 What I'm going to talk about is adaptive financing and some crypto economic primitive that we have for this as well as a use case and so that impact certificate is for a implementer of a project called Chimple which is live in India at the moment. It's distributed 500 tablet devices to 500 poor households that are using a learning program on the device and these are children who experienced not being able to go to school during the COVID epidemic over the past couple of years so the education level is really low so we have 500 kids in this pilot group and then we have a control group as well and we're collecting data on this over 12 months and so how do we finance this? So when we're talking about adaptive finance it's really in the context of crypto economic mechanisms. In general crypto economic mechanisms are financial systems that are built on verifiable stateful data and generally most of the DeFi crypto economic mechanisms you know the data is all just on-chain data so liquidity pool etc. What we want to do is enable DeFi to move to ReFi and to build crypto economic systems that are suitable for funding real projects and real state change in the world and so you know this is a sort of general principle that blockchains enable reality to exist within software and what we would like to do is bring in this layer of enabling this reality as verified earth state to be existing within the software as validated chain state and I'm gonna show how that kind of kind of works and so this can apply to finance you know so it can be I have made an investment in the real world and it's reflected on chain. It can be services so I have provided a service and I'm getting a payment both the record of having provided the service as well as the payment transaction can be stored in some form statefully on chain data the same thing and ultimately it's about reflecting changes in the real world and changes in the blockchain and when we bring this together we have a cyber physical system and with the state of the virtual system the blockchain and its related technologies reflects or was linked to the change in the state of the physical world. Now a very powerful idea here is that when changes in the state of the real world create changes in the state of the chain and when can do that okay and when changes in the state of the chain happen they can feedback as incentives to get people to change the state of the world and so this feedback loop is super important and this is where we need to look at okay well how do we make it adaptive so that it is linked to reality so what's the interface between state changes on chain state changes in the real world so the basis of this I'm going to talk a little bit of the data side of things in this feed in really powerfully to the impact certificates story and so we've been working since like 2015 on a protocol using decentralized identifiers verifiable claims and credentials to have a way of representing what's happening in the real world with identification identification verifiability creating these as verifiable claims assigned signed claims and then verification with the verification proofs which leads to the ability to to issue credentials and you can package those credentials together and put them into non-fungible impact tokens now impact tokens you know they're basically NFTs but what we've worked on is a much more robust metadata standard around this which uses decentralized identifiers and and and the did documents if for those of you are familiar with the W3C specification and so this has an identifier and it has the crypto controls and so on of a did document but we've added in some additional properties here and so this makes these these impact tokens programmable because we can now embed services and services can be chain services you know data data stores web to services whatever services you'd like to put in there as endpoints essentially and then writes so the rights can define what you can do with those services or they can be rights that are expressed in smart contracts or as legal contracts or delegated authorization Z caps and embedded in these tokens is the data the verifiable credentials the claims and and general data as well and you can link any resources to this so now we've got a package of of digital assets that is not only an information carrier but it's programmable and you can do really interesting things with it and I'll get to that in the context of the educational program in a minute and this really can be used for any classes so as Juan was saying earlier you know we have the ones that are obvious that are currently being digitized certificates such as carbon certificates renewable energy certificates and so on but really we can represent any outcomes in the certified form so whether it's nature-based health-based industrial IP etc basically anything that we care about and want to work towards investing and and spend our money on and because of the composability this can be built into any financial transaction any capital allocation government expenditures business processes consumer products and so on so you can for instance embed one of these tokens into a product as a proof that the product has a sustainable source for instance so coming back to our use case example of the temple pilot so we've implemented the script economic primitive that we've developed called an alpha bond and that's the adaptive financing bit and I'm gonna go into a little bit of detail there and this is a project that we we're running currently collecting the data in partnership with temple and UBS Bank so if we just take a step back and say cool what are we trying to achieve in this project it's from from an investor perspective investors want to achieve alpha and if it's their impact investors it's impact alpha so it's we want to have better than market average returns in terms of the impact that's achieved as well as financial returns and the balance between the impact and the finance kind of gives you the the the alpha but in addition to this when we think of this in the context of a whole system we're wanting to reduce the negative externalities such as the dependency that gets created through these these types of financing mechanisms and we want to increase positive externalities such as strengthening the education services and this can be done by decentralizing actually enabling the financing to happen at a local level enable local stakeholders to get involved rather than this being in the way that actually what this what the specific bond is is demonstrating a replacement for is a very large development impact bond for 300,000 kids in India which UBS is is implementing which is very centralized the capital is formed in Zurich and everything is kind of done from the lots of intermediaries very high costs and so on and then we want to ensure that that the project failure rate is low and having worked in in the development sector for most of my career so many projects fail you know they just don't deliver what they were supposed to deliver and generally we find out after the fact and there's no course correction and so on and so we really want to make this a more successful way of funding in terms of the outcomes that get achieved now there's this interface that we need between the real world and and the blockchain and so what is that interface you know so it's like the interface in our car is basically the the accelerator and we can see some data about how fast we're going and so on and so the interface is a token bonding curve now for those of you up to most of you know about bonding curves how they work you know you buy into the curve it means the tokens you burn to sell out of the curve is parametric pricing depending on the curve parameters and what this does is regulate supply and and demand in a sort of