Infinite Games for the Modular Future

Transcript

welcome welcome we have an expanded guest list today to a very very big guest list full of Giga brains ready to talk about infinite games really what that means um it's modular March I'm here with Robbie of of course um and we have our good friends here to discuss how infinite games are going to lead to coordination in the mod future so let's jump right into it and we'll just do a quick brief round of intros let's start with the two marks from optimism and then move to allejo and the igen layer guys welcome everyone hey my name is Mark and I'm a co-founder of

optimism um I'm really interested in building the super chain to help self coordination problems we're building more than a blockchain we're building a world government yeah and hey everyone my name is Mark uh Mark squared I'm on the Partnerships team at optimism and what I focus on really is just being a steward for the super chain so I help uh many of the chains launch to the super chain um and I'm also an observer of the defi ecosystem and seeing how the chains can really plug into the defi ecosystem cool my name is aljo I'm the founder of lumio and we're building way to credibly neutrally uh deploy alternative VM

applications onto ethereum on all the major l2s uh this enables a way to deploy move VM uh Solana VM uh in a way where we minimize uh trusted third parties so that uh all the major applications from out those Su salana can come in the pl on ethereum and Teddy hey guys uh I'm Teddy Knox I'm a research engineer at igen labs and uh I'm working on Igan da and um Igan da if you don't know is a decentralized data availability ABS built on top of aen layer and um I'm working with Mark and

various other people to integrate I da into rollups uh everybody I'm sham I uh started the L project nearly three years back uh excited to be here and talk about infinite games coordination and open Innovation awesome um I think uh I think it's cool to preface this conversation by saying that uh blockchains are almost like this distributed Network in very in general um and and they are a mechanism to coordinate across this distributed network of nodes often times there's

humans running those nodes sometimes there's machines running those nodes But ultimately these blockchains are coordination mechanisms um so I'd love to pose an open question um and and start out by talking about uh how infinite games run and how we coordinate these games using a blockchain um open question anyone uh feel free to take a stab uh I think I think it's important to kind of Define what an infinite game is and obviously you know this is subjective but to a certain extent uh you know a finite defining finite game with a set of rules and winners or losers and an infinite game being one that doesn't have uh predefined rules

therefore there's no winners or losers and the sole outcome is to continue surviving indefinitely and you know you can think of still uh you know as adversarial but I think an interesting way to think about within the blockchain space is how can we win is how is there an outcome where you know there's growth and we enable everybody to win not just one L2 not just one data availability solution so I think blockchain's enabl this and yeah super super excited to you know hear what what others here in the room have to think about that there is this article that Virgil Griffith wrote a like years ago at this point where it talks about the game theory um like academic Game Theory right where uh they've kind of figured out like all the different standard

games where the equilibria LIE how actors should act right but then if we introduce blockchain as a new actor is kind of this like trusted Oracle system where there's like this hardness to the information right it can it can change the um outcomes in all of these like Game Theory games right and this is like a very under explored area and it's going to be really exciting to kind of see how um the equilibria and all these games change when we add this new actor and I think one of the most promising paths to find like product Market fit right is is figuring out you know these

uh abstract you know academic games how they play out in the real world right and finding those Industries where they're playing out and then and then if we can model the differences in the behavior right when we add in these new actors the blockchains as oracles and and then that that could uh you know help us figure out where blockchain can be added into the real world and actually solve real world problems where there are coordination failures I think you were actually the one you can go for it yeah I was just

going to add you know the first thing I think of when I hear in infinite some games is uh just the the anecdote about the prisoners dilemma where if you're in like a a single prisoners dilemma game um there's uh no incentive to cooperate and be honest um and then once you repeat it which sort of introduces this idea of you know the infinite game um suddenly there actually is the dominant strategy of cooperating and working together with other people I think to me uh you know beyond just the specifics of um how various projects are working towards this I think you know this infinite games meme is about um switching from like a very Zero Sum

mindset where we have to you know act Cutthroat and um more to a a cooperation mode cool yeah the only thing I want to add here is that what makes all this super interesting is that what blockchain and blockchains have enabled and what ethereum has enabled is the ability to you know essentially program um value right and so when you think about games like incentives like the kind of the Bedrock and the foundation of all games are incentives and we've created really interesting incentive mechanisms within ethereum with proof of stake right no one would want to collude and essentially um try to take down ethereum because essentially once you do that

like your ethereum is going to be worthless so there is such strong incentives within crypto and within blockchains to create these infinite games um so yeah I think it's like a really relevant discussion that be having um and I'm really excited to see all the different types of infinite games that we can program within AUM so um yeah just like there's a lot of design space around this for sure and also just one more thing to add is like it's all trustless right it's like the whole ether like the whole one of the principles of ethereum is that it's like a trustless system so we don't have to trust that someone else is going to play the game according to certain rules or how we expect them to play

there's going to be incentives that make sure that they are acting in their own self-interest that benefits the entire um system so those are like just kind of the principles of games that kind of build up into this infinite game idea um when you think about um you know there are two concepts here one is infinite game and an infinite sum game an infinite game has is temporarily lasts forever right and an infinite sum game you know there's there's a zero sum game where like you know two players play and the total amount of Pi is fixed

and you have to fight for who gets how much there is a nonzero some game where the total amount of P can be expanded right and we call it an infinite some game where newer and newer forms of the game is created over time and people the the thing just keeps ever expanding so there is a notion of time of being long and infinite but also in there being lots of interesting things happening and the pie growing and you ask why would why would nonzero sum games exist so now like what what you need is if time is infinite and there is a non-zero sum game to be played at each

time then you get an infinite sum game where like there is potentially a lot of things to happen positive things to happen over the long Horizon and if you start with this as the starting point and then ask like why do we have non-zero Su games to begin with there are really only two foundational substates I I've found on on this number one is Innovation right Innovation is taking something which is a non-resource and converting it into a resource take oil and convert it into energy I'm taking air and converting it into Spectrum I'm taking sand and converting it into silicon right Sand by

