer right zks are kind of like a virtual circuit board it's not actually zeros and ones that goes through there but essentially you have logic gates those logic gates are in the circuit and you can show that the way they use shuffled or cut the Sudoku Rose is is correct and you didn't cheat so we can zoom a bit more into the ZK proof so this example is let's say that you want to show that you have a valid passport you're over a certain age but you don't want to reveal some of the

other data for example I don't know your date of birth or sorry your expiry of the document right so what you can do is you can hide some of the inputs in your passport assuming it's a digital passport then the public parts are your name or the the date right and then we have something called witness data so you essentially you take all of that together you do some calculations on it that's the witness data right and here essentially you stick into the prover the private inputs the public inputs and then you do some uh you have a relational circuit and maybe you get a digital certificate from your government showing this is a real passport stick that into a prover it

creates a proof and then the verifier can just check true or false like is it a real passport or do these conditions match and if that didn't make a lot of sense that's okay because my hope is that ZK will become kind of like dos right and the explanation behind that is before we had Mac OS or Windows for every single different type of computer architecture that you had uh different type of CPU different type of GPU you needed a new custom operating system right and then Microsoft stole um invented does and now it works General as a general basis on any computer with any architecture right and with some things like ZK evm or circum

or Noir which are languages like a programmer does not actually need to know how the math underneath Works they can still write zero knowledge programs and this is a common question that I get what is the difference between snark and Stark and honestly it doesn't really matter for you uh the reality is um snorics are much cheaper to verify on the theorem they're harder to prove but they're cheaper to verify and that's a good thing because transactions on ethereum are very expensive right Starks are more expensive to verify but they're cheaper to prove and the important thing that you really need to know is snarks use something called elliptic curves those are not Quantum resistant or

quantum computer resistant Starks essentially use Collision resistant hashes which at least for now we've not figured out a way to like essentially break with quantum computers historic star scalable transparent arguments of knowledge sonar exercise to sync non-interactive arguments of knowledge so here's some applications for ZK and how it'll actually change the world so the first and most important thing that you really need to know about ZK in general is it's verifiable computation right and here's just kind of explanation on that so here's how current blockchains work like ethereum or Bitcoin or some of the shittier ones

um every operator in that Network needs to re-execute every single transaction that comes in right if they don't we can't form consensus around what the truth of the blockchain is right so let's say you have a network with 3000 nodes each node takes five seconds to process a block uh that means there's around the world there's computers doing four hours re-executing the same information over and over and over which is honestly just wasteful right ZK is a new paradigm in Computing where essentially no you don't need to re-execute it you can just verify that the one person did their job they cannot lie so if you check the proof you know for a fact that they did the computation

correctly without consensus right so you could actually have Bitcoin instead of having 16 000 nodes re-executing over and over again and playing this really power hungry game uh you could have one node that verifies and you get the same level of security right maybe you want more than one maybe you could have ten or more in case because if the one fails essentially there's aliveness failure there's not a security failure but aliveness failure and everyone else on the network just verifies they did it they did it okay fine so the most common use that people on blockchain know for ZK scaling right and here's kind of how current blockchain works right the more

transactions you get into a block each of these white dots is another transaction that you linearly have more work to do to process that that block right but the problem is we we want like a MacBook to be able to run this network so it can be decentralized right so ethereum and Bitcoin propose or enforce something called a block limit right and ethereum is around 30 million gas in this way in every block you can make sure that a specific type of computer or a cheap enough computer can actually execute that block in zika it's a bit different right the one prover the more transactions you get is scales quasi-linear it's close to linear it's actually a bit more but

essentially you can have one super computer or just a normal server doing blocks right and executing all the transactions and then every single other node just needs to verify those transactions so you still stay under that limit maybe you can have a mobile phone actually verifying this network right and that difference between the verification and the proving is essentially the scalability limit the benefit of ZK for scaling right and for ethereum to become the the back end of the internet of value we at least need more than what visa processes right and Visa processes 20 to 60 000 transactions per second

so here's sort of how much ZK can scale ethereum for example and this is this is a bit exaggerated it's actually uh take into consideration a lot of optimizations in the future but essentially today ethereum process is 15 to 30 transactions per second with uh one zkl2 you essentially get a 10 to 100 x scale on top of it right maybe you get 3 000 transactions after we have uh something called Proto dank sharding which is ethereum's plan for essentially providing more data in each block you get another 10 to 100 x scalability so now you have like 3 000 transactions per second right and with full tank shorting we'll get actually probably around 10 to

