Why ZK Proving Is The New BTC Mining with Tim Carstens & Jeremy Bruestle

Transcript

so one person proves it everyone can verify it The verification is cheap You don't need to you know um recmp compute the same thing on every validator right So we're going to be building this really simple experience for folks who really do just want to grab a binary run it on their Mac and get started Just generate proofs right You're sitting in class you're taking notes You don't need that GPU Let it cook How exactly do you expect this adoption kind of flywheel to to take place with ZK mining Maybe I wasn't there in 2013 to mine some Bitcoin but I am here I have a MacBook I'm ready to dedicate some resources to mining some ZK proofs and get involved in this news in this new gold rush Welcome back to the rollup your home for

technical talks and open conversations that shape the modular future Nothing on today's show is to be construed as financial advice You can see our full disclosures list at the rollup.co/ventures What's going on guys Welcome back to the rollup Today's episode is a deep dive into ZK with our friends Tim and Jeremy from Risk Zero and Boundless Uh today's a special day We're going to be going deep into some of the recent white paper release as well as learning about the future of proof of verifiable work CKVMs and more Jeremy welcome back It's good to have you on and Tim first time So uh welcome on Tim Welcome to the roll Thank you Andy It's happy uh it's good time to be here brother

What is that uh that is uh behind your head there that's that's moving Is is that a GPU or something running in the background Yeah So you're probably referring to where is it It's right this this little guy here right there Yeah So so that's my cadler That's uh it's a live ticker It's uh reporting the current stats for the boundless test net Uh orders going through it cumulative cycles Uh of course we got the Wagi banner up top and bottom scrolling left and right Um but like I said I call it my tattler because every morning I walk in here and I expect to see those numbers up from last night and when they're not you find me in Slack asking why isn't the number going up So there it is It's dutifully doing its job I

think it was saying 4.3 trillion cycles on the current version of test net Not half bad We like big numbers around here So trillions sounds big enough for a test net Um Tim it's your first time on Uh we've had uh several uh conversation with Jeremy It's always a pleasure Maybe you could start by introducing yourself and how you ended up finding not only ZK technology but also Boundless and the Risk Zero team Yeah So that's a that's a long story I'll do my best to keep it tight Uh so my name is Tim Carson principal software engineer here at Risk Zero Uh Jeremy and I go way back um into the uh early Seattle hack scene actually

is how I know Jeremy Uh so ultimately the way that I know about ZK if you really pull on the threads goes back to a thing called the buffer overflow Um but anyway uh several years ago uh Jeremy Frank uh Brian they were starting up this company and Jeremy hits me up asked if I've been paying attention to what's been going on in the realm of ZK Uh I have a background in cryptography Um but I hadn't been following it Uh and then the brother weighs it on me that he's built general purpose ZKVM that can just run software written in Risk 5 Uh sounded unrealistic but I went ahead and I dove in and as it happens well at that point

the previous five now I'd say the previous eight years has just been uh ridiculously good for ZK Anyway sell them here Absolutely And now um you guys are commemorating this uh the this Boundless um release with the release of the white paper Um we're going to go deep into everything that's happening on Boundless I think the overarching theme here is the modern era of compute which comes along with the financialization of compute This is one of the things that has been financialized over the past few years and this goes along with kind of the the transition into cloud and then

ultimately out of cloud into a new modern era And so ZK and verifiable compute is an important piece of that And then the financialization of this being able to request trade all of these different forms of financialization um leads to a a more efficient market for compute So I think broad strokes here you know maybe we can tell the story of how how boundless was conceived from risk zero the ZKVM from Zeth right the first ZK EVM equivalent and ultimately into Boundless to solve some of the problems that you guys saw as you were building this out Um kind of like trace out the problem statement for us

generally Why is compute being financialized and why is that important Yeah So I'll I'll I'll do a quick version and maybe Tim can add some things in here So I think you know when we built the ZKVM we knew that this would have a huge impact on sort of scalability and privacy implications on on blockchain But but more generally it really is an alternative mechanism for verifiable compute right Like fundamentally what is a blockchain doing It's running a computation publicly with a huge amount of you know redundancy so that you can be sure that it's run correctly even without having to trust any third party right when I run compute on AWS I assume

