The Future of Public Goods Funding - Buenos Aires 2025

Let's do this. Thank you everyone for joining. We now have a amazing panel for to talk about the future of public goods funding. We have a really interesting group of people here today. A very wide breath of things that we look at.

Um so why don't we start with very brief intros and I'm going to tee up the first question because I think this is something that we in the in this space disagree on all the time. How do you define what a public good is a as a part of your intro? So I'll let you Arun uh start.

Hey everyone, I'm Arun. I'm the blockchain lead with the UNISF office of uh innovation. Uh what is a public good is a excellent question. I was actually thinking about it when I was sitting there in the in the morning. So um there are of course digital public goods which we talk about a lot.

um like the Roger was here with the tour project of course the the entire blockchain system Ethereum itself uh can be considered a digital public good uh which you know serves people um but from a UNICEF perspective um we are of course huge on the digital public goods we are part of the DPG uh A that's a digital public goods alliance but majority of our work concerns public goods that uh affect everyone like things like air, water, um educ education, healthcare. So um these are also public goods and u u for example we are doing a air quality monitoring project in Lao where we install air quality monitors in schools there and then it it captures the data and we engage children. So um the quest for me is to see how web 3 can not only um help incentivize and grow digital public goods but all public goods. Yeah. And just so people in the audience, I'm gonna double click into that if they're unfamiliar with your work at the UNICEF.

Could you just share a little bit more about the types of programs that you work on and your focus?

Sure.

Um, so I work in the office of innovation. So here uh my role in specific is about exploring the applications of blockchain and web 3 to um internal efficiency questions as well as how we deliver to uh children and as well as incentivizing digital public goods. Uh some examples for internal efficiency could be how can we apply blockchain in supply chains uh because UNICEF has a huge um supply chain. So we are running a big project in UNICEF Bangladesh for example. Um example about um how we can serve children could be we uh ran a pilot in UNICEF Bundi uh where we um made a version of a uh savings group onchain.

So this would be where groups of youth get together and they save together. Um this one we didn't use crypto because crypto was not legal in Burundi. So they still used a box where they threw in the physical cache but we tracked all the transactions on chain. So this uh we did this in this the C chain of in avalanche and then once this was um tracked properly we could then create a credit score and then pass over this data to micro finance institutions. So this is an example of how we can serve people directly and then how we can incentivize uh digital public goods.

So we have a drips list which we made as of yesterday. Amazing team at you know drips. Uh so uh that's fantastic. Uh we want to see if uh we can use karma gap to uh to um for the startups that we incubate to report on and then UNICEF will attest on it. Um then we have actually ran a DAO uh in the style of a protocol DAO and we are looking to expand that work and see how we can engage our donors in a I don't want to call it a DAO but sort of like an engagement mechanism.

So like like an example of donor would be like we have ARM uh from the UK it's a corporate donor we have Ethereum foundation of course then we have the government of Finland. How can all three of these extremely different personas come on a platform and then influence us in a certain way in how we deliver programs. So so yeah cross-section of what we what we have.

That's amazing. I'm so glad you're here. All right Julian.

So I'm Julian Zistoski. I started Colin Foundation. Um and Octant uh which is our main uh project at the moment and Octant is an experiment and vehicle in funding public good things uh mostly in web 3 and my background is in economics. Actually before uh moving to web three I uh I worked as a researcher and consultant mostly in um public economics. So like exactly in the field which is only partially coincidence because with that background of course I had some ideas of um running octant.

As for my definition, um so I think of course like a textbook definition of uh public goods is is very restrictive and impractical. I believe that at least for um web 3 u use cases the very good definition is to fund things that cannot be funded uh by pure market forces but are still useful. But also I think this definition can be extended uh not only web three but also beyond to things that perhaps are funded but market forces but are are creating negative effects that are greater than the benefits that are created and then perhaps there's also sometimes um a need to to step in with like a non-market forces that of course not necessarily has to be always funding that might be other things.

Could you give me an example of that? That's really interesting concept. Well, you know this uh this this this is one of like a economics 101 basically. So so this is about uh negative external externalities of of things you do. So um you know like a pure free market does not care about for example um green houses emissions.

