New Applications for Quadratic Funding by Connor O'Day | Devcon Bogotá

foreign [Music] I work for git coin or I am a member of the Bitcoin Dao and we're going to be talking about a Nifty thing called quadratic funding so first off I just want to take a quick poll uh who here has heard of quadratic funding okay I see a lot of hands that's good that's good who actually understands it okay okay maybe half as much I'll take it uh I also you know I'm still learning so uh who has actually funded a project in a quadratic funding program a lot of hands that's good and who only did it because you thought you'd get an airdrop in the future I see one two three okay I like the honesty I like the honesty so the too long didn't watch uh for the people that are heading out to happy hour right now uh quadratic funding has proven to be a powerful tool but primarily in Niche communities like web3 and with few new advancements over the past few years uh which is basically qf is good as adoption has grown problems are becoming more apparent scale is hard and things like decentralized identity Community curation tools will boost scalability and then tweaks to the mechanism design can help Drive adoption as communities tailor qf to fit their specific needs and we're going to call that Innovation and experimentation equals Transcendence so that's kind of the the high level overview if you need to go you can go but I think it'll be pretty interesting diving into the details here uh so first WTF is qf spoiler alert for the whole presentation that is qf uh basically what it means is funding for any given project is calculated by taking the square root of each donation adding these values together and then taking the square of that result we'll get into that a little bit further on the technical side but basically the widely accepted definition of quadratic funding is uh it is the mathematically optimal way to fund public goods in a democratic Community there's always kind of been this emphasis on public goods and Democratic communities which covers many of the use cases but perhaps not all which I'll get into later and qf was kind of first proposed by uh Zoe hitzig vitalik buterin and Glenn whale in the paper you see on the screen right now and what it really does is it let us harness the collective intelligence of a community helps ecosystems signal and fund their shared needs and what it looks like in practice is your donations are Multiplied on this logarithmic curve and I really like this about a lot of implementations as you can actually see the impact of your donation as you do it and you can kind of see how you know a small donation can get a large match a bigger donation can get a bigger match but it really tapers off because the higher value is in the number of unique contributors rather than the total amount contributed by any one person so quadratic funding came from this idea of quadratic voting which basically allows you to Signal preferences but most importantly the strength of those preferences quadratic voting as a concept existed before crypto but was popularized by Glenn whale which is also kind of independent of crypto but kind of at the same time the ecosystem was maturing but compared to you know one dollar one vote or one person one vote systems it really it really allows you to Signal preference in a much better way and so qf and public goods this is kind of the common theme you hear everything here is about funding public goods everything quadratic funding is usually in relation to public goods and so what even is a public good uh it is a non-rival non-rival risk non-excludable good uh for example you know clean air or open source software which is one that we focused on a lot these are often discussed as a pair but I don't think that will always be the case uh we're kind of starting to stray from this public good Goods pigeonhole where the types of goods is more of a gradient that people are understanding and there are things like Network Goods Community Goods plural Goods even private Goods that could benefit from qf and it doesn't mean we are funding public bads but it means we just have to kind of rethink the the framing that we're using for uh for you know what we're funding with this mechanism so where are we today uh get coin in our history since we started in 2017 has funded 69 million dollars uh worth of Open Source software and public goods which is a pretty nice number to Round out for this presentation that just kind of worked out nicely and this includes grants bounties hackathons tips a bunch of different things that can happen on the gitcoin platform not specifically just grants uh and for the git coin grants program we've seen kind of round over round growth now for over three years uh we started in 2019 uh with the first round had a 25 000 matching pool I think it got like thirteen thousand dollars of community contributions uh and two 200 contributions total so not a whole lot fast forward to Grants around 14 in June uh 600 000 contributions and over five million dollars raised for these projects these public goods these open source software projects uh which is really cool to see and these visualization graphs you see are kind of the the Network and 3D visualized of all these donations and all of these connections between users and projects and the most recent round grants around 15 just ran in December or September sorry and what I really want to call out here is that this wouldn't be possible without the the matching pool funders so these are you know Dows organizations you know private companies individuals that want to donate large amounts to the matching pool in order to incentivize the community to also donate to public goods and that last round had over 450 000 contributions over four million dollars raised and almost 40 000 unique donors but yeah so this wouldn't be possible without all these logos you see up there right now and so just want to give them a quick shout out because it's super important to what we're doing and so we're you know experimenting with all different kind of types of quadratic funding rounds we kind of started with one main pool we broke it into categories and now we're doing what we call ecosystem rounds so we've done I think 20 plus unique ecosystem rounds in the last year or so and over 40 total a lot of those are recurring um and some of them are kind of crowdfunded and more focused on a theme like the open gaming round or the ZK Tech round uh versus many others