AWS does it right and if they lied to me I would never know because uh I can't double check what they're doing blockchain in many ways there's many many impacts of blockchain but one of those impacts is that it created for the first time sort of publicly available able verifiable computing Um and you know really in some ways ZK is an alternative mechanism to do verifiable computing one that doesn't have the same issue of redundant recomputation of the everything So one person proves it everyone can verify it The verification is cheap You don't need to you know um recomputee the same thing on every validator right So in and at this point now is ZK is actually less expensive of

a way to do verifiable computing uh than consensus right uh so I think that was a huge uh unlock but you know what we were seeing is just in terms of the when you have something like this capacity there's a question of how do you how does a user actually make use of this right like how do how do I actually deploy an application either onchain or offchain that makes use of this verifiable computing capability I mean one option is I can go run my own server and you know do the the the ZK computing myself and then send the proofs to people to verify But for a lot of use cases that doesn't make sense right Um and so one questions is how do we

actually sort of bring this to the next level Boundless is really sort of this universal ZK protocol that allows uh for everyone to be able to tap into this new mechanism for verifiable compute on all of the different chains right So I can I can I can ask for a proof um from you know ETH and get the proof delivered there I can ask on you know Salana right Like you really try to make this mechanism where we brought this new capability everywhere Um and then I think in terms of you know how this relates to financialization right like we're basically building a new market um for this new mechanism for computing and it has a lot of interesting properties

and I think you know Tim can can go into some some more of them but like I think the biggest one for me is really in some sense for the first time you can you can talk about like a spot market for computing that makes sense right if you think about a normal spot market you have to have some level of fungeability but The reality is it matters whether AWS did the computation or you know um Azure did the computation because I have to trust those parties and so my level of trust relates to the on the other hand it it doesn't matter who did the ZKP compute it's self-verifying so it actually brings new capacities to the ability to sort of make a commod something something fungeible which makes it more commodifiable which makes

you be able to build spot markets and so a lot of what the paper's really about is how this uh uh sort of innovation impacts the way that people use compute verifiable comput particular Yeah Tim anything to add there Yeah like uh I think that Jeremy did a fantastic job explaining it Um you know my own my own kind of path to Boundless I think it's it's very much the same kind of story Um but I'll put a little bit of a different spin on it Um if you look at the room I'm in I think it's pretty evident that I really like computers like a hell of a lot Um I got this machine over here Its primary job is you know media editing and you know maybe a little Factorio here and there

Um right now though it's functioning as a space heater It's on uh but it's not doing anything other than keeping me like cozy Um and you know like that's that's chill and all but like how to put this A little bit more than 10 years ago my friends and I were hanging out on Bitcoin Talk forum Nobody knew where any of this stuff was headed and it was just a bunch of computer science really geeky people a few interesting very philosophical people really small scene Everything seemed possible And one of the things that was possible back then is we could warm our homes by mining which my friends and I did And

that was some of the most fun I've had on the internet um before or after And as I mentioned about how I know Jeremy I've had a lot of fun on the internet Uh and so one of the things that you know has really inspired it's part of why I joined Risero and it's absolutely a big driver in why we built Boundless the way that we did is we're really trying to in some sense get back to our roots uh get back to this idea that Satoshi had envisioned where it is possible for anyone who's you know online to be a part of these networks in a way that they choose right not in a way that someone else has gatekept from them or anything else like that Um there is a real freedom in this Those freedoms were

enshrined in his design Uh they weren't just a terms of service agreement or something And I think that a lot of this field ever since then um has has really been driven by the desire to expand the realization of that vision and to cover more of our internet experience in that way Hard to do that of course Um and you know ZK I think is a significant step forward to realizing some of those visions I don't think it'll be the last significant step forward for this space uh but you know we're going to get there one step at a time and and I think that boundless and and the way that it alters the execution story uh for decentralized applications I think it's an important

part of that and I think you know just to to bring about sort of the next topic with regard to this idea of proof of verifiable work you know um one of the great things about you know Satoshi's design was that it basically allowed an open network of people to to work together to generate consensus Um and you know the more people that were part of the network the stronger the security of the network was But one of the unfortunate things about the design and one of the reasons that even though I I loved Satoshi's papers I was a little skeptical of it when I first read the the design was that it it lacked any sort of scalability It was going to be doing four transactions a second max because even though you could throw more and more compute at it that

compute wasn't actually going to anything that helped the end users It was really just increasing the security of the network Um what's interesting is is that with ZK there's this new possibility which we you know really talk about in the paper of proof of verifiable work and the idea there is that you're again making an economic incentive mechanism to participate and to produce work for the network but that work is not just going to secure the consensus but that work is actually going to doing useful work for end users to actually running real compute that's actually for user transactions right And so I think that in some ways ZK unlocks the scaling not only of like verifiable

