Dilutive DAOs: Community Building For The Long Term - Mark Beylin

cool well thank you uh everyone for coming just to give you guys some context uh my name is mark uh i've been building kind of like daps and uh different incentive mechanisms for communities uh in ethereum for like five years now uh so i've seen a lot of the kind of different ways uh different community models have gone uh very well and and the way that some of them haven't worked um and i'm very excited to be sharing kind of this new approach we have towards building tokenized communities uh in particular with a focus on kind of like digital communities social clubs uh these sorts of uh entities these types of dows um and trying to like figure out this problem of uh fixed supply so um i will pause it and i think some of you in the room might recognize this that we have this fixed supply problem um it started uh way back in you know 2008 with satoshi inventing uh bitcoin and of course it's it's very magical now that we have these blockchains that can ensure uh scarcity across the world with no uh nation state or you know centralized actors i think it's awesome um and of course it's been you know wonderful for the charts you know uh speculative uh assets which are fixed supplied typically do much better than uh dilutive assets um and so a lot of you know tokens which have launched uh since bitcoin have also followed suit and and have seen some of that appreciation as well because ultimately you know fixed supply being non-dilutive helps with your you know pumpanomics um but there's a sort of another half of this uh which is less frequently discussed especially in the context of bitcoin uh which is around kind of inflation over time or studying the network across time you know how do these things actually work and what we see with bitcoin is that you know such a majority i think it's like 80 something percent of you know outstanding bitcoin have already been mined there are certain people especially in the ethereum community who posit that bitcoin actually kind of like won't be able to sustain itself in perpetuity because it's fee uh you know its fees won't be high enough the dilution will be high enough and so um whether or not that's true for for bitcoin i think remains to be seen but um in particular i think we're starting to see that this is a major problem for uh kind of like other projects that have launched uh since then and especially ones which are very human focused so the the kind of central theme i'm gonna bring everything back to is this idea that communities exist across time communities are incredibly easy to launch but are very very difficult to sustain um it's it's very easy to to start a new you know facebook club or or facebook group excuse me or telegram room it's a lot harder to keep people coming back um and and communities are very sort of like people based businesses right they're not like a technology company where you build something at the beginning and you know the the product just runs itself uh you know in perpetuity like craigslist right uh communities are kind of the exact opposite of that you have a bunch of people um putting in a lot of energy on a continuous basis and you need people to be carrying that torch over time right you need them to keep coming back and to have a real sort of incentive for for doing that and so one of the models that we look at around communities that i think is really meaningful is um the orbit model of communities right this idea that the community at the center is kind of like the star and you have different contributors who are in the orbit of that community um who are kind of entering and exiting um based on their sort of ability to contribute and this is what we imagine this looks like across time right where you have different contributors who are sort of flowing in ebbing and flowing and all this is kind of constantly changing um but over time people and contributors are kind of like proving themselves and so when we look at like these fix supply tokens when they're used for communities we see kind of like a number of of problems start to uh jump out fairly quickly um the first one and kind of the most obvious one is just that like it's very difficult to fix mistakes um oftentimes dows uh will have like a particular plan um they'll launch their token they'll start giving it out and they'll make some sort of mistake with their distribution and it's basically impossible to fix it and and you know talking to people who've done this oftentimes the best solution is just to fork the token leave it behind and launch a new one and that's like fine i guess in the early days but is in general just like not very um scalable across time and and i think it kind of ignores the the central sort of glaring factor which is that people make mistakes communities operate with imperfect information and their token model should reflect that and and should be able to uh kind of adjust uh in the face of their their um you know lack of sure sureness over their decisions um the primary mistake that they make is that they typically give away too much ownership too quickly um again like i said most people think about when they think about giving away ownership they think about startups or these other kind of like uh high capital based businesses where you need a lot of investment upfront and you know these businesses scale over time communities don't scale communities are the opposite they they're sort of like meant to be uh very people-oriented and and if you give away too much ownership too early um you kind of screw people over later um who don't have enough incentives to kind of like keep it going um and then finally the sort of more nuanced one is just around the idea that you know if i own like one percent of a token um and it's so important as we recognize for these communities to persist across time um then i can just keep owning my one percent and coasting while sort of the rest of the community does all the work um and this is of course also just like not very fair uh for the people who you know it's sort of this like free writer