are kind of a single funder trying to grow their own Community their own protocol uh for example you know ens polygon and some of the others you see up there and the way I view it is this could eventually be part of the Dow tooling stack where it's better than like a yes no governance vote it's faster more democratic than traditional Grant programs uh so you know you can imagine a dow that needs to decide what work streams what hubs whatever they call their projects within their data what they need to fund they could use a qf tool or QV tool and with quadratic funding it has the added benefit of raising additional funds for those projects and then cause rounds I think are what are really interesting as well where we're kind of trying to break out of this web3 bubble and you know make an impact in meat space in the real world and we've done rounds for climate diversity uh crypto advocacy uh DSi longevity we did around for Ukraine and there's still a lot of overlap between the web 3 communities and these communities you know for example in climate like a lot of the climate projects we see are kind of crypto focused climate initiatives which isn't a bad thing but what we're trying to do is really onboard you know non-cryptonatives to the the blockchain the web3 ecosystem which is difficult for everyone but I think we've been making good progress there and here's some examples just from the climate round I think these are really interesting because it really has like a global widespread impact and it's very cool to see in this case we are kind of onboarding new users to web3 which you know comes with a lot of manual hand-holding and and onboarding but uh it's really cool to see the the progress that's been made in in these areas and the last one I want to mention is these uh kind of geographic UF rounds so the regional and local rounds um I want to call out specifically the the Devcon uh Latin America focused CLR fun round so that's happening now uh git coin is not involved but you know we work with CLR fund we like them a lot and everyone should go check it out because they're also utilizing some of the mechanisms that I'll mention later in this presentation uh but yeah so that's that's happening now uh and as get coin we've done a few rounds like we did a an Oakland Youth uh charity focused round uh we did a China crypto ecosystem focused round and the downtown stimulus one was actually during covid and it was focused on Boulder Colorado and Reviving the local businesses there and I I think the thing what I want to say on this is just like we're all experts at what's right in front of us in our communities you know it can be difficult to fund projects that are across the world from you but this this mechanism can really help in in kind of you know shared shared public goods or localized public goods that maybe don't have an impact around the world but have an impact in in your neighborhood or your town so Growing Pains um civil attacks are the single greatest threat to quadratic funding if you haven't heard a civil attack is basically one human making you know many accounts um with quadratic funding you can imagine if I can make a grant I can donate it to my donate to myself from 100 different accounts that totally excuse the mechanisms steals funds uh and it's not good and so it's a constant kind of battle uh between you know mitigating civil attacks but also you know trying to respect you know personal privacy and identity systems uh and it's this game of kind of Red Team Versus Blue Team that's constantly evolving so we've seen civil attacks really increase as of late I'm not sure if it's like more awareness you know the bigger matching pools or just the bear Market people are running out of things to do to make money so they decide they're going to civil attack grants but it's an issue and we see you know we see the self donors who are trying to cheat by donating to themselves we see airdrop Farmers uh which you know are are typically not uh they're not trying to do bad necessarily they're just trying to get as many accounts as they can you know eligible for possible air drops which is also not good because it skews the mechanism and I think that's partly in due to a lot of the airdrops we've seen that have went to public goods funders in the past so we're working on mitigating that and then just general collusion between different real humans it's really hard to tell what's collusion between legitimate people when you don't know their full Identity or you know what is actually a simple attack but in both cases they're trying to mitigate that and then we see fraudulent grants so you know impersonations Grant emulations which is what we call making up a project and making it look like it's real but you know the GitHub is empty the Twitter has five followers but they really write it up to make it look real so we're dealing with a lot of that and the solution to all of this is identity so how can we Channel this airdrop farmer energy into raising more funds for good causes while not compromising the mechanism's Integrity um it's it's inevitable that people are going to donate to things because they think they'll get something out of it eventually and as long as we can kind of channel that energy you know in such a way that it doesn't you know hurt the mechanism and other grants then I think we're kind of secretly converting these uh these degens into regions without them even knowing it so Bitcoin grants right now the one the platform that most of you know or some of you may know it's still fairly centralized but but that's okay you know we can't go to governance for every single Grant decision so governance helps direct eligibility requirements and then we have a centralized team that must be trusted to act on those decisions uh but you know this is a high bandwidth uh task for our team uh humans are prone to error it's centralized and you know at the end of the day it's not what we want to be doing with with our time um and so as we kind of move towards the path to decentralization which I'll get to more later uh we're thinking about launching this permissionless protocol uh we have Brazilian identity systems in the works uh community curation and really self-serve product features so people can kind of choose how they run their own rounds uh and one of the one of The Primitives I'm most excited about is around decentralized identity and that is Bitcoin passport so this is an open source uh dap uh preserves