itself was not a resource so it was didn't factor into the sum anywhere now in know silicon and intelligence emerging out a silicon is something which is you know net additive so Innovation is one one dimension which is nonzero some the other one is coordination which is you know the seven of us here could have gone our own way this hour instead we come together and have this discussion about this thing there is an there is a super additivity to it there is a and it's possible because we coordinate so these are the two things coordination and Innovation are the two foundational substrates that enable and Empower infinite sum games so if these are the two things like what does what

do these two have to do with blockchains and the way we think about it is if you look at uh you know I already Mark alluded to Virgil Griffith's writings and many others in the ethereum community if you look at the kind of cultural uh canvas look at the Bitcoin Community the idea is hey you know sovereignty and I don't trust anybody else like you know how can I survive in a kind of like a nuclear breakdown like all this kind of stuff the ethereum community created this very rich culture of thinking about coordination as the

foundational thing that the blockchain is doing and you know Virgil is one of the you know leading thinkers there of course you know lots of other people starting from vitalic contributed to this um so you know as as you know I was entering into the blockchain space kind of understanding these strains of thought was really uh powerful particularly you know Virgil's idea is that once you have a commitment device how does that transl that modulate the structure of the game itself because you know if in a prisoner dilemma I can make a pre-commitment that if I defect then I'll get slashed then what sets of things can you now do that you cannot do

a priority because I could not make that commitment so we think about this question a lot so blockchains as commitment devices enabling more coordination so that's that's one broad broad category but the particular way we think about it is when you when you look at like coordination there are many different types of coordination like you can think of like an authoritarian can actually build like powerful Kings existed who actually created very strong conditions of coordination right you you can have an authoritarian who can actually help create a coordination uh system you can have an elitist coordination system where the elites like uh CCP or the SEC or other three-letter four-letter agencies we can

talk about create you know coordination systems you can have you know deeper than that majoritarian coordination systems where a majority of people can come together and and you know agree on certain things and a pure democracy where a majority can do anything would be a majoritarian coordination system even that even that majoritarian coordination system is subject to the tyranny of majority right a majority can do anything to the minority and this is something I was mind blown when I kind of tried to understand the etherum culture Thinking Beyond systems of majoritarian coordination because you

know all the academic work on blockchain for example was oh if the majority of the nodes are honest you know AB BCD E F G H right this is what we used to think about and you know coming and reading for example one of vitalic article back from 2014 uh talking about subjective ocracy which is the idea that you know anybody can Fork you know away and create a new state if something wrong happens so you know in in my framework I call this last tier of coordination as self enforcing coordination systems self- enforcing coordination systems are systems which are not subject to the turny of majority like because we all pre agree on

something now and it is self verifiable what the thing that we agreed on is there's no need for me to like you know look at anybody else to give a historic example gold as a value allocation mechanism is a self- enforcing coordination mechanism why is that because you know imagine I need to figure out how to allocate value across individual uals I could put up like a big Blackboard and like keep accounts of everybody like you know in all different religions there's this conception of either God has this like Ledger or you know some individual Dy has this Ledger of record for all of what like good and bad anybody did the but gold as a value allocation mechanism is self- enforcing in the

sense that you know you give me a piece of gold I can self verify that you're giving me a piece of gold because that as is biged into the piece of gold and it's self verifiable for me that I actually attained it and this is not subject to the tyranny of any majority because no majority can come together and say declare by Fiat that the gold is not valueless because Mark and I or Alo and I can still trade with gold nobody can come and prevent us so the gold standard of coordination mechanisms is or the most robust among all these coordination mechanisms is self enforcing coordination mechanisms and that is what blockchains are accelerating and that is what we here to

build self enforcing coordination mechanisms where you know once you're bound into the coordination mechanism just like we all agreed on gold if through an evolutionary process and now you know nobody has a simple way to move away from it we bind ourselves into coordination mechanisms that are self enforcing and blockchains basically help us turbocharge the era of self- enforcing coordination mechanisms so that that's our thesis one you know last comment here is if you think about Innovation and coordination Innovation runs on top of coordination mechanisms and the more rigid the coordination mechanism is the better you can actually go and innovate and a simple example is

like you know I mentioned democracy as an example of uh majoritarian coordination but a constitution is an example of self- enforcing coordination the bill of rights is something that we all hold individually like and if any update to the to a bill actually breaks my Bill of Rights then people rise up and say hey my constitutional rights getting violated so Constitution is an example of a self- enforcing coordination mechanism on top of which innovation happens on the free market and what we see is over the last like you know 300 years technology has massively accelerated the free market without giving any acceleration to the coordination mechanisms we still

have the same you know Constitution we still have the same uh you know LLC we still have the same double Ledger bookkeeping we still have the same uh you know uh Securities loss from 100 years back all of this are you know and the question is how can technology accelerate the arrow of coordination and that's what we here yeah that's super interesting it's almost like you know beyond you know physical or digital gold being Bitcoin like the blockchain is that digital way or digitizing this social coordination system um I think one one interesting thing that you said was how uh the value

itself can be infinite in these infinite games or uh infinite sum games and uh would be interesting to see it's almost like the law of thermodynamics that I think it's the second one that just uh you know goes towards entrop uh doesn't mean that you know it has to be finite things can continue to be infinite um and I wonder how that you know it's kind of like a metap metaphysical concept that we're now turning physical by writing into blockchains which are you know bytes uh you know zeros and ones but it's kind of interesting how this then can play Maybe into tokens or values of tokens so an interesting one is securing securing this social coordination layer right so etherum obviously a very important value like the the literally the security of ethereum is now you know dependent on

the value of ethereum so yeah I mean it's super interesting and idea like seems like theoretically the value of ethereum should be infinite right as we go towards Infinity um but yeah I mean it it's super interesting that you know all these you know all these theoretical Concepts have existed but now we're just digitizing them and allowing us to better coordinate um you know in a digital world uh just adding something a little bit to that the when you you know shinger wrot a famous piece called what is life where he defined living things as basically locally entropy reducing so life is basically because you know entropy increases is the second law of Thermodynamics but life living things