So, so, so that of course cannot be solved with public goods funding but that can be solved with some mechanism designs uh as we would say in the decentralized space or with some regulations as would as that is try to be solved in a let's say like a real world or like a legacy world which I prefer. Um so that's a good example but another example for example of u creating a lot of like a uh negative u effect on a societal level um much outweighing benefits are I think modern social media so so so which of course like people use them they they bring certain benefit they are like some they have utility but at the same time they both on individual and aggregated level are creating massively negative effects both like a just consuming your time when you could do so many beautiful things that are not related to you know like a destroying your spine and eyes not mentioning brain rot and they are also destroying our societies in a way so yeah I I will stop right here because I'm getting too long I believe

it's a great answer thanks all right Eliza is your mic No.

Oh, no.

Wait. Yeah, it's working.

Okay.

Hi. And so my name is Elisa and I work at Ethereum Foundation and working with the multiple initiatives, but some of them are related to like creating mechanisms and ways to like to create in public goods sustainable in the future. So where we don't need to like for example do like fundraising but also like mechanism that can allow and just do like a massive way of like and understanding like the impact and how they can how they so from funding perspective how it could be like how it could be solved in in different ways. And so then Julian said like a lot about like what what public what sorry what public goods mean and but I would also like to to say that it's like about like this creating impact and like very powerful organizations that we we use every day and for like when a lot of like protocols and a lot of organizations are using these public goods which are creating a lot of impact and they are non-commercial so it's they most of them are struggling struggle when it comes to like okay how can it can they be visible how can they monetize some of that and so it's yeah from short shortly it's and about that and also like I'm I'm sorry that I need to leave at 2:30 and because I have a hard stop and Vitalik made it made us difficult to start earlier so and yeah I'm apologies for that and but the guys are going continue with the panel.

We're lucky to have any time with you that we can.

Yeah. And please introduce yourself as well.

Oh, yeah. Okay. So, I I'm Raymond. I'm I guess moderator participant uh uh serving dual roles today. I'm one of the co-founders of Open Source Observer where we measure the impact of public goods in different ecosystems.

To date, mostly it's been in software supply chain and digital public goods like open source software. um we had take a pretty neutral definition to public goods in the sense that I mean certainly the economic strict definition is easier and when we can apply that we will um but I think uh you know as people on Twitter say public goods are good I think you know good is in the eye of the beholder so we you know I think different communities embody different value systems and I think that's an amazing thing um and what good looks like and is unsolvable by market forces is going look very different in different context specific settings whether it's in software or climate or whatever else. Um and we just want to measure what that looks like and just help different foundations or different organizations or different mechanisms um figure out whether or not the me the mechanisms that they're using the capital that are deploying are actually having the intended impact or the intended outcomes and if not how can we use data to help them understand iterate and design better experiments to more effectively leverage that capital. So that's what we do. Um, and I'm happy to be here on this panel with you guys uh to to talk about mechanisms.

I think one of the things that's most interesting about this diversity of panelists here is that we have all different very thesises I think of what's going to be most effective and how and and different experiences of experiments that we've run in the past. I am curious out of all the different public goods mechanisms that you've seen so far and and especially things that you've deployed and have hard evidence and and war torn stories about um if you could tell us what you feel like are the most promising uh public goods mechanisms that you've seen so far. What are the things that didn't work or what are things that you find you know underexplored that you think needs a little bit of double clicking into? I'll start with with you. Um I mean I'm I'm not really an expert in mechanism design but I really hope to learn from everyone.

Uh but so far uh what we have seen so uh we have an impact staking pool a lio impact staking pool. So u that's an interesting mechanism I think so where people can stake their ETH and uh choose to donate x percentage of the staking rewards to causes that they care about. Uh in fact I believe the talk after this one is by Jade from launch notes who helped us do that. I don't know if you're here Jay. Yeah.

Oh there you are. Yeah. So uh that's a super cool mechanism. Um then we really like drips as a distribution mechanism. So um the question of how do we maximally utilize what drips offers right?

So u uh there are so many digital public goods that you know unis supports uh should we all put them on the same drips list? Should we cascade them in some way? Um I also want to see if we can add a like like an NFD um to certify onchain that X project is a digital public goods and see if drips list can read the status of that NFD and then choose to turn on or off because um the DPG certification lasts only a year. It's not lifelong. So if you're committed to digital public goods, then you have to commit constantly.

Constantly it's a never- ending process, right? So you have to keep on committing and you have to keep on showcasing proof of that. So how can we get that on chain and then connect it back to the back to the trips list? So that's something that I had a chat with the trips team about. So let's see.