your privacy it's fully decentralized the data is on surround Network gitcoin uses it as you know to print sybils and provide trust bonuses to donors and it's all open source and anyone can integrate it with their own dap so we already have actually 60 000 passports created on on git coin which is really exciting and some of the Integrations that we've seen are uh beef Staker bankless Academy Rabbit Hole snapshot Guild are already using passport uh you know mainly for civil protection or quadratic voting or similar use cases like that and once we kind of nail this simple resistance with passport or you know through many combined means it unlocks this spectrum of possibilities of quadratic funding quadratic voting Fair launches Universal basic income one person one vote Dows data collectives and plural funding uh and then the other the other mechanism that I'm I'm getting more excited about is you know different uh different ways to incentivize community curation so you know curation is creation and we're experiment team with things like Grant and identity staking where you know you actually put value on your Grant or on your identity and personally I think the the Revival of token curated Registries is going to happen so if you're not familiar super bullish on this but uh tcrs were kind of popularized you know a long long time ago in crypto land in 2017 but it's an incentivized voting game that helps create trusted lists that are maintained by the very people that use them so essentially using the wisdom of the crowd principles people collectively vote on what submissions are valid what should be removed and they're rewarded in tokens for doing so uh this is this is during the 2017-2018 cycle there is a lot of hype around this and I think it kind of died out but I'm slowly starting to see it kind of come back into the the main web 3 discussion so even if it's not tcrs themselves some type of community community curation is is uh absolutely needed for you know a decentralized grants program all right so we're gonna go into some of the new flavors of quadratic funding um lots of people are doing really dope stuff with uh public goods funding quadratic funding and we're excited to work with all of them to do so uh you'll see some of the logos there on that unnamed bag of candy uh but you know these new flavors we're really going to try to Taste the rainbow here so I first want to established that almost all qf implementations kind of stray from the math uh when proposing you know adjustments or new experiments like we often get the reply like oh well that's not real qf um hence kind of the the meme I dropped in there but it took me a while to kind of fully grasp like what the image you see uh on the side there means basically the green is donations the yellow is matching and as we said before the matching is calculated by taking the square root of each contribution so those those blue lines you see there adding it up and then taking that square which is the full the full square of yellow and green that you see there and so that would be like pure qf right and that would happen if we well so so it doesn't happen in this case because what gitcoin does is we have a fixed matching pool and a fixed time limit uh which means matches will scale up or down based on the saturation so if a round is oversaturated or under saturated uh that means all the matching is allocated or all the matching is not allocated um then you don't have this perfect square that you see here and most of our rounds end up getting saturated and what you see after that point is it's kind of a zero-sum game which means you know there's no additional matching to provide but as more donations come in it just shuffles around where that matching goes so you know some people have argued that it should kind of end after it's saturated and that would make that would be what's kind of like a fixed pool only where as soon as the round is fully saturated it's over which I don't think would be a good user experience because people want to know how long they had to donate and when it's going to end but it's one thing we could experiment with uh the other thing is that there's a fixed time limit but a variable pool um you could run it that way but you essentially need infinite money uh to make sure that you're not going to run out and that's not really feasible either so we kind of have this Middle Ground here and then there's other things too the trust verification bonuses uh we have matching caps on certain rounds where there's a certain percentage of the total pool that you can take um pairwise bounding which I'll talk about in a second uh and many other mechanisms that kind of take the initial qf you know math formula and then add some tweaks to it to kind of help the the experience so pairwise bounding I'll talk about first uh this is a mechanism designed to proactively prevent collusion uh muting the impact of groups groups of contributors who contribute to the same grants as each other so if you look at the image uh behind me here you see those big clusters uh those are all donors that are donating to similar grants and that are you know have connections on their network path and so those donations are going to be are going to earn less of a match because of that kind of assumed collusion so less correlated pool donors is more matching more correlated is less matching and then we could also experiment with like plural pairwise qf which which would you know with new identity systems like VCS did IDs sbts we can sort of better figure out how close or far a single individual is to your network and you know kind of take the the pairwise bounding a step further uh Macy's and other really popular mechanism this one's actually used by CLR fund and it's all about privacy so you know privacy is a public good privacy is a human right and Macy voting or funding is private helping to avoid collusion and vote buying uh so in quadratic funding users could not cryptographically prove uh to others how much they've donated or who they've donated to some of the caveats there I think is that it relies heavily on civil resistance assumptions as everything kind of does you can only make donations on one chain you can only contribute once to around and you know if there are trusted parties you know each other in person like they could still potentially include but in any case I think you know privacy is really important and this is a good thing that uh we'll be trying to experiment