computing but also this scaling of this sort of incentive mechanism for computing and the ability to really bring online massive amounts of computational you know there's all kinds of GPUs just sitting there idle um you know and I think that if we can bring those online in a way that's not just an economic security sort of um purpose but also to actually be able to do real work for end users then we can build you know a real world computer that has the computing capability that is significant enough to actually really expand the the use cases that are possible in blockchain So so the the first iteration of a lot of these blockchain protocols

was obviously cemented after Bitcoin around proof proof of work Feathercoin Mastercoin there was like Litecoin Blackcoin this coin Dogecoin you know there was all this all these proof of work chains So the original idea of blockchain as you described Jeremy from Bitcoin's papers was obviously proof of work That is what everybody knew as the origination of these really tough math puzzles essentially uh only to to be able to be computed by expensive hardware which would then be essentially the the security parameter if you will of these networks We evolved that into uh proof ofstake networks purecoin Um there was there was others of eventually

we had a lot of the Ethereum killers Uh even though Ethereum was proof of work at the time we had a lot of these proof ofstake Ethereum killers Um and then even Ethereum transition from proof of work to proof of stake which I think hot take is is the people think that that was a big mistake for Ethereum uh when it comes to differentiating against other proof ofstake networks Don't need to get in into that That's kind of up for debate We we we understood why we had a lapse in performance from proof ofstake networks and proof of work networks which is essentially the redundancy of passing of information from nodes to a achieve consensus Right And now we're moving into this proof of

verifiable work paradigm um as you were describing Is this the logical endgame for this next era of onchain interactions and experiences Um h like how does this how does this evolution shape up in your mind Yeah I mean from our perspective this is the next step in the evolution of blockchain sort of incentive mechanisms I mean it is worth noting that proof of verifiable work is more of an incentive mechanism than like a traditional consensus mechanism in that it's not picking what the longest chain fork is or what have you But but I think that you know the the two things that are that I think are significantly different than either proof of stake or

proof of work is that one um you are you are rewarding work but this time the work is actually meaningful useful work for the end users and two that the total amount of work being done per unit time isn't limited because of the fact that ZK has this amazing magical sort of um you know um succinctness property which means that that it's equally easy to verify an immense amount of work and a tiny amount of work and no matter how much work you have the verification times don't go up What that really means is like historically any consensusbased verifiable computing has a limit which is the amount that you can

fit in a block and still comfortably distribute to everyone for redundant verification And fundamentally ZK doesn't have this limit And so we really you know I mean I think that what's very interesting is proof of verifiable work is sort of an has not been done in this sense uh before And I think we're very excited about the potential of that and really I mean obviously we believe it's the next sort of evolution in the in this sort of um history of blockchain Um I mean whether it's the end end end game I mean like there's there's a that you know there continues to be innovation in this industry But I I think for at least the next couple of years it's going to have a it's going to be the the way to go

And in hindsight I think this is so incredibly obvious that I'm surprised no one else has done it Like I I think you know if if we kind of start with the proofof work um paradigm and we look at even something as simple as Bitcoin you know the the the parallel of proof of verifiable work in Bitcoin would be you know suppose we have two miners um half of them mine you know half of that Shaw 256 algorithm the other one mines the other half and then they combine their computing power and boom you you just mined a Bitcoin block like that I think would be such a good idea but we just never got that sort of that sort of capacity And so now you

know you fast forward to today and you get ZK proofs You and and just to be very specific this is proof of verifiable work You're able to essentially have one prover mine or generate um you know some say call it 20% or 25% of a ZK proof Someone else beats them to the rest of the 75% you can now that and the usefulness Jeremy that you describe is the ability to combine these two pieces into a single proof and therefore submit that as the useful work towards the end user Yeah And I think that the key thing to note is like if you look at the Bitcoin case what the real issue with having half of the you know nodes do X and half the nodes do Y is how do they know the other

guys weren't lying right And I think that really ZK needed to exist for this to function And and also you need the breakthrough in costs that and and usability that have come from the modern ZKVMs to make this actually a practical thing Like the reality is if you look even two years back the cost of ZK was such that even though in theory the scaling works better in practice it's was just cheaper to verify the same thing a thousand a million whatever it is times on all these computers because the the effective cost of all that redundancy was still less than the cost of doing the ZK proving but ZK has dropped so

precipitously in price um and you know Like I I you know if you go back you know eight years or something like that this just didn't even remotely exist right like ZK was a a nice like you know theory lecture in like your cryptography class at the you know very end uh you know because no it was completely unusable Um so so I think that like I think that the the the the dream or the vision of this is certainly something that's been discussed over and over again Like back in the day right when people were doing proof of work mining there was always a question of well maybe we can make it do something useful like we can do some kind of study at home thing like and there's a lot of attempts to like make it fit in Um but at the end of the day uh without the ZK

