We Asked John Adler About The Future of Celestia

Transcript

People who aren't validators aren't losing much because the validators just make a small profit, right? Obviously, if you pay a validators 100x profit, you know, it's kind of unfair to gatekeep who is a validator. >> Is there a risk to having too many hard forks or do you think this could actually bolster the social consensus and kind of rally the community on a more frequent basis and we're not thinking about hard forks the right way? >> Does proof of governance restrict assets from having this kind of like monetary premium? Is that even a real thing? Does that matter?

>> Welcome back to the Rowup. We are your hosts. I'm Andy. I'm Robbie. And this is your home to learn from the biggest guest, the brightest minds at the tip of innovation. Our goal is to help spread the collective knowledge of everyone in our community. We are breaking crypto out of its echo chamber and taking it to the real world. John, welcome to the show, man. It's good, good, good to have you back. >> Thank you for having me again. >> Good to see you. Welcome, welcome. >> Um, you definitely >> the feathers. >> Yeah. >> Yeah. You've definitely been ruffling some feathers in the uh proof ofstake camp. Maybe just like just quick quick uh, you know, rewind a month ago, you're, you know, somewhere writing up this proposal, thinking about the

dynamic of proof of stake. As Zucky says, some of these ideas came from a hot tub, uh, from the Cosmos space. you're writing this proposal around proof of governance being the endgame for LSTs. What's going on in your mind? What kind of caused this thought? What like series of thoughts led to you starting this uh idea? And then we're going to dive into some of the more technical stuff uh and and go through it. >> Yeah. So, to kind of uh well, first of all, I enjoyed watching your your podcasts with uh various guests. I think I believe it was a permissionless around the time that you were you were talking around the time that you mentioned uh a bunch of people were talking about proof of governance or

reducing staking issuance that staking issuance is not yield is not is not real yield it's just attacks things like that uh very I feel like very few people cited what inspired them to start talking about this uh narrative at this ex at that exact time uh but I like I'm a bit biased but I like to think that it was my my ized all that >> you didn't get the credit that you deserve but nevertheless it did proliferate the uh zegeist >> yeah it's okay I'm used to it by now uh so the what's the thinking behind this so it started with uh the thought of we have tia as a blockchain we have tia we would like tia to be used in defi and all the things like that uh and whenever you have a proof ofstake token and you

have a defi that is using that token you end up in a situation where uh staking derivatives and LSDs and that kind of stuff might interfere with the yield in DeFi. Oh, sorry. Let me let me let me take that back. Uh it's not that the staking derivatives interfere with the yield, but rather the staking interferes the staking yield interferes with the DeFi yield. Uh and the way to counteract that is by the existence of staking derivative derivatives like LSTs. And this is what kind of Tun and his co-authors and co-conspirators uh kind of posited posited back in the day of when E2 was starting. Uh you know, Tyrone and and his uh and his co-authors basically said, you know, that

regardless of what you do on the primary chain, uh you're going to have staking derivatives. Uh you're going to have LSDs even if there's nothing even if there's no in protocol delegation. and therefore it's kind of a losing battle to not have in protocol delegation might as well embrace it. uh this turned out to be true much to the dismay if the Ethereum consensus designers implementers you know Lido and a bunch of other LSTs ended up existing on Ethereum and today you know they control the majority if not the super majority of control is a strong word don't don't read too much into that they control the majority of you know stake on voting power on Ethereum that's not a bad thing uh just kind of the natural consequence

of staking yield kind of will compete with DeFi yield And if you have a staking derivative, you can get the you can get both at the same time. Okay, those you can get your staking yield through a yield bearing uh staking derivative like st as it were or you know staked TF or exist on on Celestia and you can get the yield from DeFi lending, gambling, speculating what whatever that whatever the heck you want to do. Uh so we're now in a situation where we have that's kind of the context background context of how we got to how this applies to Celestia. We have Celestia. We would like LSDs for Celestia. If Tia is going to be used in DeFi, we'd like LSDs for TIA. uh while doing

