From Complexity to Resilience: MakerDAO's Journey towards Decentralized SubDAOs / Sam MacPherson

Transcript

oh ah there we go all right hello everybody um so this is a talk about maker Dao and the end game and it's from complexity to resilience maker Dao's Journey towards decentralized sub-dials so let's get started so who am I um I'm Sam I've had kind of a unique Journey with regards to maker um so I started uh with ethereum and uh joined the whole crypto space in the 2017 uh Bull Run just as a hobbyist investor like most I think um I was hunting around for projects that I liked Beyond ethereum and stumbled across maker in January of

2018. if you guys remember the projects back then you know there's a lot of crazy stuff and maker was like the first things like this is actually something it's not Insanity so I joined I joined the project I really liked it I was um a long time sort of forum lurker back then uh the found the maker Foundation was entirely running everything the Dow was just kind of um it was there but it wasn't really deciding anything so uh single collateral die had just launched um I believe that was December of 2017 so I messed around with the protocol you know opened the CDP um it was a lot of fun and then for a couple years just lurking in the Forum observing what was going on and just uh

you know appreciating it from a distance so um fast forward to uh 2020 um this was still the maker Foundation was mostly in charge um I started I've been at 20 plus year software engineer um but I picked up solidity I programmed my first thing which was the flash mint module which was later onboarded to maker this was August of 2020 and then later on in that year I developed the PSM which is um the First Community contributed um smart contract previously the foundation was building everything this was the first thing that was spearheaded by the Dow and at that time we were bootstrapping the Dow process and I was

involved in that process so the second this transitions to the second phase where in um when was it a summer of 2021 the maker Foundation fully dissolved and gave complete control including their portion of the maker back to the Dow so the Dow was fully in charge and the structure at that time it looked like um maker holders were voting on every single issue and the workforce was in what are called core units which are they were specialized based on their um uh field of work so I was in the protocol engineering uh core unit so we built smart contracts maintain the protocol that kind of stuff but you know there was like BD and growth core unit there's a risk core unit there's all the all the different types of core units

and the workforce was kind of divided by like this so where were now is uh the next phase so fully Dao it was sort of a good first iteration um but it's not it's not going to work and I'll get into the more why that is uh in a bit but so now we're entering into the next phase which you know is a decentralized beast mode sub-dials so dividing things up and I'll get into that now actually so uh let's go over the challenges of what the first iteration of the Dow structure uh was so um Dao's the structure is very flat and um so you know like flax structures uh this this doesn't really it's nice in

theory you know everything's all communal and stuff like that but in practice you kind of get unofficial power structures uh politic politicking that kind of stuff just fills the gaps um so in particular uh Dao's the Dow challenge's efficiency wasn't very quick at executing everything had to be debated politically had to go through um long periods of governance which um having worked with several Partners at the time people were often got frustrated with the process and they would just be throw their hands up and give up so this was a huge problem uh resilience another issue as I said uh the politicking was a big problem with maker it didn't really present itself until the bear Market kicked in Bull

markets are very good at hiding the issues with your projects and protocols so that was kind of hidden from view but once the bear Market started kicking in and we had like the Dow had to start making hard choices like firing and layoffs and stuff like that the whole thing was just grinding to a halt so this to me is not resilient at all and the other issue is voter apathy um there's the free rider problem is an ongoing issue with Dows and that if you're a maker holder like why do I lock my maker in and like you know spend my time uh learning everything you know these Protocols are extremely complex so like what is what is the incentive to do that there's not much so what I'll just

throw my maker in uh an LP or something to get a better return so this is a this is a big issue and it's ongoing so in order to overcome these challenges the strategy is to divide and conquer instead of having everything in this single uh Dao where really it's a political Beast where uh factions you know some people want maximum decentralization some people want maximum growth and they all kind of fight in the same structure the idea is to instead of having everybody fight in a single Dow you split up into multiple competitive sub-dials and have the people who are like-minded join into the subdial that they prefer so there's a sub-dial for growth orientation there's a sub-dial for decentralization maximum