actually reduce entropy locally because they're able to reproduce give rise to something which looks identical which in a random World should never happen so you know if you look at how life or the direction of evolution itself I would say right on exactly these two principles coordination and Innovation and Innovation is the law of mutation and natural selection each entity undergoes mutations and then natural selection which and forces them to survive better otherwise they get extinct but also coordination is inbuilt into the tapestry of like Evolution for example when we went from single cellular life to multicellular life

right one of the core things that happened was this transformation from procaryotes to ukar where two independent cells fused together one became mitochondria and the other became the nucleus and the nucleus specialized in information storage and mitochondria specialized in energy production and these have independent genes so they're actually two distinct you know units you know when you think of a selfish Gene there are two independent genes like fighting or coordinating co-evolving inside a cell and this was the first coordination mechanism that was kind of built and this coordination mechanism was very very very important in giving rise to multicellular life because you

could never have a lot of information storage you could never have a lot of energy to actually go and grow a multicellular organism without the coordination and so we see this you know at a fractal structure in like Evolution itself starting from you know mutation and you know selection operating on individuals but coordination arising very you know coordination mechanisms are rare to arise and even RAR to mutate because coordination mechanisms are R toise why because it requires binding commitment from many many individual entities to come in and they have to somehow sacrifice self-interest to come into a coordination mechanism right so

it is a very very difficult thing to arise but even more difficult to mutate because once you got locked in you can't unlock that easily and it's true as much for gold um as it is for other self- enforcing coordination mechanisms we are seeing today and some of them are good and some of them are not very good and I'll give a you know being an ethereum myself I'll give a not very good example of from ethereum because these coordination mechanisms because they're binding create a certain tyranny to them um like for example if you want to move away away from gold none of us have the ability to move away from gold we're just stuck with it you know as though it

has evolved its own agency we stuck with it without our power to actually go in and say oh we don't want gold anymore you know you may laugh at it and say like girl's useless but it's not going anywhere and so what happens is coordination mechanisms have to be outco competed by other coordination mechanisms otherwise you know and that requires an independent ution of another coordination mechanism to actually arise which all of this makes it very very difficult but an example on in ethereum that I wanted to give is hack wa we just saw a hack like two days back on you know one of the layer tools right and we can all see from the outside it's obvious to us that it's a hack and we sit here like without

agency being able to act anything because we have bound ourselves into a coordination mechanism which doesn't give us that expressivity so so when we thinking about like coordination mechanisms the the principle that at least I use and we use at I is karma karma is action re you know you you give an action into the commons you need to get a response which is proportional to your action into the commment and that's the guiding principle for how to design coordination systems you do a good action into the commons it should reflect it back at you you do like a bad action into the commons it should reflect it back at you if I go and hack and the system says welcome you know more hacks right now

you know some country creates a whole division to go and do hacks right that's the natural consequence of having this inside the system so you know there are ways in which we've done a great job as a community there are ways in which we absolutely have to improve and this is one example on the topic of coordination mechanisms evolving right I think an important thing to think about is the cost of coordination right because we have all these different kind of actors and the cost of them coming together and making a collective decision and acting together will greatly influence whether or not coordination

happens and something that I like to imagine is I like to think about like the cost of coordination over time right I like to think about maybe uh before cell phones were invented let's say that you wanted to meet up with a friend at the park um you know you just have to tell them a time and a place and just hope that they're there and if they're not like you know it's really difficult the cost of getting a hold of them is like a lot higher than say today where cell phones exist where you can just reach into your pocket and the cost to coordinate is much lower because you can just instantly message them right through your phone right so um what I'm what I'm very curious about is you know

what is the next 100x reduction in the cost of cordination and what does that feel like to live in right is our blockchains going to be the substrate for Designing this right and and trying to imagine this world it's almost like going through a singularity because it's very hard to to think of like what does this look like but you know the closest that I've gotten to is I kind of been imagining this like Global calendar system that it it acts on your behalf to figure out who would be good for you to coordinate with depending on your goals and then it would automatically populate

your calendar and kind of give you recommendations you can kind of imagine it being this like global system that is directing people like an invisible hand to like meet up and create value and exchange ideas and this is one non-financial application that I would love to see be built on blockchain it's one of the reasons why I'm so interested in cryptocurrency because of this you know potential future where we can greatly reduce the cost of uh coordination more than today just to kind of answer your question quick one one that comes to

mind and not to get like too abstract is maybe like humans evolving into machines or like kind of like the you know eventual you know merging with machines whereas like you know our brains might be eventually interconnected through each other and through machine learning algorithms we're just going to be optimizing like the best ways to communicate and and do stuff with each other and kind of like an interesting question to pose for the next person is you know if we are using blockchains as this coordination mechanism that's programmable and we're you know where we are abiding by the social construct that you know code is law like what happens when we can do human interv I anymore so for example the Dow hack the roll back period right that's something that you know we can look back in history and be maybe as like an exemption but you know

what happens when machines take over and you know we humans can't do anything more about it like are we okay with that I'll try that one um I would argue we should never reach that point I mean there's a lot of questions around how AI will play out but ideally AI ends up um being the Giants that we stand on the shoulders of and not you know the feet of and um yeah I would hope that uh blockchain's depart this is a bit of a you know um an idea that you don't hear a lot but I I would argue that blockchain should depart from a purely libertarian view of their role to play in the world um and this is something

you know s kind of alluded to a little bit with his reference to hacks and you know you as well Alo um we don't need to live in this sort of political economic sphere that's virtual um but has no sense of um of Justice or of of you know intentionality about what should happen I mean in general we have validators running ethereum um we have several different uh infrastructure services that uh support it and operators and actors Within These systems are slashed if they do any um behavor that they didn't commit to do uh beforehand so you know we can expand

this definition into what ends up recreating something looking like a legal system and the thing that makes me so excited about blockchain even today when I think like you know things might get a little bogged down in mem coins and things that are a little bit less big brained um is this idea that we can um set up this Sphere for competition between governances and this sort of gets into this idea of um exit or voice which is like a idea that was introduced by this economist in the 70s but essentially gets into the idea that like if you have a if you have a you know membership in some system let's say you're um a consumer for various uh