ability to quickly verify the correctness it only really suited a very very small number of problems Um yeah I mean maybe prime number search could be done but like doing user transactions is another thing entirely They didn't have the right structure to have that sort of thing and now with proof of verifiable work all computation has that structure which is I think really impactful Yeah absolutely And Tim I kind of want to hear from you because you were there like you said 2015 Bitcoin talk forums You were like heating your house using these Bitcoin miners and and it really was like this like digital gold rush where you you know it it was like you can get some picks and shovels head over to California and you can start mining for Bitcoin you know digitally of course

Now fast forward to today we have a very similar concept of essentially like mining these ZK proofs right And so I think you know a question or thought that might be you know in in our audience's mind is like well how do I get a piece of this action I was maybe I wasn't there in 2013 to mine some Bitcoin but I am here I have a MacBook I'm ready to dedicate some resources to mining some ZK proofs and get involved in this new in this new gold rush if you will Could you kind of trace some similarities What was it like back then And is that a is that a fair analogy to today I think it is a fair analogy Uh you know I was I was out at ETH Denver uh just like a month or so ago and I was hanging

out in line to get into the Bitcoin uh show and I was talking to some of the zoomers you know waiting in line and you know one of them brought up that like you know she feels horrible that she was in like 8th grade back then and like couldn't have been in the scene She feels like she's too late I was like no cuz like you're not too late you were right on time Like everything that's happened on chain up to this point has been like boom out on the horizon when we're looking back Everything that's happened up till now is we haven't hit that hockey stick yet right People get excited about coin caps you know in the ballpark of a trilly Trilly ain't nothing my friend right Like a trilli is

like one business We're talking about the whole economy here the whole internet Orders of magnitude more opportunity to be had So like to me it's like yes I I totally understand the feeling of growing up and being inspired by a thing and and not quite being able to be a part of that scene at that time That happened to me and operating systems my first love but whereas operating systems like a done story right like Unix was the last one we needed Uh all this technology that we're talking about right here very much the opposite um you know the classic question right what's the use case historically it's been moving tokens because that's what the technology has historically been well suited for uh

here though we are doing more and we're doing it with ZK in the sedative way okay bringing it back on so you asked this question all right so for those folks who are saying like okay Tim says it's like I'm not too late I'm right on time how do I get in uh it really is the case that we're talking about the computer you already have a MacBook is great A Mac Mini is great You do not need an expensive Nvidia PC although if you happen to have one it's also great Um and we're we're in this situation right It's already the case right now You can be joining in on the test net You can be contributing to the numbers on my ticker uh if you if you want to run the software it's the earliest version which is not going to be the

most userfriendly but you see how passionate I am about you know all this stuff They got me on the task We're going to be building this really simple experience for folks who really do just want to grab a binary run it on their Mac and get started Just generate proofs right You're sitting in class you're taking notes You don't need that GPU Let it cook right like you know laptop is back at the apartment you're out It's Friday night You are done with that thing Set the keyboard aside Make the machine work for you Like that's that's what we're talking about here And that's that's what we're going to be opening up And it's not going to be this kind of thing where it's like you're getting squeezed out by some other player who's got cheap

power because they're next to the river or who's got cheap power because they can burn all the coal in the world without filtering it Like you're not you're not competing with that because it's not an either or It's a yes and right like that big player with the data center You bet they're rocking volume That's what they're doing right But the demand for execution on chain is insatiable Everyone knows what Jeban's paradox is now thanks to AI and like everyone getting a a good lesson there in utility and such So I think it's going to be a really exciting time for folks I think that this network here we really are trying to do something that's very cipher punk Uh Jeremy and I like true to the roots Frank true to the

roots All of us true to the roots So that's the pitch to Tim's point you know I mean I think I think the one thing that's really highly relevant is the right now a lot of the value the use cases for chain is a lot of sort of DeFi kinds of applications And partially the reason that is is because at the current cost of the consensus based verifiable computing it's the only thing that makes sense right Yeah I'll pay a dollar transaction fee to like make some trade that's going to net me you know 100 bucks or more hopefully right But like I'm not going to pay a dollar transaction fee to like click the like button on like a social network or you