100 x more right and the thing to note is that's for 1l2 right we also have arbitrim and optimism ZK sync and scroll and all of these other things right and then there's this idea of fractal scaling right or for uh optimism they call it hyper chain right and the idea is we can keep building layers on top of that and with each layer you get another order of magnitude more scalability right and that's like at that point with a couple of l2s and some l3s we're already over 60 000 transactions per second and the important thing to note about this is unlike some blockchains Solana

um some blockchains this is sustainable scaling right um you only need that one prover to actually execute those transactions and even if the proofer goes down it's it's not a security incident it's a liveness failure right and with some optimizations like data and transaction compression I was making the provers more efficient will actually scale the blockchain even more and I do have a speech coming out probably in Paris for scaling youth supposed to have quotes the infinitely but it didn't fit on the slide um how do we essentially scale eat infinitely without having issues with the ux across all these l3s and without having liquidity fragmentation

uh another example for ZK is for the purpose of interoperability right so example this is ZK like climb Bridge right and essentially it provides you like zkibc it provides you minimal trust assumptions outside of the consensus of the two blockchains right and it essentially allows you slightly safer bridging between cross chain applications an important thing to note on this though is it's still not secure essentially you need to verify what state the state of ethereum on Solana for example for it to be a secure transaction and to do that to make sure that someone didn't just fork ethereum and say this is the truth you need to verify every single validator signature which is like 400 000 signatures good

luck submitting that onto Solana right so people take shortcuts and or have multi-cigs or other things testifying which is the correct version of ethereum uh in the fractal landscape we also have something called hyper Bridges and the idea of every one of those l2s l3s l4s use exactly the same circuit for bridging so now you actually have as long as the L2 bridge is secure all of those other bridges are completely secure without having any additional trust assumptions so it's much safer bridging and the nice thing is like for example on ZK saying there's one common Bridge so if you want to get from like oops somewhere up here to over here you don't need to bridge it all the way down

you can essentially do some math math like accounting and just change the location down here where the assets are so it just appears from here to here uh important part of blockchains one of the main reasons why we have ethereum is called Atomic composability the idea is you can one contract can trigger another contract in the same transaction right and with cross-chain world of having Salon or ethereum we can't really have Atomic composability right there's always going to be additional trust assumptions with ZK in the L2 L3 world we can essentially have from our arbitrim to ZK sync a sort of pseudo-atomic composability where you

bridge through L1 and you can trigger a transaction asynchronously with something called shared sequencers which I'm working on with projects right now maybe in ZK sync you could essentially have Atomic composability between all of the l3s so you could trigger one contract from one uh L3 directly on another one in the same block I do have a whole speech on the next generation of safer cross-chain Bridges and how all bridges still suck which is actually surprisingly important for the community because they don't understand the security assumptions they're taking um ZK is also very useful for identity So currently for signing up for any

application random website random porn site like any app you have to provide all of your information or at least your email address a password and this isn't great because if any one of those websites or applications gets hacked or sells your information your docs not even in like the anonymous way just like this affects everyone in the world and currently the way that you prove for example you're over 18 is you send them the users 32 right and this isn't great because again any one of those gets hacked or they just choose to sell the information people know this person is 32.

with ZK essentially we can create a proof of a digital um like driver's license showing just the users over 18. and still now this website will still store the fact that they're over 18 but you could be a hundred you could be 72 so it's a lot less useful information same thing for a country maybe you have to show that you're a Finnish citizen so you can use that exchange with ZK we can essentially just show that this person is not from a sanctioned country again less useful information there's a lot of countries in the world or maybe you go to a board day party right now the way that you do it is you actually show which board ape nft you have which isn't great because maybe you

bought that on the wrong wallet and now everyone sees you all of your financial history with ZK you can just show a proof of membership that you have an nft which belongs to a set and now you can enter the party and you can choose whether you want to actually give that to your friends or not which one you have and like all of these identity things kind of lead to this example of one kyc one identity to rule them all right and currently like if you sign up for coinbase you sign up for a bank like you sign up as an Uber driver you have to provide them all of your information your your tax documents your passport like literally like a 3D facial scan and again that's really bad if any of them