grocery items and you you know don't like the fact that the tomato sauce is now you know made with worse ingredients then you can just go and buy a different tomato sauce and so this is represents exit like you have this option of just choosing a different um option and that's more of the economic option the more like invisible hand option um the other option is voice which is to write a strongly worded letter to the tomato sauce company and say hey you should make your Tomato Sauce better again and as it is today we generally prefer systems where exit is the way to go like we like the fact that restaurants are you know not necessarily democratically run we can choose a different restaurant

if the restaurant isn't that good and this results in restaurants um which generally are better because they're supported by you know their good food or whatever um but this is not at all how government governments work because we're you know totally landlocked and and attached to them um I think like it's time for us as an industry to start to acknowledge that you know blockchains are more than just Financial rails or application rails but we're developing these political uh economic spheres where we can uh give people this sort of more exit uh based approach where people can pick the sphere that operates the best um

yeah things like Network States come to mind and this this idea of taking the onchain and and moving it offchain um and another kind of Point here that that that TI's made not only on the cost of coordination and that's that's the the most important one but the cost of Corruption of these systems is another kind of thought process to think through on this coordination topic so I I'm curious to understand how you guys are thinking about cost of corruption from a from a a cryptoeconomic perspective but also like a game theoretical incentive designed from like a human behavior perspective and um it's

it's it's an open floor yeah um I can talk about this a little bit um the when when you think about how what is the uh properties of the system I was alluding to to this a little bit like why is a blockchain secure and you asked this question for for a chain and then the first answer you get is oh you know you need all these distinct parties to come to agreement and you know they are um you know they're incentivized to do the right thing and then you ask what

does it mean to incentivize to do the right thing um you know there are at least I would say two distinct models there so one model is what is very common in the academic world called uh bi b bantin fault tolerance bft and bent fall tolerance means some fraction of nodes can behave arbitr badly and still the system should work correctly and that's one kind of a model to reason about these systems another model is what you know uh people normally talk about when they're talking about incentives it's called the Nash equilibrium model which is you look at the incentives and then you see whether

a given node will behave in this way if all other nodes are behaving in that way so then it's said to be an ash equilibrium so if all other nodes are doing this then will you also do it right like if everybody else follows the traffic rules then maybe you follow traffic rules is good actually if everybody else doesn't follow traffic rules you following traffic rules is anyway no good so you also don't follow the traffic rules so there are those two are equilibria the traffic game but in general you can ask for any given like system what are the equilibria and uh analyze that but even that is actually dissatisfactory for building a rigid system like you know if if we say that the main thing that blockchains are

doing is uh building these rigid systems of coordination because they're not the most agile systems of coordination a king or a president or an emperor coordinat ating us is actually the most agile system of coordination in including maybe Mark AI that coordinates us to actually show up at the right time uh you know they're more agile than anything you know uh so there's this agility versus rigidity tradeoff and blockchains are really optimizing for rigidity so that is not rigid enough either the the first one is not good enough bent and fall tolerance not good enough because both of these both the bis and fall tolerance and the incentive models don't tell you what happens if

everybody collon okay you want a system like if you want to protect against the Terran of majority you want a system which which works even if everybody colls or pretty much everybody colls and that's a very high bar and it's very very very difficult I would say like it was pretty much impossible to build a system which had that property till you kind of like you know build some of these really powerful blockchains and the the model so what would the model be then if you if you want to actually say it's unconditionally correct right so that if it's not majority honest if it's not incentivized what is it it should be unconditionally correct that should be the bar right when I get a piece of gold

from Alo I should be able to verify it and I just know it's correct right like I'm not not trusting anybody that it's correct um so that's the bar so how do we build systems which are unconditionally correct and and the one answer which I think was pioneered in the ethereum community is the idea that there is enough at stake from the you know participants that you know whatever they can gain even if all of them collude whatever they can gain is going to be smaller than what they lose so if you can make this equation true that the profit from corruption by actually going and attacking the system is smaller than the cost of corruption

somehow I'm going to find the attacker slash them this is why slashing was a core and integral part of ethereum is thinking through the rigidity of this this is one model where you say oh you know the profit from corruption should be now um greater or less profit from corruption should be less than the cost of corruption and then you say okay now the system is economically secure even if everybody colludes but then we were thinking about this and then we found a problem even with this system imagine some adversary and they don't care about monetary returns they say I'm going to kind of inflict my wrath on you you know even though I lose and I and there are lots of other problems in actually how

do you calculate the profit from corruption how do you ensure that that actually this equation holds then you you dig into it and then you find like nobody has really thought through this whole system carefully because you know you have to trace through every unit by which you can actually make profit who's thinking through this is ethereum thinking through this is it like the ef's job is it the core protocol development we don't know right so this is an important like but as you can see like you know through this system we're asking harder and harder and harder questions right we we started with majority honest and then we said oh we're going to make it incentivized that you be honest and then now you're saying oh even if everybody colludes it's still bad they're going to lose money and then

finally now you're saying even if so we call this strong cryptoeconomic security which BAS basically means an honest client should never lose money okay and that's such an insane ridiculous High bar right say an honest client never loses money and how do you build these systems know we've done some work on it we didot a paper called Stak sure which is actually thinking about exactly this question but the high level question of how do we think about cost of corruption is I think super important and I I think I'm just using this as an opportunity to raise Collective you know attention to this problem it's a super important problem we need to think about this really really carefully if we want to build rigid systems of global

coordination which I guess all of us here are aligned on that that's what we are doing we're not building like some you know I we can have fun with systems of coordination as well but that that is not the only reason for which these are built they're built to withstand really really high pressure and so thinking through that carefully and just to give a gist of our solution the high level is normally the the way uh ethereum works today is when you slash people for false take the money and then just burn it and instead if you had a mechanism so what would be better like the again going going back to my favorite principle the Kera here is not complete because I

