Proof-of-Stake Validator Responsibilities in Progressive Decentralization / Daniel Hwang
EthPrague 2023 · June 2023 · 18:45
Daniel Hwang outlines the fundamental responsibilities of the validator infrastructure layer in proof-of-stake blockchains. Speaking at ETHPrague 2023, he looks at the role this staking layer plays as a blockchain network progresses toward decentralization.
Speakers
Topics
From EthPrague 2023, hosted by ETHPrague.
Transcript
hello [Applause] um yeah my name is Daniel I'll be talking about validator responsibility and self-regulation so a little bit about myself I I uh I used to work at stake fish and let their protocols team and um I used to head the ecosystem team at Terra before I Rage Quit in 2020 because of all the issues I saw there and that's how I got into the validator ecosystem um so this is one of my favorite quotes by Philip monk from the urban ecosystem if you don't understand a system you're using you don't control it and if nobody understands a system the system is in control I think the validator ecosystem
is a particularly misunderstood ecosystem in terms of its agency and Independence to guide Progressive decentralization and manifest that and I think we're in this sort of pivotal moment especially with regulation and the West sort of hating on us for us to be able to make something of ourselves and guide us into a more decentralized system and so right what is this what is the System supposed to be right we have delegators validators the centralized foundations the core teams let's say for ethereum specifically we have like the seven or so client teams and then um the the variety of connections
between each of them so I'll be talking about some of the responsibility I'll be talking about some of the self-regulation and then perhaps some of what we might be able to look forward to doing um maybe one way to sort of visualize um the the cycle is if from the start of POS maybe five-ish years ago we were maybe children in terms of these validator entities that had an understanding of what we were supposed to be doing from the design that was passed on to us from core teams that were building these l1s with various POS mechanisms maybe right now we're in the sort of adolescent phase um where right there's enough money
going around we've had a few years of experience under our belts and I think we're at this sort of Junction where in order for us to mature uh and be taken seriously uh we have to figure out a lot of these problems that we've been seeing and that's been resurfacing time and time again um right more specifically we don't want to have like these these governments say like you know play time's over we'll take it from here and have like these centralized panopticons so responsibilities right I think the most obvious example is infrastructure management like the devops side let's let's imagine that's very well developed strong right arm the second is perhaps
governance which is maybe a lesser developed left arm and then right the purpose of what we exist for we as the validators should exist to manifest that type of progressive decentralization that a lot of blockchains or like their white papers say like eventually it will be a decentralized system um and so what does that look like and I'll use a few examples from other POS networks like Cosmos and polkadon um but I think it also applies to ethereum maybe the analog to ethereum is that we have these service providers that are sort of these pools allow like the mining pools and the Bitcoin mining Farms we had in pow
so there's there's fiduciary fiduciary responsibility in governance responsibility the fiduciary part is perhaps where we get to a lot of this these uh head budding with governments right these sovereignties that want us to protect their their citizens and the governance aspect is sort of where we're kind of driving uh these l1s as an example right this this this governance responsibility is right once these hypothetically once the centralized Foundation core teams disperse and then right everything gets handed off to validators and we support we're supposed to take on
that responsibility of guiding These Chains um and so some this is ideally what we would have but sometimes they're they're not necessarily aligned I think one of the really good examples and more glaring examples we had was when I was at Terra um right validators knew a lot of these problems even before I had joined when I found them and I started poking around and seeing all these skeletons be Unearthed right validators chose not to do anything and that's largely the reason why I think I've been so passionate about the validator space and evolving it and then when I uh when I Rage Quit in end of 2020 so right even though validity is new they chose not to do anything about it
which is quite contradictory to the existence of validators being these separate independent entities that are supposed to exist alongside these blockchains and these core Foundation teams right basically these foundations were saying jump in all the validators were saying how high which is like then why do values exist independently if they should just all follow the follow the leader so communication is important I'll run a few through a few of these pretty quickly so a really good example with this right like the self-regulation that we hope to see in the validator ecosystem that may be the appropriate complement to existing alongside Sovereign Nations