a lot of while doing background you know uh literature review on the subject uh it I was told about John Sherbano's uh post on proof of governance well I was told I I guess I had known about it way back when when it was first originally posted and he did the presentation uh on this topic way back when uh but I had kind of like memory hold it for for a while because LSDs hadn't come up for a while. So, uh, I was I was presented with this. I read it and I'm like, "Oh, this kind of makes sense." Uh, for what it's worth, I should say that while I wrote the post on the proof of governance proposal, it was primarily driven internally by Nashq, our head of

research at Celestia Labs, who was driving our internal LSD design efforts. Uh, I just kind of I just kind of published it uh as as as one as one does. just publish published research post/blog posts to get ideas out there. Uh but most of the effort was driven uh and kind of this notion of proof of governance and filling in some of the missing details in in John Charau's post was primarily driven by our head of research uh Nashq uh just to give credit where credit is due. So how does uh what does John Sharbon argue? He basically argues that in a sufficiently mature LST landscape, you end up with a way of selecting validators that isn't the same

way that you do in kind of a naive immature proof of mistake way. Is not just I have some stake and then I delegate to a particular validator and then that's it, right? It's going to be something a little more uh let's say intermediated. For example, uh let's consider the let's consider kind of a spherical cow model where you have one single staking staking uh uh liquid staking operator. Let's say let's say you had just lido and just this just lido had 100% of stake staking with it. Uh this makes sense because you know why would you

leave yield on the table if you can liquid stake and also get defi yield and also do like looping and stuff like why would you leave stake why would you leave yield on the table that's kind of you would not in any sufficiently mature system with little friction and you know we should work to remove the frictions any sufficiently mature system you would you would have you would collect all the yield that you can collect you wouldn't leave free money on the table so okay you have 100% of uh stake stick to a single uh liquid staking operator. So this does a few things. Uh the first thing is the validators that this operator selects to actually be the validators of the system. Uh that's how validators are selected not through

delegations to individual validators. So this is a type of governance thing where you have you can say like you kind of put your stake into this you know lido as an example uh and then lio is the one that actually selects the validators in this example. >> Introducing asset runes the future of decentralized finance stable coins natively on bitcoin. Hedge against volatility diversify your portfolio all directly from your bitcoin wallet. secured by ZK technology.

This is asset runes available now exclusively on Starknet. Uh another thing this does is that if 100% of tokens are staked which again if you have this mature system where people don't leave yield on the table then 100 you close or you know either 100 or close to 100 would be staked the staking issuance is actually zero. the net issuance. The reason is that let's suppose that you know you have you have Ethereum uh it has some issuance I forget what it was let's say it's 2% a year right they issue 2% of the ETH supply a year this is just an example I don't know I don't know the exact numbers if a 100% of ETH is staked that

issuance just goes back to the ETH holders in other words it's not actually real issuance because everyone gets more ETH but ETH also gets diluted by 2% so the yield is 2% the inflation is 2% it cancels out and the real yield is zero and this is why you have you know various people including like Anatoli and Twitter is very vocal about this point uh but never kind of took the leap of faith of proposing actually making a change based on this point. He keeps talking that proof ofstake issuance is not real issu. It's not real yield. It's just a tax because you get yield like fake yield but then you get taxed and then you have to pay taxes on this fake yield. Uh kind it's kind of a giant scam like capital gains tax. Exactly. It's is it's just a tax you you you know if

you're in a jurisdiction that tax taxes you then you pay taxes and that's where the money goes. It just goes out of the system. No one's no one's getting it. So if you're in the situation in this spherical cow model 100% of tokens are staked uh real issuance is actually zero. Uh so if real is zero it's arguable why would anyone stake uh and also you have this LST operator that essentially decides the validator set can you just kind of get to this end system but make it even better than that and that's kind of where proof of governance comes in which is we say okay well let's skip over all the staking stuff because if you have a 100% staked and you don't actually get any yield so why would anyone stake let's just skip

over all this let's just do the valid validator selection part. Uh, and the reason this is fine is that validators aren't meant to secure the system and they aren't meant to define what is valid. Validators are cucks. Uh, a common uh a this in different words is a common refrain uh in the Ethereum crowd when they compare and contrast how uh Bitcoin miners aren't supposed to control the system, right? And Ethereum miners back in the day when Ethereum is proof of work don't control Bitcoin and Ethereum.