slash from the bad guy but I don't redistribute to the harmed party I just put it in a bonfire and make the ultrasound money people have happy uh is not enough right we need to do something better we need to find the people who got harmed and then redistribute it to them but then who's going to do this is it like Mark or Andy or vitalic who's going to do this job of finding who the bad part harmed parties are can we bring it in protocol somehow when you're transacting like when you go ship a packet of in in FedEx right you can say take a self Insurance you can say oh this packet worth like $10,000 so if it gets lost or like you know destroyed in

transit you have to give me $10,000 and the reason they have that I you know the incentive for me to not lie is I have to pay a policy proportional to $10,000 and if it doesn't get lost you know I I just lose the money so the incentive for um this kind of a system is you self you you let people self declare the the value of their transaction and the system sells Insurance less than the total amount slash and then keeps the system in tight cryptoeconomic you know lock so that if you're an honest user and you got insurance we call this attributable security if you got attributable security then you are never

never losing money and this is true at like a core blockchain like ethereum level you can actually upgrade to use this kind of a protocol but you can also use it on igen layer like on a restake platform on top of ethereum that when people try to to when you build an ABS actively validated service which is a service that you can build on top of wagon layer you can actually take a portion of attributable security just for yourself for your abs and you know that whatever happens to the rest of the universe if you trigger a slashing condition you will be able to slash and redistribute that portion of money so we think about this problem very carefully and I think it's it's imperative for us as a community

something something that's interesting here that came to mind as you were saying that is like a lot of these Technologies are more so uh abstract Game Theory as opposed to actually you know implement it into hard Cod well they have to be in hard could but um like one of them you know we can we can talk about more but like one of them that is maybe even relatively untested is proof of stake itself so like we were talking about the cost of corruption that same cost of corruption could be a profit for a bad actor um and you know this is this is kind of like my own my own thoughts and opinions on it is like maybe proof of sake really only works for the one the one chain that is most valuable because there could also there could always be an exit that you can

flee to right so um you know for example if a alternative L1 is hacked maybe that value can still flee to ethereum you know that whole thing can collapse but then the value can be you know stolen onto ethereum um and then you also need the the thing at stake to have a monetary premium for it to you know continue uh know existing in in perpetuity so uh yeah I mean it's just like super interesting that you know we're kind of testing the stuff in prod right because you know we're like putting it out there and there's really no way to kind of like test things in test net when there's economic value or this game theory at at play so yeah I mean the the kind of like uh you know alternative to to slashing where you know instead of the money being burned

the value being burn is giving to the a good actor is like very interesting but it's like how do we you know how do we test this stuff in test net maybe it's just like kind of like I mean ethereum bootstrap with proof of work right so um you know in essence they did kind of like you know test uh proof of stake and and you know more than a a Sandbox but yeah it's like super interesting a lot of the stuff is you know we're pushing the boundaries not just on technology but also Game Theory so yeah and I wanna good I was just gonna say um yeah when thinking about um corruption really we have two tools in the mechanism designer toolbox we have carrots and we have

sticks right so we can either put a carot out there for someone to go after and you give them like this like uh you know incentive to like do something positively or we have a stick where we can like punish people right so those are kind of like the two tools that we have and we have to kind of know when to use them to create the incentives that we want to to drive the system and one other point that I just want to raise is um you know in the context of corruption we usually think of corruption as like a a morally bad thing right like we don't like Byzantine actors because they're trying to you know steal our money um but I wonder if there exists a world maybe far in the

future where you know blockchains are you know adopted by the whole world and the you know institutions are built into them right and and maybe in this world um corruption could be an attempt at trying to break out of like a Walled Garden and it could be like an attempt at like you know the next kind of wave of creative destruction of the economy trying to like you know inactive corruption from the point of view of the protocol could be you know the next new hot startup disrupting the industry right so um I think that is why at the end end of the day like kind of like layer zero in the community is really really important in these systems because um no matter what

it's the community that's deciding what is uh what software is running what is true right and that's why um staying true to the values is really important and there's kind of this like constant tension between trying to build systems that can survive um you know hundreds of years in the future no matter who is operating them versus um you know making sure that the ideals and um Community kind of lives on that um is kind of living within the systems maybe you could touch this briefly on that about um optimism and kind of how you have thought about

designing incentive mechanisms in specific there um because I think that you guys have done a great job of fostering that Community um there's a lot to learn from the kind of setup that you guys have yeah totally I think retroactive public goods funding is really like the big thing here right because the idea with retroactive he Goods funding is to create a mechanism that funds an economy that is able to produce a lot of innovation right and that's uh that's why the blockchain exists right we're here to build mechanisms on top that help to grow the economy and help to produce Innovation so the way that

retroactive public goods funding works is basically um you look in the past and then you reward things that are impactful right and this might sound kind of weird but um if you if you do funding where you give someone a grant you have no idea if that Grant is going to be successful or not right and we can say that this is a capital allocation problem and it's all about allocating Capital most efficiently right and if we're allocating grants some percentage of those grants are going to not produce impact right so if we are retroactively

funding then we can make sure that the um money the value is Flowing to those that actually created impact and the theory is that this will greatly um enhance the capital allocation of like the whole system right and it's not an end all Beall right there's obviously problems with this because now there's you know a risk because if you're not being paid up front you have some amount of risk if you're not receiving retr pgf you have no idea it's like you know so this is a mechanism that we need to prove that it works to more and more people to get them to buy into

it and I think something that is really important for everyone to know is that there's no Monopoly on retr pgf right if you don't agree with the way that optimism is doing retro pgf like I highly highly encourage you to you know do your own retro pgf or work with optimism governance to become like a sub da managing a sub portion of the retr pgf right because one important thing is that there is no Universal source of Truth for what is impactful Right everyone has different opinions and just the reason why I'm bringing this up is there was some controversy recently with optimism's latest

announcement in the next few rounds of retr pgf because the next few rounds of retr pgf are very focused on particular things right and then the people that feel like they're creating impact that are not included in those particular things are upset because they feel like they should be eligible right so it's important that we decentralize this idea of retr pgf and we enable as many people as possible to you know build it out so that everyone can have a say in funding what is impactful amazing yeah I think that that the tier like the section approach that