decide to sell it or get hacked with ZK maybe your bank or your government could attest to the fact that this is a legal person they pay their taxes and then you just send the proof to coinbase or binance or to buy a house or maybe you could have a tornado cash where you actually have privacy and you could actually show the fact that you're not a hacker or money launderer right and that brings us to privacy right important note here is ZK does not necessend yeah ZK does not essentially mean privacy it can be used for privacy but most of these implementations like Star square or polygons are not actually private blockchains right and the in

this case think of ck0 launch as the fact that it's verifiable computation you don't have to show all of your work you can just see that the state is correct but it can be used for privacy right so an example of this again if you sign up for any website most of us don't change passwords for every website you should you really should but most of us don't right so if any one of those websites gets hacked they can take your email and password and literally like log into all of your other services and the reality is 81 of all hacking in incidents is this they hack the random website you signed up for 15 years ago and now they can get into your bank and everything else so this is the way we currently do it

right you send your password to the website server uh if they're dumb they store it in plain text happens more than you'd think if they're a little bit smarter they store it as a hash right if it's plain text anyone can hack their database they get your password if it's a hash they can essentially hack into the server and just wait as people log in and they collect everyone's passwords with ZK essentially they just store like a user ID and essentially an output of a CK proof and you can essentially sign on your own computer your email and password combination just send the proof to the server and the server never knows your password or email or maybe if you want their like if they want your email you can send that but

not the password um it can also be used as a privacy blockchain right and essentially privacy blockchain unlike ethereum or Bitcoin you can't really view the balances or the transactions of other people it does this by obfuscating either the amounts that you're sending who you're sending to and it kind of depends on the implementation right and example of this is as technique right they actually they close their current version they're coming out with a new version and there are other ways to do private blockchains such as like fully homomorphic encryption or Intel sgx or even an MPC but at the moment zika is much better in in the future maybe fhe in combination with ZK will be the solution

and this is again how public blockchains work so an ethereum or Bitcoin you have to send your transaction who you're sending it to your amount and every single node on that Network processes it so it's public right in a private blockchain you can essentially create a proof on your computer showing that you're not double spending you have enough money to spare send and then every single note on that Network just verifies your proof instead of actually executing anything and I do have actually this is a speech I'll be doing today at like 2 45 in the crypto Anarchy stage but you can also watch it later uh privacy and web3 uh why should you care and how do we improve privacy protocols to prevent hackers and criminals from using them

uh and the last sort of example or set of examples is just understanding our world in general right and every time we discover a new field of mathematics we start understanding the world around us better right we start understanding maybe the sun doesn't rotate around us or how a shell is formed or how to make medicine right and they're essential to the to the development of the world everything we've come up with any sort of modern technology we have this completely based on mathematics and a silly example of this is we could not make a pencil without knowing algebra and chemistry and a bit of physics because that led took hundreds of years to develop through chemical

compounds the wood is a very specific formulation of wood the rubber is incredibly complex chemical products right and the idea is like it doesn't just apply to blockchain it applies to everything in the world right like maybe four cameras in this modern day where there's deep fakes right and maybe Trump grabs another [ __ ] and you can't heal you know fake news deep fake right you can have a camera with a trusted sensor a testing to where it was taken the time was taken right but that's not the image we're going to put on because that's going to probably be a raw image maybe you need to scale that down or crop it or put a filter over it you can create a ZK proof of any of these transaction transitions proving that this is still

the original image and this actually happened right or maybe you could have some sort of proof of storage and clouds I won't go into that one but it's actually quite an exciting field or maybe we could for speed up computers by making like a cache layer so the CP or GPU never have to re-execute the same transaction or same calculation over and over or machine learning models like maybe Google doesn't have to trust their data center providers to actually run the machine learning model correctly they can actually just prove that they ran it correctly and they're not just being lazy and lying or we could prove that data for a machine learning model is not copyrighted material right um the last two examples are a bit more

theoretical uh there was a paper essentially where cryptography might actually help us understand black holes rather than just normal mathematics like actual cryptography maybe it makes us understand black holes right or maybe makes us understand quantum entanglement the point of this speech is math is very cool like yay no just kidding um whether you like math or not Zike will be around it'll change the world around you so it's worth paying attention to uh thanks for caring about ZK hopefully he did not fall asleep like most people explaining zika and my again you can follow me on Twitter at ZK Lumi