that have certain um right have certain like things that they want us to be doing right protecting their citizens is right like we have regulatory bodies from these centralized organizations right this is exact picture of the SEC which has been like pushing out all those filings against coinbase and finance and specifically in some cases highlighting the the earn related products um right but like this this centralized example evolved alongside right hundreds of years of governance history and I think what's funny is that right none of us claim to be Alexander
Hamilton but we've tried to create like a system right without without these hundreds of years of experience maybe we've had five um if we're lucky and so right then who is responsible for keeping us in check I think one of the most misused mantras that a lot of us have have echoed to the public or these non-cryptonatives is like that whole don't trust verify Mantra which is is completely false right there's very few of us even those who are technically inclined know how to check the code or know where to check the code or even bother to check the code and so if it's not the people then right what what happens right who's
supposed to deal with these Bad actors in a decentralized system and then who's supposed to Define what is bad behavior there's obviously you're right like you know no double spends but that's from the protocol and Technical level that's important but when you get into the more ideological perspectives then then then you have some issues right this is a funny picture I asked actually used to work at a law firm that was defending HSBC in the 2012 Congressional hearing where HSBC got caught for money laundering and terrorist financing I actually quit the law firm and that's how I got into Bitcoin mining because I went to you know I didn't like the banking system
um but right this enforcement thing this is a very the most glaring example of even in centralized environments right no one went to jail slap on the wrist um and then HSBC continues to to exist with all their bad bad doings um right there's other issues that we've seen in the existing POS Community right Juno is a cosmos based blockchain that completely violated self-sovereignty by yanking uh almost well at the time at the peak it was 134 million dollars with the Juno tokens out of someone's wallet um through governance right and the cool thing was it was like 99 of all stake Juno voted on this
um so right there there are these particular assumptions that we have from these decentralized systems that expect these systems to behave appropriately maybe what we think as like an efficient market or an efficient governance system because it's decentralized people will natively want to kind of participate but when there's like this type of decentralization we also have this bystander effect and these Assumption of rationality and assuming like what we think is right is going to be upheld because it's open source and people are going to be looking clearly that wasn't the case and we have examples like Tara now um right most don't go care this is a I think these slides will be up sometime
but this is a really good example of um this sort of like paternalistic attitude that we had all expected centralized organizations or regulatory regimes to protect us with and then right we maybe assume that some of this would be taken up by others um in a decentralized system but that is also very false false in the sense of right like there's not really good mechanisms for that to to to occur and the type of people let's say for example the infrastructure service providers AKA validators who have the opportunity to do something literally on chain and coordinate most of them are barely surviving right they're right maybe the
usual suspect validators who are the larger ones have the ability and the time and the money to direct towards initiatives like this the rest of the validators that are outside the top 10 for example they're just struggling to keep the lights on and so examples are right we can have activist shorting that come from outside infrastructure providers we can have cancel culture um right and then this is like post post uh crash cancel culture or post exploit cancel culture and then right like rely right if you go back to here like the the types of people that then if we don't have these types of mechanisms from sophisticated actors to support
these ecosystems we have like a uh we have to deal with like the people right because the community makes up the uh the the ecosystem in terms of like the governance or applying pressure to particular teams to change the chain this is a really good example um I don't know if you recognize him but his name is Luciano Maji who is a Juventus director who had been caught um right intimidating rafts like locking them up like you know fixing games the interesting thing with about this was that like the fans loved him still even though there was clearly wrong doing and so that's I think some parts of the Dilemma which I think we're familiar
with like the sort of bag holders dilemma even though there's like a token or you know it's horrible or you know there's something wrong going on you have so much invested in the success of that project or in this case the team that right maybe what is right or wrong it becomes subjective obviously not to the outside world that's again mob rule that we have this is a example of what we had when in 2014 when we had no strict or very clarified controls or standards for how we dealt with catastrophic failure in our systems jihash was a mining pool at the time that had exceeded 51 percent in its mining hash power that was directed to it and they