you guys just launched is um is going to facilitate more impactful precise work um the it is modular March um we have a couple minutes left and I I have to pose the question how does this idea of modularity uh of specialized use cases of of this unbundling of blockchains how does this accelerate or contribute to coordination and better coordination across the crypto World um totally open question but me be mindful that we have five or eight eight minutes left um I can I can go quickly I think it enables Innovation because uh you

know one of the trade-offs to you know quote-unquote mon monolithic or you know just things that don't change is that uh we don't leave room for Innovation so I think alternative da layers is an perfect example of this where you know if we're all using ethereum you know potentially we might stuck into something that can be um you know further improved so yeah I think I think that's like one really good example to me modularity means small interfaces or simple interfaces and uh that's all about reducing the cost of coordination if an interface is really complex or it doesn't exist at all then it's just you know it takes more time or expertise to build on top of and then

that's potentially not as stable of a platform so um for interfaces to exist um to you know make it possible for various da layers to work with various rollup sdks this is uh something we think is important um going back to my favorite refrain of um From Evolution so the evolutionary Arc always bends towards greater specialization and greater specialization is offered by coordination right you just take human civilization we do very very specialized things because we can coordinate our trade and other things to get consume

the things that we don't produce and you know modity is just a like a specific manifestation of this where you can specialize in producing something because you know you're unique and you know you have ideas that are specialized and useful to solve that one problem for everybody and one thing I want to add to modular when people think about it is they're thinking very narrowly they're thinking about data availability and consensus and execution like to me if there are three modules it's not an interesting World um if you look at like the U Cloud era like the software as a service is

the analogy that I think about there are thousands of really successful SAS companies which hyper specialize and you know somebody's building a streaming database for Real Time AI inference like that is the level of specialty that people operate at and this evolutionary Niche you know realtime database for you know whatever AI inference is only possible when they co-evolve with other similarly well well specialized entities so the fact that you know some modules exist which specialize and solve certain specific problems then create other evolutionary

niches for other things to evolve and this is the our vision is there will be similarly thousands or tens of thousands of thriving modules and by sharing common trust they can actually and and creating mechanisms by which they can interoperate with each other you actually create a very driving interesting uh you know crypto economy this is why in igen L this the systems built on IG L are called AVS actively validated Services we don't think of when people thought of IG L they think you know we are letting other people build chains chain is a very narrow view of the world it's like in an ordered

Ledger which does all these bunch of things that need to work together it's a service it can do something very very very Atomic for example somebody could just build a service for censorship resistance somebody could just build a service for reorg resistance somebody could just build the the narrowness of what a service is I think is not yet understood and what a module is is also not yet understood there is a wide wide canvas for people to paint on that's my uh you know call out for this modular March and or or Mark g mark G we're gonna call

on you man it's been a quiet one for you here today yeah I mean oh my gosh we have had some deep thinking on this podcast today um we really have appreciated all the the thought and really abstract ideas I mean the only thing I'll say is that like as lejo was saying having more modules means more Innovation and then we could see more corners of the internet start interacting together in unique ways um so I just think that there's going to be more diversity of applications that are going to start kind of having new interesting branches of all interacting together and what's interesting as well as like having many modules assume that there's a Central Foundation like for optimism that Central Foundation is the op stack and in the future right we

imagine that people will be really innovating on what these modules can really be and how they can plug into the op stack um and they all because they're built on the op stack be able to um communicate to each other so that's just like my general position on like modularity Innovation communication Etc and T we'll give you a a quick one if you have any thoughts as well to close us out here man totally I mean you know we're talking about modularity of the protocols I think that you know we're going to continue to modularize I mean you know there's this this kind of pendulum between um you know horizontal scaling and vertical scaling

right where vertical scaling is more like the salana style like you know beef up your single node right and then the modular horizontal scaling you know we've seen Celestia and igen da all these dat data availability layers like the roll up Centric road map right and I think what the future holds is a world where we modularize the smart contracts themselves where we evolve to a place where we can actually you know take these like monolithic smart contract systems and basically Shard them across many chains and then if we're able to have like an interoperability protocol be able to you know move the data around

where it's required and we we uh can you know horizontally scale just like how web 2 can then you know we're going to unlock an entire new way of innovation and I think that we're going to you know redesign the way that we think about smart contract applications right because we still have designed smart contracts in a very monolithic way right so I think the next few years I mean this is March this is all not going to happen in March but this is going to happen you know over the next few years we're going to be going modular with the smart contracts it's it's all about you know horizontally scaling the gas I would say that as a based response um and and of a variety of different

responses from each of you guys so thank you guys for coming out and being a part of of Maj March and sharing your thoughts about coordination uh Rob and I really enjoyed and um hope you guys all took something away from this amazing thank you so much really enjoy it yeah thank you thank you thanks thanks everyone you're on mute dude how do you feel after that one Rob I mean I I mean I I'm I'm still I'm still thinking and I think I'll I'll be thinking for quite a long time um so many more questions lots of

answers but so many more questions I know I I I could just see you boiling up wanting to ask more and more but like there's just not there was like no time I think I had one question and then it was just over yeah yeah yeah I think same I mean I think I I the the uh kind of opening it up and then that that was about it um so what was your takeaways yeah so I mean I think I'll I'll frame my takeaway as the question that I still have like rattling around in my in my brain which is that um blockchains as these as these coordination mechanisms like that's the monolith like that's the B

and so you break a block so if a blockchain is a coordination mechanism you break coordination into its modular parts and so you got you got consensus execution and settlement and so I've been tangling with this idea of like whether ethereum is modular for a little while you've heard me ask like multiple people this question and so as we kind of like chip away at the parts of the ethereum cordination stack you know replace the evm with modular VMS replace da with modular da and we're left with settlement but we're starting to hear that settlement isn't as valuable as we may think that we may think settlement is and so I I I also take into account the the