with soon um retroactive public goods funding is another concept that's not necessarily qf it's more quadratic voting but the core idea is that it's easier to agree on what was useful than what will be useful um so optimism is not trying to be a grants program they don't want to fund what they want to see built in the future they want to reward what was already built so it's more export expert driven and proportional to kind of the past value generated from a project in their iteration they've kind of went for a whitelisting approach which reduces civil attacks and fraud but is less inclusive as a whole and the benefit of this is it can kind of incentivize the initial seed funding of a project so you know right now it's very hard to get initial seed funding for an open source project or a public good but if retropgf kind of gains traction and uh you know part of that retroactive funding goes to the initial funders then you might see you know private investors investing in open source uh with the idea that there's going to be an exit eventually so kind of working backwards it could sort of help fund the initial stage of many open source projects and I want to call out we we're working on something very exciting with optimism for retro pgf using the new grants protocol that I think I think we've sort of announced but uh you'll be hearing more about that soon uh it's super exciting and the other thing I kind of wanted to mention is just that like when I donate to normal quadratic funding rounds I'm often donating to projects that I think have done really good work like I'm also donating projects that I want to do good work but but I certainly think UF in general it's a good mix between retro funding and you know future funding but it's a lot easier to donate to a project that you know the name you know they're doing good work versus someone who's brand new um so all funding is sort of retro in some capacity uh but you know that's kind of a personal preference thing that we haven't fully explored uh okay we're running low on time but dominance Assurance contracts are another mechanism uh that we could potentially use uh it's a product of Game Theory meant to incentivize everybody in the community to contribute to funding the public good and avoid the free rider problem which essentially offers a refund to your donation if a threshold is not met whether that's the amount raised or the number of donors and with dominance the dominance part of it the the person offering the public good could actually would actually have to pay you some bonus on top of that so it kind of turns into a win-win for donors um you know your public good either gets funded and everyone's happy it's what they wanted or you earn a profit and it doesn't get funded and ideally in game theory that might be the dominant strategy for Free Riders that don't want to fund public goods uh and then back to quadratic voting so you know do we really need the F and qf um there's plenty of great quadratic voting experiments happening um and it might better serve the needs of some communities uh you really need strong civil protections or you know a whitelist system um and I find it kind of more useful for single Choice decisions like a verse B uh versus deciding you know on what scale to fund something and of course it lacks the benefit of raising additional funds via user contributions so what is next so git coin as I kind of alluded to we're launching a new grants protocol that's going to be fully decentralized um and we're kind of at a Crossroads because we want to you know decentralize qf uh using the latest and greatest web3 tooling but we also want to drive adoption of qf in new new areas in the real world which means making web 3 more accessible to the average non-cryptonative person so you know make an impact in meat space uh you know onboard normies and be grammar friendly so you know anyone can use it and we're kind of bouncing the scales here um it's a bit difficult to do both at one time and this is a problem for uh you know the entire web3 ecosystem is is usability and user experience but I think we'll be doing some experiments both on the the new protocol and you know in maybe a non-block chain you know qf based system so that's actually what the downtown stimulus project that I mentioned for the local rounds section um that was totally done in Fiat uh in the boulder local community and we might do more of those more social network qf experiments but finally what we wanted to get to is the get coin grants protocol uh so this consists of a suite of overlapping products uh passport round manager Grant Hub it's going to be permissionless modular and decentralized so anyone can try to uh can anyone can try different public goods funding mechanisms to kind of find what fits best for their community and this this really differs from our centralized programming platform now because we we have to manually kind of help run people's programs it's not self-serve you can't just spin one up by yourself and a lot of the a lot of the work uh is kind of you know based on our centralized team doing this this manual uh grinding to to get it done um and so this is going to be open to anyone and what we envision here is that you know project owners will use the grants Hub to kick start their project uh use it as a vehicle for growth and then you know at a point far in the future there could be hundreds or thousands of programs running at any time um all kind of designed and operated in a self-sovereign way specifically to serve their communities all right and this is the last slide I promise um so get coin grants 1.0 was a monolith it's on this centralized stack super outdated we're excited to move on uh the grants protocol is really going to take the power of Bitcoin grants and quadratic funding and make it available for any evm based chain and Community uh so this is going to be permissionless open decentralized uh you can take any of those mechanisms that I mentioned earlier or new mechanisms that haven't been thought of and Implement those into your program you can kind of self-serve and Choose Your Own Adventure so we're really excited to launch this we have a few test rounds that are going to be going live in the next few months and I think in q1 of 2023 we'll be moving the entire get coin grants program over to this new protocol which again will be modular decentralized and hopefully civil resistant so that's I think the talk that's the talk um [Applause]