know like you know move like you know two units in my like game I'm doing or whatever it is right And in order for that type of but when you bring the costs for verifiable computing down and when you get past this sort of scarcity like look there's only so much space in the block right and so if you have more demand for compute than fits in the block you just got to keep raising that price uh until like it fits again right and as a result you'll never get more compute than a certain amount and only the highest value transactions are going to get through you actually open that up now you're in a realm where the cost of the compute is just the cost of the compute If more people want to do computing more people can spin up

provers and um and and you know what we've seen is when you make computing unbounded in that regard right There's there's almost no limit to the appetite people have for compute right The cheaper you make compute the more money gets spent on compute right That's the you know Jeff's paradox thing It's very surprising You think "Oh I'm going to lower the costs so people are going to spend less money." No the answer is you lower the cost there's more things to do People actually spend more money in aggregate right Even though they're so that these efficiencies as DK brings that kind of efficiency gains to verifiable computing you know I just think it's going to be a huge um boom to everyone Yeah So maybe bit more uh

concretely Jeremy how exactly do you expect this adoption kind of flywheel to to take place with ZK mining Um and perhaps you could also just explain um ZK mining uh in terms of like the like like how that process works uh a a bit deeper and then kind of how how you see that kicking off to enable perhaps some of these net new application and ideas or just use cases that you're excited about Innovating finance global institutional and onchain deep liquidity to drive capital efficiency Now maximize yields on institutionalgrade liquid staking with

M Unlock trillion dollars of capital via F Bitcoin All of this secured by Mantle's treasury of billions in stables and ETH capital pools This entire ecosystem is coordinated by a vast network of seasoned VCs that are aiming to create the next generation of banking and financial technology This is Mantle Ladies gentlemen anons and dens It is stable coin season brought to you by your favorite the best in stable coins Fracks USD dollar stable coin backed by the best in the business the Bidd Fund

by Black Rockck and Superstate This is stable coin season It is the yieldbearing stable coin by Fracks coming soon to a chain near you Redstone is DeFi's modular oracle Supporting over 100 clients on over 50 chains and offering 1,000 assets Redstone is the leader for Bitcoin staking and yieldbearing assets They offer the most robust customizable price feeds available in any model Join the Redstone exhibition today [Music] Yeah Yeah Yeah So I mean I think the the

so we we lay a little bit out there's really sort of two major parts of what makes ZK mining work one is sort of the structure of the network and the economics and all of the sort of work that Tim has done uh in terms of you know the auction model and all the other kinds of things that actually make it so that there's you know within the marketplace component of boundless that there's a way for someone to request work for provers to be able to you know do the work get paid so there's there's there's that sort of component so you built like a spot market rather than like service agreement oriented cloudesque era essentially So the idea is is that we're going to do a spot market What's interesting and this is

where the financialization comes in You if you build a spot market for compute you can actually construct things that look a lot like SLAs's as derivatives of that spot market Effectively an SLA is actually just a a computing future on the compute spot market Uh which is like bounded by some like Yeah Yeah Yeah Exactly Exactly I want to be able to make sure I can get this amount of compute for this cost for some period of time And so I maybe that cost is a little higher than I could get at the spot market right this minute but I'm willing to make that agreement with someone else who's willing to agree that they're going to produce it you know on threat of you know say uh slashing if they don't And uh now suddenly we

actually have a a contract that basically is an SLA but it basically is built as a derivative on top of this this sort of spot market that exists So so if you just to break down the question of how does proof of verifiable work under the hood right Well so there's these two components right There's this sort of network component which is this sort of spot market for compute and how do we how do we how do people lock in the requests How do they settle You know a lot of that's explained very much in the paper And then the other uh really relevant point is because it's sort of this open network and we want to have a reward model that rewards people for the amount of computing that they have done as part of the network we need a way to like verify that the this this amount of

computing is done and the so one of the other parts of the proof of verifiable work is this mechanism to cryptographically link the to basically make it so that if I generate a zero knowledge proof I can prove the amount of work that went into it One of the things about the succinctness of zero knowledge proofs is I could be proving a huge amount of work or I could be proving a little bit amount of work The proofs look the same So the first thing is you need to do is actually have a way to cryptographically uh sort of uh link and connect the total amount of work being done into the final proof so that um in a way that isn't falsifiable so that I can say oh look if I show you this proof you know that I did x amount of work Um and then the second thing is

you need to make sure that you can't get credit for that same work twice So you need to make sure that there's basically unique identifier that goes into every proof that's associated with the prover that they can't double spend it right Um which again goes back to sort of part of the the network part of it Um but basically what we've figured out is an efficient mechanism to make that both those things happen right And so the design of boundless sort of solves sort of both those issues which is one is like how do you measure the sort of computing that's being done and then how do you and and and thus know how to properly incentivize it and then how do you also make sure that the spot market functions efficiently and more importantly that the spot market doesn't become your scaling limit right which is