coordination stack as sirom describes it which is that first you try to um get the majority you get the 50% and that's your that's your coordination mechanism and then you realize that like all right maybe we can do a little bit better we do incentive mechanisms to to help coordinate across these actors and then you say oh maybe we can get a little bit better let's try to get unconditional which is like he said an insane bar or like like standard to try to meet nearly impossible so when you look at ethereum as this coordination mechanism and you try to think like what is the valuable module that ethereum has left after you chip chip chip away these like other parts of its coordination stack you're left with the social consensus and the

social consensus is this thing that falic says that we shouldn't overuse we shouldn't stretch out social consensus and have it apply to all of these different like avss or rollups or like you know we can't the cost and and I think it was Mark from optimism who mentioned the cost of coordination and so we can't rely on this ginormous cost of coordination when we're going to go take a snapshot Vote or poll every time we need to ask some question about whether or not the ethereum stake that's staked in all these validators is going to coordinate some decision we did that once and it was and it was born to ethereum Classic and ethereum classic I checked during this live stream is still trading at five billion doll market cap

which is ridiculous in terms of the cost of coordinating that mistake if you want to call it a mistake or that decision if you want to be a little bit more nice about it five billion dollar cost of coordinating that thing so that is something that is so expensive and I only like to use that word but it's a very high price to pay for coordinating that and so you can't rely on ethereum social consensus for making the decisions and and pushing all of the coordination that has to do with ethereum onto the social consensus layer of ethereum that is too high of a cost of coordination and so that's my takeaway is like how are we going to secure all of these abss and all of

these rollups and the rollup Centric road map with the ethereum Stak in all these validators when the cost of coordinating across all of those all of those validators stake is so high that that that's my question that I still have is an open question that's my takeaway that I would ask if I'm just posing to the universe at the moment yep yeah I think um I think there was several really really interesting takeaways all of which like what you've just said makes sense uh from a from not only like a cryptoeconomic level but just from a a kind of uh incentive design level and thinking about how we're going to try to build a system which is almost like lightweight but

extremely secure and so you know the the guys touched on that quite a bit and I thought that that this this kind of evolution model that Shon brought up uh of like this nucleus and and the mitochondria being kind of two separate distinct units of a cell however they they were they contribute to the broader core of the cell and that was in a single cellular design and then I I almost think of like that being like the integrated or monolithic stack and us moving towards like this maer stack is almost like this move towards this multicellular organisms um which kind of goes into Evolution and Innovation but

another thing that that that was brought up was was was uh mutation and that if you can if you build a rigid coordination ler that facilitates Innovation on top meaning you know the example was the Constitution you could say uh Cod is law is is is almost a rigid enough coordination me um mechanism and then that facilitates you know the Industrial Revolution you know the technological era Etc and so I feel like I feel like this uh this this shift from the single cellular to multicellular uh organisms for innovation is is is possible because of the coordination that we've created amongst

this crypto Community specifically ethereum um and and is now gonna kind of push push new new new uh kind of doors open for kind of what's possible but the one thing that I wanted to return to on that was that uh man shiram touched on such a good point as well and and the guy Mar oh that that the corruption is currently seen in a negative light because of exploits and so perhaps code is law isn't the best way for us to to think about us as a as a community perhaps we can we can extrapolate our values into our own set of laws Network States Nation States and we can facilitate a

process of democracy outside of nation states and and and what the my my my final thing on this is Mark from optimism is like well if we do that or if we have these systems and and we're thinking about corruption what if corruption isn't necessarily the worst thing what if it's actually what causes the next era of innovation and it's just kind of like whoa optim very optimistic perspective for sure like but Satoshi corrupted like corrupted the the Fiat world and said it I'm gonna launch my own coin I'm G launch my own my own monetary system and so like that is that is interesting

thought so kind of putting it putting that all together I think that that we see this that we see coordination games being a extremely more uh important and being focused on by Builders but B I think that there's some of these people that are working on network States um especially as we see regulations hit um and for this for the actual value of of kind of Cor ordinating I think Network States could be a real thing in the next five years I think they're already happening I I mean at a at a at a high level like a lot of us are citizens of ethereum or or generally citizens of of the onchain world the onchain network State we

participate in that economy and I think we participate in that governance um and we have a question here from the audience about uh oh uh this is our friend AJ on democrac what are our thoughts on political meme coins contributing to elections it's funny I was going to bring up elections because if you're unsatisfied with the people who are running in in in in election you you exit yourself well you have the the choice right which is what Teddy from igen lay was saying you have the choice you can either express your voice run your run your run yourself I if you know you really have that passionate of a voice or you can go up you can ask questions you can try to shift the the political voice that candidates representing or you can totally exit

right from the system and not participate in that state geographically it's not very easy but mentally we tend to spend a lot of time and energy in this network state so maybe we can't eject ourselves to a certain extent but addressing the question what are our thoughts on political meme coins contributing to the election I'm gon to hit on my favorite use case I hope to see it as an ABS which is prediction markets because I think prediction markets bring in this very interesting thing of a shelling point and I think there is a point at which when when we go past this idea of a prediction and it becomes a self-fulfilling prophecy and the wisdom of the crowd starts to influence the outcome of such a market and I think this is something that Trump

is actually using in his I'm not even going to get I'll I'll try to stay away from like mentioning specific names but prediction markets could impact an election to the point where if there is enough money and enough substance behind a certain outcome winning then you could see more people jump on that bandwagon and you can see that used as a tool to try to show the uh the campaign strength and uh how how uh how strong a base is so I think political meme coins are one thing and I think these are another uh another tool that uh candidates can use to express their the strength of their position in in an election they probably have their own shelling point I think even more importantly we see candidates

that are literally using their prediction Market to express how much uh how much uh strength they have in the polls and why like why why are you going to fight the tide right just accept it so I think we're already seeing we're already seeing these uh political on these onchain mechanisms be used in political ways um and have impact on on real life outcomes of these elections yeah and I agree I think it's it is a completely a prediction Market game I think it is a self-fulfilling prophecy game as well which is quite interesting there's a certain Threshold at which one of these I I kind of look at presidential election mean coins which is mostly Trump and Joe Bowden um

as as uh like there's they have to cross a certain monetary threshold before they have a a serious Val like impact on on anything in the real world and the price action on these tokens are actually more of a speculative prediction Market based odds so in my view I look at prediction market so you have odds as more liquidity fills in one side the odds are rebalanced or adjusted based on effectively the the order book um so I look at the Trump and Bowden coins and other kind of uh political meme coins as almost like a a social token for achieving true consensus of opinions of