Uh they might you know build additional blocks on top of the previous longest chain but they don't define what Bitcoin and Ethereum blocks are valid. Uh they don't define hard forks etc etc. All of this is done through social consensus through uh deploying hard forks, getting communities and exchanges and users and developers and everyone to coordinate around upgrades etc etc is not done through miners or in the case of brew of stake validators deciding on the governance of the chain. The validators and miners are effectively the are downstream of governance not upstream of governance. So therefore if you imagine a very simple and naive system I'm not saying this is how proof of governance

should be done but you could imagine a very simple and naive system where you have in the software you have a list of validators and then those are your validators no staking no delegation no nothing there's your validators and then if any validator decides to misbehave well then you can penalize them in some way exercise left for the reader and then change them in the node software through a hard fork blockchain communities coordinate through hard forks anyways. Uh and then the the selection of the validator is automated through the node software uh as as by hard coding them. So again, this is like a very naive like minimally viable proof of governance implementation uh way you way you would do it. Potentially something in

production would be more sophisticated than this. Uh but that's just an example to get to get to get the listeners you kind of a mental model of how it could work. Uh and in this way you now no longer need staking. you no longer need staking issuance. Therefore, you can cut issuance. Uh and you can do various other things that make the token more the the tokconomics and the kind of value thesis of the token more obvious and more concrete and less confusing because of the existence of issuance as a fake yield. uh for example what I proposed in my uh what I proposed uh well I shouldn't say what what Nashq proposed that I then transcluded in my in my post on proof of governance is you

can burn all transaction fees currently in a lot of chains transaction fees go to the validator that uh created the block uh cosmos chains including do it a bit differently which is always weird but they distributed it to all validators but you can do even better which is in the model where 100% uh sorry not all validators, they distribute it to all stake. Uh so in the model where you know 100% of uh tokens are staked, well then if you're going to distribute it to all stake, that's the same thing as just burning it. So instead of like from an accounting account from an accounting perspective, it's easier to burn it than distribute it to everyone because it's like a single state change to just burn a token as it's paid for fees versus if you had

like distribute a small amount of dust to everyone. would be substantially more expensive from a state like from a pro programming and state uh transition point of view. So you just burn the fees. So this makes now the value acral thesis of the token much more obvious which is that it's just directly and only and concretely well the DCF really uh it's you know how how many fees do you expect the blockchain to have in the future? That's how much it's going to burn and therefore that's how much is value is going to be driven to the token. you don't have to worry about the complexities of the fake uh the fake issuance the real issuance being fake yield >> uh and things like that. So this system is uh substantially simpler than proof of stake systems. Uh it allows you to

reduce issuance substantially because now you essentially decouple the issuance you have to pay to validators because even though validators are cucks they do create blocks that that is a real cost to them right and they need to be paid for this cost. So now instead of having this weird thing where you couple validator commission which is some variable with voting power and they may have variable voting power therefore they may have varying commissions but their fixed costs for each validator is really the same like why like if you if you have you know 10% of voting power versus.1% voting power you still have to run the same full node to be a validator.