very interesting because as part of that what we had to do is design the spot market to not run on like an existing consensus based chain so the the actual overall All boundless is works across all the chains and the provers are just basically a pool of computing and they talk to these market contracts on all of the chains but there isn't a central sort of like chain that runs the whole market It's actually done in a decentralized way you know and and by doing that what means that we're not like if we introduced a new consensus domain and every transaction had to go through that new consensus domain Well even though we have an unlimited amount of computing available to us the the spot market wouldn't function past a certain volume right So I think like

there's a there's a lot of very important challenges on this You know we spent a lot of time looking at a lot of different designs for how to build this next step of the company I think at one point we had did this project trident where we actually had three completely different designs We actually built out full prototypes of them all just to really understand what the the the problem space of this whole thing looks like Um and I think the one we landed on you know we made is like a very no compromised solution in terms of truly decentralized Um and yeah the the an you know boundless in the part of the reason for the name So yeah Okay Okay so ZKVMs allowed for generalizable ZK proofs

rather than having to write circuits and have to rely on very specialized proof use cases This then changed the way people are thinking about ZK applications and coding and and writing ZK applications right Then introducing boundless and proof proof of of verifiable work Then effectively I mean the big breakthrough here is is removing this uh need to have prover decentralization again with this with this me plus you kind of idea versus this kind of me versus you Um and then this these two components together obviously summarized um then bring us to a world where we can have much more

efficient applications on chain Well whether they're they're they're onchain as we know it today or or kind of in this newer version This is this has allowed us to basically make computational computation verifiable at scale which which again is the big breakthrough that we're all kind of pushing for in the ethos of of why we're here And I think I think there's I think there's still more advancements to be made Like I think one thing that's worth noting is that this really does solve the problem of how do you do an arbitrary amount of compute I think there's still an interesting question around like how do you make sure that like the data availability problem is solved How do you make sure that like the the sort of you know updates to

shared state gets done So I think that this is not the like the end of all things but I think in terms of this this limitations that have always existed on blockchain in terms of compute yeah I think this will basically really change the game dramatically So yeah I'm going to add on that too You know it's going to do these things It's going to scale the execution It is going to be a big deal uh a lot of the application developers that I'm working with right now the designs that they are pursuing the feature sets that they are pursuing are things that you know they've been contemplating for years uh but have been simply prohibitively expensive to do in the kind of traditional way of just writing a smart

contract and running it directly on chain and so for these teams it's like you know it's like the whole organization is waking up and realizing uh that you they can think big that suddenly it's okay if you need to run colossal aggregating computations at high frequency across tons of ch like all the stuff that we're used to doing in the DC in the data center you know you're going to be able to do with your onchain applications and not only are you going to be able to do it like in principle but like you're going to be able to do it like in practice and the experience is going to be very familiar to the user Um that's the thing we haven't really talked about here yet is it's like this is like one of these subtle points

that's like another massive motivator behind the design of boundless Like okay so earlier you brought up like some of the other projects we did We did things like ZETH you know the first Ethereum block proofs in in a ZKVM right Um we did a number of these other things We had um one of our first demos we did battleship like the old classic board game Amazing game for ZK right We did these things and for each of them it's like oh man wouldn't it be cool if and then you envision what amounts to like a 12 month project to like build a really cool demo around the thing and and where where that time goes is basically that yes the ZK lets you prove the computation but now you have to integrate that into like this whole ecosystem that was designed before ZK

was a thing right so all sorts of just like the most uh kind kind of surprisingly hard problems because they're so low in the stack right Like they emerge So who's going to generate this proof and how are they going to get the data What's their incentive What happens if they don't deliver it Like these are all things that suddenly matter You spin a transaction on chain you just accept that someone's either putting it in a block or not you know but once you're saying that there's more than just that suddenly the user experience has more types of outcomes you have to contemplate Um there's also the question about what actions is the user taking to drive this process We're all very familiar with you know connect my wallet sign and send the transaction That's a

flow that that folks in this space already know You don't want to mess with that You don't want to say by the way you need another browser extension You don't want to say uh by the way you need to sign up link your Google account to the you don't you don't want any of that you know You want that experience again for the user that's as simple as I'm using an application I'm authorizing this action send it out to the network and most of the time it works out fine That's that's what you want And to do that with ZK as a as a if you're an application developer if you're a business trying to provide an experience to the user it's not enough to download a ZKVM It's not enough for there to be a smart contract on chain that verifies these proofs You need a whole thing to manage

