Rewiring the Commons: Hypercerts on AT Protocol - Buenos Aires

Um before I actually start jumping into the topic, um ah there we go. Um I want to briefly give a shout out to finding the comments because finding the comments and hypers have been basically a partner since day one actually since day zero. Um because hypers project was started at a meeting um in New York after David um gave one of the major talks about hyperserts at funding the comments. um in 2022. So a couple of years back already.

Uh and then also to the residencies on you see some pictures from the recent residency just uh north from Bonis Iris. Um and shout out to uh Ken Shafi and Adam for doing amazing work there and also besides just residencies um and a lot of the stuff that I'm presenting today is actually based on on the work that happened in the residencies. So uh let's get started with the topic. Um I feel like I'm preaching to the choir here when I kind of give the framing of on why we actually built hypersert and why we think that collective value is not valued and provided enough in our society. Um but um for everybody who is not familiar with the public goods and commons space um our economic system basically is blind on that eye because it just um aggregates individual preferences and willingnesses to pay into prices, supply and demand and that's really powerful but that's not powerful for public goods and common goods just for private goods.

Um and um what I mean with that is basically everything from digital public goods to regenerative land projects, research and devel development in the valley of death, um investigative journalism, community organizing and so many other things that actually make life um like kind of worth living. believe um and um of course we have policy systems that are also intended to solve this but they were just not built for a complex world at the pace that we experiencing right now. So that's wow that slide looked different. Um what this slide actually said is I believe we can do much better and if you're interested in the partopian concept from Eric Drexler just Google it. Um it's an interesting concept.

Um cool. Um so I believe we can do much better and I believe that because we have now the technology to actually coordinate among ourselves of what we actually value together and then make it happen. And this is kind of like what a lot of funny the comments is about. So what's next? Um so now I basically want to merge kind of these our work on hyperserts um also with a new technology that we are using um at proto the decentralized um social media protocol that blue sky is built on but it's much more flexible and it's an exciting stack to add to kind of the blockchain um stack that we have been using before because it adds much more flexibility in organizing the data that we need to also close kind of the evaluation loop that many speakers before have uh kind of criticized that we don't have it and that we um need it.

Cool. Um so we actually call this kind of value recognition networks and networks because in the network that's where the knowledge is. We have to actually identify who are the players for a given set of problems that know really um what is kind of the value that this provided that that can be on the ground um communities that can be experts in open source software or that can be AI algorithms that know it better in some cases than the traditional um evaluation uh mechanisms. So, hyperserts um serves a function of anchoring this discussion in a digital artifact um that can look like this, can also look differently, but it really just describes the work if that is an open source software in events or in the in the environmental space. And uh for the comparison with kind of social media, when you think of a post with replies and likes, you can also think about an impact claim that is as easy as making a post.

So now you're you're making an impact claim of who did what when and where that can be going in the past can be can go into the future. Um and then we can add measurements, evidence and verification of that evidence. And of course um what's important is also who says that this is a measurement. Can we trust that person? So we actually want to have kind of a social graph.

Who do I trust to make the evidence to make to make the impact um claims? And then um interestingly we can also combine all of this into evaluations um about the impact claim so that this kind of post about what you did actually become becomes something that is valuable um in itself. So we want to make impact claims and evaluations as simple as posting on social media which of course doesn't mean that evaluations the everything the work that is behind it is easy but it should be that easy to actually bring that and make a public statement about it. Um this is a little bit the infrastructure that we are now using um that most of this is already provided by at proto. So we start at the left um with kind of um lexicons that in the case of social media define what is a post, what is a like, what is a reply.

In our case it is what is an impact claim, what is a measurement, what is an evaluation. So then the second step is that data is stored on a personal data server that can be um a server from blue sky can be a server from the hypersets foundation or it can be self-hosted by the user so by the project by the evaluator so it's your data if you want to um take on the burden of um self-hosting but you can also use kind of the foundation as a hosting provider then it gets aggregated And then when it's get when it gets aggregated um actually multiple different platforms can use it. So here it can be used by m earth can be used by gainforest can be used by your own project where you race directly on your project's website. Okay. So and this is the nice thing that the projects then can independent of the platforms have their data according to that schema and raise funds according to different funding mechanisms because not one funding me mechanism is kind of the right thing.