>> So you have like all all these weird things going on because you you've coupled these things in proof of stake that don't need to be coupled. So okay, you just decouple them and then the protocol can just pay validators what they need to be paid to be slightly potentially very profitable who whatever the whatever the social consensus of the chain decides to be profitable in running their validator operations and you can do so uniformly among all validators because you just know what that number is and you just pay them uh and therefore you don't have it's it's much simpler in how validators uh are paid which is actually a nice thing. Now you might say okay well what's the point of doing this? The reason it's nice is that currently you have this awkward situation where validators that are larger have economies of scale. This

applies to both Cosmos chains and to Ethereum. Uh for example, Ethereum says uh you need to run one validator node per uh per validator, right? It turns out and you can run like one beacon node uh for many validators, but you need one validator node that does some sort of amount of work per validator. uh and that work is you know verifying some subset and attesting to some subset right the problem is that it turns out beyond some threshold which is I think like a few thousand validators it turns out you don't actually need to run additional validator nodes you can just spin up additional keys in other words at a certain point you just get economies of scale where you can spin up

you know let's say a million validators not that many whatever a large number of validators without incurring any additional like physical costs to running nodes. Uh so there are economies of scale there. The same applies to economies of scale uh in Cosmo style and tenderman style chains which is that the large more voting power a validator has the lower commission it can charge while still being more profitable while a very small validator needs to charge a higher commission to stay profitable. Proof of governance lets us do this very nice thing where all validators in the chain all effectively have the same voting power. So it's all uniform which means you know exactly you know you know that everyone has the same costs and they have the same the same voting power in

the consensus protocol and therefore they can get the same rewards. Everything is completely uniform and that makes a lot of things much easier when determining the economics of how much validator should should be getting paid. There's also some like further down things that I won't talk about like if you know that all the validators have equal voting power it makes it easier to like roundroin and do some other things uh between them because you know that there's uniformity uh various things to do with gossiping transaction transaction and block propagation that I won't get into because it's too low level but let's just say there's a huge number of secondary benefits to having these validators have uniform voting power. So anyways, that's kind of the TLDDR that there's a huge number of

benefits and a huge simplifies uh how these consensus protocols work by changing the civil resistance mechanism from this delegated staking to governance directly because ultimately even if you do delegated staking or staking with you know staking derivatives which is still delegated staking just with extra steps you still end up in a situation where you have some sort of governance uh deciding on the validators just again with more steps. So might as well skip the steps and do it directly. We don't want a very centralized um uh stake staking provider laid out 100% exchanges etc too you know the the entire idea of like okay we have all of this yield um

but if the you know the price of these different assets that we're staking is variable you have to pay taxes you have to deal with issuance as an expense from the protocol level end users um you know they don't really know what they could get if the assets down 10% you get 6% you're actually losing then you have tax etc um third and final thing is that you know we have a set of of distributed 100 200 validators with variable um you know amounts of stake in each but the same fixed cost. Um what's the difference between having you know a lot of these chains have no slashing right it's just kind of like been something that needs to be implemented you know is it is it better just to have a a a you know set of validators that are effectively almost like a security council per se

trusted uh amongst you know the best of the best and and use a process of removing these um validators rather than having to you know slash now uh I I do this recap and the summary for second time. Um, for the reason because I I want to push back on something that or at least I want to provide a a a chance for you to respond to what Ryan Sean Adams said when he screenshotted a news post and said, you know, it it's of you explaining this proposal. And what he says is that if you do this to your L1 network, you're no longer in the non-s sovereign store value game. Uh, and if you do this to your network, you you probably never were were right. Then he

goes on to say that the only two are Bitcoin and ETH. You know, does proof of governance restrict assets from having this kind of like monetary premium? Is that even a real thing? Does that matter? Uh what what would you respond to this take? >> Sure. Well, I mean, it's you could I mean, you could argue monetary premium is is a meme, just like economic security was a meme. Uh it's it's unclear to me uh that Ethereum and Bitcoin Well, it's unclear to me that Ethereum exceptionalism is isn't anything beyond a meme. Bitcoin exceptionalism definitely is. Uh it's a meme, but in the good way, not a meme in