Sentiments of crowds without having to do surveys where people can just like and say yes or no it's like a real hard answer to your question and I and it would be very interesting to do a thought experiment from today onwards until the elections Trump and Bowen coin which performs better yeah and then if that is actually aligned with the end result it's almost guaranteed that it is yeah I think I think it is is um uh you know it's the old old homage of put put your money where your mouth is and I

think uh people vote with their mouths but uh they also um you can also put your money where your mouth is and I think people are are doing that in the onchain economy one one like quick note is is to to just kind of like bring this full circle to the coordination conversation I think we're be becoming even more a like the onchain world the offchain world are becoming even more aware of one another and I think as a result these two of these two worlds converging um we end up with more accurate coordination we end up with more accurate prediction so I think maybe five years ago that may that you know maybe one of them has a super high surge market cap

right maybe that happens in a vacuum five years ago or last election cycle but now because these worlds are becoming more and more aware of one another I think it does have even more truth accuracy and hence the social the social consensus once again as the fallback mechanism of coordination yeah so kind of circling this all back together I think there was a lot of abstract and metaphysical Vibes today uh for those still listening we had an agenda of questions and I think we started with the first question um and then we kind of just totally just went for it but the first question was honestly one of the most informational ones that we heard

before it kind of got uh you know a bit more abstract and and uh we we we we were answered that infinite games and infinite sum games are two different things two different uh uh Concepts and um the idea of of infinite games is is is a Time Time based uh Concept in meaning that this game will go on forever there's no winners or losers there's simply survivors but it's not just one-dimensional it's a two-dimensional concept where there's this time out um aspect and there's this idea of the possibility of infinite games within the broader infinite game

so infinite possibilities of coordination Innovation and uh uh creation um and the time scale is is infite so that is infinite d and yeah some games but that's infinite games right right right right because you have you can have nonzero games happen consecutively moment after moment and and so therefore if you take a non a nonzero sum game over an infinite timeline you end up with an infinite sum game and this is the idea of uh uh one more time one more time for the uh not math for sure so

I yeah no I'm glad you mentioned math because I could even think of that as a as a as a series but getting away from that is basically like think about it uh like you have a number line that's your timeline and so you draw your numbers on your number line at each number you can have a game and this game at each number whether you want to however you want to measure whether it's minutes days seconds hours decades centuries you have a game at each each number and so that at that game is a non nonzero sum which means that it's potentially infinite right and so if you have a nonzero sum game and things are additive and you can continue to add value at at each of those games as you as you take an

infinite number of those games because each of them is non-zero sum you can have potentially have an infinite sum game over an infinite timeline yeah and this is like this is this is this is like this is what I like my my friends buy Litecoin and I'm like dude Litecoin is gonna underperform the market but Litecoin isn't going anywhere ever unless the miners stop Mining and and and crypto is like this system of blockchains that are that we don't really realize yet because it's so early it's only been around for 10 12 years but these systems are meant to be effectively Perpetual infinite in time

which is like it's hard to grasp our heads around now but in 20 years we might look back and be like okay this has been around for 30 years what's another 30 right right right and so I think this is a really important way for the the the the industry to think because we got here we were 10 years in we can't think like we're 10 years in to a a 50-year journey we're 10 years in to the rest of time that's true that's true and and and when you put it like that it feels very early all right guys thank you so much

for coming out as Rob says we are still early I really hope that you guys enjoyed this post I think I just like decompressed all of that into into into actionable like uh understandable components I'll leave you with one little nugget which is that another one of these takeaways is the idea of local entropy and how yes the the second law of Thermodynamics does say that we Trend towards infinite entropy entropy goes up and to the right but you get it's a it's it that's if you smooth the curve out you get these little pockets every once in a while where entropy drops and you get these low entropy things I just encourage

everyone to look into the idea of of entropy and really consider low entropy as a mechanism for bringing about some change to the world um that is a little bit different from this inevitable Trend that that uh is is inherent to this universe um I also encourage everyone to check out the rest of the content on the rollup check out Andy and uh a Andy that's three wise Banger Twitter content hit me up on Twitter hit me up on telegram it's Robbie rollup um yeah we're in the middle of modular March uh we have so much content coming out guys if you look in the in if you click on on the rollup on Twitter go ahead and click

smash the follow check out our our modular March content um we have our YouTube our Spotify of course our personal twitters um we try to be as credibly neutral and as and as curious for for knowledge as possible growing our audience growing our our our community um and being being able to do these things is very important uh for us and and we believe for the broader space so please smash the roll up on Twitter smash a follow smash a comment check out my account check out Rob's account and uh thank you for we want to hear from you guys like ask us questions like what do you guys want to learn about what do you find interesting so you know we'll we'll read all the comments hit us up I think DMS are open especially on the

rollup you can you can fire away um you guys make all this possible so yeah smash that follow button butt click on the rollup click on our accounts give us a give us a follow and uh keep track of all the modular March content um and it doesn't end with March there is plenty more modularity to come and uh conversations for the hyper curious in general so Absolut we've got some Alpha coming next week uh for Wormhole Wednesday we've got some Alpha coming we we've got some really cool videos lined up um and if you're curious about Bitcoin going modular we you can stay tuned because we've got a really really cool content series coming up with some of the best Bitcoin rollups really

trying to to to learn more about that and try to kind of separate the noise from what's real ultimately that's what that's what we're trying to do broadly if you if you've heard about modular if you've heard about modularity uh you should definitely pay attention to what we are doing um you know we are kind of trying to uh push this entire space forward um because we are big Believers and um thank you so much for being here and uh to infinite coordination and Beyond like Buzz Lightyear should have said but Disney's made for kids and this is the world blockchain baby cyanara got a blast Alpha April baby see you