Um because like a land project and an open source project they need really different funding mechanisms and that's why kind of this infrastructure still works for all the different types of funding mechanisms which we are not building but we are working with the projects that are building the funding mechanisms on top of this infrastructure. So then um extending this um in some cases we want to have more economic security around like the claims and the data. So then we can anchor the data on chain or we can tokenize the impact claim which allows you to sell your impact to funders and it allows funders to actually not only say I funded a project with a 100 bucks but you can actually say my 100 bucks actually funded 1% of this activity and this activity was evaluated as having the impact of X. So we are actually moving away from just a input measurement to actually um claiming what you funded and the impact of it. So ultimately um this is a shift away from data silos towards user control data.

Um the donation receipts are replaced with proof of funded activities and outcomes. Um we don't have just one centralized impact evaluator. We actually have many different evaluators with different approaches. um which is also great in order to incentivize um kind of the right and robust evaluations. So this could be that if we have already like three um three evaluations that go into one direction um and the fourth evaluation just says exactly the same thing, that's not very informative.

But if the fourth um evaluation is actually saying something else and turns out that it's right so that the project for example didn't actually have the impact that it claimed then um it can be a very informative evaluation and we can incentivize exactly those kinds of evaluations. Um yeah, I already um said that we need many different funding mechanisms. Um and what we really want to enable is the verified data and trustbased uh funding. Cool. Um so the second to last point here is when we actually look at the um grants applications today everybody uses AI to fill out the grants application today but in the future um this will be even further auto automated.

So you will just have your AI funding agent who runs out into the world and looks for funding applies automatically and you might just press a button to confirm. On the other hand, the funders will actually also get a lot of different um applications and they use AI to sort it to select it and ultimately just AI agents talking to each other. This leads to a kind of hopeful case because what it means is that everybody has a beautiful narrative and nobody cares about the narrative anymore because it doesn't distinguish you from other projects. What could happen then is ultimately the best thing for somebody fundraising is actually doing the work and allowing the right evaluation of it. So you have verified data that you can plug into your AI funding agent who then goes out and does all the all the work that you didn't want to do before because you actually just wanted to build open source software or take care of land or organize a community event.

Um and then the topic of the evaluation function we had already before. I think this needs to work in a really fast decision loop but we need to add then slower feedback loops that evaluate how we made the decisions. So if we just make decisions based on satellite imagery and um and community input on how to distribute funds to regenerative land projects. Then we want to evaluate if those are actually good metrics to use or if we should add others like bio acoustics or local NOS's that verify that the projects were actually doing impactful work. I think I'm running out of time a little bit but this is probably the biggest question that that I'm getting.

Who will actually buy hyperserts? Um and there are kind of in all impact domains at least these four types. Um so basically people and organizations that just want to fund impact directly, people who need to demonstrate um that they had impact um communities that actually coordinate in order to fund together and then also local and national governments that want to prove that they use the public funds effectively. So and in different impact domains this will look differently and a lot of people are saying to this like that's not realistic but I think we need to push all of us towards actually aligning in order to fund these public goods together and we can use new mechanisms to do so. Um I skip over this.

So to just tell you also what we are very concretely doing because I came in from kind of the the high level introduction. Um right now we are working with my earth and gainforest. These are two projects that want to fund regenerative land projects and um they are building on top of hyperserts and um we really this is kind of the focus um on making the protocol work work let let the protocol work for them and then we can also expand to other projects. research is um currently looking at hypersets for science. funding the comments is experimenting for uh with it for events and then FAD um is doing project financing in Taiwan with it but there are also many others and you can also build your own platform on top of hypers to build new funding mechanisms because in the combination of at proto and kind of the blockchain elements that um we retain um it is extremely easy to for you to build.

So the road map is that we um will release kind of the new SDK with the at proto integration um and then hope to have kind of the examples ready in the beginning of 2026. And if you are interested in just like seeing what we are building um scan the QR code this is a link to our announcement channel on telegram where we have linked kind of um to the GitHub stuff to another article where you can see um what is uh kind of what's happening in the hypersuits world um and then yeah get in contact reach out there are a bunch of people uh here also in the audience Adam Ken and Chaffy um feel free to to talk to us uh what you can do with this um Yeah, and I will be around as well. Thank you.