the bad way. in the sense that Bitcoin exceptionalism has proven time. Bitcoin has proven time and again that is it is exceptional and it does have monetary premium despite not having good properties uh or properties that we would consider good. Uh Ethereum you know doesn't doesn't really have uh any anything exceptional going for it uh beyond well doesn't really have anything exceptional going for it. uh with now with respect to you know if you can do this or you can't do this ultimately it's up to the community to decide. Uh, one thing that people have a hard time mental modeling around is they may see this and say, "Hey, this is just a consortium chain from back in the day. It's a, you know, enterprise is enterprise consortium chain. How that's any different?" Uh,

the difference between this and consortium chains is the same difference between Ethereum and consortium chains. It's around the social consensus of what is the chain like what is the given to the given you have a chain and a hard fork of the chain with different rules. uh which one is the valid one? That's like what the so that's ultimately what social consensus decides. You can have various processes automating. You can argue that one is more automated than the other. These are all fine things to argue and we can argue about them technically. Uh but saying because one but these are the the differences in automation are differences on a spectrum. They're not binary. There's not one thing is one thing and one thing is another thing. It just there's slight

differences by degrees. Uh now uh you know you can say well how is Ethereum uh store of value when they did the Doward fork right Ethereum just goes and tells everyone to stop trading now you can say oh that was a long time ago sure well the bitcoiners don't think it's that long ago right uh there's various criticisms criticisms that can be made but ultimately the difference between an enterprise chain and a consortium chain and something like a proof of governance or Ethereum today even is around the social consensus deciding which hard fork is valid. Because as long as you can hard fork, okay, imagine you have a consortium chain that has whatever five, seven validators. Uh the social expectation around the consortium chain

and all the centralized exchanges that will accept deposits from this consortium chain. All of these the social expectation is the seven validators run the consortium chain and they are the only ones running the consortium chain. Maybe the code is closed source etc etc. But if the code is open source and the social consensus of the community is whatever is in the node, whatever's in the software and whatever validators are decided by the governance, the decentralized governance deciding on the validators, that's who the validators are and that's the blockchain. You can't have someone centralized show up and just say I control the chain, right? The Celestia Foundation could show up and say here are the validators. Uh then 99% of the validators are just the Celestia

Foundation and the Celestia Foundation with the mustache, right? And the Celestia Foundation with like the the the glasses and the mustache, right? And then the community would say no. Uh the ability of the community to say that and more importantly the social expectation and consensus that the community can and should do this. That's the differentiator between a consortium chain and a permissionless chain with proof of governance. So proof of governance does not imply permission that it's permissioned and it does not imply that it's a consortium chain. It's still permissionless. It's still decentralized. It still has all the nice properties of blockchains uh because all those properties are downstream of social consensus and not uh that are downstream of social consensus. And social consensus is kind of what defines

the blockchain to be what it is. And that exists whether your consensus protocol is tenderint or whatever the heck E2 uses. I forget I forget what name they ended up going with. Uh and it doesn't matter if your civil resistance mechanism is proof of work or proof of stake or proof of governance. None of those things matter because all of those are downstream of the social consensus. >> So John, if I'm understanding this correctly, the the uh acceptance of new validators in into the consensus mechanism would be determined by token holders, stakeholders that would vote on accepting these new validators into the system. And each time that happens, there's a hard fork uh and effectively there's uh kind of the social consensus

that decides, okay, we're going to accept the new validator into the system and and continue the chain on with this validator included in the selection process to build a new block. Every time, you know, there's a theory that says every time there's a new hard fork, it splinters the community. That's the same reason Bitcoin Cash still has $10 billion of market cap and Ethereum Classic still has $3 billion of market cap. There's kind of the splinter of the Bitcoin community and the splinter of the Ethereum community that for, you know, whether even if it's just a market maker on some centralized exchange somewhere, you know, that that value is continuing on that chain and not on the consensus chain. And so would you agree that each time one of these hard forks