this flow from the user and to integrate it in chain on chain and make all that work You need to do it in a way that also makes sense for the info providers you know like if I'm proposing a block I get it I'm going to maximize you know my extra you know right I'm going to get my value out right I'm going to sequence some transactions in intelligent I'm going to do like all these things software exists for all that if you're if you're approver engaging in zk proof generation you kind of have a version of this too you're going to have to pick out which tasks you want to you know pursue how aggressively you want to pursue them you know are you going to race to get get it done faster than everyone I'm rambling the bottom line is there's a ton of problems that are

entailed in realizing this many years long promise of ZK's ability to like improve the onchain scaling situation and in some sense like we we've talked about how boundless is exciting for home mining we've talked about these how it's exciting for scaling but on a practical level the day-to-day thing that's going to really excite developers in particular and ultimately users in the end is going to be that we've identified and solved a tremendous number of integration problems that most developers in in blockchain haven't even contemplated yet because they're they haven't tried ZK yet So you should come and check out Boundless I think that you'll find that we've solved a lot of problems you didn't even realize existed and you might not even notice that there

are things that needed to be solved because I think we did a pretty good job making it smooth And and that's partially because of the of the slower progression of ZK and then the accelerated progression the last two years Um but also just you know whether it be funding or just uh people I I I you know we find that that people oftentimes do what the market wants which which is is often like somewhat of a band-aid solution right Like like we you know you see that things in the market get rewarded because they're fast cheap and good um but perhaps you know that they're not on the same ethos of the amount of validators or the amount

of nodes or you know how many you know what's actually kind of happening and so like we've seen we've seen these like solutions come and and they will continue to come but again it just keeps coming back that that the endgame is is ZK now I would like to pose kind of a more pointed question here when we think about something like uh the proof supply chain right Kind of making an analogy to the MEV supply chain There has been a consistent addition and subtraction of certain actors in the MEV supply chain A and there's a notion that that as you add more actors to the ME supply chain there there becomes more and more centralization choke points it is that

similar or is there a centralization choke point or a choke point in this design that we ought to be intellectually honest about and think about how to improve on So I think one thing that's really interesting is there's this idea I think it's really important to separate the idea of decentralization of the market making process from the uh sort of distribution or like having a diversity of provers because they're not actually the same problem and I think there's a little bit of a confusion because most people have built consensus systems in the past wherein those two things are the same right like fundamentally if all of your stakers or all of your miners in the

Bitcoin sense if if if if 51% of them or 80% of them are all the single entity then your whole consensus system is controlled by a single entity right now what's interesting is let's talk about this in the sort of sense of the way that the market design for boundless works so in the boundless market design the markets themselves operate on existing chains and they inherit the um the the censorship ship resistance um properties and the availability properties of those chains And so the markets are open to everyone so much as anyone can post a transaction to those chains So as much as you believe in the censorship resistance of a given chain that's how censorship resistant the

market making mechanism itself is What's interesting is is let's imagine that some huge prover like that uh is is in some megalith like AWS decides they're going to get into the ZK proving business and they're going to undercut everyone and like you know all to the wall like let's say they for a moment in time end up with 80% of the proving volume Well what's interesting is what do they what can they do with this Well let's say they want to not prove your transaction Well you put your transaction on the market the market's on an actual existing chain like say Ethereum Amazon's going to ignore it Now if the 20% of the market is rest of the 20% of the market is controlled by

honest users they can still take that up They can still take that Now maybe the truth is their prices are higher because Amazon's getting some you know crazy you know cost savings So maybe you have to pay slightly more for one of those players to be able to do your proof because Amazon's going to ignore it because they don't like what you're proving It's fine right like it's you've actually not censored the end user right Or let's say Amazon's like "You know what Actually now that we've controlled 80% of the market we're going to just jack up prices right?" Well the second that they go above the price that like I can run a machine for at my home I'm just going to come in and undercut them right Like it's a free market It's you know it's a it's a free market And as

long as the market itself cannot be distorted which in this case like the these markets exist on chains that do have a censorship resistance properties you're relying on a free market And yes there will probably be some consolidation right But that consolidation doesn't damage the availability It doesn't damage the censorship resistance properties of the actual end users And look I mean I think you know most of us have a strong faith in free markets And the truth of the matter is is that like McDonald's makes really inexpensive food but not everything is a McDonald's because that you know like the the it's it's it's so I think that one thing to look at is like if you design the actual mechanism