literally just like shut down the front end of their website so that no mining mining Farms could contribute to it and that was like a really bad case of what the hell are we doing um and and so this case right we have tyranny the devs versus tyranny of the majority um and uh right this is the type of governance that we have I'm going to go quickly through here so I I think um what we can kind of look for in terms of like this sort of maturation step for our industry as infrastructure providers or validators who have the opportunity or even what we would say a mandate to guide these these ecosystems is that we have to figure out what to do when when hits a fan and this is a really
clear example um of what hitting the fan right the West is completely antagonistic towards crypto and we have here right Gary Gensler declaring war on all these centralized exchanges and um you know coincidentally him highlighting staking and earning yields as a a point of of contention for what is deemed a security or not and so right we have certain options and this is a very abstracted one right like three very like obstructive examples like we die right like crypto dies um right and we all go fall under some like centralized panopticon chain right like like usdc uh we we go dark we all
go Anonymous like right like the last line of let's say defense in terms of who's going to power these blockchains are not necessarily the token holders but people like us who power these nodes right where do we power them do we power them in data centers bare metal in our in our basements in Cloud servers that are ultimately beholden to particular regulatory regimes where they're located or we find a way to which is not necessarily the most ideal option but perhaps the most immediate option we have to self-regulate this is like a funny picture I made with mid Journey where it's like you know the adults are coming and telling us like what have you done in your sandbox and if we have nothing to show for it then it's like
you know they're gonna take it from there right and so this is an example of fed chain um uh well this is supposed to be a the video but I don't know how to play it but this is a scene from Mr Robot are you able to print oh yeah control The Ledger and the mining servers we are the authority I will make sure you have visibility into every single wallet that's open every loan every transaction which this is a clip from I think like the fourth season of Mr Robot where they were talking about equine where it's like a centralized constantially
controlled Ledger that everyone has access they have access to right back doors Etc which is pretty antithetical to like why crypto exists as a permissionless system um and so that's example if we're dead this is example when we go dark are we going to all live under caves and um you know have like machines strapped to our backs and you know carry on the good fight but right like people like me I I don't want to live like that you know I like living in normal society and then there's an opportunity this is like a corny slide but like self-regulate and what that potentially means is um we have to figure out a way of right no offense against like the network State thesis but no everyone lives somewhere
we all have to as infrastructure providers host our infrastructure in some legal territory around the world we're not see studying on you know on these like abandoned oil rigs um right like we we are beholden to someone and so in order for us to make sure that we protect some of these principles as infrastructure providers of where we decide to host like Cloud providers data centers Etc we need to figure out a way of like not just being obliterated right and so right when it comes to the responsibilities and engaging some of these institutions or regulatory authorities we have right delegators have to take apart validators
and core teams and where in particular like the validators are the ones that continue to exist right l1s will come and go right and we've seen them come and go but validators are always there kind of learning those lessons right communication has failed right allowing people to engage has failed right like trying to validate or trying to engage with um the community on Twitter is like yelling into a storm that's failed right and so so like we're beginning to see like validators rise up as a class um right and finra is a good example not the perfect example of what potentially we might be able to see when there is this type of self-regulation that exists to it to allow like this sort of like
interplay finra is a non-governmental organization that is a Consortium of financial institutions in the United States that has the ability to self-regulate and enforce each other that the government allows sort of like hands-off approaches maybe that might be uh one way but right before we can get to that we need a lot of work to be done right there's a crypto political parties as an example example right activism and so this is the last slide but right something what we need is like this type of standardization Frameworks right classification enforcement formalized governance and the and most particularly maybe at the protocol level these types of
incentivization mechanisms that crypto is so Adept at at creating um so I think this particularly is the opportunity for validators themselves perhaps to work with core teams to understand how we move forward together in an environment where there's Clear Eyes On Us and clear action happening where we need to figure out what to do um so yeah that's that's it I only have a minute but if anyone has questions maybe um we can we can do some or maybe we're just we're just done no no questions okay thank you what no what not there no not there
okay thank you anyway [Music]