happens and it sounds like there's a lot greater velocity of hard forks, a lot more hard forks that happen in a short amount of time, it splinters the community and therefore loses some of the value here of of the chain and the value even if it's just denominated in market cap. I is there a risk to having too many hard forks or do you think this could actually bolster the social consensus and kind of rally the community on a more frequent basis and we're not thinking about hard forks the right way? >> Yeah. So first of all to address your first point around the specifics of the governance process I don't think I discussed anything around the specifics although I did give an example of one where you could just hardcode it in the

node software. the specifics. We're we're working on some proposals internally. Other people can propose things externally. Uh but there's various ways in which you can execute this governance with various uh with participation from various parties including token holders. So just because I say it's governance doesn't mean the token holders get no say whatsoever. Uh not necessarily they have no decision-making power but that they have no say. No, that's of course wrong. Of course they can have a say. uh of course you know the votes of token holders can be surfaced it just the mechanism which is done is not staking it's governance so that's the first point to clarify uh the now onto the hard for point uh I guess we'll find out uh Ethereum has

done a bunch of hard forks throughout the years like there was the contentious hard fork around the DAO and that splintered the community but Ethereum has had many non-contentious hard forks and those basically did nothing with respect to splintering and they did a lot to bolster Ethereum as you know being the success that it is today because you know the the Ethereum classic chain that refuses to hard fork I guess the Ethereum chain that you know does hard fork continuously to improve itself one of these is worth you know whatever x billion of dollars the other one is worth whatever y million of dollars uh so one one approach is you know more is better than the other but it really I don't think hard forks necessarily splinter things unless

there's contention, but if there's contention, there would be contention anyways. >> Innovating finance, global institutional and onchain deep liquidity to drive capital efficiency. Now maximize yields on institutional grade liquid staking with me. Unlock trillion dollars of capital via Fitcoin. 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

>> and the way in which that contention is resolved now is basically like one one kind of one of those you know camps loses and it's effectively like the way that we elect a political president. It's like you know just kind of half the country is unhappy for the next four years and it in the the proposal of DAO or it effectively feels like Dow governance but effect it it you know has a lot more significance on the outcome of the chain in proof of governance. It it's not just the case that kind of this camp is unhappy. they can continue in the case that they would like to uh or they can kind of just defect, right, and join join the the uh the winning the winning fork that takes place. But I

guess what you're saying is you don't expect too many of these contentious hard forks to occur. It's really just a matter of accepting new validators into the system um and paying those validators more uh more cost plus model rather than coming up with an arbitrary value in which we're going to inflate the supply and pass those on to validators. >> Yeah. Because currently it's this awkward situation where validators that are and this kind of I touched on this earlier is that validators that are in the top of the validator set are potentially highly profitable and those in the bottom are less. So uh so it becomes this game where you know validators want to be in that very highly profitable top part of the set but then the protocol is to figure out

how to pay everyone to make sure people aren't too underpaid but then you're overpaying the top percentage etc etc. it gets it gets complicated to model but also it can result in overpayment to validators because those at the top are overpaid. So it becomes this thing where validators try to get paid a lot uh in the the current model and that's that's not necessarily what you want. It would be nicer to have a model where validators are paid to be cuffs to run the pro to run validator software that produces blocks. Uh and they get paid by the protocol uh you know their costs plus some profit. Uh not you know not 100x profit some respectable profit so that people will do it. Uh but because

of this change in how validators can be compensated and will be compensated uh they'll make a profit don't get me wrong but it means they won't be having they won't have to search for the 100 they won't have to hunt and play games and stuff for the 100x profit. >> Uh what that means is also that there won't necessarily be huge competition to run a validator. Uh and therefore I don't expect people to I don't expect you know a million people wanting to run a validator. uh the people who want to run validators may have some secondary services that they that they can provide along with running a validator node uh that can they can advertise cost on and stuff like that right maybe they run a roll up as a service and they also run a validator on the side things like that so because of that I don't expect you