Um but um so far whatever I've seen is super exciting and uh from the UNICEF perspective we are happy to experiment and learn and then yeah hopefully do some do some good.

I love that. Um speaking of yield assets, have you heard of Octin? Uh and

of course uh I've had few chats with with Octant. I know about the wallet uh the V2 vision and that's something that we are looking into and we still haven't figured out how to uh work with you you guys said I mean I haven't met you till now so I've spoken to everybody else I think

so uh we must work together obviously so I think it's sort of a no no-brainer so it's only about when not if yeah

and so I it's we had a like some conversation and before this this talk and it was very interesting to see like how everyone has a different perspective when it comes to funding mechanisms and for example from the the foundation perspective we see there's a lot of appetite from different organizations to donate and and to to fund public goods they don't understand exactly like what exactly so how to allocate the funds so it's like okay we give we give you guys a bunch of um funds and then you will But the thing is like from our perspective is like okay how I mean the thing foundation is a steward of the process so we cannot receive the money the funds directly so we always need to fund find other mechanisms to make that possible through um yeah it could be like oxen there's also like some other like ways to make it possible so I think like in the future and there's going to be like some mechanisms and designs and to support that and like while creating and maybe like holds and like people can or like organizations can locate funds there and through like any governance process then they're going to be res redistributed in a way where like some I don't know I feel like some it could be like deep funding it could be like there's there's nowadays there's some mechanisms and but I still I feel like there's a lot more to be evolved there when it can when it comes to allocation different funds and governance will also play a an important role there. And yeah, and then another one that I would love to see more is some way to for example to create donations in for example when when you do a transaction on when you for example let's say when you purchase your ENS. So what if like people nowadays people don't have an optin to donate like anything at all. So I feel like there's going to be like that's also like another interesting and perspective when it comes to like and funding public goods in the future and yeah and I would like to see a lot of projects working on that.

Can I come in quickly on um uh I I forgot to mention that uh we did a pilot with Gitcoin in 2022. So that's what the so we tried out quadratic funding as a as a mechanism and uh since you mentioned deep funding. So the so the question of um allocating efficiently keeps coming into my mind. So if I if I make a drips list for example and and if I have like hundreds of digital public goods then ultimately everybody gets like 0.0 whatever percent.

So that I don't know from the face of it it doesn't sound that great. So um how to allocate it using AI is is you know something that um we are also looking into of course the the open- source way is to look at dependencies and you know stuff like that. So for me the challenge is how do I learn from this like like deep funding and then how do I bring it back to the analog real world messy messy world of you know UNICEF. Can we measure what happens in the field well enough to allocate funding in a way that makes sense for whatever is happening in the field. So of course I'll run squarely into the oracle problem, right?

How do we really certify something that's happening in the real world like a child being vaccinated or or a toilet being constructed or you know whatever, right? So um that's sort of a a vision that we are working towards to to learn from all this from the digital world and then take it into the analog world in a meaningful way. So we still don't have the answer but uh that's the like a big idea. Something that I would like I can give you. Something that I would like to add there is also like educational towards which mechanism is the best for each one of the use cases or applications and because like because nowadays there could be like oh like actually this mechanism would work work well for for us but in reality there is another one which works better but like it's like still in kind of like in the unknown.

So like a lot of like education towards like what's available in the pre like in the in the current scenario like in the present then that's also like something interesting that's going to happen hopefully we'll make that possible please I'm sorry.

Okay thanks. So um I'm not known so I probably it will not surprise you that um I think we we don't really have mechanisms that work great. we don't have them yet. Also, I think for the sake of clarity, we need to differentiate between like a diff different element of the system. So, so and in the context of the things that were mentioned, there are like um methods of getting money and methods of allocating that money where you want to spend.

They are of course totally independent. So, so you can like a pluck many methods of getting money into other methods of spending money. Now, in a way getting money in that space should be easy because there is a lot of money in that in that space in theory at least. Uh but of course the space is allergic to taxation. So everything needs to be more or less donation based which is fine.

But what we learned. So maybe Ethereum Foundation um uh has a good times uh with people like asking to take the take take their money but Ethereum Foundation is this enlightened technocrat. It has a lot of legitimacy which is great and Ethereum foundation makes this space running. So, so this is great that we have Ethereum Foundation is this enlighted technocrat. But in general, when you want people to give you your money, they ask, you know, why is it going to work?