maximalism or whatever like it could be just memes it doesn't really matter it's like you get depict to be a part and contribute to the change you want to see in the world is the idea and the other piece that I think kind of gets ignored is the a in Dao's so automation uh gets overlooked uh it's unfortunate we kind of have to have humans involved but as we all know humans in positions of Power are prone to corruption and centralization all these bad effects that we don't want so we really need to start pushing forward with the a in douse which is the Automation and this is the AI portion so um with uh simplifying the maker core into this very simple structure we can

have AIS that are enforcing and uh enacting the rules eventually right now humans need to still do this process but this is this is the goal um so yeah uh sub-dials are basically decentralized uh specialized divisions within maker Dow they can build their own unique governance process their own memes their own whatever it's like it's a build your own subdial but it has this special privileged relationship to maker if if they follow the rules that are set by maker core which are these guiding principles they get access to maker liquidity which is um there's no comparison in D5 to maker liquidity this is what you want to build the protocols that you want

um so there's three things um three kind of important pieces with the maker course so there's What's called the lineman artifacts these are the rules that make a core set so these are um kind of General um principles as maker is underwriting uh all of the credit going out right now um like there needs to be due diligence and stuff like that at high levels so this will be rules about maintaining the peg so there's like a stability in liquidity uh alignment artifact where you have to maintain the Peg and you like there's rules on returns and you have to like post Junior capital I mean it's I'm not going to get into it right now but this these are basically just the rules um there is governance AI tools and again this is uh moving towards the a in

Dao where we want to replace the humans right now we need the humans as um kind of like uh in maker core as more bureaucrats that are just say does this follow the rules yes or no therefore it's accepted no uh creative thinking and this is important to prevent uh these sort of politicking and uh there's no been no corruption but eventually it will lead to corruption if the status quo is kept so this is an important uh thing to work towards and then the atlas this is the Constitution so this will be um just the values the core values of maker Dao that people can align behind um much like a constitution of a country sort of more at a soft uh human level so another interesting thing that's

happened um so or that is happening now maker traditionally um the balance sheet has been entirely operated by maker core so uh let me run through uh the the three main things that people are probably used to now so cdps was the first um uh asset side of the balance sheet so this is loans against crypto collateral um against your eth against your stake teeth I think most people are familiar with this uh by now um the second uh primary thing that was added uh in 2020 was the addition of um stable coins and this is to maintain the peg um it's uh yeah so that was the second thing and then the third thing that was added is uh the die direct deposit module first to uh Ave and then later to

compound so this was basically just you meant the die and then you just stick it in these protocols and then they act as the lending engine as opposed to doing everything inside uh maker core so the new structure is a little bit different um the asset side of the balance sheet is going to be entirely managed by the subdials this means nothing in maker core there's no decisions being made in maker core about what to allocate these assets side to so there is uh four what are called allocator Dows and they can basically are free they're privileged they're free to Mint die at will and then stick it into any investment opportunity that presents itself so I have some examples on the right of

Investments um some of these they're like treasury Investments You See Spark protocol at the bottom this is the protocol that I work on so this is like just uh another investment you can look at it that earns some return so for the purposes of the allocator Dows their job is fairly simple they just need well it's not simple but the rules are simple but in acting this is actually quite difficult they basically need to balance between maximizing return and maintaining the peg so uh there is going to be a market driven everything is going to convert into what maker charges a market-driven base rate it charges this to the allocator Dows and this will uh adjust based on the price of die so die is

reintroducing something that looks like the target rate feedback mechanism going to a soft Peg no longer hard pegged one-to-one with the psms the stable coins and instead if the price goes up or below a dollar then the base rate will uh the rate that is charged to the allocator Dows will increase or decrease as necessary and this this enforces that the allocator and the allocator dials can keep a hundred percent of the spread that they earn on top of the base rate so what this means in practice is that uh they will be allocating to the highest yield safest Investments and there is other caveats you need to have Junior capital and stuff for riskier Investments but let's say treasury is fairly safe they allocate to treasury's