[Applause] I think we have time for questions as well I'm not actually sure how much time we have ah 10 minutes cool any questions for the session Daniel awesome talk thanks so much you helped me a lot to understand the space better how far are we away from the doors what's your gut feeling yeah so when do we see browser Integrations OS Integrations when when this is more than blockchain so

cryptography is always about an incentive model right so in blockchain we have a lot of incentives to use zero knowledge proofs because it provides us privacy your skills something that really can't be scaled currently right so there are some other applications honestly it's a community push right like chat GPT is already in trouble or some of these like AIS are in trouble for making a fake version of Drake's song right and he's pissed he's probably going to sue them so maybe they can create a ZK proof of the source of each query to show okay we didn't train it on Drake's songs it's a derivative work it's a legal version it'll be some time before it becomes mainstream adopted that being said it's it's progressed the incredible amount in the last few years

because of the billions of dollars that are in blockchain it really provides really good incentive to develop the Tooling in the math Primitives behind ZK so adaptation can happen a lot faster now than it could have a year ago two years ago thank you anyone else who would like to ask a question and I would also like to say you can also do the questions on live.eth prague.com so if you're shy asking here you can just post your question and I will speak it for you hi there uh thanks for the talk great turnout at 10 AM Saturday morning um I was wondering if you know a little bit like if you can tell a little bit

more about the history how ZK world and blockchain World actually partnered because this is really interesting that the concept has been already for some time in the space um cryptography is even older and it just like applied for ethereum World um so how did that actually happen if you can answer so the gay proofs have actually been the concept has been studied since the 80s right or late 70s and the problem with the versions back then was the fact that it had to be interactive so the prover had to be online when the verifier was checking that this computation was done correctly so that's not great because maybe they

did the proof 10 years ago right um in around 2016 to 2018 Eli Ben sassen essentially invented the first non-interactive version that's generalized and he was looking around he was thinking okay where can I get funding to actually develop it where does it make sense to develop it and he thought of blockchain right you can add a you can make a private blockchain right you can have public State that's private which is just mind-blowing that's when I started getting into ZK right and he developed zcash so Z cash was the first blockchain application of ZK thank you for a great talk this is fantastic so I tried some of these

things like specifically you know zcash the private transactions back in 2018 it would like take five minutes to do the proof your computer would heat up the fans would go on same thing now even five years later you know when you use like the railgun guys you know railgun day you can wrap your coins into SDK space and do your unit swap changes in a folded space but it's still very unreliable when you actually try to generate a proof on your desktop again it heats up fans come on takes three minutes or whatever it takes extremely long and by the time you produce the proof uh sometimes like there's an expiry through for the relayers or whatever it's it's it's to me it still

feels very experiment experimental what do you think is the current state of the maturity of the technology with respect to a practical day-to-day use case yes so there's a couple things on that first of all if you're using a zkevm you never have to worry about generating a proof right verifying a proof can be done on a mobile phone um if you use a privacy protocol like Aztec or zcash you do have to generate a proof right um right now like like Aztec for on my MacBook and I do have a relatively high-end MacBook took a couple seconds to generate that's not most people in the world though right but we're making incredible advancements with the billions of dollars that are behind this and it's getting faster and faster to

prove like some plunky three is very exciting or there's a lot of new um hyper blanket there's too many plugs uh they're getting a lot faster to prove and we're sticking stupid amounts of money into it and it'll get a lot cheaper to prove and then at the same time Hardware like consumer Hardware devices keep getting more and more powerful right so maybe today it requires a Macbook maybe in five years you can have a shitty Dell that does the same thing okay we have one question which you have partially also answered here where would you see the ZK getting real adoption in

businesses and industries any Trend example that are your favorite that you would like to share so I like um blockchain is the short answer the longer answer is um uh the CEO of uh Onyx which is JPMorgan Chase's blockchain Department uh said that if we don't have privacy in a blockchain no institution or bank will ever use blockchain right these companies have Trade Secrets they have customer data they have all these processes that they don't want to make public to people because this is how they are um they don't really believe in open source uh and so honestly like that that would be my first sort of mainstream adoptation of things for like like

um essentially tracking the supply chain it gets much more efficient in blockchain and like current applications and that's like a 10x Improvement rather than a 1.1 x Improvement like metallic said um and I think that's kind of institutional adoptation will be the first uh major use of ZK and then the other thing is like again Twitter showing that they actually use the algorithm that they posted rather than posting a random algorithm and then using something else behind with yeah thanks for a great talk I think you have a bit less people than vitalik yesterday so we are building the Eco Village one