I don't think grants work like uh what what do you guys do to make them work? And and so on and so on. And then we come to the second part which is how you decide how money is spent. And this is really hard question. You can of course do that as enlightened technocrat as Ethereum Foundation does it very often.

But then first you need tons of legitimacy to get away with whatever your decision is. And by the way, like uh this is not to be, you know, uh rude or whatever, but Ethereum foundation very often do not make like a funding decisions public exactly for that reason to not be challenged with this decisions because they are like a smart technocrat decision who knows better but does not want to spend like a all life on explaining why this is better and so on. So I understand that. I'm not sure this is like the best decision, but I understand I accept that and you know this is their money after all. But then if we want to gather funds from other sources, we need much better allocation mechanisms than what we have right now because what we have right now uh I would say massively doesn't work.

So it kind of works of course like so so I think quadring funding for for example is like a massive improvement towards just like a linear donations that like decide how matching pool is distributed. I think that the fact that we know how to involve community into that process is also massive improvement towards not being able to involve community in that process and so on and so on. But just to cut it short here, this is like a long way and we and and by the way this is unsolved problem like this is this is unsolved problem in public choice theory. So this is not that like you know like all the world is smart and we are dumb. We are much smarter than the rest of the world because we are trying to figure that out.

We have uh skills, we have technology and we have programmable money and we will get it right but we need some more time. Well, hold on. Let me double click into that because I I feel like work can't mean solve all problems and unicorns. I think what does work mean for you? Because work could mean a lot of different things in different contexts as well, right?

Work could mean sustainability. Work could mean you achieve some kind of outcome or some kind of goal. Work could like when you say it's not working well, what are the satisfaction criteria? We we don't know. And this is why it is not working because um you know the what what what you what you want to do in public goods funding or you know funding with so so so so here is exactly the problem.

So so um public trade theory and how you fund things is about uh finding out what are the preferences of the people. Now this is this is very hard like Kenetaro proves at some point that it's impossible for many reasons like one of them is that you know some people prefer bananas some other prefer apples then you cannot pre compare apples and bananas because apples are not bananas and some people prefer steaks or eggs and um on top of that many many people do not know what what they want. So, so, so if you for example were to fund Ethereum progress with a public vote on which AIPs should be uh included in the next protocol upgrade, that would be obviously like massive disaster. Massive disaster. But um uh we don't even know uh who is the Ethereum community and and whether Ethereum community wants more privacy or more adoption or more interoperability for example.

So if you don't know those things, how you measure the effects of what what you do against them?

This is why I mean we don't even know if they work.

So you're saying there's a tight interlinking between the political problem and the economic problem.

Well, I I like to say that uh funding is governance is ideology. So So because they there there is no objectivity here, right? like if you if you um fund scalability first or privacy first, this is not like a technocrat objective decision. This is this is governance or which is ideology. So what you what you believe is more by the way I use word ideology in a positive sense here.

So you always when you do things you always do that against like a set of beliefs which can which can be defined as as ideology. I mean, let me let me ask you this question because I'm impressed at how much you've been able to name drop all the different mechanisms in the web 3 ecosystem. Um, but UNICEF has done a lot of things outside of the web 3 ecosystem. Could you share your impression of what does it mean for something to work? What are lessons that you can share about things that the UNICEF has done that we shouldn't lose that we should make sure that we learn from th that past history and all those mistakes that have been made before and what are the things that uh we should avoid in in what we're doing here.

Okay. I'll give a specific example when you were talking uh what came to my mind is can um localization work to derive the needs of people in a smaller subset and then you repeat that to cover the masses. So uh we do uh we do a it's not digital but we do an analog thing. So uh so our humanitarian portfolio has this project called kits that fit. So um in humanitarian emergencies whether it's climate induced or you know human induced uh UNICEF is on the field delivering um life-saving essentials.

So what I think used to happen traditionally is something like the technocratic like somebody used to make up their mind and now what's what's happening is um we literally ask the people what they want. they say it in a in whatever way uh in in in a certain modality. We collect that information and then we prepare local supply chains to be able to deliver that and then once it's delivered we ask them again.

Then we repeat, we repeat, we repeat, we repeat and then the and then it just iterates again and again and again and of course if the crisis changes then the needs also change. Then we also change. So it's like a never- ending process um where we have to keep taking the feedback from the people in a small local environment and then multiply that across different environments. I don't know if that's uh uh if that's useful. Yeah.