high yield safe fairly safe investment right now and then they have to balance it with maintaining some amount of on-chain liquidity Market making the Peg and stuff like that so this is a very interesting new system uh that's coming online shortly um so for in particular for um our sub Dow uh we are um we Phoenix Labs is my company we are building the spark protocol and we are uh trying to foster a culture of decentralization defy in um the sub in one of the sub-dos so we're bootstrapping this culture we want uh to invite others to join us so um what we have is the subdial treasuries they're each bootstrapped

with a bunch of um new gov token but it's essentially maker to give it some fundamental value all the revenue earned by the subdial goes into a sub-dow treasury and acts as it's it's really like a mini version of maker it can allocate to various Investments and you know has a treasury and this is a junior Capital uh as this Surplus buffer builds you are free to do more risky Investments as maker core is no longer underwriting the majority of the credit um yeah Junior Capital to protect first losses so our subdial as I said is very defy oriented we want to focus on scalable decentralized collateral we're also building a decentralized front end so that getting back to retail we want

to go multi-chain we want to go where the users are I don't know um so if you guys remember back in the day opening up your first CDP and not spending a thousand dollars to have to do that this is invaluable for new users and getting people excited about this stuff so we need to get back to that and so we will be going uh multi-chain to get the scalability that we need we are partnering this is not our subdial is as open to everybody so we're partnering with arrakis gnosis and we're wanting to build up this portfolio of people who are like-minded get involved in this sub-dial and contribute in whatever way you specialize you know it's not all us it's it's very open um and then yeah the fourth thing is

multi-chain as I said and um we're looking at uh Advanced allocation strategies too so I think this is going to have a very big impact on defy I think previously maker and I think rightly so has had this uh perception from the outside of being very closed off very Insider and just very frustrating to deal with and this is being flipped entirely on its head now with the sub-dow system so I I think this is going to be very big projects and you know stablecoin liquidity is King like this is how every project lives or dies and this is what maker has that is in excess right now so you have your great project you can then uh ignite ignite it with uh maker liquidity get him like Get It by getting

involved with the sub-dow system I think this is going to be a very big change um so to get more involved uh we have the Twitter um on the maker Forum we have a sub like a forum that's like dedicated to our sub-dial there's other sub-dials too so some are more focused on real world assets um so if that interests you get involved there um but yeah thank you [Applause] thank you very much Sam ah any questions for some to start yeah great talk mate thanks for that

um so a question I find it interesting that for you in the a for Dao it was automation um I obviously Dao's you know the term is still hotly contested and for many people this autonomy I see the two being very different though um so how does that where are you guys at in the scale of automation versus autonomy I guess it is a tonic I think I just uh misremembered that um so yeah autonomy is important uh for like uh running yourself I guess um yeah I think they're they're both important automation I view is important um maybe it's not one of the letters but uh

as I said humans who are involved in positions of power it just tends to lead to rent seeking corruption and uh the things we don't want the things we're trying to get away from and defy so uh like it it to me the long-term goal with all this stuff is ungovernance right you um where you can't get rid of humans you make it you make competitive markets so it's all just part of the process of doing what we're here to do in my view so yeah it's like trust minimization exactly yeah cool and in terms of where you guys are at if you look at D being a one to ten so decentralization like one being completely centralized so that the

foundation for example and then 10 being the sub-dos being completely decentralized and the same for a and the same for o how would you rate yourselves one to ten in each of those subjects like right now yeah yeah I think most are pretty most Protocols are pretty bad right now in terms of most of these metrics I I think uh I mean I I don't think it's I think it's worthwhile to be uh like building towards something and have a goal but I don't view most uh most Dows is actually being uh any of these letters right now I think this is like a goal we're working towards a lot of if you look behind almost everything in D5 right now there's centralization vectors and everything multi-cigs whatnot but doesn't mean that the uh the