know a large churn in the validator set and therefore there's not really any reason to change it you know people people who aren't validators aren't losing much because the validators just make a small profit right obviously if you pay validators there's 100x profit. You know, it's kind of unfair to gatekeep who is a validator. If a validator just does their thing, they're on as validators. They make a small profit, small but respectable profit. >> Sure, I I guess it's nice to be a validator, but at the same time, you could just not be a validator and put your money into DeFi and make potentially more yield if you want to take some risks, right? Uh so the the how validators will be compensated will be more optimized and therefore you won't have people playing political

games to try and be a validator like the the proiles will be much more mechanical. It'll be much more you know okay I'm a validator I just run my software and that's it. I get paid some profit and therefore anyone can >> yeah it is a political game today. I mean yeah >> I I've just got one more followup. Um, and you you mentioned that optimizing the value equation to pay validators is going to bring the cost that it it to them way way way down or or excuse me to the protocol way down, right? Um, and you mentioned that opens up a variety of use cases that you can do. You can, for instance, burn all of the transaction fees. That's an amazing value acral thesis for a for a gas token. Um, you could also lower the rate of inflation

potentially all the way down to zero. This is kind of where my question is. I guess it my question is who even if it's you know the cost of the protocol right protocol kind of in quotations here is much much lower because you've optimized the value equation who is the one that pays whatever cost that there still is to validators if it's not coming from transaction fees is it coming from still inflating the supply albeit much less or is it really just like a servicing agreement with these validators in which could be paid in fiat or stables or from venture funding really just as a software as a service doesn't necessarily even have to be tokens.

>> Yeah. So the way the mo the way that's uh most intrinsically incentivized is the protocol issues a small amount of new tokens just for the validators. So this wouldn't be given to stakers because there's no stakers. It would just be given to the validators. But it could be a very small amount of tokens. Uh uh we've also been thinking internally again by we. By we, I mostly mean Nascu since he's the one driving this around how you could do things like pay them in stable coins because they get they have their costs in stable coins and it's awkward when they get paid uh they get paid costs in some fluctuating token which may go up or down. So it could be end up the protocol overpays or underpays them either of which scenarios are good. If they're

paid in a stable coin then they know exactly how what they're going to be paid which is potentially very nice and is also nicer from the protocol perspective. This requires you know knowing the exchange price potentially because you can't issue new stable coins. You can't issue US dollars for example but you can issue if you know it's Celestia protocol it can issue TIA. So we need to know the exchange rate blah blah blah. So there's some some things to be done but regardless you can uh the issuance wouldn't be dropped to zero. it would be dropped substantially like we're talking 20 potentially 40x that >> yeah I think I saw it was like it was like uh 20 like a fourth of a percent annually or something like that >> last thing for you

>> and then at and then at such prices uh it actually uh the value cruel thesis of the token becomes very compelling and very interesting because it goes from having to uh burn fees that beat the issuance and then more to if you want the token to be deflationary uh and have you know effectively net buybacks of the token. uh now you only need to burn a quarter quarter of a percent or less and suddenly that is much more in the much more feasible much more feasible range like uh I think someone did some math internally and we're looking at numbers like you know if you get uh 10 or $20,000 in fees a day which is substantially less than you know protocols like hyperlquid like you

know $10,000 of fees a day if there's large DA demand and stuff like that is quite easily achievable you get that and suddenly t is deflationary. >> So the value cruel thesis of uh these tokens that will adopt proof of governance becomes much more it becomes much more obvious because burning transaction fees etc. There's no interaction. just okay here's a here's our value the thesis thesis becomes obvious and then the actual amount also becomes better because you don't have to you don't have to surpass inflation and issuance uh in terms of the fees burnt and again the issuance is okay you set you set some high number and then you pay half of it in taxes what's the point >> taxes man don't get us started on that