Yeah. Absolutely. Um which Okay. Which are the mechanisms that are could you share also some examples of things that you feel like you know you in your past history know do not work and things that we should steer clear of

um you mean from the broader UNICEF

oh I don't know actually um what has failed I can share at a very high level one thing which is if we are experimenting with the mechanism. Do we know what we what the hypothesis is and what metrics we should measure in order to learn effectively? Sometimes the enthusiasm to try something new takes over. Something new and shiny. Let's, you know, try it out.

This is cool. You know, I can get PR, I can whatever that takes over and then um that leads to no good. So, I think before trying something new, at least do you have have you have you thought it through? uh if it if it's something large do you have a theory of change if it's something small do you at least have a hypothesis a metric and what I want to learn and what is success and failure as per whatever definition we have

the definition itself could be wrong but then the point is to discover that it's wrong at the end of the pilot I think that's something that I uh I made this mistake myself and uh I hope not to u do this in the future so now now now I want to have these things um done ahead of time so that later on I can say proudly that uh this didn't work because of this and this reason or this worked because of this and this reason and then we want to scale it up. Yeah.

Yeah. I love that. I mean at open source observer we are all about measurement. So uh we we definitely want to bring that kind of rigor. I think particularly in the crypto space there's been you know a lot of experiments that have been done without any kind of retrospective analysis.

And so there can be a lot of excitement but then you don't you know you see the talk of this interesting new concept but then when you actually deploy it in the field uh people lose like because the novelty's worn off and so you don't come back and you actually see okay this is did this actually work should we iterate again how would we improve it if we do it again so we love that we're all about that I think one of the things that I'd be curious to hear your perspective on as well as you Julian is how do you like you were talking about the preferences of the people right there's some nonquantitative there's some qualitative desire that this pe that the people want and then of course there's good arts law and you know you want to make sure that you're not measuring things counter to actually achieving the spirit of what these people want and so how do you capture that and make sure that you keep these programs aligned and not microoptimizing you know ramen cups into you know children's mouths right um I'm curious yeah how do you guys think about balancing that and making sure that what are the right mechanisms to keep that dialogue healthy and not uh hyper optimized.

Uh yes, that's uh that's a very good question. So first let's start with the understanding that you know like um people want good things. Uh they just don't usually agree what what is good. So so so when they want to do um America great for example or or great again like the chances are that the definition of great is very different for for for for everyone. So there there are many groups with different definition of great.

So and I I think it is possible to to to aggregate that somehow through like a learning about the preferences and once you know that you you can you can try to do like a impact analysis of like what was achieved was what not was not achieved. Now uh how to make sure that the um funding and then there's like a feedback loop of impact that preferably goes through some kind of u um impact market uh is really efficient is like a beyond time we have but I think it is possible and would be happy to to discuss that in in details at some point but I I think it is possible and and we are actually launching like a bigger research in initiative like we published like a a post about that on E3 search today where we invite people to to join that that institution to to find out exactly like that kind of things.

Um regarding the qualitative so um in the context that UNICEF uh works in uh by the way the answer to UNICEF's work in digital public goods is super aligned with whatever you guys are talking about um when I'm referring to the other sorts of public good um are our field work. So there of course um it depends on the context. If we are in a Sudan for example, it is such a harsh context that we come back to the most minimum basic things that need to be delivered with which has scientific backing to it. So there the qualitative feedback would be are you okay like does this work for you like you know it's something super basic right and and and it kind of works it's good enough because after everything we do we do a serious uh post distribution monitoring third party monitoring blah blah so all of that so we have a lot of qualitative data within which then goes back into our programming so it's that feed it's that feed feedback loop right um but when we go to certain other contexts where it's not so lifethreatening and when things are fine it becomes harder and and that's something that we also see in the digital world to some to some extent right so that becomes much more difficult and um if if we can if we can get a in-person discussion or something I think that will like really help because I think I mean Vitalik mentioned earlier that people's online persona is two to three standard deviations away and and it gets worse and so it I mean is it really represent ing what's actually going on like uh can it just be a purely online thing or do we have to bring some offline element to it in in order to really sense

better so I think it has to be a combination yeah

yeah mechanisms measurement but it all comes back to the people I love that message I think that makes a lot of sense that's why we're here um I want to give a round of applause to these panelists thank you so much for joining