problem it's not very fair for the people who are doing all the work so we as we first started building mico which is a platform for uh digital communities uh we kind of realized that this would be a huge problem so we initially swung the pendulum the opposite direction um and said okay well what if we had no limit on dilution um ethereum technically has no limit on dilution although it has like a very specific way tokens get minted so dilution is generally very um restricted um but basically we asked ourselves you know was was uncapped dilution any better and and what we found especially talking to users um was that it wasn't that much better uh basically you run into this huge problem especially in the context of tokens which aren't meant to be kind of traded and sold it and have these like very sort of uh high liquidity markets um is it becomes like really hard to price anything right you're pricing tasks without really knowing kind of the context within which you're um you're pricing them everything kind of has to be rewarded uh relative to each other right like i know um the person who's leading this dow should be getting i know 10 so you know maybe we'll measure new contributors relative to that but um it's like just a very difficult way to to measure contributions when there's um when there's like no market supply or there's no like denominator for a fraction um so we kind of realized this wasn't necessarily ideal and and really like what we realized we needed a limit on dilution right unlimited dilution is just not uh not scalable across time and and um also generally speaking is like very very attackable especially for social dows um which aren't like ethereum uh who can potentially uh push kind of protocol updates uh or updates to their dilution and basically just dilute away anyone who was there previously um so there needed to be kind of some assurances around the tokens that were being given out so anyone who received tokens would kind of be able to at least predict okay you know here are these you know 10 000 tokens okay this will be 0.1 this year and maybe .01 the next year and so on um so then we say okay well we know we need a limit of some sort but we don't know what it should be we don't know how to determine it and so to approach this problem we kind of started from first principles we start with our primary principle that communities exist across time the second kind of important principle is this idea that communities have churn that communities are constantly kind of growing and shrinking and that people are leaving on a regular basis and then finally we recognize that communities go through cycles and attention cycles in particular when lots of people are churning and growing simultaneously that you know the individuals aren't just like autonomous actors and so what we thought about was okay well what if we think about this like community across time and we slice it into chunks right we slice it into these little pieces um and we imagine let's say you know a month is the right period to slice into um and we think about this asset as like of of community time and so you imagine that every month you know a million shares or a million tokens or whatever uh of community time shares of community time are created um and that those shares are created like in perpetuity basically every month that a community exists it mints another million shares of community time um and that this happens you know in perpetuity forever and um what what we imagine and and this is sort of like the you know provides a limit in the sense that uh at any given moment you know that the total number of tokens is a million times the total number of months in which it's existed um which is a pretty great and relatively simple way to think about this um but it makes sense again thinking about communities across time and what you end up with is basically that um over time contributors ownership percentage is constantly in flux right as you're as you're kind of entering the orbit and exiting the amount of tokens you own won't change but the percentage that that ends up you know converting into uh will be very dynamic um and we imagine that this is actually like a virtue um because it again sort of a forces you to keep coming back um but also because it allows the community to be super responsive basically to its environment and to changing conditions changing contributors and all that stuff and so yeah like i said total ownership is kind of accruing to people or diluting away from people over time based on how they contribute not at any given moment but across time so then we start to think about okay we know we have this like limit on dilution let's say a million shares a month then we say okay well maybe it could change right maybe some months it's a million some months it's 500 000 and some months it's you know 2 million right is that maybe something that's meaningful or relevant um and again going back to our earlier question of these like cycles of attention right we kind of know that maybe there's this variability over time so maybe the token model should also kind of reflect that and um just to answer this question uh we'll we'll start with the worst case scenario because this is kind of the most important one right we imagine we're at that uh if we go back we imagine we're at that sort of like low point in attention the community's productivity is is waning um people aren't you know contributing as much the you know attention is kind of going down and you know the community's uh whole point is to bring people back they want to reverse this trend and if we're thinking about like shares of community time uh being sort of like more of them being minted when more values being created and less being minted when less value's being created uh what we would end up with is that you know as soon as the community starts to shrink the number of rewards available for contributors goes down um and this essentially kind of like leads this leads us excuse me to this um you know this like kind of death spiral uh you know path right where fewer contributors um leads