before yeah true uh we're building a security layer and uh so you mentioned the the Privacy uh like the application of decay in privacy space but like what do you think about the Security application so I thought you have like the uh Decay proof of the machine learning so which is the applicable but so what other cases you see uh for the in the security space for example the security space is actually I worked with uh two guys um FedEx at East Denver we made something called ZK proof of exploit right and essentially prove that you can hack a contract without actually showing how you did it and that's actually a major security implication because currently teams have so many bug Bounty

submissions they have hundreds of them what are they going to take their main developers at checking every single bug Bounty and if they do like they check them very quickly and maybe they missed something that's what happened with Wormhole that was actually a bug that had been submitted to them before and then they just like looked at it very quickly they're like that's not critical uh with ZK proof of exploit you could essentially uh verify oh [ __ ] this guy can steal all of our money I guess we should probably work on this right that's one example of security applications yeah thanks I think that's a very cool example as well you can create exploit projects and not tell how you are exploiting them on the D5 track

uh thanks a lot for your speech I wanted to ask you if you could give your personal opinion about uh whether Starks are a safer bet for the future than snarks because of quantum resilience rather than verification speed yeah I mean preferably we'd use Starks uh they're quite expensive to essentially verify in a blockchain I asked that question a lot from teams of how easy is it for you to switch circuits and the reality is first of all quantum computers while they are coming it's a bit of a meme when you get a quantum computer we have quantum computers today right but the process of actually getting them more and more powerful takes a very long time and for them to be powerful enough to even break

a single snark in not years or decades is still far away right and but maybe China is working on something in the background we don't know um that they are I mean the U.S is who everyone is um I do think that purely from a point of view of building for the future it is better to use the arcs that being said Bitcoin is not Quantum resistant right ethereum can be but it's not yet right literally every single bit of cryptography Google cloud like your password all of them use quantum uh like algorithms that are breakable by quantum computers right so we have a lot bigger problems at that point too I don't

disagree that blockchain would maybe be make the most sense financially to attack first but honestly I think it's going to be a very long time before we actually get to the point that we really need to be scared of that but I do encourage people like investing more in Starks doing some of the advancements in stocks that we can do with elliptic curves and Starks and just kind of progressing that field to the point that it can use mainstream adoptation in blockchains thank you thank you so just one last question otherwise I will move on we'll yeah let's go there thank you for the talk

um you have been very brief about CK Bridges and sometimes it's sold as a Panacea that we're not seeing any any Bridge attack anymore in the future if we're going to CK and connect some we're definitely going to see more of them so uh basically um if we're having layer one blockchains connected through CK Bridges uh do we basically solve the bridging problem or do we really have to build everything on top of ethereum for that one I do recommend to see the Next Generation Bridge one and why all bridges suck I go into much more detail on it but the idea is okay ZK can prove the state of ethereum in a very small verifiable

computation so now you can submit that on to Solana right and you can check okay this guy really did lock up some funds in this bridge let's give it to them the problem is you don't know whether that's the real ethereum blockchains are have four Choice rules they Fork all the time and um someone could literally just come and Fork ethereum on their own computer like just have one node and say like see I have a million see um and then Solana would release all of its funds or the bridge on Solana would release all of its fun so the problem is you need to verify the canonical fork or canonical state of which ethereum is this and for that you need essentially all the validator signatures to do it safely and that's not going to be

possible right so you still essentially like the zika party sound that works great the problem is for testing to the right State you need either all the attestations or they most of these ZK Bridges use like a small committee so it becomes a multi-sig bridge again but uh you could basically put the risk on the liquidity providers uh on the on the each side and basically uh have through the feeding fee mechanism cover for possible uh wrong choices on the other side right I mean that's that's true and I much prefer liquidity provider Bridges to wrap the asset Bridges the problem is like maybe you can send the native

version of usdc from ethereum to the native version of usdc on Solana you'll never be able to send ethereum to Solana right that's always has to be a wrapped asset because ethereum doesn't exist on Solana right so then you have these assets you think you're holding ethereum there someone just quickly Forks ethereum because maybe now there's billions of ethereum on Solana and they just ran away with billions right so wrapped assets are still going to be an issue and um even liquidity providers like I would not provide liquidity to these kind of bridges thank you thank you Daniel we'll move to the next session now

yes thank you [Applause] [Music] foreign