goal isn't worthwhile so like where where we are I don't know numbers but yeah probably as of right now poor on most metrics so poor is the answer honestly thanks again for the talk um you said the subdials will be competitive so are there any checks in place to prevent maybe there's some market conditions or some scenario where one sub Dow becomes extremely performant and maybe takes share from the others is there some limiting you you have in mind to stop some one from becoming too powerful yeah so there's a

whole tokenomics design that I didn't get into about how the cash flows are converted into uh value for the subdial tokens there's sub-dot tokens by the way um and maker core is um and that there is some like base rules that prevent these sort of runaway effects of like somebody just takes the entire pie because it you know it's going to be you want uh competitive um environment to for the best experience for the users and for everybody pretty much so maker does have um tokonomics rules that will uh enforce this okay and to be clear the end goal with AI involvement and governance is to completely remove the human element outside of maker core or is it also

including in maker core there is humans involved it's very human oriented at the edges so this is where the growth and Innovation and the efficiency like people move fast they build the things they want but because it's competitive it uh it prevents um you know rent-seeking uh corruption this kind of stuff because you need to you need to be actively trying to provide the best user experience otherwise you get out competed but yes so for maker core no humans for the edges very pro-human understood thanks yeah uh the question about triggers was a perfect intro for me are there any checks uh to prevent that for example some bigger or bigger holders like Rune itself which I don't know I think he

aggregated something like 20 percent of maker Supply uh that he won't be that he won't be preferring some sub-dials which he likes for whatever reason or he knows their people yeah so in general um the token holders it's still a bureaucratic system at the core so those token holders who have more uh maker are you know uh it's their right to be able to decide how the protocol goes but it's in their best interest to have the sub-dials competitive um because if you just if everything just piles into one sub-dial it's just going to have the same problems as before it's not going to actually be um any of the decentralized or

autonomous or um yeah oh yeah I didn't get into that but there's going to be uh voter incentives uh added to the system so people are going to be paid to vote now you can't just do that um without thinking about it it's a very dangerous Thing Once you turn on voter incentives um you are if you do it wrong like the ship is set to sail at that point and you can't you can't uh interrupt it so a lot of thought is going into how to make the system resilient for the long term and then you set it out to sea with these uh voter incentives being turned on and this this is designed to be um

such that the end users there will be AI driven explanations of what's going on there'll be delegates and then they'll be you'll be like do you like these core values very simple things you you're gonna stake your maker in governance because you want to earn a return and then you know all things being equal you'll pick the values you like and then um this is the idea behind it yeah so um to uh automate everything um you kind of need Ironclad um objective functions right so basically that that uh to kind of optimize for so basically how are these formulated and how are these ranked because obviously sometimes they compete

with one another so uh yeah so basically yeah uh yeah so this is uh the rules of the system the alignment artifacts um in very um General language it's outlined what the rules of maker core are and these are going to evolve a bit but it it maker core is evolving into more of the just uh it's bureaucratic is a boring word but you want it to be boring right like you want the rules to be very clear and it's just going to be um like evaluations at a high levels like what are these sub Dows doing how much credit are we giving to them um like are they doing risky Investments and stuff like that okay they need to be

posting more first loss Capital stuff like that um yeah does that answer your question um Vector functions that's really what kind of drives the organization right so basically who kind of and who um who's corrects after the fact so basically is it all done in the make a form or are you back to the politicking that you had earlier um so who decides I mean so the the idea is to set out the Constitution and the rule book so that it works uh indefinitely now obviously it's not going to be perfect right off the bat but uh it's to strive towards this um goal of deciding as minimal amount like this is the value set you want to decide

as little as possible inside the core and just continually work towards that and once you feel it's um sufficient enough you set it free and you get rid of the humans altogether okay uh yeah maker holders are ultimately the authority on everything so cool I will take just a moment to you know talk about how you can ask questions and live chat because of course there is a factory Dao setup that is done on the app and that is exactly what it's supposed to do rather than taking you know power from me as a moderator to all the people so they are

all in queue but thanks Sam for uh oh yeah and there's shirts up here yeah the shirts are actually here for grabs will have the next lineup right after it thank you [Applause] [Music] thank you