automatic automated market making kind of way the shortcomings of bonding curves as they've mostly been implemented are that you have fixed a priori assumptions so your parameters for the curve you kind of set them and then you can't change them the curve is kind of just going to be whatever shape and whatever configuration you've set out at at the start in a smart contract and they're not good at responding to new information the only new information is whether there's a buy or a sell but they have no other ways of seeing what's going on on chain or or kind of sensing what's going on in the real world and for most bonding curves they work because of Ponsonomics and to the previous speakers examples you know it's kind of that cancer curve as opposed to the sigmoidal kind of a curve that has an endpoint and so what's useful about having a pricing mechanism like this is that it can incorporate a lot of information so this is the first of two Hayek quotes that I really like and I'm gonna present here so you know prices are an instrument of communication that gives us a lot of information that that we don't necessarily directly have so knowing the price of the bond and corresponding that to the project and its success and its performance is really helpful so this is thanks to Zargham and his team so good systems design and and analysis so we did a year-long project with block science and came up with this mechanism of risk-adjusted bonding curves and so there's two parts to this mechanism the one part is a bonding curve but that changes its parameters based on an input of alpha so alpha is predicting between zero and one the probability of a future payoff payoff so if the impact bond is gonna pay back $100,000 and risk investors are putting in $80,000 to implement to fund the project over time the continuous funding that is brought in through this through this mechanism adjusts the price to whether the project is going to achieve its outcome payment or not is it going to achieve its result and then a payoff that payoff can be anything it can be through the production of impact tokens or impact certificates or it can be you know a financial agreement that if you achieve this this target or this milestone then you'll get paid the other part of this is a prediction market now we haven't implemented this in code yet and so it hasn't been tested out in any sort of real-world context but the prediction market part of this allows for the participants who hold bond tokens to take positions in success and failure pools so it works like a typical prediction market except that it's with people who have skin in the game so you know you're an investor and you're participating in the prediction market this is particularly useful for most projects that don't have secondary markets you know so if you've got a large project and you've got a lot of tokens out there and these liquidity pools and they're sort of trading on the secondary market then you've got a price indicator of essentially a prediction market providing feedback as to the value of the underlying project whereas in this way you don't have to have a secondary market you can have the primary participants in this primary issuance mechanism participating in the in the prediction market now these things do come at a cost you know so if you're going to operate a prediction market you have to have some additional capital to to operate that mechanism that can't be used for the project implementation so that's just one of the one of the downsides of this and this this updates through a through basically like updating priors through a Markov chain mechanism and so you kind of get a smoothing out over time so as new signals are coming in on alpha it's adjusting the system alpha to provide a prediction on whether the outcome is going to be achieved and so this serves as a sort of game people participating in this options market this prediction market and that provides an estimation of of alpha and so here's the other Hayek quotes that prediction markets are mechanisms for collecting vast amounts of information by individuals and synthesizing this into a useful data point data point here being alpha so this is what it looks like sort of schematically so we have an alpha bond which takes in investor capital and a continuous funding mechanism and mints tokens bond tokens as those investments are made into the reserve that sets the bond price it's one of the one of the main things that sets the price is the current reserve that's the relationship between supply and reserve and and then that the curve gets repriced based on this prediction that's coming in this alpha value and so if the project's doing well you know the price goes up the project is not doing well the price goes down and what this effectively does is dilutes the all investors because you're having to issue more tokens at a lower price in order to achieve the the funding target and then ultimately this should this can produce impact tokens or it can just produce an outcome payment and that then repays the investors so yeah this is kind of what it looks like in the in the temple example on our software so you know we've got the curve happening there in our case we're not using a prediction market we're using a signal that's coming in from the tablet devices so we've got software on the tablet device that is issuing claims signing claims it's basically the device is saying the student used used me and got to this level in the educational program and we use that as a signal for calculating the alpha on a continuous basis and those claims come in and get and get verified and so on and so here's an example of an education token that could be could be issued this is conceptual we're not actually issuing these tokens for the for the kids but here we have you know the an idea that services could be that based on this token you could get a personalized learning curriculum access to learning software etc you may have rights to get an educational bursary for instance if you've achieved this accomplishment and linked resources can be qualification credentials very verified attendance claims etc so this information unit becomes usable programmable and it's and it's also data rich and so there is a valuable asset so the vision with this is you know that this provides an autonomous an autonomous replicable adaptive and safe mechanism so it's designed that it has parameters that you can't go outside of and the big vision here is that we can enable much greater capital flows to happen through debt finance or through this mobilizing capital that can achieve regulatory approval because these are essentially securities you know so if you're issuing bonds for investing in solar solar panel installations or for education or whatever the bonds can be for essentially that's a security that you're issuing with some promise of future return and I believe that if we collect enough data and verify the mechanism is safe it should get regulatory approval that this can really just be automatically issued and the analogy here is drones so this picture is a picture taken of a drone swarm which is now permitted you can get a license to fly swarms of drones and the federal aviation authority and enable this because drones have control mechanisms and there's all these sort of safety mechanisms in place and so that's the sort of analogy for scaling yeah so this is just a kind of bonus slide just talking about adaptive finance in the context of the commons and really applying you know the principles of the commons to this that adaptive finance can be used to serve the commons to build new economies and and and use feedback systems and I won't go into all the all the little bits in here but you're welcome to get this a copy of this slide if it's interesting to you so yeah thank you very much