>> I think this is the crux of of the proposal right it's still and I think you know It's it's a very thoughtful proposal with a lot of detail. Um, and it but it is almost somewhat revolutionary in the way that you know we're thinking about proof of stake and and now thinking that there is something better out there and and it it's it's so different but yet you still have this somewhat of the same inputs and outputs. You still have some inflation and you still have some tokens that are burned and you get these validators that are providing a service. The only difference is that rather than overpaying all of the top validators and underpaying all of the lower validators, you're now optimizing the value equation so that the tokens that you burn are

effectively, you know, those would be passed to the validators if it was like, you know, Bitcoin after the the block reward is reached, right? Validator or miners in that case would get these transaction fees. But if that was the case, then in in in the current setup, they would be massively overpaying the top miners and underpaying the bottom ones. And this sort of balances the scales so that we're still using the value that is flowing through the system to to pay for the upkeep of the system. We're just balancing that equation quite a bit more by unbundling uh the validators with um the the activity going through the system.

Yeah. And and the and for what is we didn't really touch on this part, but in terms of of the security of the system, the system can actually be very secure against corruption, potentially even more so than proof of stake and proof of work systems, right? Proof of work is a bit interesting. Can't really do that to the same extent in proof of stake in some scenarios. But regardless, the point being, you know, in proof of work and proof of stake, you can corrupt a small number of operators and then because of the principal agent problem, you just corrupt the operators. You don't have to buy stake to overcome uh to overcome the amount staked. Like a lot of people when they model security, they say, "Oh, you have to buy a bunch of tokens." It's like, "No, you don't.

You just go to like if you're the government, you go up to Lido and say, "Hey, Lido, I want you to start forking Ethereum, right?" And Lido Lido Lido company will do that. Uh whether they like it or not because the alternative is to go to prison, >> right? Now, this is not to say Lido is bad. Far from it actually. They they seem to be quite good actors, right? But the reality is that you just go and corrupt operators. >> With proof of governance, you can get a uh you can have a thing where part of the governance process around deciding who is the validator not is not just looking at a blind list of validators and saying, "Oh, I'm going to choose the top one because the top one gives me the has the lowest commission because they have economies of scale, right? Is or

like if you're in a mining pool, it's not just I'm going to choose the top mining pool because they have the lowest commission again because of economies of scale, right? there's no commission to be there's no staking therefore there's no commission on the proof of governance validator selection so you can have things like geo distribution uh uh hosted uh hosted versus bare metal distribution uh you know platform distribution maybe some people use Linux some people use Mac to reduce platform dependency risk uh things like that uh that allow you to have a large number of validators that are geographically distributed in different jurisdictions and then to corrupt the proof of stake protocol you need, you know, 1/3 or 2/3 of the validator set, which in the case of Celestia is 100 validators. So to

corrupt oneird of them is 33. This would give you an Nakamoto coefficient of 33. While Ethereum's Nakamoto consensus uh coefficient for one/ird of the voting power is like two, >> right? So now you end up in a situation where you have to corrupt a substantially larger number of entities that are geographically distributed and like you know that and like you know you you can choose them such that you know there's there aren't 33 validators in any one jurisdiction. So it's you have to do things like cross jurisdictional collaboration if you want to corrupt them. So it can actually be very secure uh potentially under certain under certain attack scenarios more secure than any other proof of work or proof ofstake protocol that exists today.

That's not to say that they're more secure under all scenarios, but under certain scenarios, especially the ones that people are talking about with respect to government overreach or government reach depending on your positioning and and governments uh and you know World War II style scenarios, this geographic distribution and this Nakamoto coefficient actually becomes very important uh and that's where proof of governance uh really >> John thank you for for your time and for your work in the space. Uh, I appreciate the in-depth uh explanations and thoughts as well as the uh the spicy takes. You'll see some of those on your timeline later. I appreciate >> Oh god, help out.

>> Validators or is is definitely a bold one. I think we may see some chatter about that. Thanks, John, for uh for giving us a laugh and also sharing a lot of this really really great info. >> Awesome. Thanks for having me. Speak to you soon.