to fewer rewards leads to fewer people contributing etc so we kind of recognize okay well you know in the down like in the in the sort of worst case scenario we don't want rewards to shrink uh in those down down months right um and so then we can imagine uh that like relative rewards actually probably want to increase during these moments this is probably when we want to have the most rewards relatively speaking going to people right we want the the rewards to ramp up when the community needs those rewards the most or the the contributions the most um but we also recognize the opposite is bad as well imagine uh the community is growing everything is going really well but all of a sudden the community starts you know mining like minting less shares um that also doesn't make any sense that's going to slow the community's growth down um and so if we know that we can't shrink the number of shares being diluted in the the worst case we also can't shrink in the best case um if it is ever growing that essentially leads us back to the unlimited case and so uh we have to decide to keep it uh flat over time um but what we can do is think about this idea of splitting pools right and this is like fairly straightforward you know if you have a million shares being created every month um those shares are being split based on whoever's contributing in that month and so you have this sort of natural modulation effect that occurs where you know if your dow has like let's say 10 contributors uh one month um each of them is getting 10 let's say assuming it's being split evenly evenly um if the number of contributors goes down each of each contributor is getting relatively speaking more and so the mechanism has a way of sort of naturally modulating attention and responding to that um if you're able to track who's active um in such a way that yeah people are able to earn more and the incentive to stay stick around actually goes higher the fewer people that are there and and this is kind of what we wanted when we mentioned uh regular or higher relative uh rewards in those kind of like down moments or those moments when attention is really dropping um and so ultimately uh just to go into the sort of splitting pools there's a lot of really interesting things we can do when we start to think about like a million shares being created every month and then those shares flowing to different pools because we can then start to program these pools independently and use them for these different purposes right so we can imagine that a role in a dao might have a particular pool let's say all the treasurers or any treasurers who are active uh split 10 of the month's dilution um or you could potentially have a pool for just like a community treasury basically so that uh any sort of discretionary rewards can be given out at any at any moment um or you could have like pools for other random different types of distribution methods right you could have incentives for people who are referring other members uh to the dao or you could have incentives for people who are making posts uh or for people who are you know performing tasks uh basically these pools can be kind of like any functional domain within the dow any kind of effort they want to encourage that money can then flow to that pool and so the the dow itself can over time kind of just adjust what it you know the the the dilution pools uh based on what activities it thinks you know are most important in that moment so for instance next month maybe we want to focus more on growth and less on content creation we might shrink the pool that's going towards making posts and grow the pool that's going towards you know referral rewards um and this has this sort of like natural way of responding to the environment and lets the community kind of constantly reevaluate like what is most important right now what are we really focusing on um rather than being kind of like locked into this model that that we see a lot of communities doing uh where they set a token uh distribution plan and they basically just stick to it for you know a year and a half without without really being able to change it so ultimately if we combine all of this just to summarize um we have this very simple model right communities dilute a million tokens a month and those tokens go to whoever's contributing in that month and we get all these like really wonderful properties that we didn't get with fixed supply tokens so we have this like flexibility on a monthly basis you can recalculate um it incentivizes community sustenance it sort of recognizes that communities are people businesses and need kind of constant rewards over time it dilutes dormant contributors so if people are kind of slacking or you have this free rider problem it gets rid of that um it has this predictable dilution schedule so even though there is dilution to sustain it uh it isn't you know unlimited and there's a very sort of simple way for people to predict what it will be you know over time um we have this idea of the periodic recalculation this is again something really important for uh the kind of imperfect information does have especially early on um and then finally we have this because of the pools and the sort of splitting that occurs with people who are active you have this relative uh incentives you know sort of responding to fluctuations and attention so uh this is what we're building over at mico mico is a platform for social dowse and we're really excited to be on boarding people to our beta right now who want to kind of test out this new model uh we're kind of in the weeds on on uh building up new ways for for digital communities and dows to operate themselves and sustain themselves and so we're kind of always looking for new partners if if any of this was particularly interesting please feel free to reach out to me i'm at mark bailey on most platforms and i love to talk about this stuff happy to answer questions and if you're interested in using mico feel free to just shoot me a dm and we can we can get you on soon so thank you very much and if you have any questions i'd love to answer