by which you make the spot market operate such that it is not itself dependent on like say 51% of the provers being honest which in our case it's not then then then this idea a of you're really talking about trying to increase prover diversity which is a useful goal in certain instances but also free market will always maintain some level of that and yes there will always be some level of consolidation as long as that doesn't harm the end user um that's a great thing to have Got it Okay So um the the censorship resistance properties um derived from the individual chain themselves um we can w within the boundless design we can

be subject to those However the the system for the spot marketplace as we call it um is is it it could be subject to like essentially prover collusion if you will However it's designed in such a way that that ends up being a favorable outcome to the end user rather than to the parties who are colluding in this specific example Another way another good way to say it is like let's say provers do collude and they try to raise prices like there becomes an arbitrage which is like let's say the cost of computing let's say the prices are above the cost of computing Now I can go buy a computer you know or just rent one from

a cloud provider if it if the costs are that high and and I can just prove for less than all the colluding guys are proving for and for more than it costs me to run my computing and I can make some money and there's like an arbitrage opportunity and we're relying on the fact that in a free market with there's nothing preventing anyone from acting as a prover and if you make that market actually truly open then the free market should make sure that the prices are always as low as they can be unless the colluding parties can somehow distort the market making mechanism itself Then there isn't a way to and because the that isn't that exists across all of these chains right like uh it it there

is where once you make a market an actual proper free market the amount of useful collusion that exists is just always limited because there's always someone willing to defect and make more money right Like that's what we're relying on We're relying on the fact that there's a bunch of people colluding and one of them decides that they're going to defect and not price fix and make a cheap thing then then then the prices are going to go down to what they ought to be And and actually what's interesting is is that in a market like this there's not there's no barrier to entry The barrier to entry is I have a computer on my house right And in a market with that low of a barrier to entry collusion is not impossible right So I Yeah Faith in the free market is the the trick here Yeah So So zooming

out from this and kind of just looking at you know like the openness of this free market you know I was around in 2016 like I was in a school and they had all these computers and I just felt like this I just I was just like "Wow this is so such unused resources here." But I just didn't at at the time have like the technical ability to like set them up in such a way that they would be mining Bitcoin There was just a level of like knowledge that I was missing there What is going to be like the I have heard some sorry Rob I have heard some stories of kids doing that in their dorm room and getting like suspended and expelled and stuff My own knowledge limited me from getting expelled So thank you know I'm very grateful for that I guess could

have joined the dropouts Yeah Well that's that's another conversation Um so in this in this case like what is the front end What is you know the the entrance point You know I have a computer I want to be involved Can I just And so actually like you know back in the day it was like well you know if if you're not going to set up a whole Bitcoin you know minor you can join one of these like cloud mining like cloud pool type of situations and that was download the whole chain on your on your computer spend four days downloading the entire chain Well even that you had to have some software So so in this case like what is what is the front end what is what are

people going to like actually interact with in order to contribute to this marketplace Yeah So I mean I think right now on the test net like the bar is a little higher than we would like it to be Uh it is like just because it's new and you know uh we are currently you have to kind of do the the bidding process and so on and we have software we have software that does it but like realistically you need like a pretty pretty serious GPU and a decent internet connection but but but we are absolutely want are working aggressively at driving down that fundamental complexity I mean at a at a fundamental level our proof system runs very efficiently on normal consumer grade hardware It runs you know on on on any on on cards that are normal consumer cards gaming you know gaming

PCs and even on like you know laptops and so on Um and we're you know we are currently working on you know um projects to like try to figure out how to build things like the proving pools kinds of use cases right so I think the short the short version is our vision is to democratize the capacity to do these zero knowledge proofs um very much so um you know I think uh if people go join the test net right now uh they should expect it to be a little more like the early days of Bitcoin where you know like you had download You know it was a a bit of an IT uh uh complexity Um but you know that's the difference between uh the fundamentals on the ZK

uh you know and in particular in our proof system are very well suited for um you know getting things all the way down to a much more um grassroots level and we're very much interested in that Um so yeah Yeah Yeah I think one consistent trend we see across the entire space all a lot of our conversations is just lowering the barrier to entry in general We've been very very bullish on just accelerating the pace of experimentation by lowering the barriers to entry to to do that experimentation and and every time that we we we talk about ZK and just increasingly more excited So Tim Jeremy it's been an absolute pleasure Thank you guys for um

the work that you've done on boundless and uh congratulations on the white paper looking forward towards mainet and more of the things that are coming Um so gentlemen thanks for your time Thanks It was great talking to you guys again Thank you so much Hebas